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     qualities,
  17 )
  18
  19 # There are different sources of video in arte.tv, the extraction process
  20 # is different for each one. The videos usually expire in 7 days, so we can't
  21 # add tests.
  22
  23
  24 class ArteTvIE(InfoExtractor):
  25     _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html'
  26     IE_NAME = 'arte.tv'
  27
  28     def _real_extract(self, url):
  29         mobj = re.match(self._VALID_URL, url)
  30         lang = mobj.group('lang')
  31         video_id = mobj.group('id')
  32
  33         ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
  34         ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
  35         ref_xml_doc = self._download_xml(
  36             ref_xml_url, video_id, note='Downloading metadata')
  37         config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
  38         config_xml_url = config_node.attrib['ref']
  39         config = self._download_xml(
  40             config_xml_url, video_id, note='Downloading configuration')
  41
  42         formats = [{
  43             'format_id': q.attrib['quality'],
  44             # The playpath starts at 'mp4:', if we don't manually
  45             # split the url, rtmpdump will incorrectly parse them
  46             'url': q.text.split('mp4:', 1)[0],
  47             'play_path': 'mp4:' + q.text.split('mp4:', 1)[1],
  48             'ext': 'flv',
  49             'quality': 2 if q.attrib['quality'] == 'hd' else 1,
  50         } for q in config.findall('./urls/url')]
  51         self._sort_formats(formats)
  52
  53         title = config.find('.//name').text
  54         thumbnail = config.find('.//firstThumbnailUrl').text
  55         return {
  56             'id': video_id,
  57             'title': title,
  58             'thumbnail': thumbnail,
  59             'formats': formats,
  60         }
  61
  62
  63 class ArteTVPlus7IE(InfoExtractor):
  64     IE_NAME = 'arte.tv:+7'
  65     _VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'
  66
  67     @classmethod
  68     def _extract_url_info(cls, url):
  69         mobj = re.match(cls._VALID_URL, url)
  70         lang = mobj.group('lang')
  71         # This is not a real id, it can be for example AJT for the news
  72         # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
  73         video_id = mobj.group('id')
  74         return video_id, lang
  75
  76     def _real_extract(self, url):
  77         video_id, lang = self._extract_url_info(url)
  78         webpage = self._download_webpage(url, video_id)
  79         return self._extract_from_webpage(webpage, video_id, lang)
  80
  81     def _extract_from_webpage(self, webpage, video_id, lang):
  82         json_url = self._html_search_regex(
  83             [r'arte_vp_url=["\'](.*?)["\']', r'data-url=["\']([^"]+)["\']'],
  84             webpage, 'json vp url', default=None)
  85         if not json_url:
  86             iframe_url = self._html_search_regex(
  87                 r'<iframe[^>]+src=(["\'])(?P<url>.+\bjson_url=.+?)\1',
  88                 webpage, 'iframe url', group='url')
  89             json_url = compat_parse_qs(
  90                 compat_urllib_parse_urlparse(iframe_url).query)['json_url'][0]
  91         return self._extract_from_json_url(json_url, video_id, lang)
  92
  93     def _extract_from_json_url(self, json_url, video_id, lang):
  94         info = self._download_json(json_url, video_id)
  95         player_info = info['videoJsonPlayer']
  96
  97         upload_date_str = player_info.get('shootingDate')
  98         if not upload_date_str:
  99             upload_date_str = player_info.get('VDA', '').split(' ')[0]
 100
 101         title = player_info['VTI'].strip()
 102         subtitle = player_info.get('VSU', '').strip()
 103         if subtitle:
 104             title += ' - %s' % subtitle
 105
 106         info_dict = {
 107             'id': player_info['VID'],
 108             'title': title,
 109             'description': player_info.get('VDE'),
 110             'upload_date': unified_strdate(upload_date_str),
 111             'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
 112         }
 113         qfunc = qualities(['HQ', 'MQ', 'EQ', 'SQ'])
 114
 115         formats = []
 116         for format_id, format_dict in player_info['VSR'].items():
 117             f = dict(format_dict)
 118             versionCode = f.get('versionCode')
 119
 120             langcode = {
 121                 'fr': 'F',
 122                 'de': 'A',
 123             }.get(lang, lang)
 124             lang_rexs = [r'VO?%s' % langcode, r'VO?.-ST%s' % langcode]
 125             lang_pref = (
 126                 None if versionCode is None else (
 127                     10 if any(re.match(r, versionCode) for r in lang_rexs)
 128                     else -10))
 129             source_pref = 0
 130             if versionCode is not None:
 131                 # The original version with subtitles has lower relevance
 132                 if re.match(r'VO-ST(F|A)', versionCode):
 133                     source_pref -= 10
 134                 # The version with sourds/mal subtitles has also lower relevance
 135                 elif re.match(r'VO?(F|A)-STM\1', versionCode):
 136                     source_pref -= 9
 137             format = {
 138                 'format_id': format_id,
 139                 'preference': -10 if f.get('videoFormat') == 'M3U8' else None,
 140                 'language_preference': lang_pref,
 141                 'format_note': '%s, %s' % (f.get('versionCode'), f.get('versionLibelle')),
 142                 'width': int_or_none(f.get('width')),
 143                 'height': int_or_none(f.get('height')),
 144                 'tbr': int_or_none(f.get('bitrate')),
 145                 'quality': qfunc(f.get('quality')),
 146                 'source_preference': source_pref,
 147             }
 148
 149             if f.get('mediaType') == 'rtmp':
 150                 format['url'] = f['streamer']
 151                 format['play_path'] = 'mp4:' + f['url']
 152                 format['ext'] = 'flv'
 153             else:
 154                 format['url'] = f['url']
 155
 156             formats.append(format)
 157
 158         self._check_formats(formats, video_id)
 159         self._sort_formats(formats)
 160
 161         info_dict['formats'] = formats
 162         return info_dict
 163
 164
 165 # It also uses the arte_vp_url url from the webpage to extract the information
 166 class ArteTVCreativeIE(ArteTVPlus7IE):
 167     IE_NAME = 'arte.tv:creative'
 168     _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/(?:magazine?/)?(?P<id>[^?#]+)'
 169
 170     _TESTS = [{
 171         'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
 172         'info_dict': {
 173             'id': '72176',
 174             'ext': 'mp4',
 175             'title': 'Folge 2 - Corporate Design',
 176             'upload_date': '20131004',
 177         },
 178     }, {
 179         'url': 'http://creative.arte.tv/fr/Monty-Python-Reunion',
 180         'info_dict': {
 181             'id': '160676',
 182             'ext': 'mp4',
 183             'title': 'Monty Python live (mostly)',
 184             'description': 'Événement ! Quarante-cinq ans après leurs premiers succès, les légendaires Monty Python remontent sur scène.\n',
 185             'upload_date': '20140805',
 186         }
 187     }]
 188
 189
 190 class ArteTVFutureIE(ArteTVPlus7IE):
 191     IE_NAME = 'arte.tv:future'
 192     _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de)/(thema|sujet)/.*?#article-anchor-(?P<id>\d+)'
 193
 194     _TEST = {
 195         'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
 196         'info_dict': {
 197             'id': '5201',
 198             'ext': 'mp4',
 199             'title': 'Les champignons au secours de la planète',
 200             'upload_date': '20131101',
 201         },
 202     }
 203
 204     def _real_extract(self, url):
 205         anchor_id, lang = self._extract_url_info(url)
 206         webpage = self._download_webpage(url, anchor_id)
 207         row = self._search_regex(
 208             r'(?s)id="%s"[^>]*>.+?(<div[^>]*arte_vp_url[^>]*>)' % anchor_id,
 209             webpage, 'row')
 210         return self._extract_from_webpage(row, anchor_id, lang)
 211
 212
 213 class ArteTVDDCIE(ArteTVPlus7IE):
 214     IE_NAME = 'arte.tv:ddc'
 215     _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>.+)'
 216
 217     def _real_extract(self, url):
 218         video_id, lang = self._extract_url_info(url)
 219         if lang == 'folge':
 220             lang = 'de'
 221         elif lang == 'emission':
 222             lang = 'fr'
 223         webpage = self._download_webpage(url, video_id)
 224         scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
 225         script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
 226         javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
 227         json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url')
 228         return self._extract_from_json_url(json_url, video_id, lang)
 229
 230
 231 class ArteTVConcertIE(ArteTVPlus7IE):
 232     IE_NAME = 'arte.tv:concert'
 233     _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>de|fr)/(?P<id>.+)'
 234
 235     _TEST = {
 236         'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
 237         'md5': '9ea035b7bd69696b67aa2ccaaa218161',
 238         'info_dict': {
 239             'id': '186',
 240             'ext': 'mp4',
 241             'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
 242             'upload_date': '20140128',
 243             'description': 'md5:486eb08f991552ade77439fe6d82c305',
 244         },
 245     }
 246
 247
 248 class ArteTVEmbedIE(ArteTVPlus7IE):
 249     IE_NAME = 'arte.tv:embed'
 250     _VALID_URL = r'''(?x)
 251         http://www\.arte\.tv
 252         /playerv2/embed\.php\?json_url=
 253         (?P<json_url>
 254             http://arte\.tv/papi/tvguide/videos/stream/player/
 255             (?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
 256         )
 257     '''
 258
 259     def _real_extract(self, url):
 260         mobj = re.match(self._VALID_URL, url)
 261         video_id = mobj.group('id')
 262         lang = mobj.group('lang')
 263         json_url = mobj.group('json_url')
 264         return self._extract_from_json_url(json_url, video_id, lang)