youtube_dlc/extractor/amp.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     determine_ext,
   7     ExtractorError,
   8     int_or_none,
   9     mimetype2ext,
  10     parse_iso8601,
  11     unified_timestamp,
  12     url_or_none,
  13 )
  14
  15
  16 class AMPIE(InfoExtractor):
  17     # parse Akamai Adaptive Media Player feed
  18     def _extract_feed_info(self, url):
  19         feed = self._download_json(
  20             url, None, 'Downloading Akamai AMP feed',
  21             'Unable to download Akamai AMP feed')
  22         item = feed.get('channel', {}).get('item')
  23         if not item:
  24             raise ExtractorError('%s said: %s' % (self.IE_NAME, feed['error']))
  25
  26         video_id = item['guid']
  27
  28         def get_media_node(name, default=None):
  29             media_name = 'media-%s' % name
  30             media_group = item.get('media-group') or item
  31             return media_group.get(media_name) or item.get(media_name) or item.get(name, default)
  32
  33         thumbnails = []
  34         media_thumbnail = get_media_node('thumbnail')
  35         if media_thumbnail:
  36             if isinstance(media_thumbnail, dict):
  37                 media_thumbnail = [media_thumbnail]
  38             for thumbnail_data in media_thumbnail:
  39                 thumbnail = thumbnail_data.get('@attributes', {})
  40                 thumbnail_url = url_or_none(thumbnail.get('url'))
  41                 if not thumbnail_url:
  42                     continue
  43                 thumbnails.append({
  44                     'url': self._proto_relative_url(thumbnail_url, 'http:'),
  45                     'width': int_or_none(thumbnail.get('width')),
  46                     'height': int_or_none(thumbnail.get('height')),
  47                 })
  48
  49         subtitles = {}
  50         media_subtitle = get_media_node('subTitle')
  51         if media_subtitle:
  52             if isinstance(media_subtitle, dict):
  53                 media_subtitle = [media_subtitle]
  54             for subtitle_data in media_subtitle:
  55                 subtitle = subtitle_data.get('@attributes', {})
  56                 subtitle_href = url_or_none(subtitle.get('href'))
  57                 if not subtitle_href:
  58                     continue
  59                 subtitles.setdefault(subtitle.get('lang') or 'en', []).append({
  60                     'url': subtitle_href,
  61                     'ext': mimetype2ext(subtitle.get('type')) or determine_ext(subtitle_href),
  62                 })
  63
  64         formats = []
  65         media_content = get_media_node('content')
  66         if isinstance(media_content, dict):
  67             media_content = [media_content]
  68         for media_data in media_content:
  69             media = media_data.get('@attributes', {})
  70             media_url = url_or_none(media.get('url'))
  71             if not media_url:
  72                 continue
  73             ext = mimetype2ext(media.get('type')) or determine_ext(media_url)
  74             if ext == 'f4m':
  75                 formats.extend(self._extract_f4m_formats(
  76                     media_url + '?hdcore=3.4.0&plugin=aasp-3.4.0.132.124',
  77                     video_id, f4m_id='hds', fatal=False))
  78             elif ext == 'm3u8':
  79                 formats.extend(self._extract_m3u8_formats(
  80                     media_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
  81             else:
  82                 formats.append({
  83                     'format_id': media_data.get('media-category', {}).get('@attributes', {}).get('label'),
  84                     'url': media_url,
  85                     'tbr': int_or_none(media.get('bitrate')),
  86                     'filesize': int_or_none(media.get('fileSize')),
  87                     'ext': ext,
  88                 })
  89
  90         self._sort_formats(formats)
  91
  92         timestamp = unified_timestamp(item.get('pubDate'), ' ') or parse_iso8601(item.get('dc-date'))
  93
  94         return {
  95             'id': video_id,
  96             'title': get_media_node('title'),
  97             'description': get_media_node('description'),
  98             'thumbnails': thumbnails,
  99             'timestamp': timestamp,
 100             'duration': int_or_none(media_content[0].get('@attributes', {}).get('duration')),
 101             'subtitles': subtitles,
 102             'formats': formats,
 103         }