git.bitcoin.ninja Git - youtube-dl/blob - youtube_dl/extractor/imdb.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import json
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     compat_urlparse,
   9     get_element_by_attribute,
  10 )
  11
  12
  13 class ImdbIE(InfoExtractor):
  14     IE_NAME = 'imdb'
  15     IE_DESC = 'Internet Movie Database trailers'
  16     _VALID_URL = r'http://(?:www|m)\.imdb\.com/video/imdb/vi(?P<id>\d+)'
  17
  18     _TEST = {
  19         'url': 'http://www.imdb.com/video/imdb/vi2524815897',
  20         'md5': '9f34fa777ade3a6e57a054fdbcb3a068',
  21         'info_dict': {
  22             'id': '2524815897',
  23             'ext': 'mp4',
  24             'title': 'Ice Age: Continental Drift Trailer (No. 2) - IMDb',
  25             'description': 'md5:9061c2219254e5d14e03c25c98e96a81',
  26         }
  27     }
  28
  29     def _real_extract(self, url):
  30         mobj = re.match(self._VALID_URL, url)
  31         video_id = mobj.group('id')
  32         webpage = self._download_webpage('http://www.imdb.com/video/imdb/vi%s' % video_id, video_id)
  33         descr = get_element_by_attribute('itemprop', 'description', webpage)
  34         available_formats = re.findall(
  35             r'case \'(?P<f_id>.*?)\' :$\s+url = \'(?P<path>.*?)\'', webpage,
  36             flags=re.MULTILINE)
  37         formats = []
  38         for f_id, f_path in available_formats:
  39             f_path = f_path.strip()
  40             format_page = self._download_webpage(
  41                 compat_urlparse.urljoin(url, f_path),
  42                 'Downloading info for %s format' % f_id)
  43             json_data = self._search_regex(
  44                 r'<script[^>]+class="imdb-player-data"[^>]*?>(.*?)</script>',
  45                 format_page, 'json data', flags=re.DOTALL)
  46             info = json.loads(json_data)
  47             format_info = info['videoPlayerObject']['video']
  48             formats.append({
  49                 'format_id': f_id,
  50                 'url': format_info['url'],
  51             })
  52
  53         return {
  54             'id': video_id,
  55             'title': self._og_search_title(webpage),
  56             'formats': formats,
  57             'description': descr,
  58             'thumbnail': format_info['slate'],
  59         }
  60
  61
  62 class ImdbListIE(InfoExtractor):
  63     IE_NAME = 'imdb:list'
  64     IE_DESC = 'Internet Movie Database lists'
  65     _VALID_URL = r'http://www\.imdb\.com/list/(?P<id>[\da-zA-Z_-]{11})'
  66
  67     def _real_extract(self, url):
  68         mobj = re.match(self._VALID_URL, url)
  69         list_id = mobj.group('id')
  70
  71         # RSS XML is sometimes malformed
  72         rss = self._download_webpage('http://rss.imdb.com/list/%s' % list_id, list_id, 'Downloading list RSS')
  73         list_title = self._html_search_regex(r'<title>(.*?)</title>', rss, 'list title')
  74
  75         # Export is independent of actual author_id, but returns 404 if no author_id is provided.
  76         # However, passing dummy author_id seems to be enough.
  77         csv = self._download_webpage('http://www.imdb.com/list/export?list_id=%s&author_id=ur00000000' % list_id,
  78                                      list_id, 'Downloading list CSV')
  79
  80         entries = []
  81         for item in csv.split('\n')[1:]:
  82             cols = item.split(',')
  83             if len(cols) < 2:
  84                 continue
  85             item_id = cols[1][1:-1]
  86             if item_id.startswith('vi'):
  87                 entries.append(self.url_result('http://www.imdb.com/video/imdb/%s' % item_id, 'Imdb'))
  88
  89         return self.playlist_result(entries, list_id, list_title)