]> git.bitcoin.ninja Git - youtube-dl/blob - youtube_dl/extractor/nba.py
[brightcove:new] extract subtitles and strip video title
[youtube-dl] / youtube_dl / extractor / nba.py
1 from __future__ import unicode_literals
2
3 import functools
4 import os.path
5 import re
6
7 from .common import InfoExtractor
8 from ..compat import (
9     compat_urllib_parse_urlencode,
10     compat_urlparse,
11 )
12 from ..utils import (
13     int_or_none,
14     OnDemandPagedList,
15     parse_duration,
16     remove_start,
17     xpath_text,
18     xpath_attr,
19 )
20
21
22 class NBAIE(InfoExtractor):
23     _VALID_URL = r'https?://(?:watch\.|www\.)?nba\.com/(?P<path>(?:[^/]+/)+(?P<id>[^?]*?))/?(?:/index\.html)?(?:\?.*)?$'
24     _TESTS = [{
25         'url': 'http://www.nba.com/video/games/nets/2012/12/04/0021200253-okc-bkn-recap.nba/index.html',
26         'md5': '9e7729d3010a9c71506fd1248f74e4f4',
27         'info_dict': {
28             'id': '0021200253-okc-bkn-recap',
29             'ext': 'mp4',
30             'title': 'Thunder vs. Nets',
31             'description': 'Kevin Durant scores 32 points and dishes out six assists as the Thunder beat the Nets in Brooklyn.',
32             'duration': 181,
33             'timestamp': 1354638466,
34             'upload_date': '20121204',
35         },
36         'params': {
37             # m3u8 download
38             'skip_download': True,
39         },
40     }, {
41         'url': 'http://www.nba.com/video/games/hornets/2014/12/05/0021400276-nyk-cha-play5.nba/',
42         'only_matching': True,
43     }, {
44         'url': 'http://watch.nba.com/video/channels/playoffs/2015/05/20/0041400301-cle-atl-recap.nba',
45         'md5': 'b2b39b81cf28615ae0c3360a3f9668c4',
46         'info_dict': {
47             'id': '0041400301-cle-atl-recap',
48             'ext': 'mp4',
49             'title': 'Hawks vs. Cavaliers Game 1',
50             'description': 'md5:8094c3498d35a9bd6b1a8c396a071b4d',
51             'duration': 228,
52             'timestamp': 1432134543,
53             'upload_date': '20150520',
54         }
55     }, {
56         'url': 'http://www.nba.com/clippers/news/doc-rivers-were-not-trading-blake',
57         'info_dict': {
58             'id': '1455672027478-Doc_Feb16_720',
59             'ext': 'mp4',
60             'title': 'Practice: Doc Rivers - 2/16/16',
61             'description': 'Head Coach Doc Rivers addresses the media following practice.',
62             'upload_date': '20160217',
63             'timestamp': 1455672000,
64         },
65         'params': {
66             # m3u8 download
67             'skip_download': True,
68         },
69     }, {
70         'url': 'http://www.nba.com/timberwolves/wiggins-shootaround#',
71         'info_dict': {
72             'id': 'timberwolves',
73             'title': 'Shootaround Access - Dec. 12 | Andrew Wiggins',
74         },
75         'playlist_count': 30,
76         'params': {
77             # Download the whole playlist takes too long time
78             'playlist_items': '1-30',
79         },
80     }, {
81         'url': 'http://www.nba.com/timberwolves/wiggins-shootaround#',
82         'info_dict': {
83             'id': 'Wigginsmp4',
84             'ext': 'mp4',
85             'title': 'Shootaround Access - Dec. 12 | Andrew Wiggins',
86             'description': 'Wolves rookie Andrew Wiggins addresses the media after Friday\'s shootaround.',
87             'upload_date': '20141212',
88             'timestamp': 1418418600,
89         },
90         'params': {
91             'noplaylist': True,
92             # m3u8 download
93             'skip_download': True,
94         },
95     }]
96
97     _PAGE_SIZE = 30
98
99     def _fetch_page(self, team, video_id, page):
100         search_url = 'http://searchapp2.nba.com/nba-search/query.jsp?' + compat_urllib_parse_urlencode({
101             'type': 'teamvideo',
102             'start': page * self._PAGE_SIZE + 1,
103             'npp': (page + 1) * self._PAGE_SIZE + 1,
104             'sort': 'recent',
105             'output': 'json',
106             'site': team,
107         })
108         results = self._download_json(
109             search_url, video_id, note='Download page %d of playlist data' % page)['results'][0]
110         for item in results:
111             yield self.url_result(compat_urlparse.urljoin('http://www.nba.com/', item['url']))
112
113     def _extract_playlist(self, orig_path, video_id, webpage):
114         team = orig_path.split('/')[0]
115
116         if self._downloader.params.get('noplaylist'):
117             self.to_screen('Downloading just video because of --no-playlist')
118             video_path = self._search_regex(
119                 r'nbaVideoCore\.firstVideo\s*=\s*\'([^\']+)\';', webpage, 'video path')
120             video_url = 'http://www.nba.com/%s/video/%s' % (team, video_path)
121             return self.url_result(video_url)
122
123         self.to_screen('Downloading playlist - add --no-playlist to just download video')
124         playlist_title = self._og_search_title(webpage, fatal=False)
125         entries = OnDemandPagedList(
126             functools.partial(self._fetch_page, team, video_id),
127             self._PAGE_SIZE, use_cache=True)
128
129         return self.playlist_result(entries, team, playlist_title)
130
131     def _real_extract(self, url):
132         path, video_id = re.match(self._VALID_URL, url).groups()
133         orig_path = path
134         if path.startswith('nba/'):
135             path = path[3:]
136
137         if 'video/' not in path:
138             webpage = self._download_webpage(url, video_id)
139             path = remove_start(self._search_regex(r'data-videoid="([^"]+)"', webpage, 'video id'), '/')
140
141             if path == '{{id}}':
142                 return self._extract_playlist(orig_path, video_id, webpage)
143
144             # See prepareContentId() of pkgCvp.js
145             if path.startswith('video/teams'):
146                 path = 'video/channels/proxy/' + path[6:]
147
148         video_info = self._download_xml('http://www.nba.com/%s.xml' % path, video_id)
149         video_id = os.path.splitext(xpath_text(video_info, 'slug'))[0]
150         title = xpath_text(video_info, 'headline')
151         description = xpath_text(video_info, 'description')
152         duration = parse_duration(xpath_text(video_info, 'length'))
153         timestamp = int_or_none(xpath_attr(video_info, 'dateCreated', 'uts'))
154
155         thumbnails = []
156         for image in video_info.find('images'):
157             thumbnails.append({
158                 'id': image.attrib.get('cut'),
159                 'url': image.text,
160                 'width': int_or_none(image.attrib.get('width')),
161                 'height': int_or_none(image.attrib.get('height')),
162             })
163
164         formats = []
165         for video_file in video_info.findall('.//file'):
166             video_url = video_file.text
167             if video_url.startswith('/'):
168                 continue
169             if video_url.endswith('.m3u8'):
170                 formats.extend(self._extract_m3u8_formats(video_url, video_id, ext='mp4', m3u8_id='hls', fatal=False))
171             elif video_url.endswith('.f4m'):
172                 formats.extend(self._extract_f4m_formats(video_url + '?hdcore=3.4.1.1', video_id, f4m_id='hds', fatal=False))
173             else:
174                 key = video_file.attrib.get('bitrate')
175                 format_info = {
176                     'format_id': key,
177                     'url': video_url,
178                 }
179                 mobj = re.search(r'(\d+)x(\d+)(?:_(\d+))?', key)
180                 if mobj:
181                     format_info.update({
182                         'width': int(mobj.group(1)),
183                         'height': int(mobj.group(2)),
184                         'tbr': int_or_none(mobj.group(3)),
185                     })
186                 formats.append(format_info)
187         self._sort_formats(formats)
188
189         return {
190             'id': video_id,
191             'title': title,
192             'description': description,
193             'duration': duration,
194             'timestamp': timestamp,
195             'thumbnails': thumbnails,
196             'formats': formats,
197         }