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

   1 from __future__ import unicode_literals
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     float_or_none,
   6     int_or_none,
   7     parse_iso8601,
   8 )
   9
  10
  11 class NYTimesIE(InfoExtractor):
  12     _VALID_URL = r'https?://(?:(?:www\.)?nytimes\.com/video/(?:[^/]+/)+|graphics8\.nytimes\.com/bcvideo/\d+(?:\.\d+)?/iframe/embed\.html\?videoId=)(?P<id>\d+)'
  13
  14     _TEST = {
  15         'url': 'http://www.nytimes.com/video/opinion/100000002847155/verbatim-what-is-a-photocopier.html?playlistId=100000001150263',
  16         'md5': '18a525a510f942ada2720db5f31644c0',
  17         'info_dict': {
  18             'id': '100000002847155',
  19             'ext': 'mov',
  20             'title': 'Verbatim: What Is a Photocopier?',
  21             'description': 'md5:93603dada88ddbda9395632fdc5da260',
  22             'timestamp': 1398631707,
  23             'upload_date': '20140427',
  24             'uploader': 'Brett Weiner',
  25             'duration': 419,
  26         }
  27     }
  28
  29     def _real_extract(self, url):
  30         video_id = self._match_id(url)
  31
  32         video_data = self._download_json(
  33             'http://www.nytimes.com/svc/video/api/v2/video/%s' % video_id,
  34             video_id, 'Downloading video JSON')
  35
  36         title = video_data['headline']
  37         description = video_data.get('summary')
  38         duration = float_or_none(video_data.get('duration'), 1000)
  39
  40         uploader = video_data['byline']
  41         timestamp = parse_iso8601(video_data['publication_date'][:-8])
  42
  43         def get_file_size(file_size):
  44             if isinstance(file_size, int):
  45                 return file_size
  46             elif isinstance(file_size, dict):
  47                 return int(file_size.get('value', 0))
  48             else:
  49                 return 0
  50
  51         formats = [
  52             {
  53                 'url': video['url'],
  54                 'format_id': video.get('type'),
  55                 'vcodec': video.get('video_codec'),
  56                 'width': int_or_none(video.get('width')),
  57                 'height': int_or_none(video.get('height')),
  58                 'filesize': get_file_size(video.get('fileSize')),
  59             } for video in video_data['renditions']
  60         ]
  61         self._sort_formats(formats)
  62
  63         thumbnails = [
  64             {
  65                 'url': 'http://www.nytimes.com/%s' % image['url'],
  66                 'width': int_or_none(image.get('width')),
  67                 'height': int_or_none(image.get('height')),
  68             } for image in video_data['images']
  69         ]
  70
  71         return {
  72             'id': video_id,
  73             'title': title,
  74             'description': description,
  75             'timestamp': timestamp,
  76             'uploader': uploader,
  77             'duration': duration,
  78             'formats': formats,
  79             'thumbnails': thumbnails,
  80         }