diff options
Diffstat (limited to 'youtube_dl/extractor/turner.py')
| -rw-r--r-- | youtube_dl/extractor/turner.py | 155 |
1 files changed, 118 insertions, 37 deletions
diff --git a/youtube_dl/extractor/turner.py b/youtube_dl/extractor/turner.py index 1c0be9fc6..81229a54b 100644 --- a/youtube_dl/extractor/turner.py +++ b/youtube_dl/extractor/turner.py @@ -6,41 +6,77 @@ import re from .adobepass import AdobePassIE from ..compat import compat_str from ..utils import ( + fix_xml_ampersands, 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(AdobePassIE): + _AKAMAI_SPE_TOKEN_CACHE = {} + def _extract_timestamp(self, video_data): return int_or_none(xpath_attr(video_data, 'dateCreated', 'uts')) - def _extract_cvp_info(self, data_src, video_id, path_data={}, ap_data={}): - video_data = self._download_xml(data_src, video_id) + 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={}, fatal=False): + video_data = self._download_xml( + data_src, video_id, + transform_source=lambda s: fix_xml_ampersands(s).strip(), + fatal=fatal) + if not video_data: + return {} 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(',') - # if len(splited_rtmp_src) == 2: - # rtmp_src = splited_rtmp_src[1] + # split_rtmp_src = rtmp_src.split(',') + # if len(split_rtmp_src) == 2: + # rtmp_src = split_rtmp_src[1] # aifp = xpath_text(video_data, 'akamai/aifp', default='') - tokens = {} urls = [] formats = [] + thumbnails = [] + subtitles = {} rex = re.compile( r'(?P<width>[0-9]+)x(?P<height>[0-9]+)(?:_(?P<bitrate>[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() + video_url = url_or_none(video_file.text.strip()) if not video_url: continue ext = determine_ext(video_url) @@ -66,26 +102,10 @@ class TurnerBaseIE(AdobePassIE): 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: - query = { - 'path': secure_path, - 'videoId': content_id, - } - if ap_data.get('auth_required'): - query['accessToken'] = self._extract_mvpd_auth(ap_data['url'], video_id, ap_data['site_name'], ap_data['site_name']) - auth = self._download_xml( - secure_path_data['tokenizer_src'], video_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: - 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') @@ -96,9 +116,28 @@ class TurnerBaseIE(AdobePassIE): continue urls.append(video_url) format_id = video_file.get('bitrate') - if ext == 'smil': + if ext in ('scc', 'srt', 'vtt'): + subtitles.setdefault('en', []).append({ + 'ext': ext, + 'url': video_url, + }) + elif ext == 'png': + thumbnails.append({ + 'id': format_id, + 'url': video_url, + }) + elif ext == 'smil': formats.extend(self._extract_smil_formats( video_url, video_id, fatal=False)) + elif re.match(r'https?://[^/]+\.akamaihd\.net/[iz]/', video_url): + formats.extend(self._extract_akamai_formats( + video_url, video_id, { + 'hds': path_data.get('f4m', {}).get('host'), + # nba.cdn.turner.com, ht.cdn.turner.com, ht2.cdn.turner.com + # ht3.cdn.turner.com, i.cdn.turner.com, s.cdn.turner.com + # ssl.cdn.turner.com + 'http': 'pmd.cdn.turner.com', + })) elif ext == 'm3u8': m3u8_formats = self._extract_m3u8_formats( video_url, video_id, 'mp4', @@ -117,7 +156,7 @@ class TurnerBaseIE(AdobePassIE): 'url': video_url, 'ext': ext, } - mobj = rex.search(format_id + video_url) + mobj = rex.search(video_url) if mobj: f.update({ 'width': int(mobj.group('width')), @@ -140,11 +179,10 @@ class TurnerBaseIE(AdobePassIE): formats.append(f) self._sort_formats(formats) - subtitles = {} for source in video_data.findall('closedCaptions/source'): for track in source.findall('track'): - track_url = track.get('url') - if not isinstance(track_url, compat_str) or track_url.endswith('/big'): + track_url = url_or_none(track.get('url')) + if not track_url or track_url.endswith('/big'): continue lang = track.get('lang') or track.get('label') or 'en' subtitles.setdefault(lang, []).append({ @@ -156,24 +194,67 @@ class TurnerBaseIE(AdobePassIE): }.get(source.get('format')) }) - thumbnails = [{ - 'id': image.get('cut'), + thumbnails.extend({ + 'id': image.get('cut') or image.get('name'), 'url': image.text, 'width': int_or_none(image.get('width')), 'height': int_or_none(image.get('height')), - } for image in video_data.findall('images/image')] + } for image in video_data.findall('images/image')) + + 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': 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, } |