youtube_dl/extractor/imdb.py

   1 import re
   2 import json
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     compat_urlparse,
   7     get_element_by_attribute,
   8 )
   9
  10
  11 class ImdbIE(InfoExtractor):
  12     IE_NAME = u'imdb'
  13     IE_DESC = u'Internet Movie Database trailers'
  14     _VALID_URL = r'http://(?:www|m)\.imdb\.com/video/imdb/vi(?P<id>\d+)'
  15
  16     _TEST = {
  17         u'url': u'http://www.imdb.com/video/imdb/vi2524815897',
  18         u'md5': u'9f34fa777ade3a6e57a054fdbcb3a068',
  19         u'info_dict': {
  20             u'id': u'2524815897',
  21             u'ext': u'mp4',
  22             u'title': u'Ice Age: Continental Drift Trailer (No. 2) - IMDb',
  23             u'description': u'md5:9061c2219254e5d14e03c25c98e96a81',
  24         }
  25     }
  26
  27     def _real_extract(self, url):
  28         mobj = re.match(self._VALID_URL, url)
  29         video_id = mobj.group('id')
  30         webpage = self._download_webpage('http://www.imdb.com/video/imdb/vi%s' % video_id, video_id)
  31         descr = get_element_by_attribute('itemprop', 'description', webpage)
  32         available_formats = re.findall(
  33             r'case \'(?P<f_id>.*?)\' :$\s+url = \'(?P<path>.*?)\'', webpage,
  34             flags=re.MULTILINE)
  35         formats = []
  36         for f_id, f_path in available_formats:
  37             f_path = f_path.strip()
  38             format_page = self._download_webpage(
  39                 compat_urlparse.urljoin(url, f_path),
  40                 u'Downloading info for %s format' % f_id)
  41             json_data = self._search_regex(
  42                 r'<script[^>]+class="imdb-player-data"[^>]*?>(.*?)</script>',
  43                 format_page, u'json data', flags=re.DOTALL)
  44             info = json.loads(json_data)
  45             format_info = info['videoPlayerObject']['video']
  46             formats.append({
  47                 'format_id': f_id,
  48                 'url': format_info['url'],
  49             })
  50
  51         return {
  52             'id': video_id,
  53             'title': self._og_search_title(webpage),
  54             'formats': formats,
  55             'description': descr,
  56             'thumbnail': format_info['slate'],
  57         }
  58
  59 class ImdbListIE(InfoExtractor):
  60     IE_NAME = u'imdb:list'
  61     IE_DESC = u'Internet Movie Database lists'
  62     _VALID_URL = r'http://www\.imdb\.com/list/(?P<id>[\da-zA-Z_-]{11})'
  63
  64     def _real_extract(self, url):
  65         mobj = re.match(self._VALID_URL, url)
  66         list_id = mobj.group('id')
  67
  68         # RSS XML is sometimes malformed
  69         rss = self._download_webpage('http://rss.imdb.com/list/%s' % list_id, list_id, u'Downloading list RSS')
  70         list_title = self._html_search_regex(r'<title>(.*?)</title>', rss, u'list title')
  71
  72         # Export is independent of actual author_id, but returns 404 if no author_id is provided.
  73         # However, passing dummy author_id seems to be enough.
  74         csv = self._download_webpage('http://www.imdb.com/list/export?list_id=%s&author_id=ur00000000' % list_id,
  75                                      list_id, u'Downloading list CSV')
  76
  77         entries = []
  78         for item in csv.split('\n')[1:]:
  79             cols = item.split(',')
  80             if len(cols) < 2:
  81                 continue
  82             item_id = cols[1][1:-1]
  83             if item_id.startswith('vi'):
  84                 entries.append(self.url_result('http://www.imdb.com/video/imdb/%s' % item_id, 'Imdb'))
  85
  86         return self.playlist_result(entries, list_id, list_title)