xZero707/youtube-dl
1
0
Fork 0
mirror of https://github.com/ytdl-org/youtube-dl.git synced 2024-11-27 06:10:38 +01:00
Code Issues Packages Projects Releases Wiki Activity

[youtube] improve subtitle extraction

Browse Source
This commit is contained in:
Remita Amine 2021-02-01 18:12:35 +01:00
parent efef4ddf51
commit 65eee5a745
1 changed files with 18 additions and 25 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
youtube_dl/extractor/youtube.py
View File

@ -1664,7 +1664,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
player_response,
lambda x: x['captions']['playerCaptionsTracklistRenderer'], dict)
if pctr:
def process_language(container, base_url, caption, query):
def process_language(container, base_url, lang_code, query):
lang_subs = []
for fmt in self._SUBTITLE_FORMATS:
query.update({
@ -1674,35 +1674,28 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
'ext': fmt,
'url': update_url_query(base_url, query),
})
subtitles[caption['languageCode']] = lang_subs
container[lang_code] = lang_subs
subtitles = {}
for caption_track in pctr['captionTracks']:
base_url = caption_track['baseUrl']
for caption_track in (pctr.get('captionTracks') or []):
base_url = caption_track.get('baseUrl')
if not base_url:
continue
if caption_track.get('kind') != 'asr':
lang_subs = []
for fmt in self._SUBTITLE_FORMATS:
lang_subs.append({
'ext': fmt,
'url': update_url_query(base_url, {
'fmt': fmt,
}),
})
subtitles[caption_track['languageCode']] = lang_subs
lang_code = caption_track.get('languageCode')
if not lang_code:
continue
process_language(
subtitles, base_url, lang_code, {})
continue
automatic_captions = {}
for translation_language in pctr['translationLanguages']:
translation_language_code = translation_language['languageCode']
lang_subs = []
for fmt in self._SUBTITLE_FORMATS:
lang_subs.append({
'ext': fmt,
'url': update_url_query(base_url, {
'fmt': fmt,
'tlang': translation_language_code,
}),
})
automatic_captions[translation_language_code] = lang_subs
for translation_language in (pctr.get('translationLanguages') or []):
translation_language_code = translation_language.get('languageCode')
if not translation_language_code:
continue
process_language(
automatic_captions, base_url, translation_language_code,
{'tlang': translation_language_code})
info['automatic_captions'] = automatic_captions
info['subtitles'] = subtitles