+ def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
+ regexes = self._og_regexes('video') + self._og_regexes('video:url')
+ if secure:
+ regexes = self._og_regexes('video:secure_url') + regexes
+ return self._html_search_regex(regexes, html, name, **kargs)
+
+ def _og_search_url(self, html, **kargs):
+ return self._og_search_property('url', html, **kargs)
+
+ def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
+ if not isinstance(name, (list, tuple)):
+ name = [name]
+ if display_name is None:
+ display_name = name[0]
+ return self._html_search_regex(
+ [self._meta_regex(n) for n in name],
+ html, display_name, fatal=fatal, group='content', **kwargs)
+
+ def _dc_search_uploader(self, html):
+ return self._html_search_meta('dc.creator', html, 'uploader')
+
+ def _rta_search(self, html):
+ # See http://www.rtalabel.org/index.php?content=howtofaq#single
+ if re.search(r'(?ix)<meta\s+name="rating"\s+'
+ r' content="RTA-5042-1996-1400-1577-RTA"',
+ html):
+ return 18
+ return 0
+
+ def _media_rating_search(self, html):
+ # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
+ rating = self._html_search_meta('rating', html)
+
+ if not rating:
+ return None
+
+ RATING_TABLE = {
+ 'safe for kids': 0,
+ 'general': 8,
+ '14 years': 14,
+ 'mature': 17,
+ 'restricted': 19,
+ }
+ return RATING_TABLE.get(rating.lower())
+
+ def _family_friendly_search(self, html):
+ # See http://schema.org/VideoObject
+ family_friendly = self._html_search_meta('isFamilyFriendly', html)
+
+ if not family_friendly:
+ return None
+
+ RATING_TABLE = {
+ '1': 0,
+ 'true': 0,
+ '0': 18,
+ 'false': 18,
+ }
+ return RATING_TABLE.get(family_friendly.lower())
+
+ def _twitter_search_player(self, html):
+ return self._html_search_meta('twitter:player', html,
+ 'twitter card player')
+
+ def _search_json_ld(self, html, video_id, expected_type=None, **kwargs):
+ json_ld = self._search_regex(
+ r'(?s)<script[^>]+type=(["\'])application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>',
+ html, 'JSON-LD', group='json_ld', **kwargs)
+ default = kwargs.get('default', NO_DEFAULT)
+ if not json_ld:
+ return default if default is not NO_DEFAULT else {}
+ # JSON-LD may be malformed and thus `fatal` should be respected.
+ # At the same time `default` may be passed that assumes `fatal=False`
+ # for _search_regex. Let's simulate the same behavior here as well.
+ fatal = kwargs.get('fatal', True) if default == NO_DEFAULT else False
+ return self._json_ld(json_ld, video_id, fatal=fatal, expected_type=expected_type)
+
+ def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None):
+ if isinstance(json_ld, compat_str):
+ json_ld = self._parse_json(json_ld, video_id, fatal=fatal)
+ if not json_ld:
+ return {}
+ info = {}
+ if not isinstance(json_ld, (list, tuple, dict)):
+ return info
+ if isinstance(json_ld, dict):
+ json_ld = [json_ld]
+ for e in json_ld:
+ if e.get('@context') == 'http://schema.org':
+ item_type = e.get('@type')
+ if expected_type is not None and expected_type != item_type:
+ return info
+ if item_type == 'TVEpisode':
+ info.update({
+ 'episode': unescapeHTML(e.get('name')),
+ 'episode_number': int_or_none(e.get('episodeNumber')),
+ 'description': unescapeHTML(e.get('description')),
+ })
+ part_of_season = e.get('partOfSeason')
+ if isinstance(part_of_season, dict) and part_of_season.get('@type') == 'TVSeason':
+ info['season_number'] = int_or_none(part_of_season.get('seasonNumber'))
+ part_of_series = e.get('partOfSeries') or e.get('partOfTVSeries')
+ if isinstance(part_of_series, dict) and part_of_series.get('@type') == 'TVSeries':
+ info['series'] = unescapeHTML(part_of_series.get('name'))
+ elif item_type == 'Article':
+ info.update({
+ 'timestamp': parse_iso8601(e.get('datePublished')),
+ 'title': unescapeHTML(e.get('headline')),
+ 'description': unescapeHTML(e.get('articleBody')),
+ })
+ elif item_type == 'VideoObject':
+ info.update({
+ 'url': e.get('contentUrl'),
+ 'title': unescapeHTML(e.get('name')),
+ 'description': unescapeHTML(e.get('description')),
+ 'thumbnail': e.get('thumbnailUrl'),
+ 'duration': parse_duration(e.get('duration')),
+ 'timestamp': unified_timestamp(e.get('uploadDate')),
+ 'filesize': float_or_none(e.get('contentSize')),
+ 'tbr': int_or_none(e.get('bitrate')),
+ 'width': int_or_none(e.get('width')),
+ 'height': int_or_none(e.get('height')),
+ })
+ break
+ return dict((k, v) for k, v in info.items() if v is not None)
+
+ @staticmethod
+ def _hidden_inputs(html):
+ html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)
+ hidden_inputs = {}
+ for input in re.findall(r'(?i)<input([^>]+)>', html):
+ if not re.search(r'type=(["\'])(?:hidden|submit)\1', input):
+ continue
+ name = re.search(r'(?:name|id)=(["\'])(?P<value>.+?)\1', input)
+ if not name:
+ continue
+ value = re.search(r'value=(["\'])(?P<value>.*?)\1', input)
+ if not value:
+ continue
+ hidden_inputs[name.group('value')] = value.group('value')
+ return hidden_inputs
+
+ def _form_hidden_inputs(self, form_id, html):
+ form = self._search_regex(
+ r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
+ html, '%s form' % form_id, group='form')
+ return self._hidden_inputs(form)
+
+ def _sort_formats(self, formats, field_preference=None):
+ if not formats:
+ raise ExtractorError('No video formats found')
+
+ for f in formats:
+ # Automatically determine tbr when missing based on abr and vbr (improves
+ # formats sorting in some cases)
+ if 'tbr' not in f and f.get('abr') is not None and f.get('vbr') is not None:
+ f['tbr'] = f['abr'] + f['vbr']
+
+ def _formats_key(f):
+ # TODO remove the following workaround
+ from ..utils import determine_ext
+ if not f.get('ext') and 'url' in f:
+ f['ext'] = determine_ext(f['url'])
+
+ if isinstance(field_preference, (list, tuple)):
+ return tuple(
+ f.get(field)
+ if f.get(field) is not None
+ else ('' if field == 'format_id' else -1)
+ for field in field_preference)
+
+ preference = f.get('preference')
+ if preference is None:
+ preference = 0
+ if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
+ preference -= 0.5
+
+ protocol = f.get('protocol') or determine_protocol(f)
+ proto_preference = 0 if protocol in ['http', 'https'] else (-0.5 if protocol == 'rtsp' else -0.1)
+
+ if f.get('vcodec') == 'none': # audio only
+ preference -= 50
+ if self._downloader.params.get('prefer_free_formats'):
+ ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
+ else:
+ ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
+ ext_preference = 0
+ try:
+ audio_ext_preference = ORDER.index(f['ext'])
+ except ValueError:
+ audio_ext_preference = -1
+ else:
+ if f.get('acodec') == 'none': # video only
+ preference -= 40
+ if self._downloader.params.get('prefer_free_formats'):
+ ORDER = ['flv', 'mp4', 'webm']
+ else:
+ ORDER = ['webm', 'flv', 'mp4']
+ try:
+ ext_preference = ORDER.index(f['ext'])
+ except ValueError:
+ ext_preference = -1
+ audio_ext_preference = 0
+
+ return (
+ preference,
+ f.get('language_preference') if f.get('language_preference') is not None else -1,
+ f.get('quality') if f.get('quality') is not None else -1,
+ f.get('tbr') if f.get('tbr') is not None else -1,
+ f.get('filesize') if f.get('filesize') is not None else -1,
+ f.get('vbr') if f.get('vbr') is not None else -1,
+ f.get('height') if f.get('height') is not None else -1,
+ f.get('width') if f.get('width') is not None else -1,
+ proto_preference,
+ ext_preference,
+ f.get('abr') if f.get('abr') is not None else -1,
+ audio_ext_preference,
+ f.get('fps') if f.get('fps') is not None else -1,
+ f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
+ f.get('source_preference') if f.get('source_preference') is not None else -1,
+ f.get('format_id') if f.get('format_id') is not None else '',
+ )
+ formats.sort(key=_formats_key)
+
+ def _check_formats(self, formats, video_id):
+ if formats:
+ formats[:] = filter(
+ lambda f: self._is_valid_url(
+ f['url'], video_id,
+ item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
+ formats)
+
+ @staticmethod
+ def _remove_duplicate_formats(formats):
+ format_urls = set()
+ unique_formats = []
+ for f in formats:
+ if f['url'] not in format_urls:
+ format_urls.add(f['url'])
+ unique_formats.append(f)
+ formats[:] = unique_formats
+
+ def _is_valid_url(self, url, video_id, item='video'):
+ url = self._proto_relative_url(url, scheme='http:')
+ # For now assume non HTTP(S) URLs always valid
+ if not (url.startswith('http://') or url.startswith('https://')):
+ return True
+ try:
+ self._request_webpage(url, video_id, 'Checking %s URL' % item)
+ return True
+ except ExtractorError as e:
+ if isinstance(e.cause, compat_urllib_error.URLError):
+ self.to_screen(
+ '%s: %s URL is invalid, skipping' % (video_id, item))
+ return False
+ raise
+
+ def http_scheme(self):
+ """ Either "http:" or "https:", depending on the user's preferences """
+ return (
+ 'http:'
+ if self._downloader.params.get('prefer_insecure', False)
+ else 'https:')
+
+ def _proto_relative_url(self, url, scheme=None):
+ if url is None:
+ return url
+ if url.startswith('//'):
+ if scheme is None:
+ scheme = self.http_scheme()
+ return scheme + url
+ else:
+ return url
+
+ def _sleep(self, timeout, video_id, msg_template=None):
+ if msg_template is None:
+ msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
+ msg = msg_template % {'video_id': video_id, 'timeout': timeout}
+ self.to_screen(msg)
+ time.sleep(timeout)
+
+ def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None,
+ transform_source=lambda s: fix_xml_ampersands(s).strip(),
+ fatal=True, m3u8_id=None):
+ manifest = self._download_xml(
+ manifest_url, video_id, 'Downloading f4m manifest',
+ 'Unable to download f4m manifest',
+ # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
+ # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244)
+ transform_source=transform_source,
+ fatal=fatal)
+
+ if manifest is False:
+ return []
+
+ return self._parse_f4m_formats(
+ manifest, manifest_url, video_id, preference=preference, f4m_id=f4m_id,
+ transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
+
+ def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, f4m_id=None,
+ transform_source=lambda s: fix_xml_ampersands(s).strip(),
+ fatal=True, m3u8_id=None):
+ # currently youtube-dl cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
+ akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
+ if akamai_pv is not None and ';' in akamai_pv.text:
+ playerVerificationChallenge = akamai_pv.text.split(';')[0]
+ if playerVerificationChallenge.strip() != '':
+ return []
+
+ formats = []
+ manifest_version = '1.0'
+ media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
+ if not media_nodes:
+ manifest_version = '2.0'
+ media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
+ # Remove unsupported DRM protected media from final formats
+ # rendition (see https://github.com/rg3/youtube-dl/issues/8573).
+ media_nodes = remove_encrypted_media(media_nodes)
+ if not media_nodes:
+ return formats
+ base_url = xpath_text(
+ manifest, ['{http://ns.adobe.com/f4m/1.0}baseURL', '{http://ns.adobe.com/f4m/2.0}baseURL'],
+ 'base URL', default=None)
+ if base_url:
+ base_url = base_url.strip()
+
+ bootstrap_info = xpath_element(
+ manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
+ 'bootstrap info', default=None)
+
+ for i, media_el in enumerate(media_nodes):
+ tbr = int_or_none(media_el.attrib.get('bitrate'))
+ width = int_or_none(media_el.attrib.get('width'))
+ height = int_or_none(media_el.attrib.get('height'))
+ format_id = '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)]))
+ # If <bootstrapInfo> is present, the specified f4m is a
+ # stream-level manifest, and only set-level manifests may refer to
+ # external resources. See section 11.4 and section 4 of F4M spec
+ if bootstrap_info is None:
+ media_url = None
+ # @href is introduced in 2.0, see section 11.6 of F4M spec
+ if manifest_version == '2.0':
+ media_url = media_el.attrib.get('href')
+ if media_url is None:
+ media_url = media_el.attrib.get('url')
+ if not media_url:
+ continue
+ manifest_url = (
+ media_url if media_url.startswith('http://') or media_url.startswith('https://')
+ else ((base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
+ # If media_url is itself a f4m manifest do the recursive extraction
+ # since bitrates in parent manifest (this one) and media_url manifest
+ # may differ leading to inability to resolve the format by requested
+ # bitrate in f4m downloader
+ ext = determine_ext(manifest_url)
+ if ext == 'f4m':
+ f4m_formats = self._extract_f4m_formats(
+ manifest_url, video_id, preference=preference, f4m_id=f4m_id,
+ transform_source=transform_source, fatal=fatal)
+ # Sometimes stream-level manifest contains single media entry that
+ # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
+ # At the same time parent's media entry in set-level manifest may
+ # contain it. We will copy it from parent in such cases.
+ if len(f4m_formats) == 1:
+ f = f4m_formats[0]
+ f.update({
+ 'tbr': f.get('tbr') or tbr,
+ 'width': f.get('width') or width,
+ 'height': f.get('height') or height,
+ 'format_id': f.get('format_id') if not tbr else format_id,
+ })
+ formats.extend(f4m_formats)
+ continue
+ elif ext == 'm3u8':
+ formats.extend(self._extract_m3u8_formats(
+ manifest_url, video_id, 'mp4', preference=preference,
+ m3u8_id=m3u8_id, fatal=fatal))
+ continue
+ formats.append({
+ 'format_id': format_id,
+ 'url': manifest_url,
+ 'ext': 'flv' if bootstrap_info is not None else None,
+ 'tbr': tbr,
+ 'width': width,
+ 'height': height,
+ 'preference': preference,
+ })
+ return formats
+
+ def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, m3u8_id=None):
+ return {
+ 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
+ 'url': m3u8_url,
+ 'ext': ext,
+ 'protocol': 'm3u8',
+ 'preference': preference - 100 if preference else -100,
+ 'resolution': 'multiple',
+ 'format_note': 'Quality selection URL',
+ }
+
+ def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
+ entry_protocol='m3u8', preference=None,
+ m3u8_id=None, note=None, errnote=None,
+ fatal=True, live=False):
+
+ formats = [self._m3u8_meta_format(m3u8_url, ext, preference, m3u8_id)]
+
+ format_url = lambda u: (
+ u
+ if re.match(r'^https?://', u)
+ else compat_urlparse.urljoin(m3u8_url, u))
+
+ res = self._download_webpage_handle(
+ m3u8_url, video_id,
+ note=note or 'Downloading m3u8 information',
+ errnote=errnote or 'Failed to download m3u8 information',
+ fatal=fatal)
+ if res is False:
+ return []
+ m3u8_doc, urlh = res
+ m3u8_url = urlh.geturl()
+
+ # We should try extracting formats only from master playlists [1], i.e.
+ # playlists that describe available qualities. On the other hand media
+ # playlists [2] should be returned as is since they contain just the media
+ # without qualities renditions.
+ # Fortunately, master playlist can be easily distinguished from media
+ # playlist based on particular tags availability. As of [1, 2] master
+ # playlist tags MUST NOT appear in a media playist and vice versa.
+ # As of [3] #EXT-X-TARGETDURATION tag is REQUIRED for every media playlist
+ # and MUST NOT appear in master playlist thus we can clearly detect media
+ # playlist with this criterion.
+ # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.4
+ # 2. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3
+ # 3. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.1
+ if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is
+ return [{
+ 'url': m3u8_url,
+ 'format_id': m3u8_id,
+ 'ext': ext,
+ 'protocol': entry_protocol,
+ 'preference': preference,
+ }]
+ last_info = None
+ last_media = None
+ for line in m3u8_doc.splitlines():
+ if line.startswith('#EXT-X-STREAM-INF:'):
+ last_info = parse_m3u8_attributes(line)
+ elif line.startswith('#EXT-X-MEDIA:'):
+ last_media = parse_m3u8_attributes(line)
+ elif line.startswith('#') or not line.strip():
+ continue
+ else:
+ if last_info is None:
+ formats.append({'url': format_url(line)})
+ continue
+ tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
+ format_id = []
+ if m3u8_id:
+ format_id.append(m3u8_id)
+ last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') not in ('SUBTITLES', 'CLOSED-CAPTIONS') else None
+ # Despite specification does not mention NAME attribute for
+ # EXT-X-STREAM-INF it still sometimes may be present
+ stream_name = last_info.get('NAME') or last_media_name
+ # Bandwidth of live streams may differ over time thus making
+ # format_id unpredictable. So it's better to keep provided
+ # format_id intact.
+ if not live:
+ format_id.append(stream_name if stream_name else '%d' % (tbr if tbr else len(formats)))
+ f = {
+ 'format_id': '-'.join(format_id),
+ 'url': format_url(line.strip()),
+ 'tbr': tbr,
+ 'ext': ext,
+ 'fps': float_or_none(last_info.get('FRAME-RATE')),
+ 'protocol': entry_protocol,
+ 'preference': preference,
+ }
+ resolution = last_info.get('RESOLUTION')
+ if resolution:
+ width_str, height_str = resolution.split('x')
+ f['width'] = int(width_str)
+ f['height'] = int(height_str)
+ # Unified Streaming Platform
+ mobj = re.search(
+ r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url'])
+ if mobj:
+ abr, vbr = mobj.groups()
+ abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000)
+ f.update({
+ 'vbr': vbr,
+ 'abr': abr,
+ })
+ f.update(parse_codecs(last_info.get('CODECS')))
+ if last_media is not None:
+ f['m3u8_media'] = last_media
+ last_media = None
+ formats.append(f)
+ last_info = {}
+ return formats
+
+ @staticmethod
+ def _xpath_ns(path, namespace=None):
+ if not namespace:
+ return path
+ out = []
+ for c in path.split('/'):
+ if not c or c == '.':
+ out.append(c)
+ else:
+ out.append('{%s}%s' % (namespace, c))
+ return '/'.join(out)
+
+ def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
+ smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
+
+ if smil is False:
+ assert not fatal
+ return []
+
+ namespace = self._parse_smil_namespace(smil)
+
+ return self._parse_smil_formats(
+ smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
+
+ def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None):
+ smil = self._download_smil(smil_url, video_id, fatal=fatal)
+ if smil is False:
+ return {}
+ return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params)
+
+ def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None):
+ return self._download_xml(
+ smil_url, video_id, 'Downloading SMIL file',
+ 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source)
+
+ def _parse_smil(self, smil, smil_url, video_id, f4m_params=None):
+ namespace = self._parse_smil_namespace(smil)
+
+ formats = self._parse_smil_formats(
+ smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
+ subtitles = self._parse_smil_subtitles(smil, namespace=namespace)
+
+ video_id = os.path.splitext(url_basename(smil_url))[0]
+ title = None
+ description = None
+ upload_date = None
+ for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
+ name = meta.attrib.get('name')
+ content = meta.attrib.get('content')
+ if not name or not content:
+ continue
+ if not title and name == 'title':
+ title = content
+ elif not description and name in ('description', 'abstract'):
+ description = content
+ elif not upload_date and name == 'date':
+ upload_date = unified_strdate(content)
+
+ thumbnails = [{
+ 'id': image.get('type'),
+ 'url': image.get('src'),
+ 'width': int_or_none(image.get('width')),
+ 'height': int_or_none(image.get('height')),
+ } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')]
+
+ return {
+ 'id': video_id,
+ 'title': title or video_id,
+ 'description': description,
+ 'upload_date': upload_date,
+ 'thumbnails': thumbnails,
+ 'formats': formats,
+ 'subtitles': subtitles,
+ }
+
+ def _parse_smil_namespace(self, smil):
+ return self._search_regex(
+ r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None)
+
+ def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
+ base = smil_url
+ for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
+ b = meta.get('base') or meta.get('httpBase')
+ if b:
+ base = b
+ break
+
+ formats = []
+ rtmp_count = 0
+ http_count = 0
+ m3u8_count = 0
+
+ srcs = []
+ media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace))
+ for medium in media:
+ src = medium.get('src')
+ if not src or src in srcs:
+ continue
+ srcs.append(src)
+
+ bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000)
+ filesize = int_or_none(medium.get('size') or medium.get('fileSize'))
+ width = int_or_none(medium.get('width'))
+ height = int_or_none(medium.get('height'))
+ proto = medium.get('proto')
+ ext = medium.get('ext')
+ src_ext = determine_ext(src)
+ streamer = medium.get('streamer') or base
+
+ if proto == 'rtmp' or streamer.startswith('rtmp'):
+ rtmp_count += 1
+ formats.append({
+ 'url': streamer,
+ 'play_path': src,
+ 'ext': 'flv',
+ 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
+ 'tbr': bitrate,
+ 'filesize': filesize,
+ 'width': width,
+ 'height': height,
+ })
+ if transform_rtmp_url:
+ streamer, src = transform_rtmp_url(streamer, src)
+ formats[-1].update({
+ 'url': streamer,
+ 'play_path': src,
+ })
+ continue
+
+ src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src)
+ src_url = src_url.strip()
+
+ if proto == 'm3u8' or src_ext == 'm3u8':
+ m3u8_formats = self._extract_m3u8_formats(
+ src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False)
+ if len(m3u8_formats) == 1:
+ m3u8_count += 1
+ m3u8_formats[0].update({
+ 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate),
+ 'tbr': bitrate,
+ 'width': width,
+ 'height': height,
+ })
+ formats.extend(m3u8_formats)
+ continue
+
+ if src_ext == 'f4m':
+ f4m_url = src_url
+ if not f4m_params:
+ f4m_params = {
+ 'hdcore': '3.2.0',
+ 'plugin': 'flowplayer-3.2.0.1',
+ }
+ f4m_url += '&' if '?' in f4m_url else '?'
+ f4m_url += compat_urllib_parse_urlencode(f4m_params)
+ formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
+ continue
+
+ if src_url.startswith('http') and self._is_valid_url(src, video_id):
+ http_count += 1
+ formats.append({
+ 'url': src_url,
+ 'ext': ext or src_ext or 'flv',
+ 'format_id': 'http-%d' % (bitrate or http_count),
+ 'tbr': bitrate,
+ 'filesize': filesize,
+ 'width': width,
+ 'height': height,
+ })
+ continue
+
+ return formats
+
+ def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
+ urls = []
+ subtitles = {}
+ for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
+ src = textstream.get('src')
+ if not src or src in urls:
+ continue
+ urls.append(src)
+ ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src)
+ lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang
+ subtitles.setdefault(lang, []).append({
+ 'url': src,
+ 'ext': ext,
+ })
+ return subtitles
+
+ def _extract_xspf_playlist(self, playlist_url, playlist_id, fatal=True):
+ xspf = self._download_xml(
+ playlist_url, playlist_id, 'Downloading xpsf playlist',
+ 'Unable to download xspf manifest', fatal=fatal)
+ if xspf is False:
+ return []
+ return self._parse_xspf(xspf, playlist_id)
+
+ def _parse_xspf(self, playlist, playlist_id):
+ NS_MAP = {
+ 'xspf': 'http://xspf.org/ns/0/',
+ 's1': 'http://static.streamone.nl/player/ns/0',
+ }
+
+ entries = []
+ for track in playlist.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)):
+ title = xpath_text(
+ track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id)
+ description = xpath_text(
+ track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description')
+ thumbnail = xpath_text(
+ track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail')
+ duration = float_or_none(
+ xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000)
+
+ formats = [{
+ 'url': location.text,
+ 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)),
+ 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
+ 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
+ } for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP))]
+ self._sort_formats(formats)
+
+ entries.append({
+ 'id': playlist_id,
+ 'title': title,
+ 'description': description,
+ 'thumbnail': thumbnail,
+ 'duration': duration,
+ 'formats': formats,
+ })
+ return entries
+
+ def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, formats_dict={}):
+ res = self._download_webpage_handle(
+ mpd_url, video_id,
+ note=note or 'Downloading MPD manifest',
+ errnote=errnote or 'Failed to download MPD manifest',
+ fatal=fatal)
+ if res is False:
+ return []
+ mpd, urlh = res
+ mpd_base_url = re.match(r'https?://.+/', urlh.geturl()).group()
+
+ return self._parse_mpd_formats(
+ compat_etree_fromstring(mpd.encode('utf-8')), mpd_id, mpd_base_url, formats_dict=formats_dict)
+
+ def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}):
+ """
+ Parse formats from MPD manifest.
+ References:
+ 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
+ http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
+ 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
+ """
+ if mpd_doc.get('type') == 'dynamic':
+ return []
+
+ namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
+
+ def _add_ns(path):
+ return self._xpath_ns(path, namespace)
+
+ def is_drm_protected(element):
+ return element.find(_add_ns('ContentProtection')) is not None
+
+ def extract_multisegment_info(element, ms_parent_info):
+ ms_info = ms_parent_info.copy()
+ segment_list = element.find(_add_ns('SegmentList'))
+ if segment_list is not None:
+ segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
+ if segment_urls_e:
+ ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
+ initialization = segment_list.find(_add_ns('Initialization'))
+ if initialization is not None:
+ ms_info['initialization_url'] = initialization.attrib['sourceURL']
+ else:
+ segment_template = element.find(_add_ns('SegmentTemplate'))
+ if segment_template is not None:
+ start_number = segment_template.get('startNumber')
+ if start_number:
+ ms_info['start_number'] = int(start_number)
+ segment_timeline = segment_template.find(_add_ns('SegmentTimeline'))
+ if segment_timeline is not None:
+ s_e = segment_timeline.findall(_add_ns('S'))
+ if s_e:
+ ms_info['total_number'] = 0
+ ms_info['s'] = []
+ for s in s_e:
+ r = int(s.get('r', 0))
+ ms_info['total_number'] += 1 + r
+ ms_info['s'].append({
+ 't': int(s.get('t', 0)),
+ # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
+ 'd': int(s.attrib['d']),
+ 'r': r,
+ })
+ else:
+ timescale = segment_template.get('timescale')
+ if timescale:
+ ms_info['timescale'] = int(timescale)
+ segment_duration = segment_template.get('duration')
+ if segment_duration:
+ ms_info['segment_duration'] = int(segment_duration)
+ media_template = segment_template.get('media')
+ if media_template:
+ ms_info['media_template'] = media_template
+ initialization = segment_template.get('initialization')
+ if initialization:
+ ms_info['initialization_url'] = initialization
+ else:
+ initialization = segment_template.find(_add_ns('Initialization'))
+ if initialization is not None:
+ ms_info['initialization_url'] = initialization.attrib['sourceURL']
+ return ms_info
+
+ mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
+ formats = []
+ for period in mpd_doc.findall(_add_ns('Period')):
+ period_duration = parse_duration(period.get('duration')) or mpd_duration
+ period_ms_info = extract_multisegment_info(period, {
+ 'start_number': 1,
+ 'timescale': 1,
+ })
+ for adaptation_set in period.findall(_add_ns('AdaptationSet')):
+ if is_drm_protected(adaptation_set):
+ continue
+ adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
+ for representation in adaptation_set.findall(_add_ns('Representation')):
+ if is_drm_protected(representation):
+ continue
+ representation_attrib = adaptation_set.attrib.copy()
+ representation_attrib.update(representation.attrib)
+ # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
+ mime_type = representation_attrib['mimeType']
+ content_type = mime_type.split('/')[0]
+ if content_type == 'text':
+ # TODO implement WebVTT downloading
+ pass
+ elif content_type == 'video' or content_type == 'audio':
+ base_url = ''
+ for element in (representation, adaptation_set, period, mpd_doc):
+ base_url_e = element.find(_add_ns('BaseURL'))
+ if base_url_e is not None:
+ base_url = base_url_e.text + base_url
+ if re.match(r'^https?://', base_url):
+ break
+ if mpd_base_url and not re.match(r'^https?://', base_url):
+ if not mpd_base_url.endswith('/') and not base_url.startswith('/'):
+ mpd_base_url += '/'
+ base_url = mpd_base_url + base_url
+ representation_id = representation_attrib.get('id')
+ lang = representation_attrib.get('lang')
+ url_el = representation.find(_add_ns('BaseURL'))
+ filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
+ f = {
+ 'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id,
+ 'url': base_url,
+ 'ext': mimetype2ext(mime_type),
+ 'width': int_or_none(representation_attrib.get('width')),
+ 'height': int_or_none(representation_attrib.get('height')),
+ 'tbr': int_or_none(representation_attrib.get('bandwidth'), 1000),
+ 'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
+ 'fps': int_or_none(representation_attrib.get('frameRate')),
+ 'vcodec': 'none' if content_type == 'audio' else representation_attrib.get('codecs'),
+ 'acodec': 'none' if content_type == 'video' else representation_attrib.get('codecs'),
+ 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
+ 'format_note': 'DASH %s' % content_type,
+ 'filesize': filesize,
+ }
+ representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
+ if 'segment_urls' not in representation_ms_info and 'media_template' in representation_ms_info:
+ if 'total_number' not in representation_ms_info and 'segment_duration':
+ segment_duration = float(representation_ms_info['segment_duration']) / float(representation_ms_info['timescale'])
+ representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
+ media_template = representation_ms_info['media_template']
+ media_template = media_template.replace('$RepresentationID$', representation_id)
+ media_template = re.sub(r'\$(Number|Bandwidth|Time)\$', r'%(\1)d', media_template)
+ media_template = re.sub(r'\$(Number|Bandwidth|Time)%([^$]+)\$', r'%(\1)\2', media_template)
+ media_template.replace('$$', '$')
+
+ # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
+ # can't be used at the same time
+ if '%(Number' in media_template:
+ representation_ms_info['segment_urls'] = [
+ media_template % {
+ 'Number': segment_number,
+ 'Bandwidth': representation_attrib.get('bandwidth'),
+ }
+ for segment_number in range(
+ representation_ms_info['start_number'],
+ representation_ms_info['total_number'] + representation_ms_info['start_number'])]
+ else:
+ representation_ms_info['segment_urls'] = []
+ segment_time = 0
+
+ def add_segment_url():
+ representation_ms_info['segment_urls'].append(
+ media_template % {
+ 'Time': segment_time,
+ 'Bandwidth': representation_attrib.get('bandwidth'),
+ }
+ )
+
+ for num, s in enumerate(representation_ms_info['s']):
+ segment_time = s.get('t') or segment_time
+ add_segment_url()
+ for r in range(s.get('r', 0)):
+ segment_time += s['d']
+ add_segment_url()
+ segment_time += s['d']
+ if 'segment_urls' in representation_ms_info:
+ f.update({
+ 'segment_urls': representation_ms_info['segment_urls'],
+ 'protocol': 'http_dash_segments',
+ })
+ if 'initialization_url' in representation_ms_info:
+ initialization_url = representation_ms_info['initialization_url'].replace('$RepresentationID$', representation_id)
+ f.update({
+ 'initialization_url': initialization_url,
+ })
+ if not f.get('url'):
+ f['url'] = initialization_url
+ try:
+ existing_format = next(
+ fo for fo in formats
+ if fo['format_id'] == representation_id)
+ except StopIteration:
+ full_info = formats_dict.get(representation_id, {}).copy()
+ full_info.update(f)
+ formats.append(full_info)
+ else:
+ existing_format.update(f)
+ else:
+ self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
+ return formats
+
+ def _parse_html5_media_entries(self, base_url, webpage):
+ def absolute_url(video_url):
+ return compat_urlparse.urljoin(base_url, video_url)
+
+ def parse_content_type(content_type):
+ if not content_type:
+ return {}
+ ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type)
+ if ctr:
+ mimetype, codecs = ctr.groups()
+ f = parse_codecs(codecs)
+ f['ext'] = mimetype2ext(mimetype)
+ return f
+ return {}
+
+ entries = []
+ for media_tag, media_type, media_content in re.findall(r'(?s)(<(?P<tag>video|audio)[^>]*>)(.*?)</(?P=tag)>', webpage):
+ media_info = {
+ 'formats': [],
+ 'subtitles': {},
+ }
+ media_attributes = extract_attributes(media_tag)
+ src = media_attributes.get('src')
+ if src:
+ media_info['formats'].append({
+ 'url': absolute_url(src),
+ 'vcodec': 'none' if media_type == 'audio' else None,
+ })
+ media_info['thumbnail'] = media_attributes.get('poster')
+ if media_content:
+ for source_tag in re.findall(r'<source[^>]+>', media_content):
+ source_attributes = extract_attributes(source_tag)
+ src = source_attributes.get('src')
+ if not src:
+ continue
+ f = parse_content_type(source_attributes.get('type'))
+ f.update({
+ 'url': absolute_url(src),
+ 'vcodec': 'none' if media_type == 'audio' else None,
+ })
+ media_info['formats'].append(f)
+ for track_tag in re.findall(r'<track[^>]+>', media_content):
+ track_attributes = extract_attributes(track_tag)
+ kind = track_attributes.get('kind')
+ if not kind or kind == 'subtitles':
+ src = track_attributes.get('src')
+ if not src:
+ continue
+ lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label')
+ media_info['subtitles'].setdefault(lang, []).append({
+ 'url': absolute_url(src),
+ })
+ if media_info['formats']:
+ entries.append(media_info)
+ return entries
+
+ def _live_title(self, name):
+ """ Generate the title for a live video """
+ now = datetime.datetime.now()
+ now_str = now.strftime('%Y-%m-%d %H:%M')
+ return name + ' ' + now_str
+
+ def _int(self, v, name, fatal=False, **kwargs):
+ res = int_or_none(v, **kwargs)
+ if 'get_attr' in kwargs:
+ print(getattr(v, kwargs['get_attr']))
+ if res is None:
+ msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
+ if fatal:
+ raise ExtractorError(msg)
+ else:
+ self._downloader.report_warning(msg)
+ return res
+
+ def _float(self, v, name, fatal=False, **kwargs):
+ res = float_or_none(v, **kwargs)
+ if res is None:
+ msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
+ if fatal:
+ raise ExtractorError(msg)
+ else:
+ self._downloader.report_warning(msg)
+ return res
+
+ def _set_cookie(self, domain, name, value, expire_time=None):
+ cookie = compat_cookiejar.Cookie(
+ 0, name, value, None, None, domain, None,
+ None, '/', True, False, expire_time, '', None, None, None)
+ self._downloader.cookiejar.set_cookie(cookie)
+
+ def _get_cookies(self, url):
+ """ Return a compat_cookies.SimpleCookie with the cookies for the url """
+ req = sanitized_Request(url)
+ self._downloader.cookiejar.add_cookie_header(req)
+ return compat_cookies.SimpleCookie(req.get_header('Cookie'))
+
+ def get_testcases(self, include_onlymatching=False):
+ t = getattr(self, '_TEST', None)
+ if t:
+ assert not hasattr(self, '_TESTS'), \
+ '%s has _TEST and _TESTS' % type(self).__name__
+ tests = [t]
+ else:
+ tests = getattr(self, '_TESTS', [])
+ for t in tests:
+ if not include_onlymatching and t.get('only_matching', False):
+ continue
+ t['name'] = type(self).__name__[:-len('IE')]
+ yield t
+
+ def is_suitable(self, age_limit):
+ """ Test whether the extractor is generally suitable for the given
+ age limit (i.e. pornographic sites are not, all others usually are) """
+
+ any_restricted = False
+ for tc in self.get_testcases(include_onlymatching=False):
+ if tc.get('playlist', []):
+ tc = tc['playlist'][0]
+ is_restricted = age_restricted(
+ tc.get('info_dict', {}).get('age_limit'), age_limit)
+ if not is_restricted:
+ return True
+ any_restricted = any_restricted or is_restricted
+ return not any_restricted
+
+ def extract_subtitles(self, *args, **kwargs):
+ if (self._downloader.params.get('writesubtitles', False) or
+ self._downloader.params.get('listsubtitles')):
+ return self._get_subtitles(*args, **kwargs)
+ return {}
+
+ def _get_subtitles(self, *args, **kwargs):
+ raise NotImplementedError('This method must be implemented by subclasses')
+
+ @staticmethod
+ def _merge_subtitle_items(subtitle_list1, subtitle_list2):
+ """ Merge subtitle items for one language. Items with duplicated URLs
+ will be dropped. """
+ list1_urls = set([item['url'] for item in subtitle_list1])
+ ret = list(subtitle_list1)
+ ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls])
+ return ret
+
+ @classmethod
+ def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2):
+ """ Merge two subtitle dictionaries, language by language. """
+ ret = dict(subtitle_dict1)
+ for lang in subtitle_dict2:
+ ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang])
+ return ret
+
+ def extract_automatic_captions(self, *args, **kwargs):
+ if (self._downloader.params.get('writeautomaticsub', False) or
+ self._downloader.params.get('listsubtitles')):
+ return self._get_automatic_captions(*args, **kwargs)
+ return {}
+
+ def _get_automatic_captions(self, *args, **kwargs):
+ raise NotImplementedError('This method must be implemented by subclasses')
+
+ def mark_watched(self, *args, **kwargs):
+ if (self._downloader.params.get('mark_watched', False) and
+ (self._get_login_info()[0] is not None or
+ self._downloader.params.get('cookiefile') is not None)):
+ self._mark_watched(*args, **kwargs)
+
+ def _mark_watched(self, *args, **kwargs):
+ raise NotImplementedError('This method must be implemented by subclasses')
+
+ def geo_verification_headers(self):
+ headers = {}
+ geo_verification_proxy = self._downloader.params.get('geo_verification_proxy')
+ if geo_verification_proxy:
+ headers['Ytdl-request-proxy'] = geo_verification_proxy
+ return headers
+
Rapsys Git / youtubedl / blobdiff