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     ExtractorError,
   9     find_xpath_attr,
  10     unified_strdate,
  11     determine_ext,
  12     get_element_by_id,
  13     compat_str,
  14     get_element_by_attribute,
  15 )
  16
  17 # There are different sources of video in arte.tv, the extraction process
  18 # is different for each one. The videos usually expire in 7 days, so we can't
  19 # add tests.
  20
  21
  22 class ArteTvIE(InfoExtractor):
  23     _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html'
  24     IE_NAME = 'arte.tv'
  25
  26     def _real_extract(self, url):
  27         mobj = re.match(self._VALID_URL, url)
  28         lang = mobj.group('lang')
  29         video_id = mobj.group('id')
  30
  31         ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
  32         ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
  33         ref_xml_doc = self._download_xml(
  34             ref_xml_url, video_id, note='Downloading metadata')
  35         config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
  36         config_xml_url = config_node.attrib['ref']
  37         config = self._download_xml(
  38             config_xml_url, video_id, note='Downloading configuration')
  39
  40         formats = [{
  41             'forma_id': q.attrib['quality'],
  42             'url': q.text,
  43             'ext': 'flv',
  44             'quality': 2 if q.attrib['quality'] == 'hd' else 1,
  45         } for q in config.findall('./urls/url')]
  46         self._sort_formats(formats)
  47
  48         title = config.find('.//name').text
  49         thumbnail = config.find('.//firstThumbnailUrl').text
  50         return {
  51             'id': video_id,
  52             'title': title,
  53             'thumbnail': thumbnail,
  54             'formats': formats,
  55         }
  56
  57
  58 class ArteTVPlus7IE(InfoExtractor):
  59     IE_NAME = 'arte.tv:+7'
  60     _VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'
  61
  62     @classmethod
  63     def _extract_url_info(cls, url):
  64         mobj = re.match(cls._VALID_URL, url)
  65         lang = mobj.group('lang')
  66         # This is not a real id, it can be for example AJT for the news
  67         # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
  68         video_id = mobj.group('id')
  69         return video_id, lang
  70
  71     def _real_extract(self, url):
  72         video_id, lang = self._extract_url_info(url)
  73         webpage = self._download_webpage(url, video_id)
  74         return self._extract_from_webpage(webpage, video_id, lang)
  75
  76     def _extract_from_webpage(self, webpage, video_id, lang):
  77         json_url = self._html_search_regex(
  78             r'arte_vp_url="(.*?)"', webpage, 'json vp url')
  79         return self._extract_from_json_url(json_url, video_id, lang)
  80
  81     def _extract_from_json_url(self, json_url, video_id, lang):
  82         info = self._download_json(json_url, video_id)
  83         player_info = info['videoJsonPlayer']
  84
  85         info_dict = {
  86             'id': player_info['VID'],
  87             'title': player_info['VTI'],
  88             'description': player_info.get('VDE'),
  89             'upload_date': unified_strdate(player_info.get('VDA', '').split(' ')[0]),
  90             'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
  91         }
  92
  93         all_formats = player_info['VSR'].values()
  94         # Some formats use the m3u8 protocol
  95         all_formats = list(filter(lambda f: f.get('videoFormat') != 'M3U8', all_formats))
  96         def _match_lang(f):
  97             if f.get('versionCode') is None:
  98                 return True
  99             # Return true if that format is in the language of the url
 100             if lang == 'fr':
 101                 l = 'F'
 102             elif lang == 'de':
 103                 l = 'A'
 104             else:
 105                 l = lang
 106             regexes = [r'VO?%s' % l, r'VO?.-ST%s' % l]
 107             return any(re.match(r, f['versionCode']) for r in regexes)
 108         # Some formats may not be in the same language as the url
 109         formats = filter(_match_lang, all_formats)
 110         formats = list(formats) # in python3 filter returns an iterator
 111         if not formats:
 112             # Some videos are only available in the 'Originalversion'
 113             # they aren't tagged as being in French or German
 114             if all(f['versionCode'] == 'VO' for f in all_formats):
 115                 formats = all_formats
 116             else:
 117                 raise ExtractorError(u'The formats list is empty')
 118
 119         if re.match(r'[A-Z]Q', formats[0]['quality']) is not None:
 120             def sort_key(f):
 121                 return ['HQ', 'MQ', 'EQ', 'SQ'].index(f['quality'])
 122         else:
 123             def sort_key(f):
 124                 versionCode = f.get('versionCode')
 125                 if versionCode is None:
 126                     versionCode = ''
 127                 return (
 128                     # Sort first by quality
 129                     int(f.get('height', -1)),
 130                     int(f.get('bitrate', -1)),
 131                     # The original version with subtitles has lower relevance
 132                     re.match(r'VO-ST(F|A)', versionCode) is None,
 133                     # The version with sourds/mal subtitles has also lower relevance
 134                     re.match(r'VO?(F|A)-STM\1', versionCode) is None,
 135                     # Prefer http downloads over m3u8
 136                     0 if f['url'].endswith('m3u8') else 1,
 137                 )
 138         formats = sorted(formats, key=sort_key)
 139         def _format(format_info):
 140             quality = ''
 141             height = format_info.get('height')
 142             if height is not None:
 143                 quality = compat_str(height)
 144             bitrate = format_info.get('bitrate')
 145             if bitrate is not None:
 146                 quality += '-%d' % bitrate
 147             if format_info.get('versionCode') is not None:
 148                 format_id = '%s-%s' % (quality, format_info['versionCode'])
 149             else:
 150                 format_id = quality
 151             info = {
 152                 'format_id': format_id,
 153                 'format_note': format_info.get('versionLibelle'),
 154                 'width': format_info.get('width'),
 155                 'height': height,
 156             }
 157             if format_info['mediaType'] == 'rtmp':
 158                 info['url'] = format_info['streamer']
 159                 info['play_path'] = 'mp4:' + format_info['url']
 160                 info['ext'] = 'flv'
 161             else:
 162                 info['url'] = format_info['url']
 163                 info['ext'] = determine_ext(info['url'])
 164             return info
 165         info_dict['formats'] = [_format(f) for f in formats]
 166
 167         return info_dict
 168
 169
 170 # It also uses the arte_vp_url url from the webpage to extract the information
 171 class ArteTVCreativeIE(ArteTVPlus7IE):
 172     IE_NAME = 'arte.tv:creative'
 173     _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/magazine?/(?P<id>.+)'
 174
 175     _TEST = {
 176         'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
 177         'info_dict': {
 178             'id': '050489-002',
 179             'ext': 'mp4',
 180             'title': 'Agentur Amateur / Agence Amateur #2 : Corporate Design',
 181         },
 182     }
 183
 184
 185 class ArteTVFutureIE(ArteTVPlus7IE):
 186     IE_NAME = 'arte.tv:future'
 187     _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de)/(thema|sujet)/.*?#article-anchor-(?P<id>\d+)'
 188
 189     _TEST = {
 190         'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
 191         'info_dict': {
 192             'id': '050940-003',
 193             'ext': 'mp4',
 194             'title': 'Les champignons au secours de la planète',
 195         },
 196     }
 197
 198     def _real_extract(self, url):
 199         anchor_id, lang = self._extract_url_info(url)
 200         webpage = self._download_webpage(url, anchor_id)
 201         row = get_element_by_id(anchor_id, webpage)
 202         return self._extract_from_webpage(row, anchor_id, lang)
 203
 204
 205 class ArteTVDDCIE(ArteTVPlus7IE):
 206     IE_NAME = 'arte.tv:ddc'
 207     _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>.+)'
 208
 209     def _real_extract(self, url):
 210         video_id, lang = self._extract_url_info(url)
 211         if lang == 'folge':
 212             lang = 'de'
 213         elif lang == 'emission':
 214             lang = 'fr'
 215         webpage = self._download_webpage(url, video_id)
 216         scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
 217         script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
 218         javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
 219         json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url')
 220         return self._extract_from_json_url(json_url, video_id, lang)
 221
 222
 223 class ArteTVConcertIE(ArteTVPlus7IE):
 224     IE_NAME = 'arte.tv:concert'
 225     _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>de|fr)/(?P<id>.+)'
 226
 227     _TEST = {
 228         'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
 229         'md5': '9ea035b7bd69696b67aa2ccaaa218161',
 230         'info_dict': {
 231             'id': '186',
 232             'ext': 'mp4',
 233             'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
 234             'upload_date': '20140128',
 235             'description': 'md5:486eb08f991552ade77439fe6d82c305',
 236         },
 237     }
 238
 239
 240 class ArteTVEmbedIE(ArteTVPlus7IE):
 241     IE_NAME = 'arte.tv:embed'
 242     _VALID_URL = r'''(?x)
 243         http://www\.arte\.tv
 244         /playerv2/embed\.php\?json_url=
 245         (?P<json_url>
 246             http://arte\.tv/papi/tvguide/videos/stream/player/
 247             (?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
 248         )
 249     '''
 250
 251     def _real_extract(self, url):
 252         mobj = re.match(self._VALID_URL, url)
 253         video_id = mobj.group('id')
 254         lang = mobj.group('lang')
 255         json_url = mobj.group('json_url')
 256         return self._extract_from_json_url(json_url, video_id, lang)