Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/arte.py

   1 # encoding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..compat import (
   8     compat_parse_qs,
   9     compat_urllib_parse_urlparse,
  10 )
  11 from ..utils import (
  12     find_xpath_attr,
  13     unified_strdate,
  14     get_element_by_attribute,
  15     int_or_none,
  16     NO_DEFAULT,
  17     qualities,
  18 )
  19
  20 # There are different sources of video in arte.tv, the extraction process
  21 # is different for each one. The videos usually expire in 7 days, so we can't
  22 # add tests.
  23
  24
  25 class ArteTvIE(InfoExtractor):
  26     _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de|en|es)/.*-(?P<id>.*?)\.html'
  27     IE_NAME = 'arte.tv'
  28
  29     def _real_extract(self, url):
  30         mobj = re.match(self._VALID_URL, url)
  31         lang = mobj.group('lang')
  32         video_id = mobj.group('id')
  33
  34         ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
  35         ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
  36         ref_xml_doc = self._download_xml(
  37             ref_xml_url, video_id, note='Downloading metadata')
  38         config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
  39         config_xml_url = config_node.attrib['ref']
  40         config = self._download_xml(
  41             config_xml_url, video_id, note='Downloading configuration')
  42
  43         formats = [{
  44             'format_id': q.attrib['quality'],
  45             # The playpath starts at 'mp4:', if we don't manually
  46             # split the url, rtmpdump will incorrectly parse them
  47             'url': q.text.split('mp4:', 1)[0],
  48             'play_path': 'mp4:' + q.text.split('mp4:', 1)[1],
  49             'ext': 'flv',
  50             'quality': 2 if q.attrib['quality'] == 'hd' else 1,
  51         } for q in config.findall('./urls/url')]
  52         self._sort_formats(formats)
  53
  54         title = config.find('.//name').text
  55         thumbnail = config.find('.//firstThumbnailUrl').text
  56         return {
  57             'id': video_id,
  58             'title': title,
  59             'thumbnail': thumbnail,
  60             'formats': formats,
  61         }
  62
  63
  64 class ArteTVPlus7IE(InfoExtractor):
  65     IE_NAME = 'arte.tv:+7'
  66     _VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr|de|en|es)/(?:(?:sendungen|emissions|embed)/)?(?P<id>[^/]+)/(?P<name>[^/?#&+])'
  67
  68     @classmethod
  69     def _extract_url_info(cls, url):
  70         mobj = re.match(cls._VALID_URL, url)
  71         lang = mobj.group('lang')
  72         query = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
  73         if 'vid' in query:
  74             video_id = query['vid'][0]
  75         else:
  76             # This is not a real id, it can be for example AJT for the news
  77             # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
  78             video_id = mobj.group('id')
  79         return video_id, lang
  80
  81     def _real_extract(self, url):
  82         video_id, lang = self._extract_url_info(url)
  83         webpage = self._download_webpage(url, video_id)
  84         return self._extract_from_webpage(webpage, video_id, lang)
  85
  86     def _extract_from_webpage(self, webpage, video_id, lang):
  87         patterns_templates = (r'arte_vp_url=["\'](.*?%s.*?)["\']', r'data-url=["\']([^"]+%s[^"]+)["\']')
  88         ids = (video_id, '')
  89         # some pages contain multiple videos (like
  90         # http://www.arte.tv/guide/de/sendungen/XEN/xenius/?vid=055918-015_PLUS7-D),
  91         # so we first try to look for json URLs that contain the video id from
  92         # the 'vid' parameter.
  93         patterns = [t % re.escape(_id) for _id in ids for t in patterns_templates]
  94         json_url = self._html_search_regex(
  95             patterns, webpage, 'json vp url', default=None)
  96         if not json_url:
  97             def find_iframe_url(webpage, default=NO_DEFAULT):
  98                 return self._html_search_regex(
  99                     r'<iframe[^>]+src=(["\'])(?P<url>.+\bjson_url=.+?)\1',
 100                     webpage, 'iframe url', group='url', default=default)
 101
 102             iframe_url = find_iframe_url(webpage, None)
 103             if not iframe_url:
 104                 embed_url = self._html_search_regex(
 105                     r'arte_vp_url_oembed=\'([^\']+?)\'', webpage, 'embed url', default=None)
 106                 if embed_url:
 107                     player = self._download_json(
 108                         embed_url, video_id, 'Downloading player page')
 109                     iframe_url = find_iframe_url(player['html'])
 110             # en and es URLs produce react-based pages with different layout (e.g.
 111             # http://www.arte.tv/guide/en/053330-002-A/carnival-italy?zone=world)
 112             if not iframe_url:
 113                 program = self._search_regex(
 114                     r'program\s*:\s*({.+?["\']embed_html["\'].+?}),?\s*\n',
 115                     webpage, 'program', default=None)
 116                 if program:
 117                     embed_html = self._parse_json(program, video_id)
 118                     if embed_html:
 119                         iframe_url = find_iframe_url(embed_html['embed_html'])
 120             if iframe_url:
 121                 json_url = compat_parse_qs(
 122                     compat_urllib_parse_urlparse(iframe_url).query)['json_url'][0]
 123         if json_url:
 124             return self._extract_from_json_url(json_url, video_id, lang)
 125         # Differend kind of embed URL (e.g.
 126         # http://www.arte.tv/magazine/trepalium/fr/episode-0406-replay-trepalium)
 127         embed_url = self._search_regex(
 128             r'<iframe[^>]+src=(["\'])(?P<url>.+?)\1',
 129             webpage, 'embed url', group='url')
 130         return self.url_result(embed_url)
 131
 132     def _extract_from_json_url(self, json_url, video_id, lang):
 133         info = self._download_json(json_url, video_id)
 134         player_info = info['videoJsonPlayer']
 135
 136         upload_date_str = player_info.get('shootingDate')
 137         if not upload_date_str:
 138             upload_date_str = (player_info.get('VRA') or player_info.get('VDA') or '').split(' ')[0]
 139
 140         title = player_info['VTI'].strip()
 141         subtitle = player_info.get('VSU', '').strip()
 142         if subtitle:
 143             title += ' - %s' % subtitle
 144
 145         info_dict = {
 146             'id': player_info['VID'],
 147             'title': title,
 148             'description': player_info.get('VDE'),
 149             'upload_date': unified_strdate(upload_date_str),
 150             'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
 151         }
 152         qfunc = qualities(['HQ', 'MQ', 'EQ', 'SQ'])
 153
 154         LANGS = {
 155             'fr': 'F',
 156             'de': 'A',
 157             'en': 'E[ANG]',
 158             'es': 'E[ESP]',
 159         }
 160
 161         formats = []
 162         for format_id, format_dict in player_info['VSR'].items():
 163             f = dict(format_dict)
 164             versionCode = f.get('versionCode')
 165             langcode = LANGS.get(lang, lang)
 166             lang_rexs = [r'VO?%s-' % re.escape(langcode), r'VO?.-ST%s$' % re.escape(langcode)]
 167             lang_pref = None
 168             if versionCode:
 169                 matched_lang_rexs = [r for r in lang_rexs if re.match(r, versionCode)]
 170                 lang_pref = -10 if not matched_lang_rexs else 10 * len(matched_lang_rexs)
 171             source_pref = 0
 172             if versionCode is not None:
 173                 # The original version with subtitles has lower relevance
 174                 if re.match(r'VO-ST(F|A|E)', versionCode):
 175                     source_pref -= 10
 176                 # The version with sourds/mal subtitles has also lower relevance
 177                 elif re.match(r'VO?(F|A|E)-STM\1', versionCode):
 178                     source_pref -= 9
 179             format = {
 180                 'format_id': format_id,
 181                 'preference': -10 if f.get('videoFormat') == 'M3U8' else None,
 182                 'language_preference': lang_pref,
 183                 'format_note': '%s, %s' % (f.get('versionCode'), f.get('versionLibelle')),
 184                 'width': int_or_none(f.get('width')),
 185                 'height': int_or_none(f.get('height')),
 186                 'tbr': int_or_none(f.get('bitrate')),
 187                 'quality': qfunc(f.get('quality')),
 188                 'source_preference': source_pref,
 189             }
 190
 191             if f.get('mediaType') == 'rtmp':
 192                 format['url'] = f['streamer']
 193                 format['play_path'] = 'mp4:' + f['url']
 194                 format['ext'] = 'flv'
 195             else:
 196                 format['url'] = f['url']
 197
 198             formats.append(format)
 199
 200         self._check_formats(formats, video_id)
 201         self._sort_formats(formats)
 202
 203         info_dict['formats'] = formats
 204         return info_dict
 205
 206
 207 # It also uses the arte_vp_url url from the webpage to extract the information
 208 class ArteTVCreativeIE(ArteTVPlus7IE):
 209     IE_NAME = 'arte.tv:creative'
 210     _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de|en|es)/(?:magazine?/)?(?P<id>[^/?#&]+)'
 211
 212     _TESTS = [{
 213         'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
 214         'info_dict': {
 215             'id': '72176',
 216             'ext': 'mp4',
 217             'title': 'Folge 2 - Corporate Design',
 218             'upload_date': '20131004',
 219         },
 220     }, {
 221         'url': 'http://creative.arte.tv/fr/Monty-Python-Reunion',
 222         'info_dict': {
 223             'id': '160676',
 224             'ext': 'mp4',
 225             'title': 'Monty Python live (mostly)',
 226             'description': 'Événement ! Quarante-cinq ans après leurs premiers succès, les légendaires Monty Python remontent sur scène.\n',
 227             'upload_date': '20140805',
 228         }
 229     }]
 230
 231
 232 class ArteTVFutureIE(ArteTVPlus7IE):
 233     IE_NAME = 'arte.tv:future'
 234     _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de|en|es)/(?P<id>[^/?#&]+)'
 235
 236     _TESTS = [{
 237         'url': 'http://future.arte.tv/fr/info-sciences/les-ecrevisses-aussi-sont-anxieuses',
 238         'info_dict': {
 239             'id': '050940-028-A',
 240             'ext': 'mp4',
 241             'title': 'Les écrevisses aussi peuvent être anxieuses',
 242             'upload_date': '20140902',
 243         },
 244     }, {
 245         'url': 'http://future.arte.tv/fr/la-science-est-elle-responsable',
 246         'only_matching': True,
 247     }]
 248
 249
 250 class ArteTVDDCIE(ArteTVPlus7IE):
 251     IE_NAME = 'arte.tv:ddc'
 252     _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>[^/?#&]+)'
 253
 254     def _real_extract(self, url):
 255         video_id, lang = self._extract_url_info(url)
 256         if lang == 'folge':
 257             lang = 'de'
 258         elif lang == 'emission':
 259             lang = 'fr'
 260         webpage = self._download_webpage(url, video_id)
 261         scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
 262         script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
 263         javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
 264         json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url')
 265         return self._extract_from_json_url(json_url, video_id, lang)
 266
 267
 268 class ArteTVConcertIE(ArteTVPlus7IE):
 269     IE_NAME = 'arte.tv:concert'
 270     _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>fr|de|en|es)/(?P<id>[^/?#&]+)'
 271
 272     _TEST = {
 273         'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
 274         'md5': '9ea035b7bd69696b67aa2ccaaa218161',
 275         'info_dict': {
 276             'id': '186',
 277             'ext': 'mp4',
 278             'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
 279             'upload_date': '20140128',
 280             'description': 'md5:486eb08f991552ade77439fe6d82c305',
 281         },
 282     }
 283
 284
 285 class ArteTVCinemaIE(ArteTVPlus7IE):
 286     IE_NAME = 'arte.tv:cinema'
 287     _VALID_URL = r'https?://cinema\.arte\.tv/(?P<lang>fr|de|en|es)/(?P<id>.+)'
 288
 289     _TEST = {
 290         'url': 'http://cinema.arte.tv/de/node/38291',
 291         'md5': '6b275511a5107c60bacbeeda368c3aa1',
 292         'info_dict': {
 293             'id': '055876-000_PWA12025-D',
 294             'ext': 'mp4',
 295             'title': 'Tod auf dem Nil',
 296             'upload_date': '20160122',
 297             'description': 'md5:7f749bbb77d800ef2be11d54529b96bc',
 298         },
 299     }
 300
 301
 302 class ArteTVMagazineIE(ArteTVPlus7IE):
 303     IE_NAME = 'arte.tv:magazine'
 304     _VALID_URL = r'https?://(?:www\.)?arte\.tv/magazine/[^/]+/(?P<lang>fr|de|en|es)/(?P<id>[^/?#&]+)'
 305
 306     _TESTS = [{
 307         # Embedded via <iframe src="http://www.arte.tv/arte_vp/index.php?json_url=..."
 308         'url': 'http://www.arte.tv/magazine/trepalium/fr/entretien-avec-le-realisateur-vincent-lannoo-trepalium',
 309         'md5': '2a9369bcccf847d1c741e51416299f25',
 310         'info_dict': {
 311             'id': '065965-000-A',
 312             'ext': 'mp4',
 313             'title': 'Trepalium - Extrait Ep.01',
 314             'upload_date': '20160121',
 315         },
 316     }, {
 317         # Embedded via <iframe src="http://www.arte.tv/guide/fr/embed/054813-004-A/medium"
 318         'url': 'http://www.arte.tv/magazine/trepalium/fr/episode-0406-replay-trepalium',
 319         'md5': 'fedc64fc7a946110fe311634e79782ca',
 320         'info_dict': {
 321             'id': '054813-004_PLUS7-F',
 322             'ext': 'mp4',
 323             'title': 'Trepalium (4/6)',
 324             'description': 'md5:10057003c34d54e95350be4f9b05cb40',
 325             'upload_date': '20160218',
 326         },
 327     }, {
 328         'url': 'http://www.arte.tv/magazine/metropolis/de/frank-woeste-german-paris-metropolis',
 329         'only_matching': True,
 330     }]
 331
 332
 333 class ArteTVEmbedIE(ArteTVPlus7IE):
 334     IE_NAME = 'arte.tv:embed'
 335     _VALID_URL = r'''(?x)
 336         http://www\.arte\.tv
 337         /playerv2/embed\.php\?json_url=
 338         (?P<json_url>
 339             http://arte\.tv/papi/tvguide/videos/stream/player/
 340             (?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
 341         )
 342     '''
 343
 344     def _real_extract(self, url):
 345         mobj = re.match(self._VALID_URL, url)
 346         video_id = mobj.group('id')
 347         lang = mobj.group('lang')
 348         json_url = mobj.group('json_url')
 349         return self._extract_from_json_url(json_url, video_id, lang)