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

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..compat import compat_urllib_request
   7 from ..utils import (
   8     float_or_none,
   9     unescapeHTML,
  10 )
  11
  12
  13 class TwitterCardIE(InfoExtractor):
  14     _VALID_URL = r'https?://(?:www\.)?twitter\.com/i/cards/tfw/v1/(?P<id>\d+)'
  15     _TEST = {
  16         'url': 'https://twitter.com/i/cards/tfw/v1/560070183650213889',
  17         'md5': 'a74f50b310c83170319ba16de6955192',
  18         'info_dict': {
  19             'id': '560070183650213889',
  20             'ext': 'mp4',
  21             'title': 'TwitterCard',
  22             'thumbnail': 're:^https?://.*\.jpg$',
  23             'duration': 30.033,
  24         },
  25     }
  26
  27     def _real_extract(self, url):
  28         video_id = self._match_id(url)
  29
  30         # Different formats served for different User-Agents
  31         USER_AGENTS = [
  32             'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)',  # mp4
  33             'Mozilla/5.0 (Windows NT 5.2; WOW64; rv:38.0) Gecko/20100101 Firefox/38.0',  # webm
  34         ]
  35
  36         config = None
  37         formats = []
  38         for user_agent in USER_AGENTS:
  39             request = compat_urllib_request.Request(url)
  40             request.add_header('User-Agent', user_agent)
  41             webpage = self._download_webpage(request, video_id)
  42
  43             config = self._parse_json(
  44                 unescapeHTML(self._search_regex(
  45                     r'data-player-config="([^"]+)"', webpage, 'data player config')),
  46                 video_id)
  47
  48             video_url = config['playlist'][0]['source']
  49
  50             f = {
  51                 'url': video_url,
  52             }
  53
  54             m = re.search(r'/(?P<width>\d+)x(?P<height>\d+)/', video_url)
  55             if m:
  56                 f.update({
  57                     'width': int(m.group('width')),
  58                     'height': int(m.group('height')),
  59                 })
  60             formats.append(f)
  61         self._sort_formats(formats)
  62
  63         thumbnail = config.get('posterImageUrl')
  64         duration = float_or_none(config.get('duration'))
  65
  66         return {
  67             'id': video_id,
  68             'title': 'TwitterCard',
  69             'thumbnail': thumbnail,
  70             'duration': duration,
  71             'formats': formats,
  72         }
  73
  74
  75 class TwitterIE(TwitterCardIE):
  76     _VALID_URL = r'https?://(?:www|m|mobile)?\.?twitter\.com/(?P<id>[^/]+/status/\d+)'
  77
  78     _TESTS = [{
  79         'url': 'https://m.twitter.com/thereaIbanksy/status/614301758345490432',
  80         'md5': '8bbccb487bd7a31349b775915fcd412f',
  81         'info_dict': {
  82             'id': '614301758345490432',
  83             'ext': 'mp4',
  84             'title': 'thereaIbanksy - This time lapse is so pretty \U0001f60d\U0001f60d',
  85             'thumbnail': 're:^https?://.*\.jpg',
  86             'duration': 29.5,
  87             'description': 'banksy on Twitter: "This time lapse is so pretty \U0001f60d\U0001f60d http://t.co/QB8DDbqiR1"',
  88             'uploader': 'banksy',
  89             'uploader_id': 'thereaIbanksy',
  90         },
  91     }]
  92
  93     def _real_extract(self, url):
  94         id = self._match_id(url)
  95         username, twid = re.match(r'([^/]+)/status/(\d+)', id).groups()
  96         name = username
  97         url = re.sub(r'https?://(m|mobile)\.', 'https://', url)
  98         webpage = self._download_webpage(url, 'tweet: ' + url)
  99         description = unescapeHTML(self._search_regex('<title>\s*(.+?)\s*</title>', webpage, 'title'))
 100         title = description.replace('\n', ' ')
 101         splitdesc = re.match(r'^(.+?)\s*on Twitter:\s* "(.+?)"$', title)
 102         if splitdesc:
 103             name, title = splitdesc.groups()
 104         title = re.sub(r'\s*https?://[^ ]+', '', title)  # strip  'https -_t.co_BJYgOjSeGA' junk from filenames
 105         card_id = self._search_regex(r'["\']/i/cards/tfw/v1/(\d+)', webpage, '/i/card/...')
 106         card_url = 'https://twitter.com/i/cards/tfw/v1/' + card_id
 107         return {
 108             '_type': 'url_transparent',
 109             'ie_key': 'TwitterCard',
 110             'uploader_id': username,
 111             'uploader': name,
 112             'url': card_url,
 113             'webpage_url': url,
 114             'description': description,
 115             'title': username + ' - ' + title,
 116         }