From 9322f1162de5d5f4fc7e911acdce782ccb943712 Mon Sep 17 00:00:00 2001 From: insaneracist Date: Thu, 29 Oct 2020 14:37:06 -0700 Subject: [PATCH] [youtube] fix: extract artist metadata from ytInitialData (#49) --- youtube_dlc/extractor/youtube.py | 46 ++++++++++++++++++++++++++++++++ 1 file changed, 46 insertions(+) diff --git a/youtube_dlc/extractor/youtube.py b/youtube_dlc/extractor/youtube.py index 5fd22081a..c75a7edae 100644 --- a/youtube_dlc/extractor/youtube.py +++ b/youtube_dlc/extractor/youtube.py @@ -1406,6 +1406,44 @@ def _get_yt_initial_data(self, video_id, webpage): return self._parse_json( uppercase_escape(config), video_id, fatal=False) + def _get_music_metadata_from_yt_initial(self, yt_initial): + music_metadata = [] + key_map = { + 'Album': 'album', + 'Artist': 'artist', + 'Song': 'track' + } + contents = try_get(yt_initial, lambda x: x['contents']['twoColumnWatchNextResults']['results']['results']['contents']) + if type(contents) is list: + for content in contents: + music_track = {} + if type(content) is not dict: + continue + videoSecondaryInfoRenderer = try_get(content, lambda x: x['videoSecondaryInfoRenderer']) + if type(videoSecondaryInfoRenderer) is not dict: + continue + rows = try_get(videoSecondaryInfoRenderer, lambda x: x['metadataRowContainer']['metadataRowContainerRenderer']['rows']) + if type(rows) is not list: + continue + for row in rows: + metadataRowRenderer = try_get(row, lambda x: x['metadataRowRenderer']) + if type(metadataRowRenderer) is not dict: + continue + key = try_get(metadataRowRenderer, lambda x: x['title']['simpleText']) + value = try_get(metadataRowRenderer, lambda x: x['contents'][0]['simpleText']) or \ + try_get(metadataRowRenderer, lambda x: x['contents'][0]['runs'][0]['text']) + if type(key) is not str or type(value) is not str: + continue + if key in key_map: + if key_map[key] in music_track: + # we've started on a new track + music_metadata.append(music_track) + music_track = {} + music_track[key_map[key]] = value + if len(music_track.keys()): + music_metadata.append(music_track) + return music_metadata + def _get_automatic_captions(self, video_id, webpage): """We need the webpage for getting the captions url, pass it as an argument to speed up the process.""" @@ -2328,6 +2366,14 @@ def extract_meta(field): if release_year: release_year = int(release_year) + yt_initial = self._get_yt_initial_data(video_id, video_webpage) + if yt_initial: + music_metadata = self._get_music_metadata_from_yt_initial(yt_initial) + if len(music_metadata): + album = music_metadata[0].get('album') + artist = music_metadata[0].get('artist') + track = music_metadata[0].get('track') + m_episode = re.search( r']+id="watch7-headline"[^>]*>\s*]*>.*?>(?P[^<]+)\s*S(?P\d+)\s*•\s*E(?P\d+)', video_webpage)