diff options
| author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-08-05 09:07:33 +0000 |
|---|---|---|
| committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-08-05 09:07:33 +0000 |
| commit | 1719c758f7efdca9476d2a674b659191dee6344f (patch) | |
| tree | d378103d8967450628b7254d9ec1f4f91993bd04 /yt_dlp/extractor/tiktok.py | |
| parent | Adding debian version 2024.07.16-1. (diff) | |
| download | yt-dlp-1719c758f7efdca9476d2a674b659191dee6344f.tar.xz yt-dlp-1719c758f7efdca9476d2a674b659191dee6344f.zip | |
Merging upstream version 2024.07.25.
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'yt_dlp/extractor/tiktok.py')
| -rw-r--r-- | yt_dlp/extractor/tiktok.py | 19 |
1 files changed, 16 insertions, 3 deletions
diff --git a/yt_dlp/extractor/tiktok.py b/yt_dlp/extractor/tiktok.py index aa1dcec..9d823a3 100644 --- a/yt_dlp/extractor/tiktok.py +++ b/yt_dlp/extractor/tiktok.py @@ -23,7 +23,6 @@ from ..utils import ( mimetype2ext, parse_qs, qualities, - remove_start, srt_subtitles_timecode, str_or_none, traverse_obj, @@ -254,7 +253,16 @@ class TikTokBaseIE(InfoExtractor): def _get_subtitles(self, aweme_detail, aweme_id, user_name): # TODO: Extract text positioning info + + EXT_MAP = { # From lowest to highest preference + 'creator_caption': 'json', + 'srt': 'srt', + 'webvtt': 'vtt', + } + preference = qualities(tuple(EXT_MAP.values())) + subtitles = {} + # aweme/detail endpoint subs captions_info = traverse_obj( aweme_detail, ('interaction_stickers', ..., 'auto_video_caption_info', 'auto_captions', ...), expected_type=dict) @@ -278,8 +286,8 @@ class TikTokBaseIE(InfoExtractor): if not caption.get('url'): continue subtitles.setdefault(caption.get('lang') or 'en', []).append({ - 'ext': remove_start(caption.get('caption_format'), 'web'), 'url': caption['url'], + 'ext': EXT_MAP.get(caption.get('Format')), }) # webpage subs if not subtitles: @@ -288,9 +296,14 @@ class TikTokBaseIE(InfoExtractor): self._create_url(user_name, aweme_id), aweme_id, fatal=False) for caption in traverse_obj(aweme_detail, ('video', 'subtitleInfos', lambda _, v: v['Url'])): subtitles.setdefault(caption.get('LanguageCodeName') or 'en', []).append({ - 'ext': remove_start(caption.get('Format'), 'web'), 'url': caption['Url'], + 'ext': EXT_MAP.get(caption.get('Format')), }) + + # Deprioritize creator_caption json since it can't be embedded or used by media players + for lang, subs_list in subtitles.items(): + subtitles[lang] = sorted(subs_list, key=lambda x: preference(x['ext'])) + return subtitles def _parse_url_key(self, url_key): |