Improve output template (see desc)

* Objects can be traversed like `%(field.key1.key2)s`
* A number can be added to the field as `%(field+n)s`
* Deprecates `--autonumber-start`

README.md

@@ -395,8 +395,6 @@ ## Filesystem Options:
     --output-na-placeholder TEXT     Placeholder value for unavailable meta
                                      fields in output filename template
                                      (default: "NA")
-    --autonumber-start NUMBER        Specify the start value for %(autonumber)s
-                                     (default is 1)
     --restrict-filenames             Restrict filenames to only ASCII
                                      characters, and avoid "&" and spaces in
                                      filenames
@@ -833,7 +831,19 @@ # OUTPUT TEMPLATE
 
 **tl;dr:** [navigate me to examples](#output-template-examples).
 
-The basic usage of `-o` is not to set any template arguments when downloading a single file, like in `yt-dlp -o funny_video.flv "https://some/video"` (hard-coding file extension like this is not recommended). However, it may contain special sequences that will be replaced when downloading each video. The special sequences may be formatted according to [python string formatting operations](https://docs.python.org/2/library/stdtypes.html#string-formatting). For example, `%(NAME)s` or `%(NAME)05d`. To clarify, that is a percent symbol followed by a name in parentheses, followed by formatting operations. Date/time fields can also be formatted according to [strftime formatting](https://docs.python.org/3/library/datetime.html#strftime-and-strptime-format-codes) by specifying it inside the parantheses separated from the field name using a `>`. For example, `%(duration>%H-%M-%S)s`.
+The simplest usage of `-o` is not to set any template arguments when downloading a single file, like in `yt-dlp -o funny_video.flv "https://some/video"` (hard-coding file extension like this is _not_ recommended and could break certain postprocessing).
+
+It may however also contain special sequences that will be replaced when downloading each video. The special sequences may be formatted according to [python string formatting operations](https://docs.python.org/2/library/stdtypes.html#string-formatting). For example, `%(NAME)s` or `%(NAME)05d`. To clarify, that is a percent symbol followed by a name in parentheses, followed by formatting operations.
+
+The field names themselves (the part inside the parenthesis) can also have some special formatting:
+1. **Date/time Formatting**: Date/time fields can be formatted according to [strftime formatting](https://docs.python.org/3/library/datetime.html#strftime-and-strptime-format-codes) by specifying it separated from the field name using a `>`. Eg: `%(duration>%H-%M-%S)s` or `%(upload_date>%Y-%m-%d)s`
+2. **Offset numbers**: Numeric fields can have an initial offset specified by using a `+` seperator. Eg: `%(playlist_index+10)03d`. This can also be used in conjunction with the datetime formatting. Eg: `%(epoch+-3600>%H-%M-%S)s`
+3. **Object traversal**: The dictionaries and lists available in metadata can be traversed by using a `.` (dot) seperator. Eg: `%(tags.0)s` or `%(subtitles.en.-1.ext)`. Note that the fields that become available using this method are not listed below. Use `-j` to see such fields
+
+To summarize, the general syntax for a field is:
+```
+%(name[.keys][+offset][>strf])[flags][width][.precision][length]type
+```
 
 Additionally, you can set different output templates for the various metadata files separately from the general output template by specifying the type of file followed by the template separated by a colon `:`. The different filetypes supported are `subtitle`, `thumbnail`, `description`, `annotation`, `infojson`, `pl_description`, `pl_infojson`, `chapter`. For example, `-o '%(title)s.%(ext)s' -o 'thumbnail:%(title)s\%(title)s.%(ext)s'`  will put the thumbnails in a folder with the same name as the video.
 
@@ -992,7 +1002,7 @@ # FORMAT SELECTION
 
 **tl;dr:** [navigate me to examples](#format-selection-examples).
 
-The simplest case is requesting a specific format, for example with `-f 22` you can download the format with format code equal to 22. You can get the list of available format codes for particular video using `--list-formats` or `-F`. Note that these format codes are extractor specific. 
+The simplest case is requesting a specific format, for example with `-f 22` you can download the format with format code equal to 22. You can get the list of available format codes for particular video using `--list-formats` or `-F`. Note that these format codes are extractor specific.
 
 You can also use a file extension (currently `3gp`, `aac`, `flv`, `m4a`, `mp3`, `mp4`, `ogg`, `wav`, `webm` are supported) to download the best quality format of a particular file extension served as a single file, e.g. `-f webm` will download the best quality format with the `webm` extension served as a single file.
 
@@ -1263,6 +1273,7 @@ # DEPRECATED OPTIONS
     --all-formats                    -f all
     --all-subs                       --sub-langs all --write-subs
     --autonumber-size NUMBER         Use string formatting. Eg: %(autonumber)03d
+    --autonumber-start NUMBER        Use internal field formatting like %(autonumber+NUMBER)s
     --metadata-from-title FORMAT     --parse-metadata "%(title)s:FORMAT"
     --prefer-avconv                  avconv is no longer officially supported (Alias: --no-prefer-ffmpeg)
     --prefer-ffmpeg                  Default (Alias: --no-prefer-avconv)

test/test_YoutubeDL.py

@@ -655,6 +655,8 @@ def test_prepare_filename(self):
             'height': 1080,
             'title1': '$PATH',
             'title2': '%PATH%',
+            'timestamp': 1618488000,
+            'formats': [{'id': 'id1'}, {'id': 'id2'}]
         }
 
         def fname(templ, na_placeholder='NA'):
@@ -671,6 +673,7 @@ def fname(templ, na_placeholder='NA'):
         # Or by provided placeholder
         self.assertEqual(fname(NA_TEST_OUTTMPL, na_placeholder='none'), 'none-none-1234.mp4')
         self.assertEqual(fname(NA_TEST_OUTTMPL, na_placeholder=''), '--1234.mp4')
+        self.assertEqual(fname('%(height)s.%(ext)s'), '1080.mp4')
         self.assertEqual(fname('%(height)d.%(ext)s'), '1080.mp4')
         self.assertEqual(fname('%(height)6d.%(ext)s'), '  1080.mp4')
         self.assertEqual(fname('%(height)-6d.%(ext)s'), '1080  .mp4')
@@ -688,6 +691,12 @@ def fname(templ, na_placeholder='NA'):
         self.assertEqual(fname('%%(width)06d.%(ext)s'), '%(width)06d.mp4')
         self.assertEqual(fname('Hello %(title1)s'), 'Hello $PATH')
         self.assertEqual(fname('Hello %(title2)s'), 'Hello %PATH%')
+        self.assertEqual(fname('%(timestamp+-1000>%H-%M-%S)s'), '11-43-20')
+        self.assertEqual(fname('%(id+1)05d'), '01235')
+        self.assertEqual(fname('%(width+100)05d'), 'NA')
+        self.assertEqual(fname('%(formats.0)s').replace("u", ""), "{'id' - 'id1'}")
+        self.assertEqual(fname('%(formats.-1.id)s'), 'id2')
+        self.assertEqual(fname('%(formats.2)s'), 'NA')
 
     def test_format_note(self):
         ydl = YoutubeDL()

yt_dlp/YoutubeDL.py

@@ -99,6 +99,7 @@
     strftime_or_none,
     subtitles_filename,
     to_high_limit_path,
+    traverse_dict,
     UnavailableVideoError,
     url_basename,
     version_tuple,
@@ -796,6 +797,7 @@ def parse_outtmpl(self):
     def prepare_outtmpl(self, outtmpl, info_dict, sanitize=None):
         """ Make the template and info_dict suitable for substitution (outtmpl % info_dict)"""
         template_dict = dict(info_dict)
+        na = self.params.get('outtmpl_na_placeholder', 'NA')
 
         # duration_string
         template_dict['duration_string'] = (  # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
@@ -821,18 +823,10 @@ def prepare_outtmpl(self, outtmpl, info_dict, sanitize=None):
             elif template_dict.get('width'):
                 template_dict['resolution'] = '%dx?' % template_dict['width']
 
-        if sanitize is None:
-            sanitize = lambda k, v: v
-        template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
-                             for k, v in template_dict.items()
-                             if v is not None and not isinstance(v, (list, tuple, dict)))
-        na = self.params.get('outtmpl_na_placeholder', 'NA')
-        template_dict = collections.defaultdict(lambda: na, template_dict)
-
         # For fields playlist_index and autonumber convert all occurrences
         # of %(field)s to %(field)0Nd for backward compatibility
         field_size_compat_map = {
-            'playlist_index': len(str(template_dict['n_entries'])),
+            'playlist_index': len(str(template_dict.get('n_entries', na))),
             'autonumber': autonumber_size,
         }
         FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
@@ -844,32 +838,51 @@ def prepare_outtmpl(self, outtmpl, info_dict, sanitize=None):
                 outtmpl)
 
         numeric_fields = list(self._NUMERIC_FIELDS)
+        if sanitize is None:
+            sanitize = lambda k, v: v
 
-        # Format date
+        # Internal Formatting = name.key1.key2+number>strf
-        FORMAT_DATE_RE = FORMAT_RE.format(r'(?P<key>(?P<field>\w+)>(?P<format>.+?))')
+        INTERNAL_FORMAT_RE = FORMAT_RE.format(
-        for mobj in re.finditer(FORMAT_DATE_RE, outtmpl):
+            r'''(?P<final_key>
-            conv_type, field, frmt, key = mobj.group('type', 'field', 'format', 'key')
+                        (?P<fields>\w+(?:\.[-\w]+)*)
-            if key in template_dict:
+                        (?:\+(?P<add>-?\d+(?:\.\d+)?))?
-                continue
+                        (?:>(?P<strf_format>.+?))?
-            value = strftime_or_none(template_dict.get(field), frmt, na)
+            )''')
-            if conv_type in 'crs':  # string
+        for mobj in re.finditer(INTERNAL_FORMAT_RE, outtmpl):
-                value = sanitize(field, value)
+            mobj = mobj.groupdict()
-            else:  # number
+            # Object traversal
-                numeric_fields.append(key)
+            fields = mobj['fields'].split('.')
-                value = float_or_none(value, default=None)
+            final_key = mobj['final_key']
+            value = traverse_dict(template_dict, fields)
+            # Offset the value
+            if mobj['add']:
+                value = float_or_none(value)
+                if value is not None:
+                    value = value + float(mobj['add'])
+            # Datetime formatting
+            if mobj['strf_format']:
+                value = strftime_or_none(value, mobj['strf_format'])
+            if mobj['type'] in 'crs' and value is not None:  # string
+                value = sanitize('%{}'.format(mobj['type']) % fields[-1], value)
+            else:  # numeric
+                numeric_fields.append(final_key)
+                value = float_or_none(value)
             if value is not None:
-                template_dict[key] = value
+                template_dict[final_key] = value
 
         # Missing numeric fields used together with integer presentation types
         # in format specification will break the argument substitution since
         # string NA placeholder is returned for missing fields. We will patch
         # output template for missing fields to meet string presentation type.
         for numeric_field in numeric_fields:
-            if numeric_field not in template_dict:
+            if template_dict.get(numeric_field) is None:
                 outtmpl = re.sub(
                     FORMAT_RE.format(re.escape(numeric_field)),
                     r'%({0})s'.format(numeric_field), outtmpl)
 
+        template_dict = collections.defaultdict(lambda: na, (
+            (k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
+            for k, v in template_dict.items() if v is not None))
         return outtmpl, template_dict
 
     def _prepare_filename(self, info_dict, tmpl_type='default'):

yt_dlp/options.py

@@ -908,7 +908,7 @@ def _dict_from_multiple_values_options_callback(
     filesystem.add_option(
         '--autonumber-start',
         dest='autonumber_start', metavar='NUMBER', default=1, type=int,
-        help='Specify the start value for %(autonumber)s (default is %default)')
+        help=optparse.SUPPRESS_HELP)
     filesystem.add_option(
         '--restrict-filenames',
         action='store_true', dest='restrictfilenames', default=False,

yt_dlp/utils.py

@@ -6092,11 +6092,20 @@ def load_plugins(name, type, namespace):
 
 
 def traverse_dict(dictn, keys, casesense=True):
-    if not isinstance(dictn, dict):
+    keys = list(keys)[::-1]
-        return None
+    while keys:
-    first_key = keys[0]
+        key = keys.pop()
-    if not casesense:
+        if isinstance(dictn, dict):
-        dictn = {key.lower(): val for key, val in dictn.items()}
+            if not casesense:
-        first_key = first_key.lower()
+                dictn = {k.lower(): v for k, v in dictn.items()}
-    value = dictn.get(first_key, None)
+                key = key.lower()
-    return value if len(keys) < 2 else traverse_dict(value, keys[1:], casesense)
+            dictn = dictn.get(key)
+        elif isinstance(dictn, (list, tuple, compat_str)):
+            key, n = int_or_none(key), len(dictn)
+            if key is not None and -n <= key < n:
+                dictn = dictn[key]
+            else:
+                dictn = None
+        else:
+            return None
+    return dictn