yt_dlp/extractor/francetv.py

   1 import re
   2 import urllib.parse
   3
   4 from .common import InfoExtractor
   5 from .dailymotion import DailymotionIE
   6 from ..networking import HEADRequest
   7 from ..utils import (
   8     clean_html,
   9     determine_ext,
  10     filter_dict,
  11     format_field,
  12     int_or_none,
  13     join_nonempty,
  14     parse_iso8601,
  15     smuggle_url,
  16     unsmuggle_url,
  17     url_or_none,
  18 )
  19 from ..utils.traversal import traverse_obj
  20
  21
  22 class FranceTVBaseInfoExtractor(InfoExtractor):
  23     def _make_url_result(self, video_id, url=None):
  24         video_id = video_id.split('@')[0]  # for compat with old @catalog IDs
  25         full_id = f'francetv:{video_id}'
  26         if url:
  27             full_id = smuggle_url(full_id, {'hostname': urllib.parse.urlparse(url).hostname})
  28         return self.url_result(full_id, FranceTVIE, video_id)
  29
  30
  31 class FranceTVIE(InfoExtractor):
  32     _VALID_URL = r'francetv:(?P<id>[^@#]+)'
  33     _GEO_COUNTRIES = ['FR']
  34     _GEO_BYPASS = False
  35
  36     _TESTS = [{
  37         # tokenized url is in dinfo['video']['token']
  38         'url': 'francetv:ec217ecc-0733-48cf-ac06-af1347b849d1',
  39         'info_dict': {
  40             'id': 'ec217ecc-0733-48cf-ac06-af1347b849d1',
  41             'ext': 'mp4',
  42             'title': '13h15, le dimanche... - Les mystères de Jésus',
  43             'timestamp': 1502623500,
  44             'duration': 2580,
  45             'thumbnail': r're:^https?://.*\.jpg$',
  46             'upload_date': '20170813',
  47         },
  48         'params': {'skip_download': 'm3u8'},
  49     }, {
  50         # tokenized url is in dinfo['video']['token']['akamai']
  51         'url': 'francetv:c5bda21d-2c6f-4470-8849-3d8327adb2ba',
  52         'info_dict': {
  53             'id': 'c5bda21d-2c6f-4470-8849-3d8327adb2ba',
  54             'ext': 'mp4',
  55             'title': '13h15, le dimanche... - Les mystères de Jésus',
  56             'timestamp': 1514118300,
  57             'duration': 2880,
  58             'thumbnail': r're:^https?://.*\.jpg$',
  59             'upload_date': '20171224',
  60         },
  61         'params': {'skip_download': 'm3u8'},
  62     }, {
  63         'url': 'francetv:162311093',
  64         'only_matching': True,
  65     }, {
  66         'url': 'francetv:NI_1004933@Zouzous',
  67         'only_matching': True,
  68     }, {
  69         'url': 'francetv:NI_983319@Info-web',
  70         'only_matching': True,
  71     }, {
  72         'url': 'francetv:NI_983319',
  73         'only_matching': True,
  74     }, {
  75         'url': 'francetv:NI_657393@Regions',
  76         'only_matching': True,
  77     }, {
  78         # france-3 live
  79         'url': 'francetv:SIM_France3',
  80         'only_matching': True,
  81     }]
  82
  83     def _extract_video(self, video_id, hostname=None):
  84         is_live = None
  85         videos = []
  86         drm_formats = False
  87         title = None
  88         subtitle = None
  89         episode_number = None
  90         season_number = None
  91         image = None
  92         duration = None
  93         timestamp = None
  94         spritesheets = None
  95
  96         # desktop+chrome returns dash; mobile+safari returns hls
  97         for device_type, browser in [('desktop', 'chrome'), ('mobile', 'safari')]:
  98             dinfo = self._download_json(
  99                 f'https://k7.ftven.fr/videos/{video_id}', video_id,
 100                 f'Downloading {device_type} {browser} video JSON', query=filter_dict({
 101                     'device_type': device_type,
 102                     'browser': browser,
 103                     'domain': hostname,
 104                 }), fatal=False, expected_status=422)  # 422 json gives detailed error code/message
 105
 106             if not dinfo:
 107                 continue
 108
 109             if video := traverse_obj(dinfo, ('video', {dict})):
 110                 videos.append(video)
 111                 if duration is None:
 112                     duration = video.get('duration')
 113                 if is_live is None:
 114                     is_live = video.get('is_live')
 115                 if spritesheets is None:
 116                     spritesheets = video.get('spritesheets')
 117             elif code := traverse_obj(dinfo, ('code', {int})):
 118                 if code == 2009:
 119                     self.raise_geo_restricted(countries=self._GEO_COUNTRIES)
 120                 elif code in (2015, 2017):
 121                     # 2015: L'accès à cette vidéo est impossible. (DRM-only)
 122                     # 2017: Cette vidéo n'est pas disponible depuis le site web mobile (b/c DRM)
 123                     drm_formats = True
 124                     continue
 125                 self.report_warning(
 126                     f'{self.IE_NAME} said: {code} "{clean_html(dinfo.get("message"))}"')
 127                 continue
 128
 129             if meta := traverse_obj(dinfo, ('meta', {dict})):
 130                 if title is None:
 131                     title = meta.get('title')
 132                 # meta['pre_title'] contains season and episode number for series in format "S<ID> E<ID>"
 133                 season_number, episode_number = self._search_regex(
 134                     r'S(\d+)\s*E(\d+)', meta.get('pre_title'), 'episode info', group=(1, 2), default=(None, None))
 135                 if subtitle is None:
 136                     subtitle = meta.get('additional_title')
 137                 if image is None:
 138                     image = meta.get('image_url')
 139                 if timestamp is None:
 140                     timestamp = parse_iso8601(meta.get('broadcasted_at'))
 141
 142         if not videos and drm_formats:
 143             self.report_drm(video_id)
 144
 145         formats, subtitles, video_url = [], {}, None
 146         for video in traverse_obj(videos, lambda _, v: url_or_none(v['url'])):
 147             video_url = video['url']
 148             format_id = video.get('format')
 149
 150             if token_url := traverse_obj(video, ('token', (None, 'akamai'), {url_or_none}, any)):
 151                 tokenized_url = traverse_obj(self._download_json(
 152                     token_url, video_id, f'Downloading signed {format_id} manifest URL',
 153                     fatal=False, query={
 154                         'format': 'json',
 155                         'url': video_url,
 156                     }), ('url', {url_or_none}))
 157                 if tokenized_url:
 158                     video_url = tokenized_url
 159
 160             ext = determine_ext(video_url)
 161             if ext == 'f4m':
 162                 formats.extend(self._extract_f4m_formats(
 163                     video_url, video_id, f4m_id=format_id or ext, fatal=False))
 164             elif ext == 'm3u8':
 165                 format_id = format_id or 'hls'
 166                 fmts, subs = self._extract_m3u8_formats_and_subtitles(
 167                     video_url, video_id, 'mp4', m3u8_id=format_id, fatal=False)
 168                 for f in traverse_obj(fmts, lambda _, v: v['vcodec'] == 'none' and v.get('tbr') is None):
 169                     if mobj := re.match(rf'{format_id}-[Aa]udio-\w+-(?P<bitrate>\d+)', f['format_id']):
 170                         f.update({
 171                             'tbr': int_or_none(mobj.group('bitrate')),
 172                             'acodec': 'mp4a',
 173                         })
 174                 formats.extend(fmts)
 175                 self._merge_subtitles(subs, target=subtitles)
 176             elif ext == 'mpd':
 177                 fmts, subs = self._extract_mpd_formats_and_subtitles(
 178                     video_url, video_id, mpd_id=format_id or 'dash', fatal=False)
 179                 formats.extend(fmts)
 180                 self._merge_subtitles(subs, target=subtitles)
 181             elif video_url.startswith('rtmp'):
 182                 formats.append({
 183                     'url': video_url,
 184                     'format_id': join_nonempty('rtmp', format_id),
 185                     'ext': 'flv',
 186                 })
 187             else:
 188                 if self._is_valid_url(video_url, video_id, format_id):
 189                     formats.append({
 190                         'url': video_url,
 191                         'format_id': format_id,
 192                     })
 193
 194             # XXX: what is video['captions']?
 195
 196         if not formats and video_url:
 197             urlh = self._request_webpage(
 198                 HEADRequest(video_url), video_id, 'Checking for geo-restriction',
 199                 fatal=False, expected_status=403)
 200             if urlh and urlh.headers.get('x-errortype') == 'geo':
 201                 self.raise_geo_restricted(countries=self._GEO_COUNTRIES, metadata_available=True)
 202
 203         for f in formats:
 204             if f.get('acodec') != 'none' and f.get('language') in ('qtz', 'qad'):
 205                 f['language_preference'] = -10
 206                 f['format_note'] = 'audio description{}'.format(format_field(f, 'format_note', ', %s'))
 207
 208         if spritesheets:
 209             formats.append({
 210                 'format_id': 'spritesheets',
 211                 'format_note': 'storyboard',
 212                 'acodec': 'none',
 213                 'vcodec': 'none',
 214                 'ext': 'mhtml',
 215                 'protocol': 'mhtml',
 216                 'url': 'about:invalid',
 217                 'fragments': [{
 218                     'url': sheet,
 219                     # XXX: not entirely accurate; each spritesheet seems to be
 220                     # a 10x10 grid of thumbnails corresponding to approximately
 221                     # 2 seconds of the video; the last spritesheet may be shorter
 222                     'duration': 200,
 223                 } for sheet in traverse_obj(spritesheets, (..., {url_or_none}))],
 224             })
 225
 226         return {
 227             'id': video_id,
 228             'title': join_nonempty(title, subtitle, delim=' - ').strip(),
 229             'thumbnail': image,
 230             'duration': duration,
 231             'timestamp': timestamp,
 232             'is_live': is_live,
 233             'formats': formats,
 234             'subtitles': subtitles,
 235             'episode': subtitle if episode_number else None,
 236             'series': title if episode_number else None,
 237             'episode_number': int_or_none(episode_number),
 238             'season_number': int_or_none(season_number),
 239             '_format_sort_fields': ('res', 'tbr', 'proto'),  # prioritize m3u8 over dash
 240         }
 241
 242     def _real_extract(self, url):
 243         url, smuggled_data = unsmuggle_url(url, {})
 244         video_id = self._match_id(url)
 245         hostname = smuggled_data.get('hostname') or 'www.france.tv'
 246
 247         return self._extract_video(video_id, hostname=hostname)
 248
 249
 250 class FranceTVSiteIE(FranceTVBaseInfoExtractor):
 251     _VALID_URL = r'https?://(?:(?:www\.)?france\.tv|mobile\.france\.tv)/(?:[^/]+/)*(?P<id>[^/]+)\.html'
 252
 253     _TESTS = [{
 254         'url': 'https://www.france.tv/france-2/13h15-le-dimanche/140921-les-mysteres-de-jesus.html',
 255         'info_dict': {
 256             'id': 'c5bda21d-2c6f-4470-8849-3d8327adb2ba',
 257             'ext': 'mp4',
 258             'title': '13h15, le dimanche... - Les mystères de Jésus',
 259             'timestamp': 1514118300,
 260             'duration': 2880,
 261             'thumbnail': r're:^https?://.*\.jpg$',
 262             'upload_date': '20171224',
 263         },
 264         'params': {
 265             'skip_download': True,
 266         },
 267         'add_ie': [FranceTVIE.ie_key()],
 268     }, {
 269         # geo-restricted
 270         'url': 'https://www.france.tv/enfants/six-huit-ans/foot2rue/saison-1/3066387-duel-au-vieux-port.html',
 271         'info_dict': {
 272             'id': 'a9050959-eedd-4b4a-9b0d-de6eeaa73e44',
 273             'ext': 'mp4',
 274             'title': 'Foot2Rue - Duel au vieux port',
 275             'episode': 'Duel au vieux port',
 276             'series': 'Foot2Rue',
 277             'episode_number': 1,
 278             'season_number': 1,
 279             'timestamp': 1642761360,
 280             'upload_date': '20220121',
 281             'season': 'Season 1',
 282             'thumbnail': r're:^https?://.*\.jpg$',
 283             'duration': 1441,
 284         },
 285     }, {
 286         # geo-restricted livestream (workflow == 'token-akamai')
 287         'url': 'https://www.france.tv/france-4/direct.html',
 288         'info_dict': {
 289             'id': '9a6a7670-dde9-4264-adbc-55b89558594b',
 290             'ext': 'mp4',
 291             'title': r're:France 4 en direct .+',
 292             'live_status': 'is_live',
 293         },
 294         'skip': 'geo-restricted livestream',
 295     }, {
 296         # livestream (workflow == 'dai')
 297         'url': 'https://www.france.tv/france-2/direct.html',
 298         'info_dict': {
 299             'id': '006194ea-117d-4bcf-94a9-153d999c59ae',
 300             'ext': 'mp4',
 301             'title': r're:France 2 en direct .+',
 302             'live_status': 'is_live',
 303         },
 304         'params': {'skip_download': 'livestream'},
 305     }, {
 306         # france3
 307         'url': 'https://www.france.tv/france-3/des-chiffres-et-des-lettres/139063-emission-du-mardi-9-mai-2017.html',
 308         'only_matching': True,
 309     }, {
 310         # france4
 311         'url': 'https://www.france.tv/france-4/hero-corp/saison-1/134151-apres-le-calme.html',
 312         'only_matching': True,
 313     }, {
 314         # france5
 315         'url': 'https://www.france.tv/france-5/c-a-dire/saison-10/137013-c-a-dire.html',
 316         'only_matching': True,
 317     }, {
 318         # franceo
 319         'url': 'https://www.france.tv/france-o/archipels/132249-mon-ancetre-l-esclave.html',
 320         'only_matching': True,
 321     }, {
 322         'url': 'https://www.france.tv/documentaires/histoire/136517-argentine-les-500-bebes-voles-de-la-dictature.html',
 323         'only_matching': True,
 324     }, {
 325         'url': 'https://www.france.tv/jeux-et-divertissements/divertissements/133965-le-web-contre-attaque.html',
 326         'only_matching': True,
 327     }, {
 328         'url': 'https://mobile.france.tv/france-5/c-dans-l-air/137347-emission-du-vendredi-12-mai-2017.html',
 329         'only_matching': True,
 330     }, {
 331         'url': 'https://www.france.tv/142749-rouge-sang.html',
 332         'only_matching': True,
 333     }, {
 334         # france-3 live
 335         'url': 'https://www.france.tv/france-3/direct.html',
 336         'only_matching': True,
 337     }]
 338
 339     def _real_extract(self, url):
 340         display_id = self._match_id(url)
 341
 342         webpage = self._download_webpage(url, display_id)
 343
 344         video_id = self._search_regex(
 345             r'(?:data-main-video\s*=|videoId["\']?\s*[:=])\s*(["\'])(?P<id>(?:(?!\1).)+)\1',
 346             webpage, 'video id', default=None, group='id')
 347
 348         if not video_id:
 349             video_id = self._html_search_regex(
 350                 r'(?:href=|player\.setVideo\(\s*)"http://videos?\.francetv\.fr/video/([^@"]+@[^"]+)"',
 351                 webpage, 'video ID')
 352
 353         return self._make_url_result(video_id, url=url)
 354
 355
 356 class FranceTVInfoIE(FranceTVBaseInfoExtractor):
 357     IE_NAME = 'francetvinfo.fr'
 358     _VALID_URL = r'https?://(?:www|mobile|france3-regions)\.francetvinfo\.fr/(?:[^/]+/)*(?P<id>[^/?#&.]+)'
 359
 360     _TESTS = [{
 361         'url': 'https://www.francetvinfo.fr/replay-jt/france-3/soir-3/jt-grand-soir-3-jeudi-22-aout-2019_3561461.html',
 362         'info_dict': {
 363             'id': 'd12458ee-5062-48fe-bfdd-a30d6a01b793',
 364             'ext': 'mp4',
 365             'title': 'Soir 3',
 366             'upload_date': '20190822',
 367             'timestamp': 1566510730,
 368             'thumbnail': r're:^https?://.*\.jpe?g$',
 369             'duration': 1637,
 370             'subtitles': {
 371                 'fr': 'mincount:2',
 372             },
 373         },
 374         'params': {
 375             'skip_download': True,
 376         },
 377         'add_ie': [FranceTVIE.ie_key()],
 378     }, {
 379         'note': 'Only an image exists in initial webpage instead of the video',
 380         'url': 'https://www.francetvinfo.fr/sante/maladie/coronavirus/covid-19-en-inde-une-situation-catastrophique-a-new-dehli_4381095.html',
 381         'info_dict': {
 382             'id': '7d204c9e-a2d3-11eb-9e4c-000d3a23d482',
 383             'ext': 'mp4',
 384             'title': 'Covid-19 : une situation catastrophique à New Dehli - Édition du mercredi 21 avril 2021',
 385             'thumbnail': r're:^https?://.*\.jpe?g$',
 386             'duration': 76,
 387             'timestamp': 1619028518,
 388             'upload_date': '20210421',
 389         },
 390         'params': {
 391             'skip_download': True,
 392         },
 393         'add_ie': [FranceTVIE.ie_key()],
 394     }, {
 395         'url': 'http://www.francetvinfo.fr/elections/europeennes/direct-europeennes-regardez-le-debat-entre-les-candidats-a-la-presidence-de-la-commission_600639.html',
 396         'only_matching': True,
 397     }, {
 398         'url': 'http://www.francetvinfo.fr/economie/entreprises/les-entreprises-familiales-le-secret-de-la-reussite_933271.html',
 399         'only_matching': True,
 400     }, {
 401         'url': 'http://france3-regions.francetvinfo.fr/bretagne/cotes-d-armor/thalassa-echappee-breizh-ce-venredi-dans-les-cotes-d-armor-954961.html',
 402         'only_matching': True,
 403     }, {
 404         # Dailymotion embed
 405         'url': 'http://www.francetvinfo.fr/politique/notre-dame-des-landes/video-sur-france-inter-cecile-duflot-denonce-le-regard-meprisant-de-patrick-cohen_1520091.html',
 406         'md5': 'ee7f1828f25a648addc90cb2687b1f12',
 407         'info_dict': {
 408             'id': 'x4iiko0',
 409             'ext': 'mp4',
 410             'title': 'NDDL, référendum, Brexit : Cécile Duflot répond à Patrick Cohen',
 411             'description': 'md5:fdcb582c370756293a65cdfbc6ecd90e',
 412             'timestamp': 1467011958,
 413             'uploader': 'France Inter',
 414             'uploader_id': 'x2q2ez',
 415             'upload_date': '20160627',
 416             'view_count': int,
 417             'tags': ['Politique', 'France Inter', '27 juin 2016', 'Linvité de 8h20', 'Cécile Duflot', 'Patrick Cohen'],
 418             'age_limit': 0,
 419             'duration': 640,
 420             'like_count': int,
 421             'thumbnail': r're:https://[^/?#]+/v/[^/?#]+/x1080',
 422         },
 423         'add_ie': ['Dailymotion'],
 424     }, {
 425         'url': 'http://france3-regions.francetvinfo.fr/limousin/emissions/jt-1213-limousin',
 426         'only_matching': True,
 427     }, {
 428         # "<figure id=" pattern (#28792)
 429         'url': 'https://www.francetvinfo.fr/culture/patrimoine/incendie-de-notre-dame-de-paris/notre-dame-de-paris-de-l-incendie-de-la-cathedrale-a-sa-reconstruction_4372291.html',
 430         'only_matching': True,
 431     }]
 432
 433     def _real_extract(self, url):
 434         display_id = self._match_id(url)
 435
 436         webpage = self._download_webpage(url, display_id)
 437
 438         dailymotion_urls = tuple(DailymotionIE._extract_embed_urls(url, webpage))
 439         if dailymotion_urls:
 440             return self.playlist_result([
 441                 self.url_result(dailymotion_url, DailymotionIE.ie_key())
 442                 for dailymotion_url in dailymotion_urls])
 443
 444         video_id = self._search_regex(
 445             (r'player\.load[^;]+src:\s*["\']([^"\']+)',
 446              r'id-video=([^@]+@[^"]+)',
 447              r'<a[^>]+href="(?:https?:)?//videos\.francetv\.fr/video/([^@]+@[^"]+)"',
 448              r'(?:data-id|<figure[^<]+\bid)=["\']([\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})'),
 449             webpage, 'video id')
 450
 451         return self._make_url_result(video_id, url=url)