Split video by chapters (#158)

* New options `--split-chapters` and `--no-split-chapters`
* The output/path of the split files can be given using the key `chapter`
* Additional keys `section_title`, `section_number`, `section_start`, `section_end` are available in the output template
* Alias `--split-tracks` for parity with animelover/youtube-dl
* `--sponskrub-cut` and `--split-chapter` cannot work together

Closes:
https://github.com/blackjack4494/yt-dlc/issues/277
https://github.com/ytdl-org/youtube-dl/issues/28438
https://github.com/ytdl-org/youtube-dl/issues/12907
https://github.com/ytdl-org/youtube-dl/issues/6480
https://github.com/ytdl-org/youtube-dl/pull/25005

Rewritten from the implementation by: femaref and Wattux
https://github.com/Wattux/youtube-dl/tree/split-at-timestamps
https://github.com/ytdl-org/youtube-dl/pull/25005
https://github.com/femaref/youtube-dl/tree/split-track

yt_dlp/__init__.py

@@ -279,9 +279,14 @@ def _real_main(argv=None):
 
     def report_conflict(arg1, arg2):
         write_string('WARNING: %s is ignored since %s was given\n' % (arg2, arg1), out=sys.stderr)
+
     if opts.remuxvideo and opts.recodevideo:
         report_conflict('--recode-video', '--remux-video')
         opts.remuxvideo = False
+    if opts.sponskrub_cut and opts.split_chapters and opts.sponskrub is not False:
+        report_conflict('--split-chapter', '--sponskrub-cut')
+        opts.sponskrub_cut = False
+
     if opts.allow_unplayable_formats:
         if opts.extractaudio:
             report_conflict('--allow-unplayable-formats', '--extract-audio')
@@ -371,11 +376,7 @@ def _real_main(argv=None):
         })
         if not already_have_thumbnail:
             opts.writethumbnail = True
-    # XAttrMetadataPP should be run after post-processors that may change file
-    # contents
-    if opts.xattrs:
-        postprocessors.append({'key': 'XAttrMetadata'})
-    # This should be below all ffmpeg PP because it may cut parts out from the video
+    # This should be below most ffmpeg PP because it may cut parts out from the video
     # If opts.sponskrub is None, sponskrub is used, but it silently fails if the executable can't be found
     if opts.sponskrub is not False:
         postprocessors.append({
@@ -386,6 +387,11 @@ def _real_main(argv=None):
             'force': opts.sponskrub_force,
             'ignoreerror': opts.sponskrub is None,
         })
+    if opts.split_chapters:
+        postprocessors.append({'key': 'FFmpegSplitChapters'})
+    # XAttrMetadataPP should be run after post-processors that may change file contents
+    if opts.xattrs:
+        postprocessors.append({'key': 'XAttrMetadata'})
     # ExecAfterDownload must be the last PP
     if opts.exec_cmd:
         postprocessors.append({

yt_dlp/options.py

@@ -1183,6 +1183,17 @@ def parseOpts(overrideArguments=None):
         '--convert-subs', '--convert-subtitles',
         metavar='FORMAT', dest='convertsubtitles', default=None,
         help='Convert the subtitles to other format (currently supported: srt|ass|vtt|lrc)')
+    postproc.add_option(
+        '--split-chapters', '--split-tracks',
+        dest='split_chapters', action='store_true', default=False,
+        help=(
+            'Split video into multiple files based on internal chapters. '
+            'The "chapter:" prefix can be used with "--paths" and "--output" to '
+            'set the output filename for the split files. See "OUTPUT TEMPLATE" for details'))
+    postproc.add_option(
+        '--no-split-chapters', '--no-split-tracks',
+        dest='split_chapters', action='store_false',
+        help='Do not split video based on chapters (default)')
 
     sponskrub = optparse.OptionGroup(parser, 'SponSkrub (SponsorBlock) Options', description=(
         'SponSkrub (https://github.com/yt-dlp/SponSkrub) is a utility to mark/remove sponsor segments '

yt_dlp/postprocessor/__init__.py

@@ -13,6 +13,7 @@ from .ffmpeg import (
     FFmpegVideoConvertorPP,
     FFmpegVideoRemuxerPP,
     FFmpegSubtitlesConvertorPP,
+    FFmpegSplitChaptersPP,
 )
 from .xattrpp import XAttrMetadataPP
 from .execafterdownload import ExecAfterDownloadPP
@@ -31,6 +32,7 @@ __all__ = [
     'ExecAfterDownloadPP',
     'FFmpegEmbedSubtitlePP',
     'FFmpegExtractAudioPP',
+    'FFmpegSplitChaptersPP',
     'FFmpegFixupM3u8PP',
     'FFmpegFixupM4aPP',
     'FFmpegFixupStretchedPP',

yt_dlp/postprocessor/ffmpeg.py

@@ -10,6 +10,7 @@ import json
 
 from .common import AudioConversionError, PostProcessor
 
+from ..compat import compat_str
 from ..utils import (
     encodeArgument,
     encodeFilename,
@@ -769,3 +770,40 @@ class FFmpegSubtitlesConvertorPP(FFmpegPostProcessor):
                 }
 
         return sub_filenames, info
+
+
+class FFmpegSplitChaptersPP(FFmpegPostProcessor):
+
+    def _prepare_filename(self, number, chapter, info):
+        info = info.copy()
+        info.update({
+            'section_number': number,
+            'section_title': chapter.get('title'),
+            'section_start': chapter.get('start_time'),
+            'section_end': chapter.get('end_time'),
+        })
+        return self._downloader.prepare_filename(info, 'chapter')
+
+    def _ffmpeg_args_for_chapter(self, number, chapter, info):
+        destination = self._prepare_filename(number, chapter, info)
+        if not self._downloader._ensure_dir_exists(encodeFilename(destination)):
+            return
+
+        chapter['_filename'] = destination
+        self.to_screen('Chapter %03d; Destination: %s' % (number, destination))
+        return (
+            destination,
+            ['-ss', compat_str(chapter['start_time']),
+             '-to', compat_str(chapter['end_time'])])
+
+    def run(self, info):
+        chapters = info.get('chapters') or []
+        if not chapters:
+            self.report_warning('There are no tracks to extract')
+            return [], info
+
+        self.to_screen('Splitting video by chapters; %d chapters found' % len(chapters))
+        for idx, chapter in enumerate(chapters):
+            destination, opts = self._ffmpeg_args_for_chapter(idx + 1, chapter, info)
+            self.real_run_ffmpeg([(info['filepath'], opts)], [(destination, ['-c', 'copy'])])
+        return [], info

yt_dlp/utils.py

@@ -4182,8 +4182,10 @@ def qualities(quality_ids):
 
 DEFAULT_OUTTMPL = {
     'default': '%(title)s [%(id)s].%(ext)s',
+    'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
 }
 OUTTMPL_TYPES = {
+    'chapter': None,
     'subtitle': None,
     'thumbnail': None,
     'description': 'description',