uBlock/src/js/start.js

371 lines
13 KiB
JavaScript
Raw Normal View History

2014-06-24 00:42:43 +02:00
/*******************************************************************************
2016-08-13 22:42:58 +02:00
uBlock Origin - a browser extension to block requests.
Copyright (C) 2014-present Raymond Hill
2014-06-24 00:42:43 +02:00
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
*/
2016-08-13 22:42:58 +02:00
'use strict';
2014-06-24 00:42:43 +02:00
/******************************************************************************/
2015-02-13 18:10:10 +01:00
// Load all: executed once.
2014-06-24 00:42:43 +02:00
(async ( ) => {
// >>>>> start of private scope
2015-02-13 18:10:10 +01:00
Squashed commit of the following: commit 7c6cacc59b27660fabacb55d668ef099b222a9e6 Author: Raymond Hill <rhill@raymondhill.net> Date: Sat Nov 3 08:52:51 2018 -0300 code review: finalize support for wasm-based hntrie commit 8596ed80e3bdac2c36e3c860b51e7189f6bc8487 Merge: cbe1f2e 000eb82 Author: Raymond Hill <rhill@raymondhill.net> Date: Sat Nov 3 08:41:40 2018 -0300 Merge branch 'master' of github.com:gorhill/uBlock into trie-wasm commit cbe1f2e2f38484d42af3204ec7f1b5decd30f99e Merge: 270fc7f dbb7e80 Author: Raymond Hill <rhill@raymondhill.net> Date: Fri Nov 2 17:43:20 2018 -0300 Merge branch 'master' of github.com:gorhill/uBlock into trie-wasm commit 270fc7f9b3b73d79e6355522c1a42ce782fe7e5c Merge: d2a89cf d693d4f Author: Raymond Hill <rhill@raymondhill.net> Date: Fri Nov 2 16:21:08 2018 -0300 Merge branch 'master' of github.com:gorhill/uBlock into trie-wasm commit d2a89cf28f0816ffd4617c2c7b4ccfcdcc30e1b4 Merge: d7afc78 649f82f Author: Raymond Hill <rhill@raymondhill.net> Date: Fri Nov 2 14:54:58 2018 -0300 Merge branch 'master' of github.com:gorhill/uBlock into trie-wasm commit d7afc78b5f5675d7d34c5a1d0ec3099a77caef49 Author: Raymond Hill <rhill@raymondhill.net> Date: Fri Nov 2 13:56:11 2018 -0300 finalize wasm-based hntrie implementation commit e7b9e043cf36ad055791713e34eb0322dec84627 Author: Raymond Hill <rhill@raymondhill.net> Date: Fri Nov 2 08:14:02 2018 -0300 add first-pass implementation of wasm version of hntrie commit 1015cb34624f3ef73ace58b58fe4e03dfc59897f Author: Raymond Hill <rhill@raymondhill.net> Date: Wed Oct 31 17:16:47 2018 -0300 back up draft work toward experimenting with wasm hntries
2018-11-03 12:58:46 +01:00
const µb = µBlock;
2015-02-24 19:48:03 +01:00
/******************************************************************************/
vAPI.app.onShutdown = function() {
const µb = µBlock;
µb.staticFilteringReverseLookup.shutdown();
µb.assets.updateStop();
µb.staticNetFilteringEngine.reset();
µb.staticExtFilteringEngine.reset();
µb.sessionFirewall.reset();
µb.permanentFirewall.reset();
µb.sessionURLFiltering.reset();
µb.permanentURLFiltering.reset();
2018-09-03 20:06:49 +02:00
µb.sessionSwitches.reset();
µb.permanentSwitches.reset();
};
/******************************************************************************/
// This is called only once, when everything has been loaded in memory after
// the extension was launched. It can be used to inject content scripts
// in already opened web pages, to remove whatever nuisance could make it to
// the web pages before uBlock was ready.
const initializeTabs = async function() {
const manifest = browser.runtime.getManifest();
if ( manifest instanceof Object === false ) { return; }
const tabs = await vAPI.tabs.query({ url: '<all_urls>' });
const toCheck = [];
const checker = {
file: 'js/scriptlets/should-inject-contentscript.js'
};
for ( const tab of tabs ) {
µb.tabContextManager.commit(tab.id, tab.url);
µb.bindTabToPageStats(tab.id);
// https://github.com/chrisaljoudi/uBlock/issues/129
// Find out whether content scripts need to be injected
// programmatically. This may be necessary for web pages which
// were loaded before uBO launched.
toCheck.push(
/^https?:\/\//.test(tab.url)
? vAPI.tabs.executeScript(tab.id, checker)
: false
);
}
const results = await Promise.all(toCheck);
for ( let i = 0; i < results.length; i++ ) {
const result = results[i];
if ( result.length === 0 || result[0] !== true ) { continue; }
// Inject dclarative content scripts programmatically.
const tabId = tabs[i].id;
for ( const contentScript of manifest.content_scripts ) {
for ( const file of contentScript.js ) {
vAPI.tabs.executeScript(tabId, {
file: file,
allFrames: contentScript.all_frames,
runAt: contentScript.run_at
});
}
}
}
2015-02-13 18:10:10 +01:00
};
2014-12-20 21:28:16 +01:00
2014-08-21 01:39:49 +02:00
/******************************************************************************/
Refactor selfie generation into a more flexible persistence mechanism The motivation is to address the higher peak memory usage at launch time with 3rd-gen HNTrie when a selfie was present. The selfie generation prior to this change was to collect all filtering data into a single data structure, and then to serialize that whole structure at once into storage (using JSON.stringify). However, HNTrie serialization requires that a large UintArray32 be converted into a plain JS array, which itslef would be indirectly converted into a JSON string. This was the main reason why peak memory usage would be higher at launch from selfie, since the JSON string would need to be wholly unserialized into JS objects, which themselves would need to be converted into more specialized data structures (like that Uint32Array one). The solution to lower peak memory usage at launch is to refactor selfie generation to allow a more piecemeal approach: each filtering component is given the ability to serialize itself rather than to be forced to be embedded in the master selfie. With this approach, the HNTrie buffer can now serialize to its own storage by converting the buffer data directly into a string which can be directly sent to storage. This avoiding expensive intermediate steps such as converting into a JS array and then to a JSON string. As part of the refactoring, there was also opportunistic code upgrade to ES6 and Promise (eventually all of uBO's code will be proper ES6). Additionally, the polyfill to bring getBytesInUse() to Firefox has been revisited to replace the rather expensive previous implementation with an implementation with virtually no overhead.
2019-02-14 19:33:55 +01:00
const onCommandShortcutsReady = function(commandShortcuts) {
if ( Array.isArray(commandShortcuts) === false ) { return; }
µb.commandShortcuts = new Map(commandShortcuts);
if ( µb.canUpdateShortcuts === false ) { return; }
Refactor selfie generation into a more flexible persistence mechanism The motivation is to address the higher peak memory usage at launch time with 3rd-gen HNTrie when a selfie was present. The selfie generation prior to this change was to collect all filtering data into a single data structure, and then to serialize that whole structure at once into storage (using JSON.stringify). However, HNTrie serialization requires that a large UintArray32 be converted into a plain JS array, which itslef would be indirectly converted into a JSON string. This was the main reason why peak memory usage would be higher at launch from selfie, since the JSON string would need to be wholly unserialized into JS objects, which themselves would need to be converted into more specialized data structures (like that Uint32Array one). The solution to lower peak memory usage at launch is to refactor selfie generation to allow a more piecemeal approach: each filtering component is given the ability to serialize itself rather than to be forced to be embedded in the master selfie. With this approach, the HNTrie buffer can now serialize to its own storage by converting the buffer data directly into a string which can be directly sent to storage. This avoiding expensive intermediate steps such as converting into a JS array and then to a JSON string. As part of the refactoring, there was also opportunistic code upgrade to ES6 and Promise (eventually all of uBO's code will be proper ES6). Additionally, the polyfill to bring getBytesInUse() to Firefox has been revisited to replace the rather expensive previous implementation with an implementation with virtually no overhead.
2019-02-14 19:33:55 +01:00
for ( const entry of commandShortcuts ) {
vAPI.commands.update({ name: entry[0], shortcut: entry[1] });
}
};
/******************************************************************************/
2015-02-24 19:48:03 +01:00
// To bring older versions up to date
Refactor selfie generation into a more flexible persistence mechanism The motivation is to address the higher peak memory usage at launch time with 3rd-gen HNTrie when a selfie was present. The selfie generation prior to this change was to collect all filtering data into a single data structure, and then to serialize that whole structure at once into storage (using JSON.stringify). However, HNTrie serialization requires that a large UintArray32 be converted into a plain JS array, which itslef would be indirectly converted into a JSON string. This was the main reason why peak memory usage would be higher at launch from selfie, since the JSON string would need to be wholly unserialized into JS objects, which themselves would need to be converted into more specialized data structures (like that Uint32Array one). The solution to lower peak memory usage at launch is to refactor selfie generation to allow a more piecemeal approach: each filtering component is given the ability to serialize itself rather than to be forced to be embedded in the master selfie. With this approach, the HNTrie buffer can now serialize to its own storage by converting the buffer data directly into a string which can be directly sent to storage. This avoiding expensive intermediate steps such as converting into a JS array and then to a JSON string. As part of the refactoring, there was also opportunistic code upgrade to ES6 and Promise (eventually all of uBO's code will be proper ES6). Additionally, the polyfill to bring getBytesInUse() to Firefox has been revisited to replace the rather expensive previous implementation with an implementation with virtually no overhead.
2019-02-14 19:33:55 +01:00
const onVersionReady = function(lastVersion) {
if ( lastVersion === vAPI.app.version ) { return; }
// Since built-in resources may have changed since last version, we
// force a reload of all resources.
µb.redirectEngine.invalidateResourcesSelfie();
const lastVersionInt = vAPI.app.intFromVersion(lastVersion);
// https://github.com/uBlockOrigin/uBlock-issues/issues/494
// Remove useless per-site switches.
if ( lastVersionInt <= 1019003007 ) {
µb.sessionSwitches.toggle('no-scripting', 'behind-the-scene', 0);
µb.permanentSwitches.toggle('no-scripting', 'behind-the-scene', 0);
µb.saveHostnameSwitches();
}
vAPI.storage.set({ version: vAPI.app.version });
2015-02-13 18:10:10 +01:00
};
/******************************************************************************/
2014-09-08 23:46:58 +02:00
2015-04-07 03:26:05 +02:00
// https://github.com/chrisaljoudi/uBlock/issues/226
2015-02-13 18:10:10 +01:00
// Whitelist in memory.
// Whitelist parser needs PSL to be ready.
// gorhill 2014-12-15: not anymore
Refactor selfie generation into a more flexible persistence mechanism The motivation is to address the higher peak memory usage at launch time with 3rd-gen HNTrie when a selfie was present. The selfie generation prior to this change was to collect all filtering data into a single data structure, and then to serialize that whole structure at once into storage (using JSON.stringify). However, HNTrie serialization requires that a large UintArray32 be converted into a plain JS array, which itslef would be indirectly converted into a JSON string. This was the main reason why peak memory usage would be higher at launch from selfie, since the JSON string would need to be wholly unserialized into JS objects, which themselves would need to be converted into more specialized data structures (like that Uint32Array one). The solution to lower peak memory usage at launch is to refactor selfie generation to allow a more piecemeal approach: each filtering component is given the ability to serialize itself rather than to be forced to be embedded in the master selfie. With this approach, the HNTrie buffer can now serialize to its own storage by converting the buffer data directly into a string which can be directly sent to storage. This avoiding expensive intermediate steps such as converting into a JS array and then to a JSON string. As part of the refactoring, there was also opportunistic code upgrade to ES6 and Promise (eventually all of uBO's code will be proper ES6). Additionally, the polyfill to bring getBytesInUse() to Firefox has been revisited to replace the rather expensive previous implementation with an implementation with virtually no overhead.
2019-02-14 19:33:55 +01:00
const onNetWhitelistReady = function(netWhitelistRaw) {
if ( typeof netWhitelistRaw === 'string' ) {
netWhitelistRaw = netWhitelistRaw.split('\n');
}
µb.netWhitelist = µb.whitelistFromArray(netWhitelistRaw);
2015-02-24 19:48:03 +01:00
µb.netWhitelistModifyTime = Date.now();
2015-02-13 18:10:10 +01:00
};
/******************************************************************************/
// User settings are in memory
Refactor selfie generation into a more flexible persistence mechanism The motivation is to address the higher peak memory usage at launch time with 3rd-gen HNTrie when a selfie was present. The selfie generation prior to this change was to collect all filtering data into a single data structure, and then to serialize that whole structure at once into storage (using JSON.stringify). However, HNTrie serialization requires that a large UintArray32 be converted into a plain JS array, which itslef would be indirectly converted into a JSON string. This was the main reason why peak memory usage would be higher at launch from selfie, since the JSON string would need to be wholly unserialized into JS objects, which themselves would need to be converted into more specialized data structures (like that Uint32Array one). The solution to lower peak memory usage at launch is to refactor selfie generation to allow a more piecemeal approach: each filtering component is given the ability to serialize itself rather than to be forced to be embedded in the master selfie. With this approach, the HNTrie buffer can now serialize to its own storage by converting the buffer data directly into a string which can be directly sent to storage. This avoiding expensive intermediate steps such as converting into a JS array and then to a JSON string. As part of the refactoring, there was also opportunistic code upgrade to ES6 and Promise (eventually all of uBO's code will be proper ES6). Additionally, the polyfill to bring getBytesInUse() to Firefox has been revisited to replace the rather expensive previous implementation with an implementation with virtually no overhead.
2019-02-14 19:33:55 +01:00
const onUserSettingsReady = function(fetched) {
const userSettings = µb.userSettings;
2015-02-24 19:48:03 +01:00
fromFetch(userSettings, fetched);
2014-08-21 16:56:36 +02:00
if ( µb.privacySettingsSupported ) {
vAPI.browserSettings.set({
'hyperlinkAuditing': !userSettings.hyperlinkAuditingDisabled,
'prefetching': !userSettings.prefetchingDisabled,
'webrtcIPAddress': !userSettings.webrtcIPAddressHidden
});
}
2015-06-01 21:03:22 +02:00
2015-03-27 18:00:55 +01:00
µb.permanentFirewall.fromString(fetched.dynamicFilteringString);
2015-02-13 18:10:10 +01:00
µb.sessionFirewall.assign(µb.permanentFirewall);
2015-05-21 20:15:17 +02:00
µb.permanentURLFiltering.fromString(fetched.urlFilteringString);
µb.sessionURLFiltering.assign(µb.permanentURLFiltering);
2018-09-03 20:06:49 +02:00
µb.permanentSwitches.fromString(fetched.hostnameSwitchesString);
µb.sessionSwitches.assign(µb.permanentSwitches);
2015-02-13 18:10:10 +01:00
};
2015-02-24 00:31:29 +01:00
/******************************************************************************/
// https://bugzilla.mozilla.org/show_bug.cgi?id=1588916
// Save magic format numbers into the cache storage itself.
2015-02-24 19:48:03 +01:00
const onCacheSettingsReady = function(fetched) {
2015-02-24 19:48:03 +01:00
if ( fetched.compiledMagic !== µb.systemSettings.compiledMagic ) {
µb.assets.remove(/^compiled\//);
µb.compiledFormatChanged = true;
µb.selfieIsInvalid = true;
2015-02-24 00:31:29 +01:00
}
2015-02-24 19:48:03 +01:00
if ( fetched.selfieMagic !== µb.systemSettings.selfieMagic ) {
µb.selfieIsInvalid = true;
2015-02-24 00:31:29 +01:00
}
if ( µb.selfieIsInvalid ) {
2015-11-29 23:06:58 +01:00
µb.selfieManager.destroy();
µb.cacheStorage.set(µb.systemSettings);
2015-02-24 00:31:29 +01:00
}
2015-02-24 19:48:03 +01:00
};
/******************************************************************************/
Refactor selfie generation into a more flexible persistence mechanism The motivation is to address the higher peak memory usage at launch time with 3rd-gen HNTrie when a selfie was present. The selfie generation prior to this change was to collect all filtering data into a single data structure, and then to serialize that whole structure at once into storage (using JSON.stringify). However, HNTrie serialization requires that a large UintArray32 be converted into a plain JS array, which itslef would be indirectly converted into a JSON string. This was the main reason why peak memory usage would be higher at launch from selfie, since the JSON string would need to be wholly unserialized into JS objects, which themselves would need to be converted into more specialized data structures (like that Uint32Array one). The solution to lower peak memory usage at launch is to refactor selfie generation to allow a more piecemeal approach: each filtering component is given the ability to serialize itself rather than to be forced to be embedded in the master selfie. With this approach, the HNTrie buffer can now serialize to its own storage by converting the buffer data directly into a string which can be directly sent to storage. This avoiding expensive intermediate steps such as converting into a JS array and then to a JSON string. As part of the refactoring, there was also opportunistic code upgrade to ES6 and Promise (eventually all of uBO's code will be proper ES6). Additionally, the polyfill to bring getBytesInUse() to Firefox has been revisited to replace the rather expensive previous implementation with an implementation with virtually no overhead.
2019-02-14 19:33:55 +01:00
const onFirstFetchReady = function(fetched) {
// https://github.com/uBlockOrigin/uBlock-issues/issues/507
// Firefox-specific: somehow `fetched` is undefined under certain
// circumstances even though we asked to load with default values.
if ( fetched instanceof Object === false ) {
fetched = createDefaultProps();
}
2015-02-24 19:48:03 +01:00
// Order is important -- do not change:
2015-03-07 05:36:09 +01:00
fromFetch(µb.localSettings, fetched);
2015-02-24 19:48:03 +01:00
onUserSettingsReady(fetched);
2015-03-07 05:36:09 +01:00
fromFetch(µb.restoreBackupSettings, fetched);
2015-02-24 19:48:03 +01:00
onNetWhitelistReady(fetched.netWhitelist);
onVersionReady(fetched.version);
onCommandShortcutsReady(fetched.commandShortcuts);
2015-02-24 19:48:03 +01:00
};
/******************************************************************************/
Refactor selfie generation into a more flexible persistence mechanism The motivation is to address the higher peak memory usage at launch time with 3rd-gen HNTrie when a selfie was present. The selfie generation prior to this change was to collect all filtering data into a single data structure, and then to serialize that whole structure at once into storage (using JSON.stringify). However, HNTrie serialization requires that a large UintArray32 be converted into a plain JS array, which itslef would be indirectly converted into a JSON string. This was the main reason why peak memory usage would be higher at launch from selfie, since the JSON string would need to be wholly unserialized into JS objects, which themselves would need to be converted into more specialized data structures (like that Uint32Array one). The solution to lower peak memory usage at launch is to refactor selfie generation to allow a more piecemeal approach: each filtering component is given the ability to serialize itself rather than to be forced to be embedded in the master selfie. With this approach, the HNTrie buffer can now serialize to its own storage by converting the buffer data directly into a string which can be directly sent to storage. This avoiding expensive intermediate steps such as converting into a JS array and then to a JSON string. As part of the refactoring, there was also opportunistic code upgrade to ES6 and Promise (eventually all of uBO's code will be proper ES6). Additionally, the polyfill to bring getBytesInUse() to Firefox has been revisited to replace the rather expensive previous implementation with an implementation with virtually no overhead.
2019-02-14 19:33:55 +01:00
const toFetch = function(from, fetched) {
for ( const k in from ) {
if ( from.hasOwnProperty(k) === false ) { continue; }
2015-02-24 19:48:03 +01:00
fetched[k] = from[k];
}
};
Refactor selfie generation into a more flexible persistence mechanism The motivation is to address the higher peak memory usage at launch time with 3rd-gen HNTrie when a selfie was present. The selfie generation prior to this change was to collect all filtering data into a single data structure, and then to serialize that whole structure at once into storage (using JSON.stringify). However, HNTrie serialization requires that a large UintArray32 be converted into a plain JS array, which itslef would be indirectly converted into a JSON string. This was the main reason why peak memory usage would be higher at launch from selfie, since the JSON string would need to be wholly unserialized into JS objects, which themselves would need to be converted into more specialized data structures (like that Uint32Array one). The solution to lower peak memory usage at launch is to refactor selfie generation to allow a more piecemeal approach: each filtering component is given the ability to serialize itself rather than to be forced to be embedded in the master selfie. With this approach, the HNTrie buffer can now serialize to its own storage by converting the buffer data directly into a string which can be directly sent to storage. This avoiding expensive intermediate steps such as converting into a JS array and then to a JSON string. As part of the refactoring, there was also opportunistic code upgrade to ES6 and Promise (eventually all of uBO's code will be proper ES6). Additionally, the polyfill to bring getBytesInUse() to Firefox has been revisited to replace the rather expensive previous implementation with an implementation with virtually no overhead.
2019-02-14 19:33:55 +01:00
const fromFetch = function(to, fetched) {
for ( const k in to ) {
if ( to.hasOwnProperty(k) === false ) { continue; }
if ( fetched.hasOwnProperty(k) === false ) { continue; }
2015-02-24 19:48:03 +01:00
to[k] = fetched[k];
}
2015-02-24 00:31:29 +01:00
};
const createDefaultProps = function() {
Refactor selfie generation into a more flexible persistence mechanism The motivation is to address the higher peak memory usage at launch time with 3rd-gen HNTrie when a selfie was present. The selfie generation prior to this change was to collect all filtering data into a single data structure, and then to serialize that whole structure at once into storage (using JSON.stringify). However, HNTrie serialization requires that a large UintArray32 be converted into a plain JS array, which itslef would be indirectly converted into a JSON string. This was the main reason why peak memory usage would be higher at launch from selfie, since the JSON string would need to be wholly unserialized into JS objects, which themselves would need to be converted into more specialized data structures (like that Uint32Array one). The solution to lower peak memory usage at launch is to refactor selfie generation to allow a more piecemeal approach: each filtering component is given the ability to serialize itself rather than to be forced to be embedded in the master selfie. With this approach, the HNTrie buffer can now serialize to its own storage by converting the buffer data directly into a string which can be directly sent to storage. This avoiding expensive intermediate steps such as converting into a JS array and then to a JSON string. As part of the refactoring, there was also opportunistic code upgrade to ES6 and Promise (eventually all of uBO's code will be proper ES6). Additionally, the polyfill to bring getBytesInUse() to Firefox has been revisited to replace the rather expensive previous implementation with an implementation with virtually no overhead.
2019-02-14 19:33:55 +01:00
const fetchableProps = {
'commandShortcuts': [],
'dynamicFilteringString': [
'behind-the-scene * * noop',
'behind-the-scene * image noop',
'behind-the-scene * 3p noop',
'behind-the-scene * inline-script noop',
'behind-the-scene * 1p-script noop',
'behind-the-scene * 3p-script noop',
'behind-the-scene * 3p-frame noop'
].join('\n'),
2016-01-03 19:58:25 +01:00
'urlFilteringString': '',
'hostnameSwitchesString': [
2018-09-03 20:06:49 +02:00
'no-large-media: behind-the-scene false',
].join('\n'),
2016-01-03 19:58:25 +01:00
'lastRestoreFile': '',
'lastRestoreTime': 0,
'lastBackupFile': '',
'lastBackupTime': 0,
'netWhitelist': µb.netWhitelistDefault,
2016-01-03 19:58:25 +01:00
'version': '0.0.0.0'
2015-10-21 17:53:03 +02:00
};
2016-01-03 19:58:25 +01:00
toFetch(µb.localSettings, fetchableProps);
toFetch(µb.userSettings, fetchableProps);
toFetch(µb.restoreBackupSettings, fetchableProps);
return fetchableProps;
2016-01-03 19:58:25 +01:00
};
2017-01-26 16:17:38 +01:00
/******************************************************************************/
try {
// https://github.com/gorhill/uBlock/issues/531
await µb.restoreAdminSettings();
Refactor selfie generation into a more flexible persistence mechanism The motivation is to address the higher peak memory usage at launch time with 3rd-gen HNTrie when a selfie was present. The selfie generation prior to this change was to collect all filtering data into a single data structure, and then to serialize that whole structure at once into storage (using JSON.stringify). However, HNTrie serialization requires that a large UintArray32 be converted into a plain JS array, which itslef would be indirectly converted into a JSON string. This was the main reason why peak memory usage would be higher at launch from selfie, since the JSON string would need to be wholly unserialized into JS objects, which themselves would need to be converted into more specialized data structures (like that Uint32Array one). The solution to lower peak memory usage at launch is to refactor selfie generation to allow a more piecemeal approach: each filtering component is given the ability to serialize itself rather than to be forced to be embedded in the master selfie. With this approach, the HNTrie buffer can now serialize to its own storage by converting the buffer data directly into a string which can be directly sent to storage. This avoiding expensive intermediate steps such as converting into a JS array and then to a JSON string. As part of the refactoring, there was also opportunistic code upgrade to ES6 and Promise (eventually all of uBO's code will be proper ES6). Additionally, the polyfill to bring getBytesInUse() to Firefox has been revisited to replace the rather expensive previous implementation with an implementation with virtually no overhead.
2019-02-14 19:33:55 +01:00
log.info(`Admin settings ready ${Date.now()-vAPI.T0} ms after launch`);
await µb.loadHiddenSettings();
log.info(`Hidden settings ready ${Date.now()-vAPI.T0} ms after launch`);
2016-01-03 19:58:25 +01:00
const cacheBackend = await µb.cacheStorage.select(
µb.hiddenSettings.cacheStorageAPI
);
log.info(`Backend storage for cache will be ${cacheBackend}`);
await Promise.all([
µb.loadSelectedFilterLists().then(( ) => {
log.info(`List selection ready ${Date.now()-vAPI.T0} ms after launch`);
}),
µb.cacheStorage.get(
{ compiledMagic: 0, selfieMagic: 0 }
).then(fetched => {
log.info(`Cache magic numbers ready ${Date.now()-vAPI.T0} ms after launch`);
onCacheSettingsReady(fetched);
}),
vAPI.storage.get(createDefaultProps()).then(fetched => {
log.info(`First fetch ready ${Date.now()-vAPI.T0} ms after launch`);
onFirstFetchReady(fetched);
}),
µb.loadPublicSuffixList().then(( ) => {
log.info(`PSL ready ${Date.now()-vAPI.T0} ms after launch`);
}),
]);
} catch (ex) {
console.trace(ex);
}
// https://github.com/uBlockOrigin/uBlock-issues/issues/817#issuecomment-565730122
// Still try to load filter lists regardless of whether a serious error
// occurred in the previous initialization steps.
let selfieIsValid = false;
try {
selfieIsValid = await µb.selfieManager.load();
if ( selfieIsValid === true ) {
log.info(`Selfie ready ${Date.now()-vAPI.T0} ms after launch`);
}
} catch (ex) {
console.trace(ex);
}
if ( selfieIsValid !== true ) {
try {
await µb.loadFilterLists();
log.info(`Filter lists ready ${Date.now()-vAPI.T0} ms after launch`);
} catch (ex) {
console.trace(ex);
}
}
2016-01-03 19:58:25 +01:00
// Final initialization steps after all needed assets are in memory.
// Start network observers.
µb.webRequest.start();
// Ensure that the resources allocated for decompression purpose (likely
// large buffers) are garbage-collectable immediately after launch.
// Otherwise I have observed that it may take quite a while before the
// garbage collection of these resources kicks in. Relinquishing as soon
// as possible ensure minimal memory usage baseline.
µb.lz4Codec.relinquish();
// Initialize internal state with maybe already existing tabs.
initializeTabs();
// https://github.com/chrisaljoudi/uBlock/issues/184
// Check for updates not too far in the future.
µb.assets.addObserver(µb.assetObserver.bind(µb));
µb.scheduleAssetUpdater(
µb.userSettings.autoUpdate
? µb.hiddenSettings.autoUpdateDelayAfterLaunch * 1000
: 0
);
// Force an update of the context menu according to the currently
// active tab.
µb.contextMenu.update();
// Maybe install non-default popup document
if (
browser.browserAction instanceof Object &&
browser.browserAction.setPopup instanceof Function
) {
const uiFlavor = µb.hiddenSettings.uiFlavor;
if ( uiFlavor !== 'unset' && /\w+/.test(uiFlavor) ) {
browser.browserAction.setPopup({
popup: vAPI.getURL(`popup-${uiFlavor}.html`)
});
}
}
// https://github.com/uBlockOrigin/uBlock-issues/issues/717
// Prevent the extension from being restarted mid-session.
browser.runtime.onUpdateAvailable.addListener(details => {
const toInt = vAPI.app.intFromVersion;
if (
µBlock.hiddenSettings.extensionUpdateForceReload === true ||
toInt(details.version) <= toInt(vAPI.app.version)
) {
vAPI.app.restart();
}
});
log.info(`All ready ${Date.now()-vAPI.T0} ms after launch`);
2015-03-11 23:26:00 +01:00
// <<<<< end of private scope
})();