2019-04-14 22:23:52 +02:00
|
|
|
/*******************************************************************************
|
|
|
|
|
|
|
|
uBlock Origin - a browser extension to block requests.
|
|
|
|
Copyright (C) 2019-present Raymond Hill
|
|
|
|
|
|
|
|
This program is free software: you can redistribute it and/or modify
|
|
|
|
it under the terms of the GNU General Public License as published by
|
|
|
|
the Free Software Foundation, either version 3 of the License, or
|
|
|
|
(at your option) any later version.
|
|
|
|
|
|
|
|
This program is distributed in the hope that it will be useful,
|
|
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
GNU General Public License for more details.
|
|
|
|
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
|
|
along with this program. If not, see {http://www.gnu.org/licenses/}.
|
|
|
|
|
|
|
|
Home: https://github.com/gorhill/uBlock
|
|
|
|
*/
|
|
|
|
|
2019-10-26 19:13:53 +02:00
|
|
|
/* globals WebAssembly */
|
|
|
|
|
2019-04-14 22:23:52 +02:00
|
|
|
'use strict';
|
|
|
|
|
2019-06-19 01:16:39 +02:00
|
|
|
// *****************************************************************************
|
|
|
|
// start of local namespace
|
|
|
|
|
|
|
|
{
|
|
|
|
|
2019-04-14 22:23:52 +02:00
|
|
|
/*******************************************************************************
|
|
|
|
|
2019-06-19 01:16:39 +02:00
|
|
|
A BidiTrieContainer is mostly a large buffer in which distinct but related
|
2019-04-14 22:23:52 +02:00
|
|
|
tries are stored. The memory layout of the buffer is as follow:
|
|
|
|
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
0-2047: haystack section
|
|
|
|
2048-2051: number of significant characters in the haystack
|
|
|
|
2052-2055: offset to start of trie data section (=> trie0)
|
|
|
|
2056-2059: offset to end of trie data section (=> trie1)
|
|
|
|
2060-2063: offset to start of character data section (=> char0)
|
|
|
|
2064-2067: offset to end of character data section (=> char1)
|
|
|
|
2068: start of trie data section
|
2019-04-14 22:23:52 +02:00
|
|
|
|
2019-06-19 01:16:39 +02:00
|
|
|
+--------------+
|
|
|
|
Normal cell: | And | If "Segment info" matches:
|
|
|
|
(aka CELL) +--------------+ Goto "And"
|
|
|
|
| Or | Else
|
|
|
|
+--------------+ Goto "Or"
|
|
|
|
| Segment info |
|
|
|
|
+--------------+
|
|
|
|
|
|
|
|
+--------------+
|
|
|
|
Boundary cell: | Right And | "Right And" and/or "Left And"
|
|
|
|
(aka BCELL) +--------------+ can be 0 in last-segment condition.
|
|
|
|
| Left And |
|
|
|
|
+--------------+
|
|
|
|
| 0 |
|
|
|
|
+--------------+
|
|
|
|
|
|
|
|
Given following filters and assuming token is "ad" for all of them:
|
|
|
|
|
|
|
|
-images/ad-
|
|
|
|
/google_ad.
|
|
|
|
/images_ad.
|
|
|
|
_images/ad.
|
|
|
|
|
|
|
|
We get the following internal representation:
|
|
|
|
|
|
|
|
+-----------+ +-----------+ +---+
|
|
|
|
| |---->| |---->| 0 |
|
|
|
|
+-----------+ +-----------+ +---+ +-----------+
|
|
|
|
| 0 | +--| | | |---->| 0 |
|
|
|
|
+-----------+ | +-----------+ +---+ +-----------+
|
|
|
|
| ad | | | - | | 0 | | 0 |
|
|
|
|
+-----------+ | +-----------+ +---+ +-----------+
|
|
|
|
| | -images/ |
|
|
|
|
| +-----------+ +---+ +-----------+
|
|
|
|
+->| |---->| 0 |
|
|
|
|
+-----------+ +---+ +-----------+ +-----------+
|
|
|
|
| 0 | | |---->| |---->| 0 |
|
|
|
|
+-----------+ +---+ +-----------+ +-----------+
|
|
|
|
| . | | 0 | +--| | +--| |
|
|
|
|
+-----------+ +---+ | +-----------+ | +-----------+
|
|
|
|
| | _ | | | /google |
|
|
|
|
| +-----------+ | +-----------+
|
|
|
|
| |
|
|
|
|
| | +-----------+
|
|
|
|
| +->| 0 |
|
|
|
|
| +-----------+
|
|
|
|
| | 0 |
|
|
|
|
| +-----------+
|
|
|
|
| | /images |
|
|
|
|
| +-----------+
|
|
|
|
|
|
|
|
|
| +-----------+
|
|
|
|
+->| 0 |
|
|
|
|
+-----------+
|
|
|
|
| 0 |
|
|
|
|
+-----------+
|
|
|
|
| _images/ |
|
|
|
|
+-----------+
|
|
|
|
|
2019-04-14 22:23:52 +02:00
|
|
|
*/
|
|
|
|
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const PAGE_SIZE = 65536*2;
|
|
|
|
const HAYSTACK_START = 0;
|
2019-10-23 00:14:49 +02:00
|
|
|
const HAYSTACK_SIZE = 2048; // i32 / i8
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const HAYSTACK_SIZE_SLOT = HAYSTACK_SIZE >>> 2; // 512 / 2048
|
2019-10-28 18:57:35 +01:00
|
|
|
const TRIE0_SLOT = HAYSTACK_SIZE_SLOT + 1; // 513 / 2052
|
|
|
|
const TRIE1_SLOT = HAYSTACK_SIZE_SLOT + 2; // 514 / 2056
|
|
|
|
const CHAR0_SLOT = HAYSTACK_SIZE_SLOT + 3; // 515 / 2060
|
|
|
|
const CHAR1_SLOT = HAYSTACK_SIZE_SLOT + 4; // 516 / 2064
|
|
|
|
const RESULT_L_SLOT = HAYSTACK_SIZE_SLOT + 5; // 517 / 2068
|
|
|
|
const RESULT_R_SLOT = HAYSTACK_SIZE_SLOT + 6; // 518 / 2072
|
|
|
|
const RESULT_IU_SLOT = HAYSTACK_SIZE_SLOT + 7; // 519 / 2076
|
|
|
|
const TRIE0_START = HAYSTACK_SIZE_SLOT + 8 << 2; // 2080
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// TODO: need a few slots for result values if WASM-ing
|
2019-06-19 01:16:39 +02:00
|
|
|
|
|
|
|
const CELL_BYTE_LENGTH = 12;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const MIN_FREE_CELL_BYTE_LENGTH = CELL_BYTE_LENGTH * 8;
|
2019-04-14 22:23:52 +02:00
|
|
|
|
2019-06-19 01:16:39 +02:00
|
|
|
const CELL_AND = 0;
|
|
|
|
const CELL_OR = 1;
|
|
|
|
const SEGMENT_INFO = 2;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const BCELL_NEXT_AND = 0;
|
|
|
|
const BCELL_ALT_AND = 1;
|
|
|
|
const BCELL_EXTRA = 2;
|
|
|
|
const BCELL_EXTRA_MAX = 0x00FFFFFF;
|
|
|
|
|
|
|
|
const toSegmentInfo = (aL, l, r) => ((r - l) << 24) | (aL + l);
|
|
|
|
const roundToPageSize = v => (v + PAGE_SIZE-1) & ~(PAGE_SIZE-1);
|
2019-04-14 22:23:52 +02:00
|
|
|
|
2019-06-19 01:16:39 +02:00
|
|
|
|
|
|
|
µBlock.BidiTrieContainer = class {
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
constructor(details, extraHandler) {
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
if ( details instanceof Object === false ) { details = {}; }
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const len = roundToPageSize(details.byteLength || 0);
|
2019-10-26 19:13:53 +02:00
|
|
|
const minInitialSize = PAGE_SIZE * 4;
|
|
|
|
this.buf8 = new Uint8Array(Math.max(len, minInitialSize));
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
this.buf32 = new Uint32Array(this.buf8.buffer);
|
2019-06-19 01:16:39 +02:00
|
|
|
this.buf32[TRIE0_SLOT] = TRIE0_START;
|
|
|
|
this.buf32[TRIE1_SLOT] = this.buf32[TRIE0_SLOT];
|
2019-10-26 19:13:53 +02:00
|
|
|
this.buf32[CHAR0_SLOT] = details.char0 || (minInitialSize >>> 1);
|
2019-06-19 01:16:39 +02:00
|
|
|
this.buf32[CHAR1_SLOT] = this.buf32[CHAR0_SLOT];
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
this.haystack = this.buf8.subarray(
|
|
|
|
HAYSTACK_START,
|
|
|
|
HAYSTACK_START + HAYSTACK_SIZE
|
|
|
|
);
|
|
|
|
this.extraHandler = extraHandler;
|
|
|
|
this.textDecoder = null;
|
2019-10-26 19:13:53 +02:00
|
|
|
this.wasmMemory = null;
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
|
|
|
//--------------------------------------------------------------------------
|
|
|
|
// Public methods
|
|
|
|
//--------------------------------------------------------------------------
|
|
|
|
|
2019-10-28 18:57:35 +01:00
|
|
|
get haystackLen() {
|
|
|
|
return this.buf32[HAYSTACK_SIZE_SLOT];
|
|
|
|
}
|
|
|
|
|
|
|
|
set haystackLen(v) {
|
|
|
|
this.buf32[HAYSTACK_SIZE_SLOT] = v;
|
|
|
|
}
|
|
|
|
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
reset() {
|
2019-06-19 01:16:39 +02:00
|
|
|
this.buf32[TRIE1_SLOT] = this.buf32[TRIE0_SLOT];
|
|
|
|
this.buf32[CHAR1_SLOT] = this.buf32[CHAR0_SLOT];
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
2019-10-28 18:57:35 +01:00
|
|
|
matches(icell, ai) {
|
2019-06-19 01:16:39 +02:00
|
|
|
const buf32 = this.buf32;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const buf8 = this.buf8;
|
2019-06-19 01:16:39 +02:00
|
|
|
const char0 = buf32[CHAR0_SLOT];
|
2019-10-28 18:57:35 +01:00
|
|
|
const aR = buf32[HAYSTACK_SIZE_SLOT];
|
2019-10-29 15:26:34 +01:00
|
|
|
let al = ai, x = 0, y = 0;
|
2019-04-14 22:23:52 +02:00
|
|
|
for (;;) {
|
2019-10-29 15:26:34 +01:00
|
|
|
x = buf8[al];
|
2019-04-14 22:23:52 +02:00
|
|
|
al += 1;
|
2019-10-29 15:26:34 +01:00
|
|
|
// find matching segment
|
2019-04-14 22:23:52 +02:00
|
|
|
for (;;) {
|
2019-10-29 15:26:34 +01:00
|
|
|
y = buf32[icell+SEGMENT_INFO];
|
|
|
|
let bl = char0 + (y & 0x00FFFFFF);
|
|
|
|
if ( buf8[bl] === x ) {
|
|
|
|
y = (y >>> 24) - 1;
|
|
|
|
if ( y !== 0 ) {
|
|
|
|
x = al + y;
|
|
|
|
if ( x > aR ) { return 0; }
|
|
|
|
for (;;) {
|
|
|
|
bl += 1;
|
|
|
|
if ( buf8[bl] !== buf8[al] ) { return 0; }
|
|
|
|
al += 1;
|
|
|
|
if ( al === x ) { break; }
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
2019-06-19 01:16:39 +02:00
|
|
|
icell = buf32[icell+CELL_OR];
|
2019-10-28 18:57:35 +01:00
|
|
|
if ( icell === 0 ) { return 0; }
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
|
|
|
// next segment
|
2019-06-19 01:16:39 +02:00
|
|
|
icell = buf32[icell+CELL_AND];
|
2019-10-29 15:26:34 +01:00
|
|
|
x = buf32[icell+BCELL_EXTRA];
|
|
|
|
if ( x <= BCELL_EXTRA_MAX ) {
|
|
|
|
if ( x !== 0 && this.matchesExtra(ai, al, x) !== 0 ) {
|
2019-10-28 18:57:35 +01:00
|
|
|
return 1;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
2019-10-29 15:26:34 +01:00
|
|
|
x = buf32[icell+BCELL_ALT_AND];
|
|
|
|
if ( x !== 0 && this.matchesLeft(x, ai, al) !== 0 ) {
|
2019-10-28 18:57:35 +01:00
|
|
|
return 1;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
2019-10-28 18:57:35 +01:00
|
|
|
icell = buf32[icell+BCELL_NEXT_AND];
|
|
|
|
if ( icell === 0 ) { return 0; }
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
2019-10-28 18:57:35 +01:00
|
|
|
if ( al === aR ) { return 0; }
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
2019-10-28 18:57:35 +01:00
|
|
|
return 0;
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
|
|
|
|
2019-10-29 15:26:34 +01:00
|
|
|
matchesLeft(icell, ar, r) {
|
2019-06-19 01:16:39 +02:00
|
|
|
const buf32 = this.buf32;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const buf8 = this.buf8;
|
2019-06-19 01:16:39 +02:00
|
|
|
const char0 = buf32[CHAR0_SLOT];
|
2019-10-29 15:26:34 +01:00
|
|
|
let x = 0, y = 0;
|
2019-06-19 01:16:39 +02:00
|
|
|
for (;;) {
|
2019-10-28 18:57:35 +01:00
|
|
|
if ( ar === 0 ) { return 0; }
|
2019-06-19 01:16:39 +02:00
|
|
|
ar -= 1;
|
2019-10-29 15:26:34 +01:00
|
|
|
x = buf8[ar];
|
2019-06-19 01:16:39 +02:00
|
|
|
// find first segment with a first-character match
|
|
|
|
for (;;) {
|
2019-10-29 15:26:34 +01:00
|
|
|
y = buf32[icell+SEGMENT_INFO];
|
|
|
|
let br = char0 + (y & 0x00FFFFFF);
|
|
|
|
y = (y >>> 24) - 1;
|
|
|
|
br += y;
|
|
|
|
if ( buf8[br] === x ) { // all characters in segment must match
|
|
|
|
if ( y !== 0 ) {
|
|
|
|
x = ar - y;
|
|
|
|
if ( x < 0 ) { return 0; }
|
|
|
|
for (;;) {
|
|
|
|
ar -= 1; br -= 1;
|
|
|
|
if ( buf8[ar] !== buf8[br] ) { return 0; }
|
|
|
|
if ( ar === x ) { break; }
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
2019-06-19 01:16:39 +02:00
|
|
|
icell = buf32[icell+CELL_OR];
|
2019-10-28 18:57:35 +01:00
|
|
|
if ( icell === 0 ) { return 0; }
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
|
|
|
// next segment
|
|
|
|
icell = buf32[icell+CELL_AND];
|
2019-10-29 15:26:34 +01:00
|
|
|
x = buf32[icell+BCELL_EXTRA];
|
|
|
|
if ( x <= BCELL_EXTRA_MAX ) {
|
|
|
|
if ( x !== 0 && this.matchesExtra(ar, r, x) !== 0 ) {
|
2019-10-28 18:57:35 +01:00
|
|
|
return 1;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
|
|
|
icell = buf32[icell+BCELL_NEXT_AND];
|
2019-10-28 18:57:35 +01:00
|
|
|
if ( icell === 0 ) { return 0; }
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
2019-10-28 18:57:35 +01:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
matchesExtra(l, r, ix) {
|
2019-10-29 15:26:34 +01:00
|
|
|
let iu = 0;
|
2019-10-28 18:57:35 +01:00
|
|
|
if ( ix !== 1 ) {
|
|
|
|
iu = this.extraHandler(l, r, ix);
|
|
|
|
if ( iu === 0 ) { return 0; }
|
|
|
|
} else {
|
|
|
|
iu = -1;
|
|
|
|
}
|
2019-10-29 15:26:34 +01:00
|
|
|
this.buf32[RESULT_IU_SLOT] = iu;
|
2019-10-28 18:57:35 +01:00
|
|
|
this.buf32[RESULT_L_SLOT] = l;
|
|
|
|
this.buf32[RESULT_R_SLOT] = r;
|
|
|
|
return 1;
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
createOne(args) {
|
2019-04-14 22:23:52 +02:00
|
|
|
if ( Array.isArray(args) ) {
|
|
|
|
return new this.STrieRef(this, args[0], args[1]);
|
|
|
|
}
|
|
|
|
// grow buffer if needed
|
2019-06-19 01:16:39 +02:00
|
|
|
if ( (this.buf32[CHAR0_SLOT] - this.buf32[TRIE1_SLOT]) < CELL_BYTE_LENGTH ) {
|
|
|
|
this.growBuf(CELL_BYTE_LENGTH, 0);
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
2019-06-19 01:16:39 +02:00
|
|
|
const iroot = this.buf32[TRIE1_SLOT] >>> 2;
|
|
|
|
this.buf32[TRIE1_SLOT] += CELL_BYTE_LENGTH;
|
|
|
|
this.buf32[iroot+CELL_OR] = 0;
|
|
|
|
this.buf32[iroot+CELL_AND] = 0;
|
|
|
|
this.buf32[iroot+SEGMENT_INFO] = 0;
|
2019-04-14 22:23:52 +02:00
|
|
|
return new this.STrieRef(this, iroot, 0);
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
compileOne(trieRef) {
|
2019-04-14 22:23:52 +02:00
|
|
|
return [ trieRef.iroot, trieRef.size ];
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
add(iroot, aL0, n, pivot = 0) {
|
|
|
|
const aR = n;
|
2019-06-19 01:16:39 +02:00
|
|
|
if ( aR === 0 ) { return 0; }
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// Grow buffer if needed. The characters are already in our character
|
|
|
|
// data buffer, so we do not need to grow character data buffer.
|
2019-04-14 22:23:52 +02:00
|
|
|
if (
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
(this.buf32[CHAR0_SLOT] - this.buf32[TRIE1_SLOT]) <
|
|
|
|
MIN_FREE_CELL_BYTE_LENGTH
|
2019-04-14 22:23:52 +02:00
|
|
|
) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
this.growBuf(MIN_FREE_CELL_BYTE_LENGTH, 0);
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
2019-06-19 01:16:39 +02:00
|
|
|
const buf32 = this.buf32;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const char0 = buf32[CHAR0_SLOT];
|
2019-08-22 23:11:49 +02:00
|
|
|
let icell = iroot;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
let aL = char0 + aL0;
|
2019-08-22 23:11:49 +02:00
|
|
|
// special case: first node in trie
|
|
|
|
if ( buf32[icell+SEGMENT_INFO] === 0 ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
buf32[icell+SEGMENT_INFO] = toSegmentInfo(aL0, pivot, aR);
|
|
|
|
return this.addLeft(icell, aL0, pivot);
|
2019-08-22 23:11:49 +02:00
|
|
|
}
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const buf8 = this.buf8;
|
|
|
|
let al = pivot;
|
2019-04-14 22:23:52 +02:00
|
|
|
let inext;
|
|
|
|
// find a matching cell: move down
|
|
|
|
for (;;) {
|
2019-06-19 01:16:39 +02:00
|
|
|
const binfo = buf32[icell+SEGMENT_INFO];
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// length of segment
|
|
|
|
const bR = binfo >>> 24;
|
2019-06-19 01:16:39 +02:00
|
|
|
// skip boundary cells
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
if ( bR === 0 ) {
|
|
|
|
icell = buf32[icell+BCELL_NEXT_AND];
|
2019-06-19 01:16:39 +02:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
let bl = char0 + (binfo & 0x00FFFFFF);
|
|
|
|
// if first character is no match, move to next descendant
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
if ( buf8[bl] !== buf8[aL+al] ) {
|
2019-06-19 01:16:39 +02:00
|
|
|
inext = buf32[icell+CELL_OR];
|
|
|
|
if ( inext === 0 ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
inext = this.addCell(0, 0, toSegmentInfo(aL0, al, aR));
|
2019-06-19 01:16:39 +02:00
|
|
|
buf32[icell+CELL_OR] = inext;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
return this.addLeft(inext, aL0, pivot);
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
|
|
|
icell = inext;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
// 1st character was tested
|
|
|
|
let bi = 1;
|
|
|
|
al += 1;
|
|
|
|
// find 1st mismatch in rest of segment
|
|
|
|
if ( bR !== 1 ) {
|
|
|
|
for (;;) {
|
|
|
|
if ( bi === bR ) { break; }
|
|
|
|
if ( al === aR ) { break; }
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
if ( buf8[bl+bi] !== buf8[aL+al] ) { break; }
|
2019-06-19 01:16:39 +02:00
|
|
|
bi += 1;
|
|
|
|
al += 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
// all segment characters matched
|
|
|
|
if ( bi === bR ) {
|
|
|
|
// needle remainder: no
|
|
|
|
if ( al === aR ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
return this.addLeft(icell, aL0, pivot);
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
|
|
|
// needle remainder: yes
|
|
|
|
inext = buf32[icell+CELL_AND];
|
|
|
|
if ( buf32[inext+CELL_AND] !== 0 ) {
|
|
|
|
icell = inext;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
// add needle remainder
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
icell = this.addCell(0, 0, toSegmentInfo(aL0, al, aR));
|
2019-06-19 01:16:39 +02:00
|
|
|
buf32[inext+CELL_AND] = icell;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
return this.addLeft(icell, aL0, pivot);
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
|
|
|
// some characters matched
|
|
|
|
// split current segment
|
|
|
|
bl -= char0;
|
|
|
|
buf32[icell+SEGMENT_INFO] = bi << 24 | bl;
|
|
|
|
inext = this.addCell(
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
buf32[icell+CELL_AND], 0, bR - bi << 24 | bl + bi
|
2019-06-19 01:16:39 +02:00
|
|
|
);
|
|
|
|
buf32[icell+CELL_AND] = inext;
|
|
|
|
// needle remainder: no = need boundary cell
|
|
|
|
if ( al === aR ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
return this.addLeft(icell, aL0, pivot);
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
|
|
|
// needle remainder: yes = need new cell for remaining characters
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
icell = this.addCell(0, 0, toSegmentInfo(aL0, al, aR));
|
2019-06-19 01:16:39 +02:00
|
|
|
buf32[inext+CELL_OR] = icell;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
return this.addLeft(icell, aL0, pivot);
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
addLeft(icell, aL0, pivot) {
|
2019-06-19 01:16:39 +02:00
|
|
|
const buf32 = this.buf32;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const char0 = buf32[CHAR0_SLOT];
|
|
|
|
let aL = aL0 + char0;
|
2019-06-19 01:16:39 +02:00
|
|
|
// fetch boundary cell
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
let iboundary = buf32[icell+CELL_AND];
|
2019-06-19 01:16:39 +02:00
|
|
|
// add boundary cell if none exist
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
if (
|
|
|
|
iboundary === 0 ||
|
|
|
|
buf32[iboundary+SEGMENT_INFO] > BCELL_EXTRA_MAX
|
|
|
|
) {
|
|
|
|
const inext = iboundary;
|
|
|
|
iboundary = this.allocateCell();
|
2019-06-19 01:16:39 +02:00
|
|
|
buf32[icell+CELL_AND] = iboundary;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
buf32[iboundary+BCELL_NEXT_AND] = inext;
|
|
|
|
if ( pivot === 0 ) { return iboundary; }
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// shortest match with no extra conditions will always win
|
|
|
|
if ( buf32[iboundary+BCELL_EXTRA] === 1 ) {
|
|
|
|
return iboundary;
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
|
|
|
// bail out if no left segment
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
if ( pivot === 0 ) { return iboundary; }
|
2019-06-19 01:16:39 +02:00
|
|
|
// fetch root cell of left segment
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
icell = buf32[iboundary+BCELL_ALT_AND];
|
|
|
|
if ( icell === 0 ) {
|
|
|
|
icell = this.allocateCell();
|
|
|
|
buf32[iboundary+BCELL_ALT_AND] = icell;
|
|
|
|
}
|
2019-06-19 01:16:39 +02:00
|
|
|
// special case: first node in trie
|
|
|
|
if ( buf32[icell+SEGMENT_INFO] === 0 ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
buf32[icell+SEGMENT_INFO] = toSegmentInfo(aL0, 0, pivot);
|
|
|
|
iboundary = this.allocateCell();
|
|
|
|
buf32[icell+CELL_AND] = iboundary;
|
|
|
|
return iboundary;
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
const buf8 = this.buf8;
|
|
|
|
let ar = pivot, inext;
|
2019-06-19 01:16:39 +02:00
|
|
|
// find a matching cell: move down
|
|
|
|
for (;;) {
|
|
|
|
const binfo = buf32[icell+SEGMENT_INFO];
|
2019-04-14 22:23:52 +02:00
|
|
|
// skip boundary cells
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
if ( binfo <= BCELL_EXTRA_MAX ) {
|
|
|
|
inext = buf32[icell+CELL_AND];
|
|
|
|
if ( inext !== 0 ) {
|
|
|
|
icell = inext;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
iboundary = this.allocateCell();
|
|
|
|
buf32[icell+CELL_AND] =
|
|
|
|
this.addCell(iboundary, 0, toSegmentInfo(aL0, 0, ar));
|
|
|
|
// TODO: boundary cell might be last
|
|
|
|
// add remainder + boundary cell
|
|
|
|
return iboundary;
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
2019-06-19 01:16:39 +02:00
|
|
|
const bL = char0 + (binfo & 0x00FFFFFF);
|
|
|
|
const bR = bL + (binfo >>> 24);
|
|
|
|
let br = bR;
|
2019-04-14 22:23:52 +02:00
|
|
|
// if first character is no match, move to next descendant
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
if ( buf8[br-1] !== buf8[aL+ar-1] ) {
|
2019-06-19 01:16:39 +02:00
|
|
|
inext = buf32[icell+CELL_OR];
|
2019-04-14 22:23:52 +02:00
|
|
|
if ( inext === 0 ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
iboundary = this.allocateCell();
|
|
|
|
inext = this.addCell(
|
|
|
|
iboundary, 0, toSegmentInfo(aL0, 0, ar)
|
|
|
|
);
|
2019-06-19 01:16:39 +02:00
|
|
|
buf32[icell+CELL_OR] = inext;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
return iboundary;
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
|
|
|
icell = inext;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
// 1st character was tested
|
2019-06-19 01:16:39 +02:00
|
|
|
br -= 1;
|
|
|
|
ar -= 1;
|
2019-04-14 22:23:52 +02:00
|
|
|
// find 1st mismatch in rest of segment
|
2019-06-19 01:16:39 +02:00
|
|
|
if ( br !== bL ) {
|
2019-04-14 22:23:52 +02:00
|
|
|
for (;;) {
|
2019-06-19 01:16:39 +02:00
|
|
|
if ( br === bL ) { break; }
|
|
|
|
if ( ar === 0 ) { break; }
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
if ( buf8[br-1] !== buf8[aL+ar-1] ) { break; }
|
2019-06-19 01:16:39 +02:00
|
|
|
br -= 1;
|
|
|
|
ar -= 1;
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
// all segment characters matched
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// a: ...vvvvvvv
|
|
|
|
// b: vvvvvvv
|
2019-06-19 01:16:39 +02:00
|
|
|
if ( br === bL ) {
|
|
|
|
inext = buf32[icell+CELL_AND];
|
2019-04-14 22:23:52 +02:00
|
|
|
// needle remainder: no
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// a: vvvvvvv
|
|
|
|
// b: vvvvvvv
|
|
|
|
// r: 0 & vvvvvvv
|
2019-06-19 01:16:39 +02:00
|
|
|
if ( ar === 0 ) {
|
2019-04-14 22:23:52 +02:00
|
|
|
// boundary cell already present
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
if ( buf32[inext+BCELL_EXTRA] <= BCELL_EXTRA_MAX ) {
|
|
|
|
return inext;
|
2019-06-19 01:16:39 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
// need boundary cell
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
iboundary = this.allocateCell();
|
|
|
|
buf32[iboundary+CELL_AND] = inext;
|
|
|
|
buf32[icell+CELL_AND] = iboundary;
|
|
|
|
return iboundary;
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
|
|
|
// needle remainder: yes
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// a: yyyyyyyvvvvvvv
|
|
|
|
// b: vvvvvvv
|
2019-04-14 22:23:52 +02:00
|
|
|
else {
|
|
|
|
if ( inext !== 0 ) {
|
|
|
|
icell = inext;
|
|
|
|
continue;
|
|
|
|
}
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// TODO: we should never reach here because there will
|
|
|
|
// always be a boundary cell.
|
|
|
|
debugger; // jshint ignore:line
|
2019-04-14 22:23:52 +02:00
|
|
|
// boundary cell + needle remainder
|
|
|
|
inext = this.addCell(0, 0, 0);
|
2019-06-19 01:16:39 +02:00
|
|
|
buf32[icell+CELL_AND] = inext;
|
|
|
|
buf32[inext+CELL_AND] =
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
this.addCell(0, 0, toSegmentInfo(aL0, 0, ar));
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
// some segment characters matched
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// a: ...vvvvvvv
|
|
|
|
// b: yyyyyyyvvvvvvv
|
2019-04-14 22:23:52 +02:00
|
|
|
else {
|
|
|
|
// split current cell
|
2019-06-19 01:16:39 +02:00
|
|
|
buf32[icell+SEGMENT_INFO] = (bR - br) << 24 | (br - char0);
|
2019-04-14 22:23:52 +02:00
|
|
|
inext = this.addCell(
|
2019-06-19 01:16:39 +02:00
|
|
|
buf32[icell+CELL_AND],
|
2019-04-14 22:23:52 +02:00
|
|
|
0,
|
2019-06-19 01:16:39 +02:00
|
|
|
(br - bL) << 24 | (bL - char0)
|
2019-04-14 22:23:52 +02:00
|
|
|
);
|
|
|
|
// needle remainder: no = need boundary cell
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// a: vvvvvvv
|
|
|
|
// b: yyyyyyyvvvvvvv
|
|
|
|
// r: yyyyyyy & 0 & vvvvvvv
|
2019-06-19 01:16:39 +02:00
|
|
|
if ( ar === 0 ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
iboundary = this.allocateCell();
|
|
|
|
buf32[icell+CELL_AND] = iboundary;
|
|
|
|
buf32[iboundary+CELL_AND] = inext;
|
|
|
|
return iboundary;
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
// needle remainder: yes = need new cell for remaining
|
|
|
|
// characters
|
|
|
|
// a: wwwwvvvvvvv
|
|
|
|
// b: yyyyyyyvvvvvvv
|
|
|
|
// r: (0 & wwww | yyyyyyy) & vvvvvvv
|
2019-04-14 22:23:52 +02:00
|
|
|
else {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
buf32[icell+CELL_AND] = inext;
|
|
|
|
iboundary = this.allocateCell();
|
|
|
|
buf32[inext+CELL_OR] = this.addCell(
|
|
|
|
iboundary, 0, toSegmentInfo(aL0, 0, ar)
|
|
|
|
);
|
|
|
|
return iboundary;
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
|
|
|
}
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
//debugger; // jshint ignore:line
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
2019-10-27 13:36:17 +01:00
|
|
|
optimize(shrink = false) {
|
|
|
|
if ( shrink ) {
|
|
|
|
this.shrinkBuf();
|
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
return {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
byteLength: this.buf8.byteLength,
|
2019-06-19 01:16:39 +02:00
|
|
|
char0: this.buf32[CHAR0_SLOT],
|
2019-04-14 22:23:52 +02:00
|
|
|
};
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
serialize(encoder) {
|
2019-04-14 22:23:52 +02:00
|
|
|
if ( encoder instanceof Object ) {
|
|
|
|
return encoder.encode(
|
|
|
|
this.buf32.buffer,
|
2019-06-19 01:16:39 +02:00
|
|
|
this.buf32[CHAR1_SLOT]
|
2019-04-14 22:23:52 +02:00
|
|
|
);
|
|
|
|
}
|
|
|
|
return Array.from(
|
|
|
|
new Uint32Array(
|
|
|
|
this.buf32.buffer,
|
|
|
|
0,
|
2019-06-19 01:16:39 +02:00
|
|
|
this.buf32[CHAR1_SLOT] + 3 >>> 2
|
2019-04-14 22:23:52 +02:00
|
|
|
)
|
|
|
|
);
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
unserialize(selfie, decoder) {
|
2019-04-14 22:23:52 +02:00
|
|
|
const shouldDecode = typeof selfie === 'string';
|
|
|
|
let byteLength = shouldDecode
|
|
|
|
? decoder.decodeSize(selfie)
|
|
|
|
: selfie.length << 2;
|
2019-04-20 15:06:54 +02:00
|
|
|
if ( byteLength === 0 ) { return false; }
|
2019-10-27 13:36:17 +01:00
|
|
|
this.reallocateBuf(byteLength);
|
2019-04-14 22:23:52 +02:00
|
|
|
if ( shouldDecode ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
decoder.decode(selfie, this.buf8.buffer);
|
2019-04-14 22:23:52 +02:00
|
|
|
} else {
|
|
|
|
this.buf32.set(selfie);
|
|
|
|
}
|
2019-04-20 15:06:54 +02:00
|
|
|
return true;
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
storeString(s) {
|
|
|
|
const n = s.length;
|
|
|
|
if ( (this.buf8.length - this.buf32[CHAR1_SLOT]) < n ) {
|
|
|
|
this.growBuf(0, n);
|
|
|
|
}
|
|
|
|
const offset = this.buf32[CHAR1_SLOT];
|
|
|
|
this.buf32[CHAR1_SLOT] = offset + n;
|
|
|
|
const buf8 = this.buf8;
|
|
|
|
for ( let i = 0; i < n; i++ ) {
|
|
|
|
buf8[offset+i] = s.charCodeAt(i);
|
|
|
|
}
|
|
|
|
return offset - this.buf32[CHAR0_SLOT];
|
|
|
|
}
|
|
|
|
|
|
|
|
extractString(i, n) {
|
|
|
|
if ( this.textDecoder === null ) {
|
|
|
|
this.textDecoder = new TextDecoder();
|
|
|
|
}
|
|
|
|
const offset = this.buf32[CHAR0_SLOT] + i;
|
|
|
|
return this.textDecoder.decode(
|
|
|
|
this.buf8.subarray(offset, offset + n)
|
|
|
|
);
|
|
|
|
}
|
|
|
|
|
|
|
|
// WASMable.
|
2019-10-26 19:13:53 +02:00
|
|
|
startsWith(haystackLeft, haystackRight, needleLeft, needleLen) {
|
|
|
|
if ( haystackLeft < 0 || (haystackLeft + needleLen) > haystackRight ) {
|
|
|
|
return 0;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
2019-10-26 19:13:53 +02:00
|
|
|
const charCodes = this.buf8;
|
|
|
|
needleLeft += this.buf32[CHAR0_SLOT];
|
|
|
|
const needleRight = needleLeft + needleLen;
|
|
|
|
while ( charCodes[haystackLeft] === charCodes[needleLeft] ) {
|
|
|
|
needleLeft += 1;
|
|
|
|
if ( needleLeft === needleRight ) { return 1; }
|
|
|
|
haystackLeft += 1;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
2019-10-26 19:13:53 +02:00
|
|
|
return 0;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
// Find the left-most instance of substring in main string
|
|
|
|
// WASMable.
|
2019-10-26 19:13:53 +02:00
|
|
|
indexOf(haystackLeft, haystackEnd, needleLeft, needleLen) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
haystackEnd -= needleLen;
|
2019-10-26 19:13:53 +02:00
|
|
|
if ( haystackEnd < haystackLeft ) { return -1; }
|
|
|
|
needleLeft += this.buf32[CHAR0_SLOT];
|
|
|
|
const needleRight = needleLeft + needleLen;
|
|
|
|
const charCodes = this.buf8;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
for (;;) {
|
2019-10-26 19:13:53 +02:00
|
|
|
let i = haystackLeft;
|
|
|
|
let j = needleLeft;
|
|
|
|
while ( charCodes[i] === charCodes[j] ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
j += 1;
|
2019-10-26 19:13:53 +02:00
|
|
|
if ( j === needleRight ) { return haystackLeft; }
|
|
|
|
i += 1;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
2019-10-26 19:13:53 +02:00
|
|
|
haystackLeft += 1;
|
|
|
|
if ( haystackLeft === haystackEnd ) { break; }
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Find the right-most instance of substring in main string.
|
|
|
|
// WASMable.
|
2019-10-26 19:13:53 +02:00
|
|
|
lastIndexOf(haystackBeg, haystackEnd, needleLeft, needleLen) {
|
|
|
|
let haystackLeft = haystackEnd - needleLen;
|
|
|
|
if ( haystackLeft < haystackBeg ) { return -1; }
|
|
|
|
needleLeft += this.buf32[CHAR0_SLOT];
|
|
|
|
const needleRight = needleLeft + needleLen;
|
|
|
|
const charCodes = this.buf8;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
for (;;) {
|
2019-10-26 19:13:53 +02:00
|
|
|
let i = haystackLeft;
|
|
|
|
let j = needleLeft;
|
|
|
|
while ( charCodes[i] === charCodes[j] ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
j += 1;
|
2019-10-26 19:13:53 +02:00
|
|
|
if ( j === needleRight ) { return haystackLeft; }
|
|
|
|
i += 1;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
2019-10-26 19:13:53 +02:00
|
|
|
if ( haystackLeft === haystackBeg ) { break; }
|
|
|
|
haystackLeft -= 1;
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
}
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
2019-10-26 19:13:53 +02:00
|
|
|
async enableWASM() {
|
|
|
|
if ( this.wasmMemory instanceof WebAssembly.Memory ) { return true; }
|
|
|
|
const module = await getWasmModule();
|
|
|
|
if ( module instanceof WebAssembly.Module === false ) {
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
const memory = new WebAssembly.Memory({
|
|
|
|
initial: this.buf8.length >>> 16
|
|
|
|
});
|
|
|
|
const instance = await WebAssembly.instantiate(
|
|
|
|
module,
|
2019-10-28 18:57:35 +01:00
|
|
|
{ imports: { memory, extraHandler: this.extraHandler } }
|
2019-10-26 19:13:53 +02:00
|
|
|
);
|
|
|
|
if ( instance instanceof WebAssembly.Instance === false ) {
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
this.wasmMemory = memory;
|
|
|
|
const curPageCount = memory.buffer.byteLength >>> 16;
|
|
|
|
const newPageCount = this.buf8.byteLength + PAGE_SIZE-1 >>> 16;
|
|
|
|
if ( newPageCount > curPageCount ) {
|
|
|
|
memory.grow(newPageCount - curPageCount);
|
|
|
|
}
|
|
|
|
const buf8 = new Uint8Array(memory.buffer);
|
|
|
|
buf8.set(this.buf8);
|
|
|
|
this.buf8 = buf8;
|
|
|
|
this.buf32 = new Uint32Array(this.buf8.buffer);
|
|
|
|
this.haystack = this.buf8.subarray(
|
|
|
|
HAYSTACK_START,
|
|
|
|
HAYSTACK_START + HAYSTACK_SIZE
|
|
|
|
);
|
2019-10-28 18:57:35 +01:00
|
|
|
this.matches = instance.exports.matches;
|
2019-10-26 19:13:53 +02:00
|
|
|
this.startsWith = instance.exports.startsWith;
|
|
|
|
this.indexOf = instance.exports.indexOf;
|
|
|
|
this.lastIndexOf = instance.exports.lastIndexOf;
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
2019-04-14 22:23:52 +02:00
|
|
|
//--------------------------------------------------------------------------
|
|
|
|
// Private methods
|
|
|
|
//--------------------------------------------------------------------------
|
|
|
|
|
2019-06-19 01:16:39 +02:00
|
|
|
allocateCell() {
|
|
|
|
let icell = this.buf32[TRIE1_SLOT];
|
|
|
|
this.buf32[TRIE1_SLOT] = icell + CELL_BYTE_LENGTH;
|
2019-04-14 22:23:52 +02:00
|
|
|
icell >>>= 2;
|
2019-06-19 01:16:39 +02:00
|
|
|
this.buf32[icell+0] = 0;
|
|
|
|
this.buf32[icell+1] = 0;
|
|
|
|
this.buf32[icell+2] = 0;
|
|
|
|
return icell;
|
|
|
|
}
|
|
|
|
|
|
|
|
addCell(iand, ior, v) {
|
|
|
|
const icell = this.allocateCell();
|
|
|
|
this.buf32[icell+CELL_AND] = iand;
|
|
|
|
this.buf32[icell+CELL_OR] = ior;
|
|
|
|
this.buf32[icell+SEGMENT_INFO] = v;
|
2019-04-14 22:23:52 +02:00
|
|
|
return icell;
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
growBuf(trieGrow, charGrow) {
|
2019-04-14 22:23:52 +02:00
|
|
|
const char0 = Math.max(
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
roundToPageSize(this.buf32[TRIE1_SLOT] + trieGrow),
|
2019-06-19 01:16:39 +02:00
|
|
|
this.buf32[CHAR0_SLOT]
|
2019-04-14 22:23:52 +02:00
|
|
|
);
|
2019-06-19 01:16:39 +02:00
|
|
|
const char1 = char0 + this.buf32[CHAR1_SLOT] - this.buf32[CHAR0_SLOT];
|
2019-04-14 22:23:52 +02:00
|
|
|
const bufLen = Math.max(
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
roundToPageSize(char1 + charGrow),
|
|
|
|
this.buf8.length
|
2019-04-14 22:23:52 +02:00
|
|
|
);
|
2019-10-27 13:36:17 +01:00
|
|
|
if ( bufLen > this.buf8.length ) {
|
|
|
|
this.reallocateBuf(bufLen);
|
|
|
|
}
|
|
|
|
if ( char0 !== this.buf32[CHAR0_SLOT] ) {
|
|
|
|
this.buf8.copyWithin(
|
|
|
|
char0,
|
|
|
|
this.buf32[CHAR0_SLOT],
|
|
|
|
this.buf32[CHAR1_SLOT]
|
|
|
|
);
|
|
|
|
this.buf32[CHAR0_SLOT] = char0;
|
|
|
|
this.buf32[CHAR1_SLOT] = char1;
|
|
|
|
}
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
shrinkBuf() {
|
2019-06-19 01:16:39 +02:00
|
|
|
const char0 = this.buf32[TRIE1_SLOT] + MIN_FREE_CELL_BYTE_LENGTH;
|
|
|
|
const char1 = char0 + this.buf32[CHAR1_SLOT] - this.buf32[CHAR0_SLOT];
|
2019-04-14 22:23:52 +02:00
|
|
|
const bufLen = char1 + 256;
|
2019-06-19 01:16:39 +02:00
|
|
|
if ( char0 !== this.buf32[CHAR0_SLOT] ) {
|
2019-10-26 19:13:53 +02:00
|
|
|
this.buf8.copyWithin(
|
|
|
|
char0,
|
|
|
|
this.buf32[CHAR0_SLOT],
|
|
|
|
this.buf32[CHAR1_SLOT]
|
|
|
|
);
|
2019-06-19 01:16:39 +02:00
|
|
|
this.buf32[CHAR0_SLOT] = char0;
|
2019-10-27 13:36:17 +01:00
|
|
|
this.buf32[CHAR1_SLOT] = char1;
|
2019-04-14 22:23:52 +02:00
|
|
|
}
|
2019-10-27 13:36:17 +01:00
|
|
|
if ( bufLen < this.buf8.length ) {
|
|
|
|
this.reallocateBuf(bufLen);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
reallocateBuf(newSize) {
|
|
|
|
newSize = roundToPageSize(newSize);
|
|
|
|
if ( newSize === this.buf8.length ) { return; }
|
|
|
|
if ( this.wasmMemory === null ) {
|
|
|
|
const newBuf = new Uint8Array(newSize);
|
|
|
|
newBuf.set(
|
|
|
|
newBuf.length < this.buf8.length
|
|
|
|
? this.buf8.subarray(0, newBuf.length)
|
|
|
|
: this.buf8
|
|
|
|
);
|
|
|
|
this.buf8 = newBuf;
|
|
|
|
} else {
|
|
|
|
const growBy =
|
|
|
|
((newSize + 0xFFFF) >>> 16) - (this.buf8.length >>> 16);
|
|
|
|
if ( growBy <= 0 ) { return; }
|
|
|
|
this.wasmMemory.grow(growBy);
|
|
|
|
this.buf8 = new Uint8Array(this.wasmMemory.buffer);
|
|
|
|
}
|
|
|
|
this.buf32 = new Uint32Array(this.buf8.buffer);
|
|
|
|
this.haystack = this.buf8.subarray(
|
|
|
|
HAYSTACK_START,
|
|
|
|
HAYSTACK_START + HAYSTACK_SIZE
|
|
|
|
);
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
};
|
|
|
|
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
/*******************************************************************************
|
|
|
|
|
|
|
|
Class to hold reference to a specific trie
|
2019-04-14 22:23:52 +02:00
|
|
|
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
*/
|
|
|
|
|
2019-06-19 01:16:39 +02:00
|
|
|
µBlock.BidiTrieContainer.prototype.STrieRef = class {
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
constructor(container, iroot, size) {
|
|
|
|
this.container = container;
|
|
|
|
this.iroot = iroot;
|
|
|
|
this.size = size;
|
|
|
|
}
|
|
|
|
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
add(i, n, pivot = 0) {
|
|
|
|
const iboundary = this.container.add(this.iroot, i, n, pivot);
|
|
|
|
if ( iboundary !== 0 ) {
|
2019-04-14 22:23:52 +02:00
|
|
|
this.size += 1;
|
|
|
|
}
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
return iboundary;
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
|
|
|
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
getExtra(iboundary) {
|
|
|
|
return this.container.buf32[iboundary+BCELL_EXTRA];
|
|
|
|
}
|
|
|
|
|
|
|
|
setExtra(iboundary, v) {
|
|
|
|
this.container.buf32[iboundary+BCELL_EXTRA] = v;
|
|
|
|
}
|
|
|
|
|
|
|
|
matches(i) {
|
|
|
|
return this.container.matches(this.iroot, i);
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
|
|
|
|
2019-05-06 17:12:39 +02:00
|
|
|
dump() {
|
|
|
|
for ( const s of this ) {
|
|
|
|
console.log(s);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2019-10-28 18:57:35 +01:00
|
|
|
get $l() { return this.container.buf32[RESULT_L_SLOT] | 0; }
|
|
|
|
get $r() { return this.container.buf32[RESULT_R_SLOT] | 0; }
|
|
|
|
get $iu() { return this.container.buf32[RESULT_IU_SLOT] | 0; }
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
[Symbol.iterator]() {
|
2019-04-14 22:23:52 +02:00
|
|
|
return {
|
|
|
|
value: undefined,
|
|
|
|
done: false,
|
|
|
|
next: function() {
|
|
|
|
if ( this.icell === 0 ) {
|
|
|
|
if ( this.forks.length === 0 ) {
|
|
|
|
this.value = undefined;
|
|
|
|
this.done = true;
|
|
|
|
return this;
|
|
|
|
}
|
|
|
|
this.charPtr = this.forks.pop();
|
|
|
|
this.icell = this.forks.pop();
|
|
|
|
}
|
|
|
|
for (;;) {
|
2019-06-19 01:16:39 +02:00
|
|
|
const idown = this.container.buf32[this.icell+CELL_OR];
|
2019-04-14 22:23:52 +02:00
|
|
|
if ( idown !== 0 ) {
|
|
|
|
this.forks.push(idown, this.charPtr);
|
|
|
|
}
|
2019-06-19 01:16:39 +02:00
|
|
|
const v = this.container.buf32[this.icell+SEGMENT_INFO];
|
|
|
|
let i0 = this.container.buf32[CHAR0_SLOT] + (v & 0x00FFFFFF);
|
2019-04-14 22:23:52 +02:00
|
|
|
const i1 = i0 + (v >>> 24);
|
|
|
|
while ( i0 < i1 ) {
|
Expand bidi-trie usage in static network filtering engine
Related issues:
- https://github.com/uBlockOrigin/uBlock-issues/issues/761
- https://github.com/uBlockOrigin/uBlock-issues/issues/528
The previous bidi-trie code could only hold filters which
are plain pattern, i.e. no wildcard characters, and which
had no origin option (`domain=`), right and/or left anchor,
and no `csp=` option.
Example of filters that could be moved into a bidi-trie
data structure:
&ad_box_
/w/d/capu.php?z=$script,third-party
||liveonlinetv247.com/images/muvixx-150x50-watch-now-in-hd-play-btn.gif
Examples of filters that could NOT be moved to a bidi-trie:
-adap.$domain=~l-adap.org
/tsc.php?*&ses=
||ibsrv.net/*forumsponsor$domain=[...]
@@||imgspice.com/jquery.cookie.js|$script
||view.atdmt.com^*/iview/$third-party
||postimg.cc/image/$csp=[...]
Ideally the filters above should be able to be moved to a
bidi-trie since they are basically plain patterns, or at
least partially moved to a bidi-trie when there is only a
single wildcard (i.e. made of two plain patterns).
Also, there were two distinct bidi-tries in which
plain-pattern filters can be moved to: one for patterns
without hostname anchoring and another one for patterns
with hostname-anchoring. This was required because the
hostname-anchored patterns have an extra condition which
is outside the bidi-trie knowledge.
This commit expands the number of filters which can be
stored in the bidi-trie, and also remove the need to
use two distinct bidi-tries.
- Added ability to associate a pattern with an integer
in the bidi-trie [1].
- The bidi-trie match code passes this externally
provided integer when calling an externally
provided method used for testing extra conditions
that may be present for a plain pattern found to
be matching in the bidi-trie.
- Decomposed existing filters into smaller logical units:
- FilterPlainLeftAnchored =>
FilterPatternPlain +
FilterAnchorLeft
- FilterPlainRightAnchored =>
FilterPatternPlain +
FilterAnchorRight
- FilterExactMatch =>
FilterPatternPlain +
FilterAnchorLeft +
FilterAnchorRight
- FilterPlainHnAnchored =>
FilterPatternPlain +
FilterAnchorHn
- FilterWildcard1 =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
]
- FilterWildcard1HnAnchored =>
FilterPatternPlain + [
FilterPatternLeft or
FilterPatternRight
] +
FilterAnchorHn
- FilterGenericHnAnchored =>
FilterPatternGeneric +
FilterAnchorHn
- FilterGenericHnAndRightAnchored =>
FilterPatternGeneric +
FilterAnchorRight +
FilterAnchorHn
- FilterOriginMixedSet =>
FilterOriginMissSet +
FilterOriginHitSet
- Instances of FilterOrigin[...], FilterDataHolder
can also be added to a composite filter to
represent `domain=` and `csp=` options.
- Added a new filter class, FilterComposite, for
filters which are a combination of two or more
logical units. A FilterComposite instance is a
match when *all* filters composing it are a
match.
Since filters are now encoded into combination of
smaller units, it becomes possible to extract the
FilterPatternPlain component and store it in the
bidi-trie, and use the integer as a handle for the
remaining extra conditions, if any.
Since a single pattern in the bidi-trie may be a
component for different filters, the associated
integer points to a sequence of extra conditions,
and a match occurs as soon as one of the extra
conditions (which may itself be a sequence of
conditions) is fulfilled.
Decomposing filters which are currently single
instance into sequences of smaller logical filters
means increasing the storage and CPU overhead when
evaluating such filters. The CPU overhead is
compensated by the fact that more filters can now
moved into the bidi-trie, where the first match is
efficiently evaluated. The extra conditions have to
be evaluated if and only if there is a match in the
bidi-trie.
The storage overhead is compensated by the
bidi-trie's intrinsic nature of merging similar
patterns.
Furthermore, the storage overhead is reduced by no
longer using JavaScript array to store collection
of filters (which is what FilterComposite is):
the same technique used in [2] is imported to store
sequences of filters.
A sequence of filters is a sequence of integer pairs
where the first integer is an index to an actual
filter instance stored in a global array of filters
(`filterUnits`), while the second integer is an index
to the next pair in the sequence -- which means all
sequences of filters are encoded in one single array
of integers (`filterSequences` => Uint32Array). As
a result, a sequence of filters can be represented by
one single integer -- an index to the first pair --
regardless of the number of filters in the sequence.
This representation is further leveraged to replace
the use of JavaScript array in FilterBucket [3],
which used a JavaScript array to store collection
of filters. Doing so means there is no more need for
FilterPair [4], which purpose was to be a lightweight
representation when there was only two filters in a
collection.
As a result of the above changes, the map of `token`
(integer) => filter instance (object) used to
associate tokens to filters or collections of filters
is replaced with a more efficient map of `token`
(integer) to filter unit index (integer) to lookup a
filter object from the global `filterUnits` array.
Another consequence of using one single global
array to store all filter instances means we can reuse
existing instances when a logical filter instance is
parameter-less, which is the case for FilterAnchorLeft,
FilterAnchorRight, FilterAnchorHn, the index to these
single instances is reused where needed.
`urlTokenizer` now stores the character codes of the
scanned URL into a bidi-trie buffer, for reuse when
string matching methods are called.
New method: `tokenHistogram()`, used to generate
histograms of occurrences of token extracted from URLs
in built-in benchmark. The top results of the "miss"
histogram are used as "bad tokens", i.e. tokens to
avoid if possible when compiling filter lists.
All plain pattern strings are now stored in the
bidi-trie memory buffer, regardless of whether they
will be used in the trie proper or not.
Three methods have been added to the bidi-trie to test
stored string against the URL which is also stored in
then bidi-trie.
FilterParser is now instanciated on demand and
released when no longer used.
***
[1] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/strie.js#L120
[2] https://github.com/gorhill/uBlock/commit/e94024d350b066e4e04a772b0a3dbc69daab3fb7
[3] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1630
[4] https://github.com/gorhill/uBlock/blob/135a45a878f5b93bc538f822981e3a42b1e9073f/src/js/static-net-filtering.js#L1566
2019-10-21 14:15:58 +02:00
|
|
|
this.charBuf[this.charPtr] = this.container.buf8[i0];
|
2019-05-06 17:12:39 +02:00
|
|
|
this.charPtr += 1;
|
2019-04-14 22:23:52 +02:00
|
|
|
i0 += 1;
|
|
|
|
}
|
2019-06-19 01:16:39 +02:00
|
|
|
this.icell = this.container.buf32[this.icell+CELL_AND];
|
2019-04-14 22:23:52 +02:00
|
|
|
if ( this.icell === 0 ) {
|
|
|
|
return this.toPattern();
|
|
|
|
}
|
2019-06-19 01:16:39 +02:00
|
|
|
if ( this.container.buf32[this.icell+SEGMENT_INFO] === 0 ) {
|
|
|
|
this.icell = this.container.buf32[this.icell+CELL_AND];
|
2019-04-14 22:23:52 +02:00
|
|
|
return this.toPattern();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
},
|
|
|
|
toPattern: function() {
|
|
|
|
this.value = this.textDecoder.decode(
|
2019-05-06 17:12:39 +02:00
|
|
|
new Uint8Array(this.charBuf.buffer, 0, this.charPtr)
|
2019-04-14 22:23:52 +02:00
|
|
|
);
|
|
|
|
return this;
|
|
|
|
},
|
|
|
|
container: this.container,
|
|
|
|
icell: this.iroot,
|
|
|
|
charBuf: new Uint8Array(256),
|
2019-05-06 17:12:39 +02:00
|
|
|
charPtr: 0,
|
2019-04-14 22:23:52 +02:00
|
|
|
forks: [],
|
|
|
|
textDecoder: new TextDecoder()
|
|
|
|
};
|
Add HNTrie-based filter classes to store origin-only filters
Related issue:
- https://github.com/uBlockOrigin/uBlock-issues/issues/528#issuecomment-484408622
Following STrie-related work in above issue, I noticed that a large
number of filters in EasyList were filters which only had to match
against the document origin. For instance, among just the top 10
most populous buckets, there were four such buckets with over
hundreds of entries each:
- bits: 72, token: "http", 146 entries
- bits: 72, token: "https", 139 entries
- bits: 88, token: "http", 122 entries
- bits: 88, token: "https", 118 entries
These filters in these buckets have to be matched against all
the network requests.
In order to leverage HNTrie for these filters[1], they are now handled
in a special way so as to ensure they all end up in a single HNTrie
(per bucket), which means that instead of scanning hundreds of entries
per URL, there is now a single scan per bucket per URL for these
apply-everywhere filters.
Now, any filter which fulfill ALL the following condition will be
processed in a special manner internally:
- Is of the form `|https://` or `|http://` or `*`; and
- Does have a `domain=` option; and
- Does not have a negated domain in its `domain=` option; and
- Does not have `csp=` option; and
- Does not have a `redirect=` option
If a filter does not fulfill ALL the conditions above, no change
in behavior.
A filter which matches ALL of the above will be processed in a special
manner:
- The `domain=` option will be decomposed so as to create as many
distinct filter as there is distinct value in the `domain=` option
- This also apply to the `badfilter` version of the filter, which
means it now become possible to `badfilter` only one of the
distinct filter without having to `badfilter` all of them.
- The logger will always report these special filters with only a
single hostname in the `domain=` option.
***
[1] HNTrie is currently WASM-ed on Firefox.
2019-04-19 22:33:46 +02:00
|
|
|
}
|
2019-04-14 22:23:52 +02:00
|
|
|
};
|
2019-06-19 01:16:39 +02:00
|
|
|
|
2019-10-26 19:13:53 +02:00
|
|
|
/******************************************************************************/
|
|
|
|
|
|
|
|
// Code below is to attempt to load a WASM module which implements:
|
|
|
|
//
|
|
|
|
// - BidiTrieContainer.startsWith()
|
|
|
|
//
|
|
|
|
// The WASM module is entirely optional, the JS implementations will be
|
|
|
|
// used should the WASM module be unavailable for whatever reason.
|
|
|
|
|
|
|
|
const getWasmModule = (( ) => {
|
|
|
|
let wasmModulePromise;
|
|
|
|
|
|
|
|
return function() {
|
|
|
|
if ( wasmModulePromise instanceof Promise ) {
|
|
|
|
return wasmModulePromise;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (
|
|
|
|
typeof WebAssembly !== 'object' ||
|
|
|
|
typeof WebAssembly.compileStreaming !== 'function'
|
|
|
|
) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Soft-dependency on vAPI so that the code here can be used outside of
|
|
|
|
// uBO (i.e. tests, benchmarks)
|
2019-10-29 15:26:34 +01:00
|
|
|
if ( typeof vAPI === 'object' && vAPI.canWASM !== true ) { return; }
|
2019-10-26 19:13:53 +02:00
|
|
|
|
|
|
|
// The wasm module will work only if CPU is natively little-endian,
|
|
|
|
// as we use native uint32 array in our js code.
|
|
|
|
const uint32s = new Uint32Array(1);
|
|
|
|
const uint8s = new Uint8Array(uint32s.buffer);
|
|
|
|
uint32s[0] = 1;
|
|
|
|
if ( uint8s[0] !== 1 ) { return; }
|
|
|
|
|
|
|
|
// The directory from which the current script was fetched should also
|
|
|
|
// contain the related WASM file. The script is fetched from a trusted
|
|
|
|
// location, and consequently so will be the related WASM file.
|
|
|
|
let workingDir;
|
|
|
|
{
|
|
|
|
const url = new URL(document.currentScript.src);
|
|
|
|
const match = /[^\/]+$/.exec(url.pathname);
|
|
|
|
if ( match !== null ) {
|
|
|
|
url.pathname = url.pathname.slice(0, match.index);
|
|
|
|
}
|
|
|
|
workingDir = url.href;
|
|
|
|
}
|
|
|
|
|
|
|
|
wasmModulePromise = fetch(
|
|
|
|
workingDir + 'wasm/biditrie.wasm',
|
|
|
|
{ mode: 'same-origin' }
|
|
|
|
).then(
|
|
|
|
WebAssembly.compileStreaming
|
|
|
|
).catch(reason => {
|
|
|
|
log.info(reason);
|
|
|
|
});
|
|
|
|
|
|
|
|
return wasmModulePromise;
|
|
|
|
};
|
|
|
|
})();
|
|
|
|
|
2019-06-19 01:16:39 +02:00
|
|
|
// end of local namespace
|
|
|
|
// *****************************************************************************
|
|
|
|
|
|
|
|
}
|