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 ..utils import (
   8     find_xpath_attr,
   9     unified_strdate,
  10     get_element_by_attribute,
  11     int_or_none,
  12     qualities,
  13 )
  14
  15 # There are different sources of video in arte.tv, the extraction process
  16 # is different for each one. The videos usually expire in 7 days, so we can't
  17 # add tests.
  18
  19
  20 class ArteTvIE(InfoExtractor):
  21     _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html'
  22     IE_NAME = 'arte.tv'
  23
  24     def _real_extract(self, url):
  25         mobj = re.match(self._VALID_URL, url)
  26         lang = mobj.group('lang')
  27         video_id = mobj.group('id')
  28
  29         ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
  30         ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
  31         ref_xml_doc = self._download_xml(
  32             ref_xml_url, video_id, note='Downloading metadata')
  33         config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
  34         config_xml_url = config_node.attrib['ref']
  35         config = self._download_xml(
  36             config_xml_url, video_id, note='Downloading configuration')
  37
  38         formats = [{
  39             'format_id': q.attrib['quality'],
  40             # The playpath starts at 'mp4:', if we don't manually
  41             # split the url, rtmpdump will incorrectly parse them
  42             'url': q.text.split('mp4:', 1)[0],
  43             'play_path': 'mp4:' + q.text.split('mp4:', 1)[1],
  44             'ext': 'flv',
  45             'quality': 2 if q.attrib['quality'] == 'hd' else 1,
  46         } for q in config.findall('./urls/url')]
  47         self._sort_formats(formats)
  48
  49         title = config.find('.//name').text
  50         thumbnail = config.find('.//firstThumbnailUrl').text
  51         return {
  52             'id': video_id,
  53             'title': title,
  54             'thumbnail': thumbnail,
  55             'formats': formats,
  56         }
  57
  58
  59 class ArteTVPlus7IE(InfoExtractor):
  60     IE_NAME = 'arte.tv:+7'
  61     _VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'
  62
  63     @classmethod
  64     def _extract_url_info(cls, url):
  65         mobj = re.match(cls._VALID_URL, url)
  66         lang = mobj.group('lang')
  67         # This is not a real id, it can be for example AJT for the news
  68         # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
  69         video_id = mobj.group('id')
  70         return video_id, lang
  71
  72     def _real_extract(self, url):
  73         video_id, lang = self._extract_url_info(url)
  74         webpage = self._download_webpage(url, video_id)
  75         return self._extract_from_webpage(webpage, video_id, lang)
  76
  77     def _extract_from_webpage(self, webpage, video_id, lang):
  78         json_url = self._html_search_regex(
  79             [r'arte_vp_url=["\'](.*?)["\']', r'data-url=["\']([^"]+)["\']'],
  80             webpage, 'json vp url')
  81         return self._extract_from_json_url(json_url, video_id, lang)
  82
  83     def _extract_from_json_url(self, json_url, video_id, lang):
  84         info = self._download_json(json_url, video_id)
  85         player_info = info['videoJsonPlayer']
  86
  87         upload_date_str = player_info.get('shootingDate')
  88         if not upload_date_str:
  89             upload_date_str = player_info.get('VDA', '').split(' ')[0]
  90
  91         title = player_info['VTI'].strip()
  92         subtitle = player_info.get('VSU', '').strip()
  93         if subtitle:
  94             title += ' - %s' % subtitle
  95
  96         info_dict = {
  97             'id': player_info['VID'],
  98             'title': title,
  99             'description': player_info.get('VDE'),
 100             'upload_date': unified_strdate(upload_date_str),
 101             'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
 102         }
 103         qfunc = qualities(['HQ', 'MQ', 'EQ', 'SQ'])
 104
 105         formats = []
 106         for format_id, format_dict in player_info['VSR'].items():
 107             f = dict(format_dict)
 108             versionCode = f.get('versionCode')
 109
 110             langcode = {
 111                 'fr': 'F',
 112                 'de': 'A',
 113             }.get(lang, lang)
 114             lang_rexs = [r'VO?%s' % langcode, r'VO?.-ST%s' % langcode]
 115             lang_pref = (
 116                 None if versionCode is None else (
 117                     10 if any(re.match(r, versionCode) for r in lang_rexs)
 118                     else -10))
 119             source_pref = 0
 120             if versionCode is not None:
 121                 # The original version with subtitles has lower relevance
 122                 if re.match(r'VO-ST(F|A)', versionCode):
 123                     source_pref -= 10
 124                 # The version with sourds/mal subtitles has also lower relevance
 125                 elif re.match(r'VO?(F|A)-STM\1', versionCode):
 126                     source_pref -= 9
 127             format = {
 128                 'format_id': format_id,
 129                 'preference': -10 if f.get('videoFormat') == 'M3U8' else None,
 130                 'language_preference': lang_pref,
 131                 'format_note': '%s, %s' % (f.get('versionCode'), f.get('versionLibelle')),
 132                 'width': int_or_none(f.get('width')),
 133                 'height': int_or_none(f.get('height')),
 134                 'tbr': int_or_none(f.get('bitrate')),
 135                 'quality': qfunc(f.get('quality')),
 136                 'source_preference': source_pref,
 137             }
 138
 139             if f.get('mediaType') == 'rtmp':
 140                 format['url'] = f['streamer']
 141                 format['play_path'] = 'mp4:' + f['url']
 142                 format['ext'] = 'flv'
 143             else:
 144                 format['url'] = f['url']
 145
 146             formats.append(format)
 147
 148         self._check_formats(formats, video_id)
 149         self._sort_formats(formats)
 150
 151         info_dict['formats'] = formats
 152         return info_dict
 153
 154
 155 # It also uses the arte_vp_url url from the webpage to extract the information
 156 class ArteTVCreativeIE(ArteTVPlus7IE):
 157     IE_NAME = 'arte.tv:creative'
 158     _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/(?:magazine?/)?(?P<id>[^?#]+)'
 159
 160     _TESTS = [{
 161         'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
 162         'info_dict': {
 163             'id': '72176',
 164             'ext': 'mp4',
 165             'title': 'Folge 2 - Corporate Design',
 166             'upload_date': '20131004',
 167         },
 168     }, {
 169         'url': 'http://creative.arte.tv/fr/Monty-Python-Reunion',
 170         'info_dict': {
 171             'id': '160676',
 172             'ext': 'mp4',
 173             'title': 'Monty Python live (mostly)',
 174             'description': 'Événement ! Quarante-cinq ans après leurs premiers succès, les légendaires Monty Python remontent sur scène.\n',
 175             'upload_date': '20140805',
 176         }
 177     }]
 178
 179
 180 class ArteTVFutureIE(ArteTVPlus7IE):
 181     IE_NAME = 'arte.tv:future'
 182     _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de)/(thema|sujet)/.*?#article-anchor-(?P<id>\d+)'
 183
 184     _TEST = {
 185         'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
 186         'info_dict': {
 187             'id': '5201',
 188             'ext': 'mp4',
 189             'title': 'Les champignons au secours de la planète',
 190             'upload_date': '20131101',
 191         },
 192     }
 193
 194     def _real_extract(self, url):
 195         anchor_id, lang = self._extract_url_info(url)
 196         webpage = self._download_webpage(url, anchor_id)
 197         row = self._search_regex(
 198             r'(?s)id="%s"[^>]*>.+?(<div[^>]*arte_vp_url[^>]*>)' % anchor_id,
 199             webpage, 'row')
 200         return self._extract_from_webpage(row, anchor_id, lang)
 201
 202
 203 class ArteTVDDCIE(ArteTVPlus7IE):
 204     IE_NAME = 'arte.tv:ddc'
 205     _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>.+)'
 206
 207     def _real_extract(self, url):
 208         video_id, lang = self._extract_url_info(url)
 209         if lang == 'folge':
 210             lang = 'de'
 211         elif lang == 'emission':
 212             lang = 'fr'
 213         webpage = self._download_webpage(url, video_id)
 214         scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
 215         script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
 216         javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
 217         json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url')
 218         return self._extract_from_json_url(json_url, video_id, lang)
 219
 220
 221 class ArteTVConcertIE(ArteTVPlus7IE):
 222     IE_NAME = 'arte.tv:concert'
 223     _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>de|fr)/(?P<id>.+)'
 224
 225     _TEST = {
 226         'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
 227         'md5': '9ea035b7bd69696b67aa2ccaaa218161',
 228         'info_dict': {
 229             'id': '186',
 230             'ext': 'mp4',
 231             'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
 232             'upload_date': '20140128',
 233             'description': 'md5:486eb08f991552ade77439fe6d82c305',
 234         },
 235     }
 236
 237
 238 class ArteTVEmbedIE(ArteTVPlus7IE):
 239     IE_NAME = 'arte.tv:embed'
 240     _VALID_URL = r'''(?x)
 241         http://www\.arte\.tv
 242         /playerv2/embed\.php\?json_url=
 243         (?P<json_url>
 244             http://arte\.tv/papi/tvguide/videos/stream/player/
 245             (?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
 246         )
 247     '''
 248
 249     def _real_extract(self, url):
 250         mobj = re.match(self._VALID_URL, url)
 251         video_id = mobj.group('id')
 252         lang = mobj.group('lang')
 253         json_url = mobj.group('json_url')
 254         return self._extract_from_json_url(json_url, video_id, lang)