diff options
Diffstat (limited to 'youtube_dl/extractor/common.py')
-rw-r--r-- | youtube_dl/extractor/common.py | 122 |
1 files changed, 85 insertions, 37 deletions
diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index 2cb55d6af..74d30ec50 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -27,6 +27,7 @@ from ..compat import ( compat_urllib_parse_urlencode, compat_urllib_request, compat_urlparse, + compat_xml_parse_error, ) from ..downloader.f4m import remove_encrypted_media from ..utils import ( @@ -245,6 +246,10 @@ class InfoExtractor(object): specified in the URL. end_time: Time in seconds where the reproduction should end, as specified in the URL. + chapters: A list of dictionaries, with the following entries: + * "start_time" - The start time of the chapter in seconds + * "end_time" - The end time of the chapter in seconds + * "title" (optional, string) The following fields should only be used when the video belongs to some logical chapter or section: @@ -372,7 +377,7 @@ class InfoExtractor(object): cls._VALID_URL_RE = re.compile(cls._VALID_URL) m = cls._VALID_URL_RE.match(url) assert m - return m.group('id') + return compat_str(m.group('id')) @classmethod def working(cls): @@ -416,7 +421,7 @@ class InfoExtractor(object): if country_code: self._x_forwarded_for_ip = GeoUtils.random_ipv4(country_code) if self._downloader.params.get('verbose', False): - self._downloader.to_stdout( + self._downloader.to_screen( '[debug] Using fake IP %s (%s) as X-Forwarded-For.' % (self._x_forwarded_for_ip, country_code.upper())) @@ -642,15 +647,29 @@ class InfoExtractor(object): def _download_xml(self, url_or_request, video_id, note='Downloading XML', errnote='Unable to download XML', - transform_source=None, fatal=True, encoding=None, data=None, headers={}, query={}): + transform_source=None, fatal=True, encoding=None, + data=None, headers={}, query={}): """Return the xml as an xml.etree.ElementTree.Element""" xml_string = self._download_webpage( - url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding, data=data, headers=headers, query=query) + url_or_request, video_id, note, errnote, fatal=fatal, + encoding=encoding, data=data, headers=headers, query=query) if xml_string is False: return xml_string + return self._parse_xml( + xml_string, video_id, transform_source=transform_source, + fatal=fatal) + + def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True): if transform_source: xml_string = transform_source(xml_string) - return compat_etree_fromstring(xml_string.encode('utf-8')) + try: + return compat_etree_fromstring(xml_string.encode('utf-8')) + except compat_xml_parse_error as ve: + errmsg = '%s: Failed to parse XML ' % video_id + if fatal: + raise ExtractorError(errmsg, cause=ve) + else: + self.report_warning(errmsg + str(ve)) def _download_json(self, url_or_request, video_id, note='Downloading JSON metadata', @@ -726,12 +745,12 @@ class InfoExtractor(object): video_info['title'] = video_title return video_info - def playlist_from_matches(self, matches, video_id, video_title, getter=None, ie=None): - urlrs = orderedSet( + def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None): + urls = orderedSet( self.url_result(self._proto_relative_url(getter(m) if getter else m), ie) for m in matches) return self.playlist_result( - urlrs, playlist_id=video_id, playlist_title=video_title) + urls, playlist_id=playlist_id, playlist_title=playlist_title) @staticmethod def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None): @@ -936,7 +955,8 @@ class InfoExtractor(object): def _family_friendly_search(self, html): # See http://schema.org/VideoObject - family_friendly = self._html_search_meta('isFamilyFriendly', html) + family_friendly = self._html_search_meta( + 'isFamilyFriendly', html, default=None) if not family_friendly: return None @@ -990,6 +1010,7 @@ class InfoExtractor(object): 'tbr': int_or_none(e.get('bitrate')), 'width': int_or_none(e.get('width')), 'height': int_or_none(e.get('height')), + 'view_count': int_or_none(e.get('interactionCount')), }) for e in json_ld: @@ -997,17 +1018,17 @@ class InfoExtractor(object): item_type = e.get('@type') if expected_type is not None and expected_type != item_type: return info - if item_type == 'TVEpisode': + if item_type in ('TVEpisode', 'Episode'): 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': + if isinstance(part_of_season, dict) and part_of_season.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'): 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': + if isinstance(part_of_series, dict) and part_of_series.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'): info['series'] = unescapeHTML(part_of_series.get('name')) elif item_type == 'Article': info.update({ @@ -1017,10 +1038,10 @@ class InfoExtractor(object): }) elif item_type == 'VideoObject': extract_video_object(e) - elif item_type == 'WebPage': - video = e.get('video') - if isinstance(video, dict) and video.get('@type') == 'VideoObject': - extract_video_object(video) + continue + video = e.get('video') + if isinstance(video, dict) and video.get('@type') == 'VideoObject': + extract_video_object(video) break return dict((k, v) for k, v in info.items() if v is not None) @@ -1780,7 +1801,7 @@ class InfoExtractor(object): ms_info['timescale'] = int(timescale) segment_duration = source.get('duration') if segment_duration: - ms_info['segment_duration'] = int(segment_duration) + ms_info['segment_duration'] = float(segment_duration) def extract_Initialization(source): initialization = source.find(_add_ns('Initialization')) @@ -1887,9 +1908,13 @@ class InfoExtractor(object): 'Bandwidth': bandwidth, } + def location_key(location): + return 'url' if re.match(r'^https?://', location) else 'path' + if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info: media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time')) + media_location_key = location_key(media_template) # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$ # can't be used at the same time @@ -1899,7 +1924,7 @@ class InfoExtractor(object): 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 % { + media_location_key: media_template % { 'Number': segment_number, 'Bandwidth': bandwidth, }, @@ -1923,7 +1948,7 @@ class InfoExtractor(object): 'Number': segment_number, } representation_ms_info['fragments'].append({ - 'url': segment_url, + media_location_key: segment_url, 'duration': float_or_none(segment_d, representation_ms_info['timescale']), }) @@ -1947,8 +1972,9 @@ class InfoExtractor(object): for s in representation_ms_info['s']: duration = float_or_none(s['d'], timescale) for r in range(s.get('r', 0) + 1): + segment_uri = representation_ms_info['segment_urls'][segment_index] fragments.append({ - 'url': representation_ms_info['segment_urls'][segment_index], + location_key(segment_uri): segment_uri, 'duration': duration, }) segment_index += 1 @@ -1957,6 +1983,7 @@ class InfoExtractor(object): # No fragments key is present in this case. if 'fragments' in representation_ms_info: f.update({ + 'fragment_base_url': base_url, 'fragments': [], 'protocol': 'http_dash_segments', }) @@ -1964,10 +1991,8 @@ class InfoExtractor(object): initialization_url = representation_ms_info['initialization_url'] if not f.get('url'): f['url'] = initialization_url - f['fragments'].append({'url': initialization_url}) + f['fragments'].append({location_key(initialization_url): initialization_url}) f['fragments'].extend(representation_ms_info['fragments']) - for fragment in f['fragments']: - fragment['url'] = urljoin(base_url, fragment['url']) try: existing_format = next( fo for fo in formats @@ -1996,6 +2021,12 @@ class InfoExtractor(object): compat_etree_fromstring(ism.encode('utf-8')), urlh.geturl(), ism_id) def _parse_ism_formats(self, ism_doc, ism_url, ism_id=None): + """ + Parse formats from ISM manifest. + References: + 1. [MS-SSTR]: Smooth Streaming Protocol, + https://msdn.microsoft.com/en-us/library/ff469518.aspx + """ if ism_doc.get('IsLive') == 'TRUE' or ism_doc.find('Protection') is not None: return [] @@ -2017,8 +2048,11 @@ class InfoExtractor(object): self.report_warning('%s is not a supported codec' % fourcc) continue tbr = int(track.attrib['Bitrate']) // 1000 - width = int_or_none(track.get('MaxWidth')) - height = int_or_none(track.get('MaxHeight')) + # [1] does not mention Width and Height attributes. However, + # they're often present while MaxWidth and MaxHeight are + # missing, so should be used as fallbacks + width = int_or_none(track.get('MaxWidth') or track.get('Width')) + height = int_or_none(track.get('MaxHeight') or track.get('Height')) sampling_rate = int_or_none(track.get('SamplingRate')) track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern) @@ -2096,19 +2130,19 @@ class InfoExtractor(object): return f return {} - def _media_formats(src, cur_media_type): + def _media_formats(src, cur_media_type, type_info={}): full_url = absolute_url(src) - ext = determine_ext(full_url) + ext = type_info.get('ext') or determine_ext(full_url) if ext == '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, - preference=preference) + preference=preference, fatal=False) elif ext == 'mpd': is_plain_url = False formats = self._extract_mpd_formats( - full_url, video_id, mpd_id=mpd_id) + full_url, video_id, mpd_id=mpd_id, fatal=False) else: is_plain_url = True formats = [{ @@ -2118,15 +2152,18 @@ class InfoExtractor(object): return is_plain_url, formats entries = [] + # amp-video and amp-audio are very similar to their HTML5 counterparts + # so we wll include them right here (see + # https://www.ampproject.org/docs/reference/components/amp-video) media_tags = [(media_tag, media_type, '') for media_tag, media_type - in re.findall(r'(?s)(<(video|audio)[^>]*/>)', webpage)] + in re.findall(r'(?s)(<(?:amp-)?(video|audio)[^>]*/>)', webpage)] media_tags.extend(re.findall( # We only allow video|audio followed by a whitespace or '>'. # Allowing more characters may end up in significant slow down (see # https://github.com/rg3/youtube-dl/issues/11979, example URL: # http://www.porntrex.com/maps/videositemap.xml). - r'(?s)(<(?P<tag>video|audio)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>', webpage)) + r'(?s)(<(?P<tag>(?:amp-)?(?:video|audio))(?:\s+[^>]*)?>)(.*?)</(?P=tag)>', webpage)) for media_tag, media_type, media_content in media_tags: media_info = { 'formats': [], @@ -2144,9 +2181,15 @@ class InfoExtractor(object): src = source_attributes.get('src') if not src: continue - is_plain_url, formats = _media_formats(src, media_type) + f = parse_content_type(source_attributes.get('type')) + is_plain_url, formats = _media_formats(src, media_type, f) if is_plain_url: - f = parse_content_type(source_attributes.get('type')) + # res attribute is not standard but seen several times + # in the wild + f.update({ + 'height': int_or_none(source_attributes.get('res')), + 'format_id': source_attributes.get('label'), + }) f.update(formats[0]) media_info['formats'].append(f) else: @@ -2169,7 +2212,7 @@ class InfoExtractor(object): def _extract_akamai_formats(self, manifest_url, video_id, hosts={}): formats = [] hdcore_sign = 'hdcore=3.7.0' - f4m_url = re.sub(r'(https?://[^/+])/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m') + f4m_url = re.sub(r'(https?://[^/]+)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m') hds_host = hosts.get('hds') if hds_host: f4m_url = re.sub(r'(https?://)[^/]+', r'\1' + hds_host, f4m_url) @@ -2191,8 +2234,9 @@ class InfoExtractor(object): 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 + url_base = self._search_regex( + r'(?:(?:https?|rtmp|rtsp):)?(//[^?]+)', url, 'format url') + http_base_url = '%s:%s' % ('http', url_base) formats = [] if 'm3u8' not in skip_protocols: formats.extend(self._extract_m3u8_formats( @@ -2226,7 +2270,7 @@ class InfoExtractor(object): for protocol in ('rtmp', 'rtsp'): if protocol not in skip_protocols: formats.append({ - 'url': protocol + url_base, + 'url': '%s:%s' % (protocol, url_base), 'format_id': protocol, 'protocol': protocol, }) @@ -2284,6 +2328,8 @@ class InfoExtractor(object): tracks = video_data.get('tracks') if tracks and isinstance(tracks, list): for track in tracks: + if not isinstance(track, dict): + continue if track.get('kind') != 'captions': continue track_url = urljoin(base_url, track.get('file')) @@ -2313,6 +2359,8 @@ class InfoExtractor(object): urls = [] formats = [] for source in jwplayer_sources_data: + if not isinstance(source, dict): + continue source_url = self._proto_relative_url(source.get('file')) if not source_url: continue |