]> git.bitcoin.ninja Git - youtube-dl/blob - youtube_dl/extractor/amp.py
[brightcove:new] extract subtitles and strip video title
[youtube-dl] / 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 )
9
10
11 class AMPIE(InfoExtractor):
12     # parse Akamai Adaptive Media Player feed
13     def _extract_feed_info(self, url):
14         item = self._download_json(
15             url, None, 'Downloading Akamai AMP feed',
16             'Unable to download Akamai AMP feed')['channel']['item']
17
18         video_id = item['guid']
19
20         def get_media_node(name, default=None):
21             media_name = 'media-%s' % name
22             media_group = item.get('media-group') or item
23             return media_group.get(media_name) or item.get(media_name) or item.get(name, default)
24
25         thumbnails = []
26         media_thumbnail = get_media_node('thumbnail')
27         if media_thumbnail:
28             if isinstance(media_thumbnail, dict):
29                 media_thumbnail = [media_thumbnail]
30             for thumbnail_data in media_thumbnail:
31                 thumbnail = thumbnail_data['@attributes']
32                 thumbnails.append({
33                     'url': self._proto_relative_url(thumbnail['url'], 'http:'),
34                     'width': int_or_none(thumbnail.get('width')),
35                     'height': int_or_none(thumbnail.get('height')),
36                 })
37
38         subtitles = {}
39         media_subtitle = get_media_node('subTitle')
40         if media_subtitle:
41             if isinstance(media_subtitle, dict):
42                 media_subtitle = [media_subtitle]
43             for subtitle_data in media_subtitle:
44                 subtitle = subtitle_data['@attributes']
45                 lang = subtitle.get('lang') or 'en'
46                 subtitles[lang] = [{'url': subtitle['href']}]
47
48         formats = []
49         media_content = get_media_node('content')
50         if isinstance(media_content, dict):
51             media_content = [media_content]
52         for media_data in media_content:
53             media = media_data['@attributes']
54             media_type = media['type']
55             if media_type == 'video/f4m':
56                 formats.extend(self._extract_f4m_formats(
57                     media['url'] + '?hdcore=3.4.0&plugin=aasp-3.4.0.132.124',
58                     video_id, f4m_id='hds', fatal=False))
59             elif media_type == 'application/x-mpegURL':
60                 formats.extend(self._extract_m3u8_formats(
61                     media['url'], video_id, 'mp4', m3u8_id='hls', fatal=False))
62             else:
63                 formats.append({
64                     'format_id': media_data['media-category']['@attributes']['label'],
65                     'url': media['url'],
66                     'tbr': int_or_none(media.get('bitrate')),
67                     'filesize': int_or_none(media.get('fileSize')),
68                 })
69
70         self._sort_formats(formats)
71
72         timestamp = parse_iso8601(item.get('pubDate'), ' ') or parse_iso8601(item.get('dc-date'))
73
74         return {
75             'id': video_id,
76             'title': get_media_node('title'),
77             'description': get_media_node('description'),
78             'thumbnails': thumbnails,
79             'timestamp': timestamp,
80             'duration': int_or_none(media_content[0].get('@attributes', {}).get('duration')),
81             'subtitles': subtitles,
82             'formats': formats,
83         }