66 lines
		
	
	
	
		
			2.5 KiB
		
	
	
	
		
			Python
		
	
	
	
	
	
			
		
		
	
	
			66 lines
		
	
	
	
		
			2.5 KiB
		
	
	
	
		
			Python
		
	
	
	
	
	
| # encoding: utf-8
 | |
| from __future__ import unicode_literals
 | |
| 
 | |
| import re
 | |
| 
 | |
| from .common import InfoExtractor
 | |
| 
 | |
| 
 | |
| class AftonbladetIE(InfoExtractor):
 | |
|     _VALID_URL = r'^http://tv\.aftonbladet\.se/webbtv.+?(?P<video_id>article[0-9]+)\.ab(?:$|[?#])'
 | |
|     _TEST = {
 | |
|         'url': 'http://tv.aftonbladet.se/webbtv/nyheter/vetenskap/rymden/article36015.ab',
 | |
|         'info_dict': {
 | |
|             'id': 'article36015',
 | |
|             'ext': 'mp4',
 | |
|             'title': 'Vulkanutbrott i rymden - nu släpper NASA bilderna',
 | |
|             'description': 'Jupiters måne mest aktiv av alla himlakroppar',
 | |
|             'timestamp': 1394142732,
 | |
|             'upload_date': '20140306',
 | |
|         },
 | |
|     }
 | |
| 
 | |
|     def _real_extract(self, url):
 | |
|         mobj = re.search(self._VALID_URL, url)
 | |
| 
 | |
|         video_id = mobj.group('video_id')
 | |
|         webpage = self._download_webpage(url, video_id)
 | |
| 
 | |
|         # find internal video meta data
 | |
|         meta_url = 'http://aftonbladet-play.drlib.aptoma.no/video/%s.json'
 | |
|         internal_meta_id = self._html_search_regex(
 | |
|             r'data-aptomaId="([\w\d]+)"', webpage, 'internal_meta_id')
 | |
|         internal_meta_url = meta_url % internal_meta_id
 | |
|         internal_meta_json = self._download_json(
 | |
|             internal_meta_url, video_id, 'Downloading video meta data')
 | |
| 
 | |
|         # find internal video formats
 | |
|         format_url = 'http://aftonbladet-play.videodata.drvideo.aptoma.no/actions/video/?id=%s'
 | |
|         internal_video_id = internal_meta_json['videoId']
 | |
|         internal_formats_url = format_url % internal_video_id
 | |
|         internal_formats_json = self._download_json(
 | |
|             internal_formats_url, video_id, 'Downloading video formats')
 | |
| 
 | |
|         formats = []
 | |
|         for fmt in internal_formats_json['formats']['http']['pseudostreaming']['mp4']:
 | |
|             p = fmt['paths'][0]
 | |
|             formats.append({
 | |
|                 'url': 'http://%s:%d/%s/%s' % (p['address'], p['port'], p['path'], p['filename']),
 | |
|                 'ext': 'mp4',
 | |
|                 'width': fmt['width'],
 | |
|                 'height': fmt['height'],
 | |
|                 'tbr': fmt['bitrate'],
 | |
|                 'protocol': 'http',
 | |
|             })
 | |
|         self._sort_formats(formats)
 | |
| 
 | |
|         return {
 | |
|             'id': video_id,
 | |
|             'title': internal_meta_json['title'],
 | |
|             'formats': formats,
 | |
|             'thumbnail': internal_meta_json['imageUrl'],
 | |
|             'description': internal_meta_json['shortPreamble'],
 | |
|             'timestamp': internal_meta_json['timePublished'],
 | |
|             'duration': internal_meta_json['duration'],
 | |
|             'view_count': internal_meta_json['views'],
 | |
|         }
 | 
