+ 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, mpd_url=mpd_url)
+
+ def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}, mpd_url=None):
+ """
+ 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()
+
+ # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
+ # common attributes and elements. We will only extract relevant
+ # for us.
+ def extract_common(source):
+ segment_timeline = source.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,
+ })
+ start_number = source.get('startNumber')
+ if start_number:
+ ms_info['start_number'] = int(start_number)
+ timescale = source.get('timescale')
+ if timescale:
+ ms_info['timescale'] = int(timescale)
+ segment_duration = source.get('duration')
+ if segment_duration:
+ ms_info['segment_duration'] = int(segment_duration)
+
+ def extract_Initialization(source):
+ initialization = source.find(_add_ns('Initialization'))
+ if initialization is not None:
+ ms_info['initialization_url'] = initialization.attrib['sourceURL']
+
+ segment_list = element.find(_add_ns('SegmentList'))
+ if segment_list is not None:
+ extract_common(segment_list)
+ extract_Initialization(segment_list)
+ 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]
+ else:
+ segment_template = element.find(_add_ns('SegmentTemplate'))
+ if segment_template is not None:
+ extract_common(segment_template)
+ 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:
+ extract_Initialization(segment_template)
+ return ms_info
+
+ def combine_url(base_url, target_url):
+ if re.match(r'^https?://', target_url):
+ return target_url
+ return '%s%s%s' % (base_url, '' if base_url.endswith('/') else '/', target_url)
+
+ 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,
+ 'manifest_url': mpd_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:
+
+ 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 and 's' not in representation_ms_info:
+ segment_duration = None
+ if 'total_number' not in representation_ms_info and 'segment_duration':
+ segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale'])
+ representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
+ representation_ms_info['fragments'] = [{
+ 'url': media_template % {
+ 'Number': segment_number,
+ 'Bandwidth': representation_attrib.get('bandwidth'),
+ },
+ 'duration': segment_duration,
+ } for segment_number in range(
+ representation_ms_info['start_number'],
+ representation_ms_info['total_number'] + representation_ms_info['start_number'])]
+ else:
+ # $Number*$ or $Time$ in media template with S list available
+ # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
+ # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
+ representation_ms_info['fragments'] = []
+ segment_time = 0
+ segment_d = None
+ segment_number = representation_ms_info['start_number']
+
+ def add_segment_url():
+ segment_url = media_template % {
+ 'Time': segment_time,
+ 'Bandwidth': representation_attrib.get('bandwidth'),
+ 'Number': segment_number,
+ }
+ representation_ms_info['fragments'].append({
+ 'url': segment_url,
+ 'duration': float_or_none(segment_d, representation_ms_info['timescale']),
+ })
+
+ for num, s in enumerate(representation_ms_info['s']):
+ segment_time = s.get('t') or segment_time
+ segment_d = s['d']
+ add_segment_url()
+ segment_number += 1
+ for r in range(s.get('r', 0)):
+ segment_time += segment_d
+ add_segment_url()
+ segment_number += 1
+ segment_time += segment_d
+ elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info:
+ # No media template
+ # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
+ # or any YouTube dashsegments video
+ fragments = []
+ s_num = 0
+ for segment_url in representation_ms_info['segment_urls']:
+ s = representation_ms_info['s'][s_num]
+ for r in range(s.get('r', 0) + 1):
+ fragments.append({
+ 'url': segment_url,
+ 'duration': float_or_none(s['d'], representation_ms_info['timescale']),
+ })
+ representation_ms_info['fragments'] = fragments
+ # NB: MPD manifest may contain direct URLs to unfragmented media.
+ # No fragments key is present in this case.
+ if 'fragments' in representation_ms_info:
+ f.update({
+ 'fragments': [],
+ 'protocol': 'http_dash_segments',
+ })
+ if 'initialization_url' in representation_ms_info:
+ initialization_url = representation_ms_info['initialization_url'].replace('$RepresentationID$', representation_id)
+ if not f.get('url'):
+ f['url'] = initialization_url
+ f['fragments'].append({'url': initialization_url})
+ f['fragments'].extend(representation_ms_info['fragments'])
+ for fragment in f['fragments']:
+ fragment['url'] = combine_url(base_url, fragment['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, video_id, m3u8_id=None, m3u8_entry_protocol='m3u8'):
+ 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 {}
+
+ def _media_formats(src, cur_media_type):
+ full_url = absolute_url(src)
+ if determine_ext(full_url) == 'm3u8':
+ is_plain_url = False
+ formats = self._extract_m3u8_formats(
+ full_url, video_id, ext='mp4',
+ entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id)
+ else:
+ is_plain_url = True
+ formats = [{
+ 'url': full_url,
+ 'vcodec': 'none' if cur_media_type == 'audio' else None,
+ }]
+ return is_plain_url, formats
+
+ 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:
+ _, formats = _media_formats(src, media_type)
+ media_info['formats'].extend(formats)
+ 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
+ is_plain_url, formats = _media_formats(src, media_type)
+ if is_plain_url:
+ f = parse_content_type(source_attributes.get('type'))
+ f.update(formats[0])
+ media_info['formats'].append(f)
+ else:
+ media_info['formats'].extend(formats)
+ 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 _extract_akamai_formats(self, manifest_url, video_id):
+ formats = []
+ f4m_url = re.sub(r'(https?://.+?)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m')
+ formats.extend(self._extract_f4m_formats(
+ update_url_query(f4m_url, {'hdcore': '3.7.0'}),
+ video_id, f4m_id='hds', fatal=False))
+ m3u8_url = re.sub(r'(https?://.+?)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8')
+ formats.extend(self._extract_m3u8_formats(
+ m3u8_url, video_id, 'mp4', 'm3u8_native',
+ m3u8_id='hls', fatal=False))
+ return formats
+
+ def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]):
+ url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url)
+ url_base = self._search_regex(r'(?:https?|rtmp|rtsp)(://[^?]+)', url, 'format url')
+ http_base_url = 'http' + url_base
+ formats = []
+ if 'm3u8' not in skip_protocols:
+ formats.extend(self._extract_m3u8_formats(
+ http_base_url + '/playlist.m3u8', video_id, 'mp4',
+ m3u8_entry_protocol, m3u8_id='hls', fatal=False))
+ if 'f4m' not in skip_protocols:
+ formats.extend(self._extract_f4m_formats(
+ http_base_url + '/manifest.f4m',
+ video_id, f4m_id='hds', fatal=False))
+ if re.search(r'(?:/smil:|\.smil)', url_base):
+ if 'dash' not in skip_protocols:
+ formats.extend(self._extract_mpd_formats(
+ http_base_url + '/manifest.mpd',
+ video_id, mpd_id='dash', fatal=False))
+ if 'smil' not in skip_protocols:
+ rtmp_formats = self._extract_smil_formats(
+ http_base_url + '/jwplayer.smil',
+ video_id, fatal=False)
+ for rtmp_format in rtmp_formats:
+ rtsp_format = rtmp_format.copy()
+ rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path'])
+ del rtsp_format['play_path']
+ del rtsp_format['ext']
+ rtsp_format.update({
+ 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'),
+ 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'),
+ 'protocol': 'rtsp',
+ })
+ formats.extend([rtmp_format, rtsp_format])
+ else:
+ for protocol in ('rtmp', 'rtsp'):
+ if protocol not in skip_protocols:
+ formats.append({
+ 'url': protocol + url_base,
+ 'format_id': protocol,
+ 'protocol': protocol,
+ })
+ return formats
projects / youtube-dl / blobdiff