jon1scr
/
yt-dlc
forked from minhngoc25a/yt-dlc
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

[youtube] Simplify and extract more metadata from url_encoded_fmt_stream_map (Closes #5993)

This commit is contained in:
Sergey M․ 2015-08-30 08:07:07 +06:00
parent c9afb51cea
commit aabc2be693
1 changed files with 35 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

53
youtube_dl/extractor/youtube.py
View File

@ -1304,32 +1304,49 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
if 'ratebypass' not in url: if 'ratebypass' not in url:
url += '&ratebypass=yes' url += '&ratebypass=yes'
width = None # Some itags are not included in DASH manifest thus corresponding formats will
height = None # lack metadata (see https://github.com/rg3/youtube-dl/pull/5993).
size_str = url_data.get('size', [''])[0] # Trying to extract metadata from url_encoded_fmt_stream_map entry.
if size_str.count('x') == 1: mobj = re.search(r'^(?P<width>\d+)[xX](?P<height>\d+)$', url_data.get('size', [''])[0])
width, height = [int_or_none(x) for x in size_str.split('x')] width, height = (int(mobj.group('width')), int(mobj.group('height'))) if mobj else (None, None)
dct = {
format_url = {
'format_id': format_id, 'format_id': format_id,
'url': url, 'url': url,
'player_url': player_url, 'player_url': player_url,
# As of this writing these are only defined for DASH formats:
'filesize': int_or_none(url_data.get('clen', [None])[0]), 'filesize': int_or_none(url_data.get('clen', [None])[0]),
'tbr': float_or_none(url_data.get('bitrate', [None])[0], scale=1024), 'tbr': float_or_none(url_data.get('bitrate', [None])[0], 1000),
'width': width, 'width': width,
'height': height, 'height': height,
'fps': int_or_none(url_data.get('fps', [None])[0]), 'fps': int_or_none(url_data.get('fps', [None])[0]),
'format_note': url_data.get('quality_label', [None])[0] or url_data.get('quality', [None])[0],
} }
type_ = url_data.get('type', [None])[0]
# drop Nones so they do not overwrite the defaults from self._formats if type_:
format_url = dict((k, v) for k, v in format_url.items() if v is not None) type_split = type_.split(';')
kind_ext = type_split[0].split('/')
format_full = self._formats.get(format_id, {}).copy() if len(kind_ext) == 2:
format_full.update(format_url) kind, ext = kind_ext
dct['ext'] = ext
formats.append(format_full) if kind in ('audio', 'video'):
codecs = None
for mobj in re.finditer(
r'(?P<key>[a-zA-Z_-]+)=(?P<quote>["\']?)(?P<val>.+?)(?P=quote)(?:;|$)', type_):
if mobj.group('key') == 'codecs':
codecs = mobj.group('val')
break
if codecs:
codecs = codecs.split(',')
if len(codecs) == 2:
acodec, vcodec = codecs[0], codecs[1]
else:
acodec, vcodec = (codecs[0], 'none') if kind == 'audio' else ('none', codecs[0])
dct.update({
'acodec': acodec,
'vcodec': vcodec,
})
if format_id in self._formats:
dct.update(self._formats[format_id])
formats.append(dct)
elif video_info.get('hlsvp'): elif video_info.get('hlsvp'):
manifest_url = video_info['hlsvp'][0] manifest_url = video_info['hlsvp'][0]
url_map = self._extract_from_m3u8(manifest_url, video_id) url_map = self._extract_from_m3u8(manifest_url, video_id)