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