aboutsummaryrefslogtreecommitdiff
path: root/youtube_dl
diff options
context:
space:
mode:
authorSergey M․ <dstftw@gmail.com>2019-09-11 22:44:47 +0700
committerSergey M․ <dstftw@gmail.com>2019-09-11 22:44:47 +0700
commitbf1317d257d13188601c837c983830355c6203e5 (patch)
tree7b9e4a7f1ee6115ada5494c1fd11d6c4533c9848 /youtube_dl
parentbff90fc518d6ccadaafc26407a688dc1bbd32dff (diff)
[youtube] Quick extraction tempfix (closes #22367, closes #22163)
Diffstat (limited to 'youtube_dl')
-rw-r--r--youtube_dl/extractor/youtube.py190
1 files changed, 109 insertions, 81 deletions
diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py
index abafd5157..9d0058b2a 100644
--- a/youtube_dl/extractor/youtube.py
+++ b/youtube_dl/extractor/youtube.py
@@ -1915,6 +1915,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
return int_or_none(self._search_regex(
r'\bclen[=/](\d+)', media_url, 'filesize', default=None))
+ streaming_formats = try_get(player_response, lambda x: x['streamingData']['formats'], list) or []
+ streaming_formats.extend(try_get(player_response, lambda x: x['streamingData']['adaptiveFormats'], list) or [])
+
if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
self.report_rtmp_download()
formats = [{
@@ -1923,10 +1926,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
'url': video_info['conn'][0],
'player_url': player_url,
}]
- elif not is_live and (len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1):
+ elif not is_live and (streaming_formats or len(video_info.get('url_encoded_fmt_stream_map', [''])[0]) >= 1 or len(video_info.get('adaptive_fmts', [''])[0]) >= 1):
encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts', [''])[0]
if 'rtmpe%3Dyes' in encoded_url_map:
raise ExtractorError('rtmpe downloads are not supported, see https://github.com/ytdl-org/youtube-dl/issues/343 for more information.', expected=True)
+ formats = []
formats_spec = {}
fmt_list = video_info.get('fmt_list', [''])[0]
if fmt_list:
@@ -1941,90 +1945,105 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
'height': int_or_none(width_height[1]),
}
q = qualities(['small', 'medium', 'hd720'])
- streaming_formats = try_get(player_response, lambda x: x['streamingData']['formats'], list)
- if streaming_formats:
- for fmt in streaming_formats:
- itag = str_or_none(fmt.get('itag'))
- if not itag:
- continue
- quality = fmt.get('quality')
- quality_label = fmt.get('qualityLabel') or quality
- formats_spec[itag] = {
- 'asr': int_or_none(fmt.get('audioSampleRate')),
- 'filesize': int_or_none(fmt.get('contentLength')),
- 'format_note': quality_label,
- 'fps': int_or_none(fmt.get('fps')),
- 'height': int_or_none(fmt.get('height')),
- 'quality': q(quality),
- # bitrate for itag 43 is always 2147483647
- 'tbr': float_or_none(fmt.get('averageBitrate') or fmt.get('bitrate'), 1000) if itag != '43' else None,
- 'width': int_or_none(fmt.get('width')),
- }
- formats = []
- for url_data_str in encoded_url_map.split(','):
- url_data = compat_parse_qs(url_data_str)
- if 'itag' not in url_data or 'url' not in url_data or url_data.get('drm_families'):
+ for fmt in streaming_formats:
+ itag = str_or_none(fmt.get('itag'))
+ if not itag:
continue
+ quality = fmt.get('quality')
+ quality_label = fmt.get('qualityLabel') or quality
+ formats_spec[itag] = {
+ 'asr': int_or_none(fmt.get('audioSampleRate')),
+ 'filesize': int_or_none(fmt.get('contentLength')),
+ 'format_note': quality_label,
+ 'fps': int_or_none(fmt.get('fps')),
+ 'height': int_or_none(fmt.get('height')),
+ 'quality': q(quality),
+ # bitrate for itag 43 is always 2147483647
+ 'tbr': float_or_none(fmt.get('averageBitrate') or fmt.get('bitrate'), 1000) if itag != '43' else None,
+ 'width': int_or_none(fmt.get('width')),
+ }
+
+ for fmt in streaming_formats:
+ if fmt.get('drm_families'):
+ continue
+ url = url_or_none(fmt.get('url'))
+
+ if not url:
+ cipher = fmt.get('cipher')
+ if not cipher:
+ continue
+ url_data = compat_parse_qs(cipher)
+ url = url_or_none(try_get(url_data, lambda x: x['url'][0], compat_str))
+ if not url:
+ continue
+ else:
+ cipher = None
+ url_data = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
+
stream_type = int_or_none(try_get(url_data, lambda x: x['stream_type'][0]))
# Unsupported FORMAT_STREAM_TYPE_OTF
if stream_type == 3:
continue
- format_id = url_data['itag'][0]
- url = url_data['url'][0]
-
- if 's' in url_data or self._downloader.params.get('youtube_include_dash_manifest', True):
- ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
- jsplayer_url_json = self._search_regex(
- ASSETS_RE,
- embed_webpage if age_gate else video_webpage,
- 'JS player URL (1)', default=None)
- if not jsplayer_url_json and not age_gate:
- # We need the embed website after all
- if embed_webpage is None:
- embed_url = proto + '://www.youtube.com/embed/%s' % video_id
- embed_webpage = self._download_webpage(
- embed_url, video_id, 'Downloading embed webpage')
- jsplayer_url_json = self._search_regex(
- ASSETS_RE, embed_webpage, 'JS player URL')
- player_url = json.loads(jsplayer_url_json)
- if player_url is None:
- player_url_json = self._search_regex(
- r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
- video_webpage, 'age gate player URL')
- player_url = json.loads(player_url_json)
-
- if 'sig' in url_data:
- url += '&signature=' + url_data['sig'][0]
- elif 's' in url_data:
- encrypted_sig = url_data['s'][0]
+ format_id = fmt.get('itag') or url_data['itag'][0]
+ if not format_id:
+ continue
+ format_id = compat_str(format_id)
- if self._downloader.params.get('verbose'):
+ if cipher:
+ if 's' in url_data or self._downloader.params.get('youtube_include_dash_manifest', True):
+ ASSETS_RE = r'"assets":.+?"js":\s*("[^"]+")'
+ jsplayer_url_json = self._search_regex(
+ ASSETS_RE,
+ embed_webpage if age_gate else video_webpage,
+ 'JS player URL (1)', default=None)
+ if not jsplayer_url_json and not age_gate:
+ # We need the embed website after all
+ if embed_webpage is None:
+ embed_url = proto + '://www.youtube.com/embed/%s' % video_id
+ embed_webpage = self._download_webpage(
+ embed_url, video_id, 'Downloading embed webpage')
+ jsplayer_url_json = self._search_regex(
+ ASSETS_RE, embed_webpage, 'JS player URL')
+
+ player_url = json.loads(jsplayer_url_json)
if player_url is None:
- player_version = 'unknown'
- player_desc = 'unknown'
- else:
- if player_url.endswith('swf'):
- player_version = self._search_regex(
- r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
- 'flash player', fatal=False)
- player_desc = 'flash player %s' % player_version
+ player_url_json = self._search_regex(
+ r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")',
+ video_webpage, 'age gate player URL')
+ player_url = json.loads(player_url_json)
+
+ if 'sig' in url_data:
+ url += '&signature=' + url_data['sig'][0]
+ elif 's' in url_data:
+ encrypted_sig = url_data['s'][0]
+
+ if self._downloader.params.get('verbose'):
+ if player_url is None:
+ player_version = 'unknown'
+ player_desc = 'unknown'
else:
- player_version = self._search_regex(
- [r'html5player-([^/]+?)(?:/html5player(?:-new)?)?\.js',
- r'(?:www|player(?:_ias)?)-([^/]+)(?:/[a-z]{2,3}_[A-Z]{2})?/base\.js'],
- player_url,
- 'html5 player', fatal=False)
- player_desc = 'html5 player %s' % player_version
-
- parts_sizes = self._signature_cache_id(encrypted_sig)
- self.to_screen('{%s} signature length %s, %s' %
- (format_id, parts_sizes, player_desc))
-
- signature = self._decrypt_signature(
- encrypted_sig, video_id, player_url, age_gate)
- sp = try_get(url_data, lambda x: x['sp'][0], compat_str) or 'signature'
- url += '&%s=%s' % (sp, signature)
+ if player_url.endswith('swf'):
+ player_version = self._search_regex(
+ r'-(.+?)(?:/watch_as3)?\.swf$', player_url,
+ 'flash player', fatal=False)
+ player_desc = 'flash player %s' % player_version
+ else:
+ player_version = self._search_regex(
+ [r'html5player-([^/]+?)(?:/html5player(?:-new)?)?\.js',
+ r'(?:www|player(?:_ias)?)-([^/]+)(?:/[a-z]{2,3}_[A-Z]{2})?/base\.js'],
+ player_url,
+ 'html5 player', fatal=False)
+ player_desc = 'html5 player %s' % player_version
+
+ parts_sizes = self._signature_cache_id(encrypted_sig)
+ self.to_screen('{%s} signature length %s, %s' %
+ (format_id, parts_sizes, player_desc))
+
+ signature = self._decrypt_signature(
+ encrypted_sig, video_id, player_url, age_gate)
+ sp = try_get(url_data, lambda x: x['sp'][0], compat_str) or 'signature'
+ url += '&%s=%s' % (sp, signature)
if 'ratebypass' not in url:
url += '&ratebypass=yes'
@@ -2044,24 +2063,33 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
mobj = re.search(r'^(?P<width>\d+)[xX](?P<height>\d+)$', url_data.get('size', [''])[0])
width, height = (int(mobj.group('width')), int(mobj.group('height'))) if mobj else (None, None)
+ if width is None:
+ width = int_or_none(fmt.get('width'))
+ if height is None:
+ height = int_or_none(fmt.get('height'))
+
filesize = int_or_none(url_data.get(
'clen', [None])[0]) or _extract_filesize(url)
- quality = url_data.get('quality', [None])[0]
+ quality = url_data.get('quality', [None])[0] or fmt.get('quality')
+ quality_label = url_data.get('quality_label', [None])[0] or fmt.get('qualityLabel')
+
+ tbr = float_or_none(url_data.get('bitrate', [None])[0], 1000) or float_or_none(fmt.get('bitrate'), 1000)
+ fps = int_or_none(url_data.get('fps', [None])[0]) or int_or_none(fmt.get('fps'))
more_fields = {
'filesize': filesize,
- 'tbr': float_or_none(url_data.get('bitrate', [None])[0], 1000),
+ 'tbr': tbr,
'width': width,
'height': height,
- 'fps': int_or_none(url_data.get('fps', [None])[0]),
- 'format_note': url_data.get('quality_label', [None])[0] or quality,
+ 'fps': fps,
+ 'format_note': quality_label or quality,
'quality': q(quality),
}
for key, value in more_fields.items():
if value:
dct[key] = value
- type_ = url_data.get('type', [None])[0]
+ type_ = url_data.get('type', [None])[0] or fmt.get('mimeType')
if type_:
type_split = type_.split(';')
kind_ext = type_split[0].split('/')