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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..compat import compat_urllib_request
   8 from ..utils import (
   9     float_or_none,
  10     unescapeHTML,
  11     xpath_text,
  12 )
  13
  14
  15 class TwitterCardIE(InfoExtractor):
  16     IE_NAME = 'twitter:card'
  17     _VALID_URL = r'https?://(?:www\.)?twitter\.com/i/cards/tfw/v1/(?P<id>\d+)'
  18     _TESTS = [
  19         {
  20             'url': 'https://twitter.com/i/cards/tfw/v1/560070183650213889',
  21             'md5': '7d2f6b4d2eb841a7ccc893d479bfceb4',
  22             'info_dict': {
  23                 'id': '560070183650213889',
  24                 'ext': 'mp4',
  25                 'title': 'TwitterCard',
  26                 'thumbnail': 're:^https?://.*\.jpg$',
  27                 'duration': 30.033,
  28             }
  29         },
  30         {
  31             'url': 'https://twitter.com/i/cards/tfw/v1/623160978427936768',
  32             'md5': '7ee2a553b63d1bccba97fbed97d9e1c8',
  33             'info_dict': {
  34                 'id': '623160978427936768',
  35                 'ext': 'mp4',
  36                 'title': 'TwitterCard',
  37                 'thumbnail': 're:^https?://.*\.jpg',
  38                 'duration': 80.155,
  39             },
  40         }
  41     ]
  42
  43     def _real_extract(self, url):
  44         video_id = self._match_id(url)
  45
  46         # Different formats served for different User-Agents
  47         USER_AGENTS = [
  48             'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)',  # mp4
  49             'Mozilla/5.0 (Windows NT 5.2; WOW64; rv:38.0) Gecko/20100101 Firefox/38.0',  # webm
  50         ]
  51
  52         config = None
  53         formats = []
  54         for user_agent in USER_AGENTS:
  55             request = compat_urllib_request.Request(url)
  56             request.add_header('User-Agent', user_agent)
  57             webpage = self._download_webpage(request, video_id)
  58
  59             config = self._parse_json(
  60                 unescapeHTML(self._search_regex(
  61                     r'data-player-config="([^"]+)"', webpage, 'data player config')),
  62                 video_id)
  63             if 'playlist' not in config:
  64                 if 'vmapUrl' in config:
  65                     vmap_data = self._download_xml(config['vmapUrl'], video_id)
  66                     video_url = xpath_text(vmap_data, './/MediaFile').strip()
  67                     f = {
  68                         'url': video_url,
  69                     }
  70                     ext = re.search(r'\.([a-z0-9]{2,4})(\?.+)?$', video_url)
  71                     if ext:
  72                         f['ext'] = ext.group(1)
  73                     formats.append(f)
  74                     break   # same video regardless of UA
  75                 continue
  76
  77             video_url = config['playlist'][0]['source']
  78
  79             f = {
  80                 'url': video_url,
  81             }
  82
  83             m = re.search(r'/(?P<width>\d+)x(?P<height>\d+)/', video_url)
  84             if m:
  85                 f.update({
  86                     'width': int(m.group('width')),
  87                     'height': int(m.group('height')),
  88                 })
  89             formats.append(f)
  90         self._sort_formats(formats)
  91
  92         thumbnail = config.get('posterImageUrl')
  93         duration = float_or_none(config.get('duration'))
  94
  95         return {
  96             'id': video_id,
  97             'title': 'TwitterCard',
  98             'thumbnail': thumbnail,
  99             'duration': duration,
 100             'formats': formats,
 101         }
 102
 103
 104 class TwitterIE(TwitterCardIE):
 105     IE_NAME = 'twitter'
 106     _VALID_URL = r'https?://(?:www|m|mobile)?\.?twitter\.com/(?P<id>[^/]+/status/\d+)'
 107
 108     _TESTS = [{
 109         'url': 'https://twitter.com/freethenipple/status/643211948184596480',
 110         'md5': '31cd83a116fc41f99ae3d909d4caf6a0',
 111         'info_dict': {
 112             'id': '643211948184596480',
 113             'ext': 'mp4',
 114             'title': 'freethenipple - FTN supporters on Hollywood Blvd today!',
 115             'thumbnail': 're:^https?://.*\.jpg',
 116             'duration': 12.922,
 117             'description': 'FREE THE NIPPLE on Twitter: "FTN supporters on Hollywood Blvd today! http://t.co/c7jHH749xJ"',
 118             'uploader': 'FREE THE NIPPLE',
 119             'uploader_id': 'freethenipple',
 120         },
 121     }]
 122
 123     def _real_extract(self, url):
 124         id = self._match_id(url)
 125         username, twid = re.match(r'([^/]+)/status/(\d+)', id).groups()
 126         name = username
 127         url = re.sub(r'https?://(m|mobile)\.', 'https://', url)
 128         webpage = self._download_webpage(url, 'tweet: ' + url)
 129         description = unescapeHTML(self._search_regex('<title>\s*(.+?)\s*</title>', webpage, 'title'))
 130         title = description.replace('\n', ' ')
 131         splitdesc = re.match(r'^(.+?)\s*on Twitter:\s* "(.+?)"$', title)
 132         if splitdesc:
 133             name, title = splitdesc.groups()
 134         title = re.sub(r'\s*https?://[^ ]+', '', title)  # strip  'https -_t.co_BJYgOjSeGA' junk from filenames
 135         card_id = self._search_regex(r'["\']/i/cards/tfw/v1/(\d+)', webpage, '/i/card/...')
 136         card_url = 'https://twitter.com/i/cards/tfw/v1/' + card_id
 137         return {
 138             '_type': 'url_transparent',
 139             'ie_key': 'TwitterCard',
 140             'uploader_id': username,
 141             'uploader': name,
 142             'url': card_url,
 143             'webpage_url': url,
 144             'description': description,
 145             'title': username + ' - ' + title,
 146         }