_ Git - youtube-dl/blob - youtube_dl/extractor/ccc.py

   1 # encoding: utf-8
   2 import re
   3 import json
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     unified_strdate,
   8 )
   9
  10 class ComCarCoffIE(InfoExtractor):
  11     _VALID_URL = r'https?://(?:www\.)?comediansincarsgettingcoffee\.com/(?P<id>[a-z0-9\-]+)/?'
  12     _TESTS = [
  13         {
  14             'url': 'http://comediansincarsgettingcoffee.com/miranda-sings-happy-thanksgiving-miranda/',
  15             'info_dict': {
  16                 'id': 'miranda-sings-happy-thanksgiving-miranda',
  17                 'upload_date': '20141127',
  18                 'title': 'Happy Thanksgiving Miranda',
  19                 'description': 'Jerry Seinfeld and his special guest Miranda Sings cruise around town in search of coffee, complaining and apologizing along the way.',
  20                 'thumbnail': 'http://ccc.crackle.com/images/s5e4_thumb.jpg',
  21             },
  22         }
  23     ]
  24
  25     def _real_extract(self, url):
  26         display_id = self._match_id(url)
  27         webpage = self._download_webpage(url, display_id)
  28
  29         full_data = json.loads(self._search_regex(
  30             r'<script type="application/json" id="videoData">(?P<json>.+?)</script>',
  31             webpage, 'json'))
  32
  33         video_id = full_data['activeVideo']['video']
  34         video_data = full_data['videos'][video_id]
  35
  36         return {
  37             'id': video_id,
  38             'display_id': display_id,
  39             'title': video_data['title'],
  40             'description': video_data['description'],
  41             # XXX: the original datum is a full ISO timestamp... why convert it to a worse format?
  42             'upload_date': unified_strdate(video_data['pubDate']),
  43             'thumbnail': video_data['images']['thumb'],
  44             # XXX: what do we do with video_data['images']['poster']?
  45             'formats': self._extract_m3u8_formats(video_data['mediaUrl'], video_id),
  46         }