commit e0d42dd4b270d06a953822c091afefd946bd93f2
parent a07879d6b2edc474b0595a29932726fa7aa14b3a
Author: Remita Amine <remitamine@gmail.com>
Date:   Wed, 30 May 2018 13:21:07 +0100

[teamcoco] Fix extraction for full episodes(closes #16573)

Diffstat:
Myoutube_dl/extractor/tbs.py | 61+++++++++++++++++++------------------------------------------
Myoutube_dl/extractor/teamcoco.py | 102+++++++++++++++++++++++++++++++++++++++++++++----------------------------------
Myoutube_dl/extractor/turner.py | 47+++++++++++++++++++++++++++++++++++++++++++++--
3 files changed, 122 insertions(+), 88 deletions(-)

diff --git a/youtube_dl/extractor/tbs.py b/youtube_dl/extractor/tbs.py @@ -4,6 +4,10 @@ from __future__ import unicode_literals import re from .turner import TurnerBaseIE +from ..compat import ( + compat_urllib_parse_urlparse, + compat_parse_qs, +) from ..utils import ( float_or_none, int_or_none, @@ -38,48 +42,22 @@ class TBSIE(TurnerBaseIE): def _real_extract(self, url): site, display_id = re.match(self._VALID_URL, url).groups() webpage = self._download_webpage(url, display_id) - video_data = self._parse_json(self._search_regex( + drupal_settings = self._parse_json(self._search_regex( r'<script[^>]+?data-drupal-selector="drupal-settings-json"[^>]*?>({.+?})</script>', - webpage, 'drupal setting'), display_id)['turner_playlist'][0] + webpage, 'drupal setting'), display_id) + video_data = drupal_settings['turner_playlist'][0] media_id = video_data['mediaID'] title = video_data['title'] + tokenizer_query = compat_parse_qs(compat_urllib_parse_urlparse( + drupal_settings['ngtv_token_url']).query) - 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.vgtf.net/token/token_spe', - m3u8_url, media_id, { - 'url': url, - 'site_name': site[:3].upper(), - 'auth_required': video_data.get('authRequired') == '1', - }) - formats.extend(self._extract_m3u8_formats( - m3u8_url, media_id, 'mp4', m3u8_id='hls', fatal=False)) - - duration = float_or_none(stream_data.get('totalRuntime') or video_data.get('duration')) - - if not chapters: - for chapter in stream_data.get('contentSegments', []): - start_time = float_or_none(chapter.get('start')) - duration = float_or_none(chapter.get('duration')) - if start_time is None or duration is None: - continue - chapters.append({ - 'start_time': start_time, - 'end_time': start_time + duration, - }) - self._sort_formats(formats) + info = self._extract_ngtv_info( + media_id, tokenizer_query, { + 'url': url, + 'site_name': site[:3].upper(), + 'auth_required': video_data.get('authRequired') == '1', + }) thumbnails = [] for image_id, image in video_data.get('images', {}).items(): @@ -98,15 +76,14 @@ class TBSIE(TurnerBaseIE): }) thumbnails.append(i) - return { + info.update({ 'id': media_id, 'title': title, 'description': strip_or_none(video_data.get('descriptionNoTags') or video_data.get('shortDescriptionNoTags')), - 'duration': duration, + 'duration': float_or_none(video_data.get('duration')) or info.get('duration'), 'timestamp': int_or_none(video_data.get('created')), 'season_number': int_or_none(video_data.get('season')), 'episode_number': int_or_none(video_data.get('episode')), - 'cahpters': chapters, 'thumbnails': thumbnails, - 'formats': formats, - } + }) + return info diff --git a/youtube_dl/extractor/teamcoco.py b/youtube_dl/extractor/teamcoco.py @@ -3,7 +3,7 @@ from __future__ import unicode_literals import json -from .common import InfoExtractor +from .turner import TurnerBaseIE from ..utils import ( determine_ext, ExtractorError, @@ -15,7 +15,7 @@ from ..utils import ( ) -class TeamcocoIE(InfoExtractor): +class TeamcocoIE(TurnerBaseIE): _VALID_URL = r'https?://teamcoco\.com/(?P<id>([^/]+/)*[^/?#]+)' _TESTS = [ { @@ -110,6 +110,8 @@ class TeamcocoIE(InfoExtractor): name } duration + turnerMediaId + turnerMediaAuthToken } } ... on NotFoundSlug { @@ -123,53 +125,65 @@ class TeamcocoIE(InfoExtractor): record = response['record'] video_id = record['id'] - video_sources = self._graphql_call('''{ - %s(id: "%s") { - src - } -}''', 'RecordVideoSource', video_id) or {} - - formats = [] - get_quality = qualities(['low', 'sd', 'hd', 'uhd']) - for format_id, src in video_sources.get('src', {}).items(): - if not isinstance(src, dict): - continue - src_url = src.get('src') - if not src_url: - continue - ext = determine_ext(src_url, mimetype2ext(src.get('type'))) - if format_id == 'hls' or ext == 'm3u8': - # compat_urllib_parse.urljoin does not work here - if src_url.startswith('/'): - src_url = 'http://ht.cdn.turner.com/tbs/big/teamcoco' + src_url - formats.extend(self._extract_m3u8_formats( - src_url, video_id, 'mp4', m3u8_id=format_id, fatal=False)) - else: - if src_url.startswith('/mp4:protected/'): - # TODO Correct extraction for these files - continue - tbr = int_or_none(self._search_regex( - r'(\d+)k\.mp4', src_url, 'tbr', default=None)) - - formats.append({ - 'url': src_url, - 'ext': ext, - 'tbr': tbr, - 'format_id': format_id, - 'quality': get_quality(format_id), - }) - if not formats: - formats = self._extract_m3u8_formats( - record['file']['url'], video_id, 'mp4', fatal=False) - self._sort_formats(formats) - - return { + info = { 'id': video_id, 'display_id': display_id, - 'formats': formats, 'title': record['title'], 'thumbnail': record.get('thumb', {}).get('preview'), 'description': record.get('teaser'), 'duration': parse_duration(record.get('duration')), 'timestamp': parse_iso8601(record.get('publishOn')), } + + media_id = record.get('turnerMediaId') + if media_id: + self._initialize_geo_bypass({ + 'countries': ['US'], + }) + info.update(self._extract_ngtv_info(media_id, { + 'accessToken': record['turnerMediaAuthToken'], + 'accessTokenType': 'jws', + })) + else: + video_sources = self._graphql_call('''{ + %s(id: "%s") { + src + } +}''', 'RecordVideoSource', video_id) or {} + + formats = [] + get_quality = qualities(['low', 'sd', 'hd', 'uhd']) + for format_id, src in video_sources.get('src', {}).items(): + if not isinstance(src, dict): + continue + src_url = src.get('src') + if not src_url: + continue + ext = determine_ext(src_url, mimetype2ext(src.get('type'))) + if format_id == 'hls' or ext == 'm3u8': + # compat_urllib_parse.urljoin does not work here + if src_url.startswith('/'): + src_url = 'http://ht.cdn.turner.com/tbs/big/teamcoco' + src_url + formats.extend(self._extract_m3u8_formats( + src_url, video_id, 'mp4', m3u8_id=format_id, fatal=False)) + else: + if src_url.startswith('/mp4:protected/'): + # TODO Correct extraction for these files + continue + tbr = int_or_none(self._search_regex( + r'(\d+)k\.mp4', src_url, 'tbr', default=None)) + + formats.append({ + 'url': src_url, + 'ext': ext, + 'tbr': tbr, + 'format_id': format_id, + 'quality': get_quality(format_id), + }) + if not formats: + formats = self._extract_m3u8_formats( + record['file']['url'], video_id, 'mp4', fatal=False) + self._sort_formats(formats) + info['formats'] = formats + + return info diff --git a/youtube_dl/extractor/turner.py b/youtube_dl/extractor/turner.py @@ -9,6 +9,7 @@ from ..utils import ( xpath_text, int_or_none, determine_ext, + float_or_none, parse_duration, xpath_attr, update_url_query, @@ -23,14 +24,17 @@ class TurnerBaseIE(AdobePassIE): 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): + 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, - 'videoId': content_id, } + 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( @@ -188,3 +192,42 @@ class TurnerBaseIE(AdobePassIE): '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, + }