X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fturner.py;h=4a6cbfbb88d41dbb9d65a98eb4975e49a5ab0809;hb=HEAD;hp=0d4271f117c3185ea161fa629c72ad201a0d61e9;hpb=2982514072594b1f708abdf654b31da77c0bfa81;p=youtube-dl diff --git a/youtube_dl/extractor/turner.py b/youtube_dl/extractor/turner.py index 0d4271f11..4a6cbfbb8 100644 --- a/youtube_dl/extractor/turner.py +++ b/youtube_dl/extractor/turner.py @@ -3,22 +3,57 @@ from __future__ import unicode_literals import re -from .common import InfoExtractor +from .adobepass import AdobePassIE +from ..compat import compat_str from ..utils import ( xpath_text, int_or_none, determine_ext, + float_or_none, parse_duration, xpath_attr, update_url_query, + ExtractorError, + strip_or_none, + url_or_none, ) -class TurnerBaseIE(InfoExtractor): - def _extract_cvp_info(self, data_src, video_id, path_data={}): +class TurnerBaseIE(AdobePassIE): + _AKAMAI_SPE_TOKEN_CACHE = {} + + def _extract_timestamp(self, video_data): + return int_or_none(xpath_attr(video_data, 'dateCreated', 'uts')) + + def _add_akamai_spe_token(self, tokenizer_src, video_url, content_id, ap_data, custom_tokenizer_query=None): + secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*' + token = self._AKAMAI_SPE_TOKEN_CACHE.get(secure_path) + if not token: + query = { + 'path': secure_path, + } + if custom_tokenizer_query: + query.update(custom_tokenizer_query) + else: + query['videoId'] = content_id + if ap_data.get('auth_required'): + query['accessToken'] = self._extract_mvpd_auth(ap_data['url'], content_id, ap_data['site_name'], ap_data['site_name']) + auth = self._download_xml( + tokenizer_src, content_id, query=query) + error_msg = xpath_text(auth, 'error/msg') + if error_msg: + raise ExtractorError(error_msg, expected=True) + token = xpath_text(auth, 'token') + if not token: + return video_url + self._AKAMAI_SPE_TOKEN_CACHE[secure_path] = token + return video_url + '?hdnea=' + token + + def _extract_cvp_info(self, data_src, video_id, path_data={}, ap_data={}): video_data = self._download_xml(data_src, video_id) - video_id = video_data.attrib['id'].split('/')[-1].split('.')[0] + video_id = video_data.attrib['id'] title = xpath_text(video_data, 'headline', fatal=True) + content_id = xpath_text(video_data, 'contentId') or video_id # rtmp_src = xpath_text(video_data, 'akamai/src') # if rtmp_src: # splited_rtmp_src = rtmp_src.split(',') @@ -26,14 +61,13 @@ class TurnerBaseIE(InfoExtractor): # rtmp_src = splited_rtmp_src[1] # aifp = xpath_text(video_data, 'akamai/aifp', default='') - tokens = {} urls = [] formats = [] - rex = re.compile(r'''(?x) - (?P[0-9]+)x(?P[0-9]+) - (?:_(?P[0-9]+))? - ''') - for video_file in video_data.findall('files/file'): + rex = re.compile( + r'(?P[0-9]+)x(?P[0-9]+)(?:_(?P[0-9]+))?') + # Possible formats locations: files/file, files/groupFiles/files + # and maybe others + for video_file in video_data.findall('.//file'): video_url = video_file.text.strip() if not video_url: continue @@ -49,7 +83,7 @@ class TurnerBaseIE(InfoExtractor): # auth = self._download_webpage( # protected_path_data['tokenizer_src'], query={ # 'path': protected_path, - # 'videoId': video_id, + # 'videoId': content_id, # 'aifp': aifp, # }) # token = xpath_text(auth, 'token') @@ -60,20 +94,10 @@ class TurnerBaseIE(InfoExtractor): secure_path_data = path_data.get('secure') if not secure_path_data: continue - video_url = secure_path_data['media_src'] + video_url - secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*' - token = tokens.get(secure_path) - if not token: - auth = self._download_xml( - secure_path_data['tokenizer_src'], video_id, query={ - 'path': secure_path, - 'videoId': video_id, - }) - token = xpath_text(auth, 'token') - if not token: - continue - tokens[secure_path] = token - video_url = video_url + '?hdnea=' + token + video_url = self._add_akamai_spe_token( + secure_path_data['tokenizer_src'], + secure_path_data['media_src'] + video_url, + content_id, ap_data) elif not re.match('https?://', video_url): base_path_data = path_data.get(ext, path_data.get('default', {})) media_src = base_path_data.get('media_src') @@ -83,16 +107,22 @@ class TurnerBaseIE(InfoExtractor): if video_url in urls: continue urls.append(video_url) - format_id = video_file.attrib['bitrate'] + format_id = video_file.get('bitrate') if ext == 'smil': - formats.extend(self._extract_smil_formats(video_url, video_id, fatal=False)) + formats.extend(self._extract_smil_formats( + video_url, video_id, fatal=False)) elif ext == 'm3u8': - formats.extend(self._extract_m3u8_formats( - video_url, video_id, 'mp4', m3u8_id=format_id, fatal=False)) + m3u8_formats = self._extract_m3u8_formats( + video_url, video_id, 'mp4', + m3u8_id=format_id or 'hls', fatal=False) + if '/secure/' in video_url and '?hdnea=' in video_url: + for f in m3u8_formats: + f['_seekable'] = False + formats.extend(m3u8_formats) elif ext == 'f4m': formats.extend(self._extract_f4m_formats( update_url_query(video_url, {'hdcore': '3.7.0'}), - video_id, f4m_id=format_id, fatal=False)) + video_id, f4m_id=format_id or 'hds', fatal=False)) else: f = { 'format_id': format_id, @@ -106,29 +136,31 @@ class TurnerBaseIE(InfoExtractor): 'height': int(mobj.group('height')), 'tbr': int_or_none(mobj.group('bitrate')), }) - elif format_id.isdigit(): - f['tbr'] = int(format_id) - else: - mobj = re.match(r'ios_(audio|[0-9]+)$', format_id) - if mobj: - if mobj.group(1) == 'audio': - f.update({ - 'vcodec': 'none', - 'ext': 'm4a', - }) - else: - f['tbr'] = int(mobj.group(1)) + elif isinstance(format_id, compat_str): + if format_id.isdigit(): + f['tbr'] = int(format_id) + else: + mobj = re.match(r'ios_(audio|[0-9]+)$', format_id) + if mobj: + if mobj.group(1) == 'audio': + f.update({ + 'vcodec': 'none', + 'ext': 'm4a', + }) + else: + f['tbr'] = int(mobj.group(1)) formats.append(f) self._sort_formats(formats) subtitles = {} for source in video_data.findall('closedCaptions/source'): for track in source.findall('track'): - source_url = source.get('url') - if not source_url: + track_url = url_or_none(track.get('url')) + if not track_url or track_url.endswith('/big'): continue - subtitles.set_default(source.get('lang') or source.get('label') or 'en', []).append({ - 'url': source_url, + lang = track.get('lang') or track.get('label') or 'en' + subtitles.setdefault(lang, []).append({ + 'url': track_url, 'ext': { 'scc': 'scc', 'webvtt': 'vtt', @@ -143,21 +175,60 @@ class TurnerBaseIE(InfoExtractor): 'height': int_or_none(image.get('height')), } for image in video_data.findall('images/image')] - timestamp = None - if 'cnn.com' not in data_src: - timestamp = int_or_none(xpath_attr(video_data, 'dateCreated', 'uts')) + is_live = xpath_text(video_data, 'isLive') == 'true' return { 'id': video_id, - 'title': title, + 'title': self._live_title(title) if is_live else title, 'formats': formats, 'subtitles': subtitles, 'thumbnails': thumbnails, - 'description': xpath_text(video_data, 'description'), + 'thumbnail': xpath_text(video_data, 'poster'), + 'description': strip_or_none(xpath_text(video_data, 'description')), 'duration': parse_duration(xpath_text(video_data, 'length') or xpath_text(video_data, 'trt')), - 'timestamp': timestamp, + 'timestamp': self._extract_timestamp(video_data), 'upload_date': xpath_attr(video_data, 'metas', 'version'), 'series': xpath_text(video_data, 'showTitle'), 'season_number': int_or_none(xpath_text(video_data, 'seasonNumber')), 'episode_number': int_or_none(xpath_text(video_data, 'episodeNumber')), + 'is_live': is_live, + } + + def _extract_ngtv_info(self, media_id, tokenizer_query, ap_data=None): + streams_data = self._download_json( + 'http://medium.ngtv.io/media/%s/tv' % media_id, + media_id)['media']['tv'] + duration = None + chapters = [] + formats = [] + for supported_type in ('unprotected', 'bulkaes'): + stream_data = streams_data.get(supported_type, {}) + m3u8_url = stream_data.get('secureUrl') or stream_data.get('url') + if not m3u8_url: + continue + if stream_data.get('playlistProtection') == 'spe': + m3u8_url = self._add_akamai_spe_token( + 'http://token.ngtv.io/token/token_spe', + m3u8_url, media_id, ap_data or {}, tokenizer_query) + formats.extend(self._extract_m3u8_formats( + m3u8_url, media_id, 'mp4', m3u8_id='hls', fatal=False)) + + duration = float_or_none(stream_data.get('totalRuntime')) + + if not chapters: + for chapter in stream_data.get('contentSegments', []): + start_time = float_or_none(chapter.get('start')) + chapter_duration = float_or_none(chapter.get('duration')) + if start_time is None or chapter_duration is None: + continue + chapters.append({ + 'start_time': start_time, + 'end_time': start_time + chapter_duration, + }) + self._sort_formats(formats) + + return { + 'formats': formats, + 'chapters': chapters, + 'duration': duration, }