Miroirs/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp synced 2024-11-16 07:48:01 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 6

[ie/facebook] Fix formats extraction (#11513)

Browse source 
Closes #11497
Authored by: bashonly
This commit is contained in:
bashonly 2024-11-12 23:23:10 +00:00 committed by GitHub
parent a9f85670d0
commit bacc31b05a
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 30 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

35
yt_dlp/extractor/facebook.py
View file

@ -563,13 +563,13 @@ class FacebookIE(InfoExtractor):
return extract_video_data(try_get( return extract_video_data(try_get(
js_data, lambda x: x['jsmods']['instances'], list) or []) js_data, lambda x: x['jsmods']['instances'], list) or [])
def extract_dash_manifest(video, formats): def extract_dash_manifest(vid_data, formats, mpd_url=None):
dash_manifest = traverse_obj( dash_manifest = traverse_obj(
video, 'dash_manifest', 'playlist', 'dash_manifest_xml_string', expected_type=str) vid_data, 'dash_manifest', 'playlist', 'dash_manifest_xml_string', 'manifest_xml', expected_type=str)
if dash_manifest: if dash_manifest:
formats.extend(self._parse_mpd_formats( formats.extend(self._parse_mpd_formats(
compat_etree_fromstring(urllib.parse.unquote_plus(dash_manifest)), compat_etree_fromstring(urllib.parse.unquote_plus(dash_manifest)),
mpd_url=url_or_none(video.get('dash_manifest_url')))) mpd_url=url_or_none(video.get('dash_manifest_url')) or mpd_url))
def process_formats(info): def process_formats(info):
# Downloads with browser's User-Agent are rate limited. Working around # Downloads with browser's User-Agent are rate limited. Working around
@ -619,9 +619,12 @@ class FacebookIE(InfoExtractor):
video = video['creation_story'] video = video['creation_story']
video['owner'] = traverse_obj(video, ('short_form_video_context', 'video_owner')) video['owner'] = traverse_obj(video, ('short_form_video_context', 'video_owner'))
video.update(reel_info) video.update(reel_info)
fmt_data = traverse_obj(video, ('videoDeliveryLegacyFields', {dict})) or video
formats = [] formats = []
q = qualities(['sd', 'hd']) q = qualities(['sd', 'hd'])
# Legacy formats extraction
fmt_data = traverse_obj(video, ('videoDeliveryLegacyFields', {dict})) or video
for key, format_id in (('playable_url', 'sd'), ('playable_url_quality_hd', 'hd'), for key, format_id in (('playable_url', 'sd'), ('playable_url_quality_hd', 'hd'),
('playable_url_dash', ''), ('browser_native_hd_url', 'hd'), ('playable_url_dash', ''), ('browser_native_hd_url', 'hd'),
('browser_native_sd_url', 'sd')): ('browser_native_sd_url', 'sd')):
@ -629,7 +632,7 @@ class FacebookIE(InfoExtractor):
if not playable_url: if not playable_url:
continue continue
if determine_ext(playable_url) == 'mpd': if determine_ext(playable_url) == 'mpd':
formats.extend(self._extract_mpd_formats(playable_url, video_id)) formats.extend(self._extract_mpd_formats(playable_url, video_id, fatal=False))
else: else:
formats.append({ formats.append({
'format_id': format_id, 'format_id': format_id,
@ -638,6 +641,28 @@ class FacebookIE(InfoExtractor):
'url': playable_url, 'url': playable_url,
}) })
extract_dash_manifest(fmt_data, formats) extract_dash_manifest(fmt_data, formats)
# New videoDeliveryResponse formats extraction
fmt_data = traverse_obj(video, ('videoDeliveryResponseFragment', 'videoDeliveryResponseResult'))
mpd_urls = traverse_obj(fmt_data, ('dash_manifest_urls', ..., 'manifest_url', {url_or_none}))
dash_manifests = traverse_obj(fmt_data, ('dash_manifests', lambda _, v: v['manifest_xml']))
for idx, dash_manifest in enumerate(dash_manifests):
extract_dash_manifest(dash_manifest, formats, mpd_url=traverse_obj(mpd_urls, idx))
if not dash_manifests:
# Only extract from MPD URLs if the manifests are not already provided
for mpd_url in mpd_urls:
formats.extend(self._extract_mpd_formats(mpd_url, video_id, fatal=False))
for prog_fmt in traverse_obj(fmt_data, ('progressive_urls', lambda _, v: v['progressive_url'])):
format_id = traverse_obj(prog_fmt, ('metadata', 'quality', {str.lower}))
formats.append({
'format_id': format_id,
# sd, hd formats w/o resolution info should be deprioritized below DASH
'quality': q(format_id) - 3,
'url': prog_fmt['progressive_url'],
})
for m3u8_url in traverse_obj(fmt_data, ('hls_playlist_urls', ..., 'hls_playlist_url', {url_or_none})):
formats.extend(self._extract_m3u8_formats(m3u8_url, video_id, 'mp4', fatal=False, m3u8_id='hls'))
if not formats: if not formats:
# Do not append false positive entry w/o any formats # Do not append false positive entry w/o any formats
return return