_ 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     xpath_text,
  11     remove_end,
  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             'url': 'https://twitter.com/i/cards/tfw/v1/654001591733886977',
  43             'md5': 'b6f35e8b08a0bec6c8af77a2f4b3a814',
  44             'info_dict': {
  45                 'id': 'dq4Oj5quskI',
  46                 'ext': 'mp4',
  47                 'title': 'Ubuntu 11.10 Overview',
  48                 'description': 'Take a quick peek at what\'s new and improved in Ubuntu 11.10.\n\nOnce installed take a look at 10 Things to Do After Installing: http://www.omgubuntu.co.uk/2011/10/10-things-to-do-after-installing-ubuntu-11-10/',
  49                 'upload_date': '20111013',
  50                 'uploader': 'OMG! Ubuntu!',
  51                 'uploader_id': 'omgubuntu',
  52             },
  53         }
  54     ]
  55
  56     def _real_extract(self, url):
  57         video_id = self._match_id(url)
  58
  59         # Different formats served for different User-Agents
  60         USER_AGENTS = [
  61             'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)',  # mp4
  62             'Mozilla/5.0 (Windows NT 5.2; WOW64; rv:38.0) Gecko/20100101 Firefox/38.0',  # webm
  63         ]
  64
  65         config = None
  66         formats = []
  67         for user_agent in USER_AGENTS:
  68             request = compat_urllib_request.Request(url)
  69             request.add_header('User-Agent', user_agent)
  70             webpage = self._download_webpage(request, video_id)
  71
  72             youtube_url = self._html_search_regex(
  73                 r'<iframe[^>]+src="((?:https?:)?//www.youtube.com/embed/[^"]+)"',
  74                 webpage, 'youtube iframe', default=None)
  75             if youtube_url:
  76                 return self.url_result(youtube_url, 'Youtube')
  77
  78             config = self._parse_json(self._html_search_regex(
  79                 r'data-player-config="([^"]+)"', webpage, 'data player config'),
  80                 video_id)
  81             if 'playlist' not in config:
  82                 if 'vmapUrl' in config:
  83                     vmap_data = self._download_xml(config['vmapUrl'], video_id)
  84                     video_url = xpath_text(vmap_data, './/MediaFile').strip()
  85                     formats.append({
  86                         'url': video_url,
  87                     })
  88                     break   # same video regardless of UA
  89                 continue
  90
  91             video_url = config['playlist'][0]['source']
  92
  93             f = {
  94                 'url': video_url,
  95             }
  96
  97             m = re.search(r'/(?P<width>\d+)x(?P<height>\d+)/', video_url)
  98             if m:
  99                 f.update({
 100                     'width': int(m.group('width')),
 101                     'height': int(m.group('height')),
 102                 })
 103             formats.append(f)
 104         self._sort_formats(formats)
 105
 106         thumbnail = config.get('posterImageUrl')
 107         duration = float_or_none(config.get('duration'))
 108
 109         return {
 110             'id': video_id,
 111             'title': 'TwitterCard',
 112             'thumbnail': thumbnail,
 113             'duration': duration,
 114             'formats': formats,
 115         }
 116
 117
 118 class TwitterIE(InfoExtractor):
 119     IE_NAME = 'twitter'
 120     _VALID_URL = r'https?://(?:www\.|m\.|mobile\.)?twitter\.com/(?P<user_id>[^/]+)/status/(?P<id>\d+)'
 121     _TEMPLATE_URL = 'https://twitter.com/%s/status/%s'
 122
 123     _TEST = {
 124         'url': 'https://twitter.com/freethenipple/status/643211948184596480',
 125         'md5': '31cd83a116fc41f99ae3d909d4caf6a0',
 126         'info_dict': {
 127             'id': '643211948184596480',
 128             'ext': 'mp4',
 129             'title': 'FREE THE NIPPLE - FTN supporters on Hollywood Blvd today!',
 130             'thumbnail': 're:^https?://.*\.jpg',
 131             'duration': 12.922,
 132             'description': 'FREE THE NIPPLE on Twitter: "FTN supporters on Hollywood Blvd today! http://t.co/c7jHH749xJ"',
 133             'uploader': 'FREE THE NIPPLE',
 134             'uploader_id': 'freethenipple',
 135         },
 136     }
 137
 138     def _real_extract(self, url):
 139         mobj = re.match(self._VALID_URL, url)
 140         user_id = mobj.group('user_id')
 141         twid = mobj.group('id')
 142
 143         webpage = self._download_webpage(self._TEMPLATE_URL % (user_id, twid), twid)
 144
 145         username = remove_end(self._og_search_title(webpage), ' on Twitter')
 146
 147         title = self._og_search_description(webpage).strip('').replace('\n', ' ')
 148
 149         # strip  'https -_t.co_BJYgOjSeGA' junk from filenames
 150         mobj = re.match(r'“(.*)\s+(https?://[^ ]+)”', title)
 151         title, short_url = mobj.groups()
 152
 153         card_id = self._search_regex(
 154             r'["\']/i/cards/tfw/v1/(\d+)', webpage, 'twitter card url')
 155         card_url = 'https://twitter.com/i/cards/tfw/v1/' + card_id
 156
 157         return {
 158             '_type': 'url_transparent',
 159             'ie_key': 'TwitterCard',
 160             'uploader_id': user_id,
 161             'uploader': username,
 162             'url': card_url,
 163             'webpage_url': url,
 164             'description': '%s on Twitter: "%s %s"' % (username, title, short_url),
 165             'title': username + ' - ' + title,
 166         }