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