You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

264 lines
10 KiB

11 years ago
11 years ago
  1. # encoding: utf-8
  2. from __future__ import unicode_literals
  3. import re
  4. from .common import InfoExtractor
  5. from ..utils import (
  6. ExtractorError,
  7. find_xpath_attr,
  8. unified_strdate,
  9. determine_ext,
  10. get_element_by_id,
  11. compat_str,
  12. get_element_by_attribute,
  13. )
  14. # There are different sources of video in arte.tv, the extraction process
  15. # is different for each one. The videos usually expire in 7 days, so we can't
  16. # add tests.
  17. class ArteTvIE(InfoExtractor):
  18. _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html'
  19. IE_NAME = 'arte.tv'
  20. def _real_extract(self, url):
  21. mobj = re.match(self._VALID_URL, url)
  22. lang = mobj.group('lang')
  23. video_id = mobj.group('id')
  24. ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
  25. ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
  26. ref_xml_doc = self._download_xml(
  27. ref_xml_url, video_id, note='Downloading metadata')
  28. config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
  29. config_xml_url = config_node.attrib['ref']
  30. config = self._download_xml(
  31. config_xml_url, video_id, note='Downloading configuration')
  32. formats = [{
  33. 'forma_id': q.attrib['quality'],
  34. # The playpath starts at 'mp4:', if we don't manually
  35. # split the url, rtmpdump will incorrectly parse them
  36. 'url': q.text.split('mp4:', 1)[0],
  37. 'play_path': 'mp4:' + q.text.split('mp4:', 1)[1],
  38. 'ext': 'flv',
  39. 'quality': 2 if q.attrib['quality'] == 'hd' else 1,
  40. } for q in config.findall('./urls/url')]
  41. self._sort_formats(formats)
  42. title = config.find('.//name').text
  43. thumbnail = config.find('.//firstThumbnailUrl').text
  44. return {
  45. 'id': video_id,
  46. 'title': title,
  47. 'thumbnail': thumbnail,
  48. 'formats': formats,
  49. }
  50. class ArteTVPlus7IE(InfoExtractor):
  51. IE_NAME = 'arte.tv:+7'
  52. _VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'
  53. @classmethod
  54. def _extract_url_info(cls, url):
  55. mobj = re.match(cls._VALID_URL, url)
  56. lang = mobj.group('lang')
  57. # This is not a real id, it can be for example AJT for the news
  58. # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
  59. video_id = mobj.group('id')
  60. return video_id, lang
  61. def _real_extract(self, url):
  62. video_id, lang = self._extract_url_info(url)
  63. webpage = self._download_webpage(url, video_id)
  64. return self._extract_from_webpage(webpage, video_id, lang)
  65. def _extract_from_webpage(self, webpage, video_id, lang):
  66. json_url = self._html_search_regex(
  67. r'arte_vp_url="(.*?)"', webpage, 'json vp url')
  68. return self._extract_from_json_url(json_url, video_id, lang)
  69. def _extract_from_json_url(self, json_url, video_id, lang):
  70. info = self._download_json(json_url, video_id)
  71. player_info = info['videoJsonPlayer']
  72. info_dict = {
  73. 'id': player_info['VID'],
  74. 'title': player_info['VTI'],
  75. 'description': player_info.get('VDE'),
  76. 'upload_date': unified_strdate(player_info.get('VDA', '').split(' ')[0]),
  77. 'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
  78. }
  79. all_formats = player_info['VSR'].values()
  80. # Some formats use the m3u8 protocol
  81. all_formats = list(filter(lambda f: f.get('videoFormat') != 'M3U8', all_formats))
  82. def _match_lang(f):
  83. if f.get('versionCode') is None:
  84. return True
  85. # Return true if that format is in the language of the url
  86. if lang == 'fr':
  87. l = 'F'
  88. elif lang == 'de':
  89. l = 'A'
  90. else:
  91. l = lang
  92. regexes = [r'VO?%s' % l, r'VO?.-ST%s' % l]
  93. return any(re.match(r, f['versionCode']) for r in regexes)
  94. # Some formats may not be in the same language as the url
  95. # TODO: Might want not to drop videos that does not match requested language
  96. # but to process those formats with lower precedence
  97. formats = filter(_match_lang, all_formats)
  98. formats = list(formats) # in python3 filter returns an iterator
  99. if not formats:
  100. # Some videos are only available in the 'Originalversion'
  101. # they aren't tagged as being in French or German
  102. # Sometimes there are neither videos of requested lang code
  103. # nor original version videos available
  104. # For such cases we just take all_formats as is
  105. formats = all_formats
  106. if not formats:
  107. raise ExtractorError('The formats list is empty')
  108. if re.match(r'[A-Z]Q', formats[0]['quality']) is not None:
  109. def sort_key(f):
  110. return ['HQ', 'MQ', 'EQ', 'SQ'].index(f['quality'])
  111. else:
  112. def sort_key(f):
  113. versionCode = f.get('versionCode')
  114. if versionCode is None:
  115. versionCode = ''
  116. return (
  117. # Sort first by quality
  118. int(f.get('height', -1)),
  119. int(f.get('bitrate', -1)),
  120. # The original version with subtitles has lower relevance
  121. re.match(r'VO-ST(F|A)', versionCode) is None,
  122. # The version with sourds/mal subtitles has also lower relevance
  123. re.match(r'VO?(F|A)-STM\1', versionCode) is None,
  124. # Prefer http downloads over m3u8
  125. 0 if f['url'].endswith('m3u8') else 1,
  126. )
  127. formats = sorted(formats, key=sort_key)
  128. def _format(format_info):
  129. quality = ''
  130. height = format_info.get('height')
  131. if height is not None:
  132. quality = compat_str(height)
  133. bitrate = format_info.get('bitrate')
  134. if bitrate is not None:
  135. quality += '-%d' % bitrate
  136. if format_info.get('versionCode') is not None:
  137. format_id = '%s-%s' % (quality, format_info['versionCode'])
  138. else:
  139. format_id = quality
  140. info = {
  141. 'format_id': format_id,
  142. 'format_note': format_info.get('versionLibelle'),
  143. 'width': format_info.get('width'),
  144. 'height': height,
  145. }
  146. if format_info['mediaType'] == 'rtmp':
  147. info['url'] = format_info['streamer']
  148. info['play_path'] = 'mp4:' + format_info['url']
  149. info['ext'] = 'flv'
  150. else:
  151. info['url'] = format_info['url']
  152. info['ext'] = determine_ext(info['url'])
  153. return info
  154. info_dict['formats'] = [_format(f) for f in formats]
  155. return info_dict
  156. # It also uses the arte_vp_url url from the webpage to extract the information
  157. class ArteTVCreativeIE(ArteTVPlus7IE):
  158. IE_NAME = 'arte.tv:creative'
  159. _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/magazine?/(?P<id>.+)'
  160. _TEST = {
  161. 'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
  162. 'info_dict': {
  163. 'id': '050489-002',
  164. 'ext': 'mp4',
  165. 'title': 'Agentur Amateur / Agence Amateur #2 : Corporate Design',
  166. },
  167. }
  168. class ArteTVFutureIE(ArteTVPlus7IE):
  169. IE_NAME = 'arte.tv:future'
  170. _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de)/(thema|sujet)/.*?#article-anchor-(?P<id>\d+)'
  171. _TEST = {
  172. 'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
  173. 'info_dict': {
  174. 'id': '5201',
  175. 'ext': 'mp4',
  176. 'title': 'Les champignons au secours de la planète',
  177. 'upload_date': '20131101',
  178. },
  179. }
  180. def _real_extract(self, url):
  181. anchor_id, lang = self._extract_url_info(url)
  182. webpage = self._download_webpage(url, anchor_id)
  183. row = get_element_by_id(anchor_id, webpage)
  184. return self._extract_from_webpage(row, anchor_id, lang)
  185. class ArteTVDDCIE(ArteTVPlus7IE):
  186. IE_NAME = 'arte.tv:ddc'
  187. _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>.+)'
  188. def _real_extract(self, url):
  189. video_id, lang = self._extract_url_info(url)
  190. if lang == 'folge':
  191. lang = 'de'
  192. elif lang == 'emission':
  193. lang = 'fr'
  194. webpage = self._download_webpage(url, video_id)
  195. scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
  196. script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
  197. javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
  198. json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url')
  199. return self._extract_from_json_url(json_url, video_id, lang)
  200. class ArteTVConcertIE(ArteTVPlus7IE):
  201. IE_NAME = 'arte.tv:concert'
  202. _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>de|fr)/(?P<id>.+)'
  203. _TEST = {
  204. 'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
  205. 'md5': '9ea035b7bd69696b67aa2ccaaa218161',
  206. 'info_dict': {
  207. 'id': '186',
  208. 'ext': 'mp4',
  209. 'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
  210. 'upload_date': '20140128',
  211. 'description': 'md5:486eb08f991552ade77439fe6d82c305',
  212. },
  213. }
  214. class ArteTVEmbedIE(ArteTVPlus7IE):
  215. IE_NAME = 'arte.tv:embed'
  216. _VALID_URL = r'''(?x)
  217. http://www\.arte\.tv
  218. /playerv2/embed\.php\?json_url=
  219. (?P<json_url>
  220. http://arte\.tv/papi/tvguide/videos/stream/player/
  221. (?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
  222. )
  223. '''
  224. def _real_extract(self, url):
  225. mobj = re.match(self._VALID_URL, url)
  226. video_id = mobj.group('id')
  227. lang = mobj.group('lang')
  228. json_url = mobj.group('json_url')
  229. return self._extract_from_json_url(json_url, video_id, lang)