mirror of
https://github.com/yt-dlp/yt-dlp
synced 2024-12-25 21:59:04 +01:00
[ie/bilibili] Fix supporter-only video extraction (#11711)
Fix bug in 239f5f36fe
Closes #11702
Authored by: grqz, bashonly
Co-authored-by: bashonly <88596187+bashonly@users.noreply.github.com>
This commit is contained in:
parent
d8fb349086
commit
f05a1cd149
1 changed files with 64 additions and 61 deletions
|
@ -652,13 +652,6 @@ class BiliBiliIE(BilibiliBaseIE):
|
||||||
else:
|
else:
|
||||||
video_data = initial_state['videoData']
|
video_data = initial_state['videoData']
|
||||||
|
|
||||||
if video_data.get('is_upower_exclusive'):
|
|
||||||
high_level = traverse_obj(initial_state, ('elecFullInfo', 'show_info', 'high_level', {dict})) or {}
|
|
||||||
raise ExtractorError(
|
|
||||||
'This is a supporter-only video: '
|
|
||||||
f'{join_nonempty("title", "sub_title", from_dict=high_level, delim=",")}. '
|
|
||||||
f'{self._login_hint()}', expected=True)
|
|
||||||
|
|
||||||
video_id, title = video_data['bvid'], video_data.get('title')
|
video_id, title = video_data['bvid'], video_data.get('title')
|
||||||
|
|
||||||
# Bilibili anthologies are similar to playlists but all videos share the same video ID as the anthology itself.
|
# Bilibili anthologies are similar to playlists but all videos share the same video ID as the anthology itself.
|
||||||
|
@ -726,62 +719,72 @@ class BiliBiliIE(BilibiliBaseIE):
|
||||||
self._get_interactive_entries(video_id, cid, metainfo, headers=headers), **metainfo,
|
self._get_interactive_entries(video_id, cid, metainfo, headers=headers), **metainfo,
|
||||||
duration=traverse_obj(initial_state, ('videoData', 'duration', {int_or_none})),
|
duration=traverse_obj(initial_state, ('videoData', 'duration', {int_or_none})),
|
||||||
__post_extractor=self.extract_comments(aid))
|
__post_extractor=self.extract_comments(aid))
|
||||||
else:
|
|
||||||
formats = self.extract_formats(play_info)
|
|
||||||
|
|
||||||
if not traverse_obj(play_info, ('dash')):
|
formats = self.extract_formats(play_info)
|
||||||
# we only have legacy formats and need additional work
|
|
||||||
has_qn = lambda x: x in traverse_obj(formats, (..., 'quality'))
|
|
||||||
for qn in traverse_obj(play_info, ('accept_quality', lambda _, v: not has_qn(v), {int})):
|
|
||||||
formats.extend(traverse_obj(
|
|
||||||
self.extract_formats(self._download_playinfo(video_id, cid, headers=headers, qn=qn)),
|
|
||||||
lambda _, v: not has_qn(v['quality'])))
|
|
||||||
self._check_missing_formats(play_info, formats)
|
|
||||||
flv_formats = traverse_obj(formats, lambda _, v: v['fragments'])
|
|
||||||
if flv_formats and len(flv_formats) < len(formats):
|
|
||||||
# Flv and mp4 are incompatible due to `multi_video` workaround, so drop one
|
|
||||||
if not self._configuration_arg('prefer_multi_flv'):
|
|
||||||
dropped_fmts = ', '.join(
|
|
||||||
f'{f.get("format_note")} ({f.get("format_id")})' for f in flv_formats)
|
|
||||||
formats = traverse_obj(formats, lambda _, v: not v.get('fragments'))
|
|
||||||
if dropped_fmts:
|
|
||||||
self.to_screen(
|
|
||||||
f'Dropping incompatible flv format(s) {dropped_fmts} since mp4 is available. '
|
|
||||||
'To extract flv, pass --extractor-args "bilibili:prefer_multi_flv"')
|
|
||||||
else:
|
|
||||||
formats = traverse_obj(
|
|
||||||
# XXX: Filtering by extractor-arg is for testing purposes
|
|
||||||
formats, lambda _, v: v['quality'] == int(self._configuration_arg('prefer_multi_flv')[0]),
|
|
||||||
) or [max(flv_formats, key=lambda x: x['quality'])]
|
|
||||||
|
|
||||||
if traverse_obj(formats, (0, 'fragments')):
|
if video_data.get('is_upower_exclusive'):
|
||||||
# We have flv formats, which are individual short videos with their own timestamps and metainfo
|
high_level = traverse_obj(initial_state, ('elecFullInfo', 'show_info', 'high_level', {dict})) or {}
|
||||||
# Binary concatenation corrupts their timestamps, so we need a `multi_video` workaround
|
msg = f'{join_nonempty("title", "sub_title", from_dict=high_level, delim=",")}. {self._login_hint()}'
|
||||||
return {
|
if not formats:
|
||||||
**metainfo,
|
raise ExtractorError(f'This is a supporter-only video: {msg}', expected=True)
|
||||||
'_type': 'multi_video',
|
if '试看' in traverse_obj(play_info, ('accept_description', ..., {str})):
|
||||||
'entries': [{
|
self.report_warning(
|
||||||
'id': f'{metainfo["id"]}_{idx}',
|
f'This is a supporter-only video, only the preview will be extracted: {msg}',
|
||||||
'title': metainfo['title'],
|
video_id=video_id)
|
||||||
'http_headers': metainfo['http_headers'],
|
|
||||||
'formats': [{
|
if not traverse_obj(play_info, 'dash'):
|
||||||
**fragment,
|
# we only have legacy formats and need additional work
|
||||||
'format_id': formats[0].get('format_id'),
|
has_qn = lambda x: x in traverse_obj(formats, (..., 'quality'))
|
||||||
}],
|
for qn in traverse_obj(play_info, ('accept_quality', lambda _, v: not has_qn(v), {int})):
|
||||||
'subtitles': self.extract_subtitles(video_id, cid) if idx == 0 else None,
|
formats.extend(traverse_obj(
|
||||||
'__post_extractor': self.extract_comments(aid) if idx == 0 else None,
|
self.extract_formats(self._download_playinfo(video_id, cid, headers=headers, qn=qn)),
|
||||||
} for idx, fragment in enumerate(formats[0]['fragments'])],
|
lambda _, v: not has_qn(v['quality'])))
|
||||||
'duration': float_or_none(play_info.get('timelength'), scale=1000),
|
self._check_missing_formats(play_info, formats)
|
||||||
}
|
flv_formats = traverse_obj(formats, lambda _, v: v['fragments'])
|
||||||
else:
|
if flv_formats and len(flv_formats) < len(formats):
|
||||||
return {
|
# Flv and mp4 are incompatible due to `multi_video` workaround, so drop one
|
||||||
**metainfo,
|
if not self._configuration_arg('prefer_multi_flv'):
|
||||||
'formats': formats,
|
dropped_fmts = ', '.join(
|
||||||
'duration': float_or_none(play_info.get('timelength'), scale=1000),
|
f'{f.get("format_note")} ({f.get("format_id")})' for f in flv_formats)
|
||||||
'chapters': self._get_chapters(aid, cid),
|
formats = traverse_obj(formats, lambda _, v: not v.get('fragments'))
|
||||||
'subtitles': self.extract_subtitles(video_id, cid),
|
if dropped_fmts:
|
||||||
'__post_extractor': self.extract_comments(aid),
|
self.to_screen(
|
||||||
}
|
f'Dropping incompatible flv format(s) {dropped_fmts} since mp4 is available. '
|
||||||
|
'To extract flv, pass --extractor-args "bilibili:prefer_multi_flv"')
|
||||||
|
else:
|
||||||
|
formats = traverse_obj(
|
||||||
|
# XXX: Filtering by extractor-arg is for testing purposes
|
||||||
|
formats, lambda _, v: v['quality'] == int(self._configuration_arg('prefer_multi_flv')[0]),
|
||||||
|
) or [max(flv_formats, key=lambda x: x['quality'])]
|
||||||
|
|
||||||
|
if traverse_obj(formats, (0, 'fragments')):
|
||||||
|
# We have flv formats, which are individual short videos with their own timestamps and metainfo
|
||||||
|
# Binary concatenation corrupts their timestamps, so we need a `multi_video` workaround
|
||||||
|
return {
|
||||||
|
**metainfo,
|
||||||
|
'_type': 'multi_video',
|
||||||
|
'entries': [{
|
||||||
|
'id': f'{metainfo["id"]}_{idx}',
|
||||||
|
'title': metainfo['title'],
|
||||||
|
'http_headers': metainfo['http_headers'],
|
||||||
|
'formats': [{
|
||||||
|
**fragment,
|
||||||
|
'format_id': formats[0].get('format_id'),
|
||||||
|
}],
|
||||||
|
'subtitles': self.extract_subtitles(video_id, cid) if idx == 0 else None,
|
||||||
|
'__post_extractor': self.extract_comments(aid) if idx == 0 else None,
|
||||||
|
} for idx, fragment in enumerate(formats[0]['fragments'])],
|
||||||
|
'duration': float_or_none(play_info.get('timelength'), scale=1000),
|
||||||
|
}
|
||||||
|
|
||||||
|
return {
|
||||||
|
**metainfo,
|
||||||
|
'formats': formats,
|
||||||
|
'duration': float_or_none(play_info.get('timelength'), scale=1000),
|
||||||
|
'chapters': self._get_chapters(aid, cid),
|
||||||
|
'subtitles': self.extract_subtitles(video_id, cid),
|
||||||
|
'__post_extractor': self.extract_comments(aid),
|
||||||
|
}
|
||||||
|
|
||||||
|
|
||||||
class BiliBiliBangumiIE(BilibiliBaseIE):
|
class BiliBiliBangumiIE(BilibiliBaseIE):
|
||||||
|
|
Loading…
Reference in a new issue