[Rte] Improve extractor
[youtube-dl] / youtube_dl / extractor / nfl.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5
6 from .common import InfoExtractor
7 from ..utils import (
8     ExtractorError,
9     compat_urllib_parse_urlparse,
10     int_or_none,
11     remove_end,
12 )
13
14
15 class NFLIE(InfoExtractor):
16     IE_NAME = 'nfl.com'
17     _VALID_URL = r'''(?x)https?://
18         (?P<host>(?:www\.)?(?:nfl\.com|.*?\.clubs\.nfl\.com))/
19         (?:.+?/)*
20         (?P<id>(?:\d[a-z]{2}\d{13}|\w{8}\-(?:\w{4}\-){3}\w{12}))'''
21     _TESTS = [
22         {
23             'url': 'http://www.nfl.com/videos/nfl-game-highlights/0ap3000000398478/Week-3-Redskins-vs-Eagles-highlights',
24             'md5': '394ef771ddcd1354f665b471d78ec4c6',
25             'info_dict': {
26                 'id': '0ap3000000398478',
27                 'ext': 'mp4',
28                 'title': 'Week 3: Redskins vs. Eagles highlights',
29                 'description': 'md5:56323bfb0ac4ee5ab24bd05fdf3bf478',
30                 'upload_date': '20140921',
31                 'timestamp': 1411337580,
32                 'thumbnail': 're:^https?://.*\.jpg$',
33             }
34         },
35         {
36             'url': 'http://prod.www.steelers.clubs.nfl.com/video-and-audio/videos/LIVE_Post_Game_vs_Browns/9d72f26a-9e2b-4718-84d3-09fb4046c266',
37             'md5': 'cf85bdb4bc49f6e9d3816d130c78279c',
38             'info_dict': {
39                 'id': '9d72f26a-9e2b-4718-84d3-09fb4046c266',
40                 'ext': 'mp4',
41                 'title': 'LIVE: Post Game vs. Browns',
42                 'description': 'md5:6a97f7e5ebeb4c0e69a418a89e0636e8',
43                 'upload_date': '20131229',
44                 'timestamp': 1388354455,
45                 'thumbnail': 're:^https?://.*\.jpg$',
46             }
47         }
48     ]
49
50     @staticmethod
51     def prepend_host(host, url):
52         if not url.startswith('http'):
53             if not url.startswith('/'):
54                 url = '/%s' % url
55             url = 'http://{0:}{1:}'.format(host, url)
56         return url
57
58     @staticmethod
59     def format_from_stream(stream, protocol, host, path_prefix='',
60                            preference=0, note=None):
61         url = '{protocol:}://{host:}/{prefix:}{path:}'.format(
62             protocol=protocol,
63             host=host,
64             prefix=path_prefix,
65             path=stream.get('path'),
66         )
67         return {
68             'url': url,
69             'vbr': int_or_none(stream.get('rate', 0), 1000),
70             'preference': preference,
71             'format_note': note,
72         }
73
74     def _real_extract(self, url):
75         mobj = re.match(self._VALID_URL, url)
76         video_id, host = mobj.group('id'), mobj.group('host')
77
78         webpage = self._download_webpage(url, video_id)
79
80         config_url = NFLIE.prepend_host(host, self._search_regex(
81             r'(?:config|configURL)\s*:\s*"([^"]+)"', webpage, 'config URL'))
82         config = self._download_json(config_url, video_id,
83                                      note='Downloading player config')
84         url_template = NFLIE.prepend_host(
85             host, '{contentURLTemplate:}'.format(**config))
86         video_data = self._download_json(
87             url_template.format(id=video_id), video_id)
88
89         formats = []
90         cdn_data = video_data.get('cdnData', {})
91         streams = cdn_data.get('bitrateInfo', [])
92         if cdn_data.get('format') == 'EXTERNAL_HTTP_STREAM':
93             parts = compat_urllib_parse_urlparse(cdn_data.get('uri'))
94             protocol, host = parts.scheme, parts.netloc
95             for stream in streams:
96                 formats.append(
97                     NFLIE.format_from_stream(stream, protocol, host))
98         else:
99             cdns = config.get('cdns')
100             if not cdns:
101                 raise ExtractorError('Failed to get CDN data', expected=True)
102
103             for name, cdn in cdns.items():
104                 # LimeLight streams don't seem to work
105                 if cdn.get('name') == 'LIMELIGHT':
106                     continue
107
108                 protocol = cdn.get('protocol')
109                 host = remove_end(cdn.get('host', ''), '/')
110                 if not (protocol and host):
111                     continue
112
113                 prefix = cdn.get('pathprefix', '')
114                 if prefix and not prefix.endswith('/'):
115                     prefix = '%s/' % prefix
116
117                 preference = 0
118                 if protocol == 'rtmp':
119                     preference = -2
120                 elif 'prog' in name.lower():
121                     preference = 1
122
123                 for stream in streams:
124                     formats.append(
125                         NFLIE.format_from_stream(stream, protocol, host,
126                                                  prefix, preference, name))
127
128         self._sort_formats(formats)
129
130         thumbnail = None
131         for q in ('xl', 'l', 'm', 's', 'xs'):
132             thumbnail = video_data.get('imagePaths', {}).get(q)
133             if thumbnail:
134                 break
135
136         return {
137             'id': video_id,
138             'title': video_data.get('headline'),
139             'formats': formats,
140             'description': video_data.get('caption'),
141             'duration': video_data.get('duration'),
142             'thumbnail': thumbnail,
143             'timestamp': int_or_none(video_data.get('posted'), 1000),
144         }