Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/amp.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     int_or_none,
   7     parse_iso8601,
   8     mimetype2ext,
   9     determine_ext,
  10 )
  11
  12
  13 class AMPIE(InfoExtractor):
  14     # parse Akamai Adaptive Media Player feed
  15     def _extract_feed_info(self, url):
  16         item = self._download_json(
  17             url, None, 'Downloading Akamai AMP feed',
  18             'Unable to download Akamai AMP feed')['channel']['item']
  19
  20         video_id = item['guid']
  21
  22         def get_media_node(name, default=None):
  23             media_name = 'media-%s' % name
  24             media_group = item.get('media-group') or item
  25             return media_group.get(media_name) or item.get(media_name) or item.get(name, default)
  26
  27         thumbnails = []
  28         media_thumbnail = get_media_node('thumbnail')
  29         if media_thumbnail:
  30             if isinstance(media_thumbnail, dict):
  31                 media_thumbnail = [media_thumbnail]
  32             for thumbnail_data in media_thumbnail:
  33                 thumbnail = thumbnail_data['@attributes']
  34                 thumbnails.append({
  35                     'url': self._proto_relative_url(thumbnail['url'], 'http:'),
  36                     'width': int_or_none(thumbnail.get('width')),
  37                     'height': int_or_none(thumbnail.get('height')),
  38                 })
  39
  40         subtitles = {}
  41         media_subtitle = get_media_node('subTitle')
  42         if media_subtitle:
  43             if isinstance(media_subtitle, dict):
  44                 media_subtitle = [media_subtitle]
  45             for subtitle_data in media_subtitle:
  46                 subtitle = subtitle_data['@attributes']
  47                 lang = subtitle.get('lang') or 'en'
  48                 subtitles[lang] = [{'url': subtitle['href']}]
  49
  50         formats = []
  51         media_content = get_media_node('content')
  52         if isinstance(media_content, dict):
  53             media_content = [media_content]
  54         for media_data in media_content:
  55             media = media_data.get('@attributes', {})
  56             media_url = media.get('url')
  57             if not media_url:
  58                 continue
  59             ext = mimetype2ext(media.get('type')) or determine_ext(media_url)
  60             if ext == 'f4m':
  61                 formats.extend(self._extract_f4m_formats(
  62                     media_url + '?hdcore=3.4.0&plugin=aasp-3.4.0.132.124',
  63                     video_id, f4m_id='hds', fatal=False))
  64             elif ext == 'm3u8':
  65                 formats.extend(self._extract_m3u8_formats(
  66                     media_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
  67             else:
  68                 formats.append({
  69                     'format_id': media_data.get('media-category', {}).get('@attributes', {}).get('label'),
  70                     'url': media['url'],
  71                     'tbr': int_or_none(media.get('bitrate')),
  72                     'filesize': int_or_none(media.get('fileSize')),
  73                     'ext': ext,
  74                 })
  75
  76         self._sort_formats(formats)
  77
  78         timestamp = parse_iso8601(item.get('pubDate'), ' ') or parse_iso8601(item.get('dc-date'))
  79
  80         return {
  81             'id': video_id,
  82             'title': get_media_node('title'),
  83             'description': get_media_node('description'),
  84             'thumbnails': thumbnails,
  85             'timestamp': timestamp,
  86             'duration': int_or_none(media_content[0].get('@attributes', {}).get('duration')),
  87             'subtitles': subtitles,
  88             'formats': formats,
  89         }