Add option --embed-info-json to embed info-json in mkv

Closes #1644

yt_dlp/YoutubeDL.py

@@ -431,7 +431,7 @@ class YoutubeDL(object):
     compat_opts:       Compatibility options. See "Differences in default behavior".
                        The following options do not work when used through the API:
                        filename, abort-on-error, multistreams, no-live-chat, format-sort
-                       no-clean-infojson, no-playlist-metafiles, no-keep-subs.
+                       no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
                        Refer __init__.py for their implementation
     progress_template: Dictionary of templates for progress outputs.
                        Allowed keys are 'download', 'postprocess',
@@ -2654,6 +2654,8 @@ def process_info(self, info_dict):
         infofn = self.prepare_filename(info_dict, 'infojson')
         _infojson_written = self._write_info_json('video', info_dict, infofn)
         if _infojson_written:
+            info_dict['infojson_filename'] = infofn
+            # For backward compatability, even though it was a private field
             info_dict['__infojson_filename'] = infofn
         elif _infojson_written is None:
             return
@@ -3012,8 +3014,8 @@ def sanitize_info(info_dict, remove_private_keys=False):
         keep_keys = ['_type']  # Always keep this to facilitate load-info-json
         if remove_private_keys:
             remove_keys |= {
-                'requested_formats', 'requested_subtitles', 'requested_entries',
+                'requested_formats', 'requested_subtitles', 'requested_entries', 'entries',
-                'filepath', 'entries', 'original_url', 'playlist_autonumber',
+                'filepath', 'infojson_filename', 'original_url', 'playlist_autonumber',
             }
             empty_values = (None, {}, [], set(), tuple())
             reject = lambda k, v: k not in keep_keys and (

yt_dlp/__init__.py

@@ -290,6 +290,11 @@ def set_default_compat(compat_name, opt_name, default=True, remove_compat=True):
     set_default_compat('abort-on-error', 'ignoreerrors', 'only_download')
     set_default_compat('no-playlist-metafiles', 'allow_playlist_files')
     set_default_compat('no-clean-infojson', 'clean_infojson')
+    if 'no-attach-info-json' in compat_opts:
+        if opts.embed_infojson:
+            _unused_compat_opt('no-attach-info-json')
+        else:
+            opts.embed_infojson = False
     if 'format-sort' in compat_opts:
         opts.format_sort.extend(InfoExtractor.FormatSort.ytdl_default)
     _video_multistreams_set = set_default_compat('multistreams', 'allow_multiple_video_streams', False, remove_compat=False)
@@ -526,11 +531,14 @@ def metadataparser_actions(f):
     # By default ffmpeg preserves metadata applicable for both
     # source and target containers. From this point the container won't change,
     # so metadata can be added here.
-    if opts.addmetadata or opts.addchapters:
+    if opts.addmetadata or opts.addchapters or opts.embed_infojson:
+        if opts.embed_infojson is None:
+            opts.embed_infojson = 'if_exists'
         postprocessors.append({
             'key': 'FFmpegMetadata',
             'add_chapters': opts.addchapters,
             'add_metadata': opts.addmetadata,
+            'add_infojson': opts.embed_infojson,
         })
     # Note: Deprecated
     # This should be above EmbedThumbnail since sponskrub removes the thumbnail attachment

yt_dlp/options.py

@@ -1287,7 +1287,9 @@ def _dict_from_options_callback(
     postproc.add_option(
         '--embed-metadata', '--add-metadata',
         action='store_true', dest='addmetadata', default=False,
-        help='Embed metadata to the video file. Also adds chapters to file unless --no-add-chapters is used (Alias: --add-metadata)')
+        help=(
+            'Embed metadata to the video file. Also embeds chapters/infojson if present '
+            'unless --no-embed-chapters/--no-embed-info-json are used (Alias: --add-metadata)'))
     postproc.add_option(
         '--no-embed-metadata', '--no-add-metadata',
         action='store_false', dest='addmetadata',
@@ -1300,6 +1302,14 @@ def _dict_from_options_callback(
         '--no-embed-chapters', '--no-add-chapters',
         action='store_false', dest='addchapters',
         help='Do not add chapter markers (default) (Alias: --no-add-chapters)')
+    postproc.add_option(
+        '--embed-info-json',
+        action='store_true', dest='embed_infojson', default=None,
+        help='Embed the infojson as an attachment to mkv/mka video files')
+    postproc.add_option(
+        '--no-embed-info-json',
+        action='store_false', dest='embed_infojson',
+        help='Do not embed the infojson as an attachment to the video file')
     postproc.add_option(
         '--metadata-from-title',
         metavar='FORMAT', dest='metafromtitle',

yt_dlp/postprocessor/ffmpeg.py

@@ -28,6 +28,7 @@
     shell_quote,
     traverse_obj,
     variadic,
+    write_json_file,
 )
 
 
@@ -636,10 +637,11 @@ def run(self, information):
 
 class FFmpegMetadataPP(FFmpegPostProcessor):
 
-    def __init__(self, downloader, add_metadata=True, add_chapters=True):
+    def __init__(self, downloader, add_metadata=True, add_chapters=True, add_infojson='if_exists'):
         FFmpegPostProcessor.__init__(self, downloader)
         self._add_metadata = add_metadata
         self._add_chapters = add_chapters
+        self._add_infojson = add_infojson
 
     @staticmethod
     def _options(target_ext):
@@ -652,13 +654,23 @@ def _options(target_ext):
     @PostProcessor._restrict_to(images=False)
     def run(self, info):
         filename, metadata_filename = info['filepath'], None
-        options = []
+        files_to_delete, options = [], []
         if self._add_chapters and info.get('chapters'):
             metadata_filename = replace_extension(filename, 'meta')
             options.extend(self._get_chapter_opts(info['chapters'], metadata_filename))
+            files_to_delete.append(metadata_filename)
         if self._add_metadata:
             options.extend(self._get_metadata_opts(info))
 
+        if self._add_infojson:
+            if info['ext'] in ('mkv', 'mka'):
+                infojson_filename = info.get('infojson_filename')
+                options.extend(self._get_infojson_opts(info, infojson_filename))
+                if not infojson_filename:
+                    files_to_delete.append(info.get('infojson_filename'))
+            elif self._add_infojson is True:
+                self.to_screen('The info-json can only be attached to mkv/mka files')
+
         if not options:
             self.to_screen('There isn\'t any metadata to add')
             return [], info
@@ -668,8 +680,8 @@ def run(self, info):
         self.run_ffmpeg_multiple_files(
             (filename, metadata_filename), temp_filename,
             itertools.chain(self._options(info['ext']), *options))
-        if metadata_filename:
+        for file in filter(None, files_to_delete):
-            os.remove(metadata_filename)
+            os.remove(file)  # Don't obey --keep-files
         os.replace(temp_filename, filename)
         return [], info
 
@@ -741,14 +753,25 @@ def add(meta_list, info_list=None):
                     yield ('-metadata:s:%d' % (stream_idx + i), 'language=%s' % lang)
             stream_idx += stream_count
 
-        if ('no-attach-info-json' not in self.get_param('compat_opts', [])
+    def _get_infojson_opts(self, info, infofn):
-                and '__infojson_filename' in info and info['ext'] in ('mkv', 'mka')):
+        if not infofn or not os.path.exists(infofn):
+            if self._add_infojson is not True:
+                return
+            infofn = infofn or '%s.temp' % (
+                self._downloader.prepare_filename(info, 'infojson')
+                or replace_extension(self._downloader.prepare_filename(info), 'info.json', info['ext']))
+            if not self._downloader._ensure_dir_exists(infofn):
+                return
+            self.write_debug(f'Writing info-json to: {infofn}')
+            write_json_file(self._downloader.sanitize_info(info, self.get_param('clean_infojson', True)), infofn)
+            info['infojson_filename'] = infofn
+
         old_stream, new_stream = self.get_stream_number(info['filepath'], ('tags', 'mimetype'), 'application/json')
         if old_stream is not None:
             yield ('-map', '-0:%d' % old_stream)
             new_stream -= 1
 
-            yield ('-attach', info['__infojson_filename'],
+        yield ('-attach', infofn,
                '-metadata:s:%d' % new_stream, 'mimetype=application/json')