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         query = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
  72         if 'vid' in query:
  73             video_id = query['vid'][0]
  74         else:
  75             # This is not a real id, it can be for example AJT for the news
  76             # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
  77             video_id = mobj.group('id')
  78         return video_id, lang
  79
  80     def _real_extract(self, url):
  81         video_id, lang = self._extract_url_info(url)
  82         webpage = self._download_webpage(url, video_id)
  83         return self._extract_from_webpage(webpage, video_id, lang)
  84
  85     def _extract_from_webpage(self, webpage, video_id, lang):
  86         patterns_templates = (r'arte_vp_url=["\'](.*?%s.*?)["\']', r'data-url=["\']([^"]+%s[^"]+)["\']')
  87         ids = (video_id, '')
  88         # some pages contain multiple videos (like
  89         # http://www.arte.tv/guide/de/sendungen/XEN/xenius/?vid=055918-015_PLUS7-D),
  90         # so we first try to look for json URLs that contain the video id from
  91         # the 'vid' parameter.
  92         patterns = [t % re.escape(_id) for _id in ids for t in patterns_templates]
  93         json_url = self._html_search_regex(
  94             patterns, webpage, 'json vp url', default=None)
  95         if not json_url:
  96             iframe_url = self._html_search_regex(
  97                 r'<iframe[^>]+src=(["\'])(?P<url>.+\bjson_url=.+?)\1',
  98                 webpage, 'iframe url', group='url')
  99             json_url = compat_parse_qs(
 100                 compat_urllib_parse_urlparse(iframe_url).query)['json_url'][0]
 101         return self._extract_from_json_url(json_url, video_id, lang)
 102
 103     def _extract_from_json_url(self, json_url, video_id, lang):
 104         info = self._download_json(json_url, video_id)
 105         player_info = info['videoJsonPlayer']
 106
 107         upload_date_str = player_info.get('shootingDate')
 108         if not upload_date_str:
 109             upload_date_str = player_info.get('VDA', '').split(' ')[0]
 110
 111         title = player_info['VTI'].strip()
 112         subtitle = player_info.get('VSU', '').strip()
 113         if subtitle:
 114             title += ' - %s' % subtitle
 115
 116         info_dict = {
 117             'id': player_info['VID'],
 118             'title': title,
 119             'description': player_info.get('VDE'),
 120             'upload_date': unified_strdate(upload_date_str),
 121             'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
 122         }
 123         qfunc = qualities(['HQ', 'MQ', 'EQ', 'SQ'])
 124
 125         formats = []
 126         for format_id, format_dict in player_info['VSR'].items():
 127             f = dict(format_dict)
 128             versionCode = f.get('versionCode')
 129
 130             langcode = {
 131                 'fr': 'F',
 132                 'de': 'A',
 133             }.get(lang, lang)
 134             lang_rexs = [r'VO?%s' % langcode, r'VO?.-ST%s' % langcode]
 135             lang_pref = (
 136                 None if versionCode is None else (
 137                     10 if any(re.match(r, versionCode) for r in lang_rexs)
 138                     else -10))
 139             source_pref = 0
 140             if versionCode is not None:
 141                 # The original version with subtitles has lower relevance
 142                 if re.match(r'VO-ST(F|A)', versionCode):
 143                     source_pref -= 10
 144                 # The version with sourds/mal subtitles has also lower relevance
 145                 elif re.match(r'VO?(F|A)-STM\1', versionCode):
 146                     source_pref -= 9
 147             format = {
 148                 'format_id': format_id,
 149                 'preference': -10 if f.get('videoFormat') == 'M3U8' else None,
 150                 'language_preference': lang_pref,
 151                 'format_note': '%s, %s' % (f.get('versionCode'), f.get('versionLibelle')),
 152                 'width': int_or_none(f.get('width')),
 153                 'height': int_or_none(f.get('height')),
 154                 'tbr': int_or_none(f.get('bitrate')),
 155                 'quality': qfunc(f.get('quality')),
 156                 'source_preference': source_pref,
 157             }
 158
 159             if f.get('mediaType') == 'rtmp':
 160                 format['url'] = f['streamer']
 161                 format['play_path'] = 'mp4:' + f['url']
 162                 format['ext'] = 'flv'
 163             else:
 164                 format['url'] = f['url']
 165
 166             formats.append(format)
 167
 168         self._check_formats(formats, video_id)
 169         self._sort_formats(formats)
 170
 171         info_dict['formats'] = formats
 172         return info_dict
 173
 174
 175 # It also uses the arte_vp_url url from the webpage to extract the information
 176 class ArteTVCreativeIE(ArteTVPlus7IE):
 177     IE_NAME = 'arte.tv:creative'
 178     _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/(?:magazine?/)?(?P<id>[^?#]+)'
 179
 180     _TESTS = [{
 181         'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
 182         'info_dict': {
 183             'id': '72176',
 184             'ext': 'mp4',
 185             'title': 'Folge 2 - Corporate Design',
 186             'upload_date': '20131004',
 187         },
 188     }, {
 189         'url': 'http://creative.arte.tv/fr/Monty-Python-Reunion',
 190         'info_dict': {
 191             'id': '160676',
 192             'ext': 'mp4',
 193             'title': 'Monty Python live (mostly)',
 194             'description': 'Événement ! Quarante-cinq ans après leurs premiers succès, les légendaires Monty Python remontent sur scène.\n',
 195             'upload_date': '20140805',
 196         }
 197     }]
 198
 199
 200 class ArteTVFutureIE(ArteTVPlus7IE):
 201     IE_NAME = 'arte.tv:future'
 202     _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de)/(thema|sujet)/.*?#article-anchor-(?P<id>\d+)'
 203
 204     _TEST = {
 205         'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
 206         'info_dict': {
 207             'id': '5201',
 208             'ext': 'mp4',
 209             'title': 'Les champignons au secours de la planète',
 210             'upload_date': '20131101',
 211         },
 212     }
 213
 214     def _real_extract(self, url):
 215         anchor_id, lang = self._extract_url_info(url)
 216         webpage = self._download_webpage(url, anchor_id)
 217         row = self._search_regex(
 218             r'(?s)id="%s"[^>]*>.+?(<div[^>]*arte_vp_url[^>]*>)' % anchor_id,
 219             webpage, 'row')
 220         return self._extract_from_webpage(row, anchor_id, lang)
 221
 222
 223 class ArteTVDDCIE(ArteTVPlus7IE):
 224     IE_NAME = 'arte.tv:ddc'
 225     _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>.+)'
 226
 227     def _real_extract(self, url):
 228         video_id, lang = self._extract_url_info(url)
 229         if lang == 'folge':
 230             lang = 'de'
 231         elif lang == 'emission':
 232             lang = 'fr'
 233         webpage = self._download_webpage(url, video_id)
 234         scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
 235         script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
 236         javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
 237         json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url')
 238         return self._extract_from_json_url(json_url, video_id, lang)
 239
 240
 241 class ArteTVConcertIE(ArteTVPlus7IE):
 242     IE_NAME = 'arte.tv:concert'
 243     _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>de|fr)/(?P<id>.+)'
 244
 245     _TEST = {
 246         'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
 247         'md5': '9ea035b7bd69696b67aa2ccaaa218161',
 248         'info_dict': {
 249             'id': '186',
 250             'ext': 'mp4',
 251             'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
 252             'upload_date': '20140128',
 253             'description': 'md5:486eb08f991552ade77439fe6d82c305',
 254         },
 255     }
 256
 257
 258 class ArteTVEmbedIE(ArteTVPlus7IE):
 259     IE_NAME = 'arte.tv:embed'
 260     _VALID_URL = r'''(?x)
 261         http://www\.arte\.tv
 262         /playerv2/embed\.php\?json_url=
 263         (?P<json_url>
 264             http://arte\.tv/papi/tvguide/videos/stream/player/
 265             (?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
 266         )
 267     '''
 268
 269     def _real_extract(self, url):
 270         mobj = re.match(self._VALID_URL, url)
 271         video_id = mobj.group('id')
 272         lang = mobj.group('lang')
 273         json_url = mobj.group('json_url')
 274         return self._extract_from_json_url(json_url, video_id, lang)