_ Git - youtube-dl/blob - youtube_dl/extractor/tv2.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     determine_ext,
   9     int_or_none,
  10     float_or_none,
  11     js_to_json,
  12     parse_iso8601,
  13     remove_end,
  14     try_get,
  15 )
  16
  17
  18 class TV2IE(InfoExtractor):
  19     _VALID_URL = r'https?://(?:www\.)?tv2\.no/v/(?P<id>\d+)'
  20     _TEST = {
  21         'url': 'http://www.tv2.no/v/916509/',
  22         'info_dict': {
  23             'id': '916509',
  24             'ext': 'mp4',
  25             'title': 'Se Frode Gryttens hyllest av Steven Gerrard',
  26             'description': 'TV 2 Sportens huspoet tar avskjed med Liverpools kaptein Steven Gerrard.',
  27             'timestamp': 1431715610,
  28             'upload_date': '20150515',
  29             'duration': 156.967,
  30             'view_count': int,
  31             'categories': list,
  32         },
  33         'params': {
  34             # m3u8 download
  35             'skip_download': True,
  36         },
  37     }
  38
  39     def _real_extract(self, url):
  40         video_id = self._match_id(url)
  41
  42         formats = []
  43         format_urls = []
  44         for protocol in ('HDS', 'HLS'):
  45             data = self._download_json(
  46                 'http://sumo.tv2.no/api/web/asset/%s/play.json?protocol=%s&videoFormat=SMIL+ISMUSP' % (video_id, protocol),
  47                 video_id, 'Downloading play JSON')['playback']
  48             items = try_get(data, lambda x: x['items']['item'])
  49             if not items:
  50                 continue
  51             if not isinstance(items, list):
  52                 items = [items]
  53             for item in items:
  54                 if not isinstance(item, dict):
  55                     continue
  56                 video_url = item.get('url')
  57                 if not video_url or video_url in format_urls:
  58                     continue
  59                 format_id = '%s-%s' % (protocol.lower(), item.get('mediaFormat'))
  60                 if not self._is_valid_url(video_url, video_id, format_id):
  61                     continue
  62                 format_urls.append(video_url)
  63                 ext = determine_ext(video_url)
  64                 if ext == 'f4m':
  65                     formats.extend(self._extract_f4m_formats(
  66                         video_url, video_id, f4m_id=format_id, fatal=False))
  67                 elif ext == 'm3u8':
  68                     formats.extend(self._extract_m3u8_formats(
  69                         video_url, video_id, 'mp4', entry_protocol='m3u8_native',
  70                         m3u8_id=format_id, fatal=False))
  71                 elif ext == 'ism' or video_url.endswith('.ism/Manifest'):
  72                     pass
  73                 else:
  74                     formats.append({
  75                         'url': video_url,
  76                         'format_id': format_id,
  77                         'tbr': int_or_none(item.get('bitrate')),
  78                         'filesize': int_or_none(item.get('fileSize')),
  79                     })
  80         self._sort_formats(formats)
  81
  82         asset = self._download_json(
  83             'http://sumo.tv2.no/api/web/asset/%s.json' % video_id,
  84             video_id, 'Downloading metadata JSON')['asset']
  85
  86         title = asset['title']
  87         description = asset.get('description')
  88         timestamp = parse_iso8601(asset.get('createTime'))
  89         duration = float_or_none(asset.get('accurateDuration') or asset.get('duration'))
  90         view_count = int_or_none(asset.get('views'))
  91         categories = asset.get('keywords', '').split(',')
  92
  93         thumbnails = [{
  94             'id': thumbnail.get('@type'),
  95             'url': thumbnail.get('url'),
  96         } for _, thumbnail in asset.get('imageVersions', {}).items()]
  97
  98         return {
  99             'id': video_id,
 100             'url': video_url,
 101             'title': title,
 102             'description': description,
 103             'thumbnails': thumbnails,
 104             'timestamp': timestamp,
 105             'duration': duration,
 106             'view_count': view_count,
 107             'categories': categories,
 108             'formats': formats,
 109         }
 110
 111
 112 class TV2ArticleIE(InfoExtractor):
 113     _VALID_URL = r'https?://(?:www\.)?tv2\.no/(?:a|\d{4}/\d{2}/\d{2}(/[^/]+)+)/(?P<id>\d+)'
 114     _TESTS = [{
 115         'url': 'http://www.tv2.no/2015/05/16/nyheter/alesund/krim/pingvin/6930542',
 116         'info_dict': {
 117             'id': '6930542',
 118             'title': 'Russen hetses etter pingvintyveri - innrømmer å ha åpnet luken på buret',
 119             'description': 'md5:339573779d3eea3542ffe12006190954',
 120         },
 121         'playlist_count': 2,
 122     }, {
 123         'url': 'http://www.tv2.no/a/6930542',
 124         'only_matching': True,
 125     }]
 126
 127     def _real_extract(self, url):
 128         playlist_id = self._match_id(url)
 129
 130         webpage = self._download_webpage(url, playlist_id)
 131
 132         # Old embed pattern (looks unused nowadays)
 133         assets = re.findall(r'data-assetid=["\'](\d+)', webpage)
 134
 135         if not assets:
 136             # New embed pattern
 137             for v in re.findall(r'TV2ContentboxVideo\(({.+?})\)', webpage):
 138                 video = self._parse_json(
 139                     v, playlist_id, transform_source=js_to_json, fatal=False)
 140                 if not video:
 141                     continue
 142                 asset = video.get('assetId')
 143                 if asset:
 144                     assets.append(asset)
 145
 146         entries = [
 147             self.url_result('http://www.tv2.no/v/%s' % asset_id, 'TV2')
 148             for asset_id in assets]
 149
 150         title = remove_end(self._og_search_title(webpage), ' - TV2.no')
 151         description = remove_end(self._og_search_description(webpage), ' - TV2.no')
 152
 153         return self.playlist_result(entries, playlist_id, title, description)