Merge branch 'master' of github.com:rg3/youtube-dl into weibo
[ytdl] / youtube_dl / extractor / turner.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5
6 from .adobepass import AdobePassIE
7 from ..compat import compat_str
8 from ..utils import (
9     xpath_text,
10     int_or_none,
11     determine_ext,
12     parse_duration,
13     xpath_attr,
14     update_url_query,
15     ExtractorError,
16     strip_or_none,
17 )
18
19
20 class TurnerBaseIE(AdobePassIE):
21     _AKAMAI_SPE_TOKEN_CACHE = {}
22
23     def _extract_timestamp(self, video_data):
24         return int_or_none(xpath_attr(video_data, 'dateCreated', 'uts'))
25
26     def _add_akamai_spe_token(self, tokenizer_src, video_url, content_id, ap_data):
27         secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*'
28         token = self._AKAMAI_SPE_TOKEN_CACHE.get(secure_path)
29         if not token:
30             query = {
31                 'path': secure_path,
32                 'videoId': content_id,
33             }
34             if ap_data.get('auth_required'):
35                 query['accessToken'] = self._extract_mvpd_auth(ap_data['url'], content_id, ap_data['site_name'], ap_data['site_name'])
36             auth = self._download_xml(
37                 tokenizer_src, content_id, query=query)
38             error_msg = xpath_text(auth, 'error/msg')
39             if error_msg:
40                 raise ExtractorError(error_msg, expected=True)
41             token = xpath_text(auth, 'token')
42             if not token:
43                 return video_url
44             self._AKAMAI_SPE_TOKEN_CACHE[secure_path] = token
45         return video_url + '?hdnea=' + token
46
47     def _extract_cvp_info(self, data_src, video_id, path_data={}, ap_data={}):
48         video_data = self._download_xml(data_src, video_id)
49         video_id = video_data.attrib['id']
50         title = xpath_text(video_data, 'headline', fatal=True)
51         content_id = xpath_text(video_data, 'contentId') or video_id
52         # rtmp_src = xpath_text(video_data, 'akamai/src')
53         # if rtmp_src:
54         #     splited_rtmp_src = rtmp_src.split(',')
55         #     if len(splited_rtmp_src) == 2:
56         #         rtmp_src = splited_rtmp_src[1]
57         # aifp = xpath_text(video_data, 'akamai/aifp', default='')
58
59         urls = []
60         formats = []
61         rex = re.compile(
62             r'(?P<width>[0-9]+)x(?P<height>[0-9]+)(?:_(?P<bitrate>[0-9]+))?')
63         # Possible formats locations: files/file, files/groupFiles/files
64         # and maybe others
65         for video_file in video_data.findall('.//file'):
66             video_url = video_file.text.strip()
67             if not video_url:
68                 continue
69             ext = determine_ext(video_url)
70             if video_url.startswith('/mp4:protected/'):
71                 continue
72                 # TODO Correct extraction for these files
73                 # protected_path_data = path_data.get('protected')
74                 # if not protected_path_data or not rtmp_src:
75                 #     continue
76                 # protected_path = self._search_regex(
77                 #     r'/mp4:(.+)\.[a-z0-9]', video_url, 'secure path')
78                 # auth = self._download_webpage(
79                 #     protected_path_data['tokenizer_src'], query={
80                 #         'path': protected_path,
81                 #         'videoId': content_id,
82                 #         'aifp': aifp,
83                 #     })
84                 # token = xpath_text(auth, 'token')
85                 # if not token:
86                 #     continue
87                 # video_url = rtmp_src + video_url + '?' + token
88             elif video_url.startswith('/secure/'):
89                 secure_path_data = path_data.get('secure')
90                 if not secure_path_data:
91                     continue
92                 video_url = self._add_akamai_spe_token(
93                     secure_path_data['tokenizer_src'],
94                     secure_path_data['media_src'] + video_url,
95                     content_id, ap_data)
96             elif not re.match('https?://', video_url):
97                 base_path_data = path_data.get(ext, path_data.get('default', {}))
98                 media_src = base_path_data.get('media_src')
99                 if not media_src:
100                     continue
101                 video_url = media_src + video_url
102             if video_url in urls:
103                 continue
104             urls.append(video_url)
105             format_id = video_file.get('bitrate')
106             if ext == 'smil':
107                 formats.extend(self._extract_smil_formats(
108                     video_url, video_id, fatal=False))
109             elif ext == 'm3u8':
110                 m3u8_formats = self._extract_m3u8_formats(
111                     video_url, video_id, 'mp4',
112                     m3u8_id=format_id or 'hls', fatal=False)
113                 if '/secure/' in video_url and '?hdnea=' in video_url:
114                     for f in m3u8_formats:
115                         f['_seekable'] = False
116                 formats.extend(m3u8_formats)
117             elif ext == 'f4m':
118                 formats.extend(self._extract_f4m_formats(
119                     update_url_query(video_url, {'hdcore': '3.7.0'}),
120                     video_id, f4m_id=format_id or 'hds', fatal=False))
121             else:
122                 f = {
123                     'format_id': format_id,
124                     'url': video_url,
125                     'ext': ext,
126                 }
127                 mobj = rex.search(format_id + video_url)
128                 if mobj:
129                     f.update({
130                         'width': int(mobj.group('width')),
131                         'height': int(mobj.group('height')),
132                         'tbr': int_or_none(mobj.group('bitrate')),
133                     })
134                 elif isinstance(format_id, compat_str):
135                     if format_id.isdigit():
136                         f['tbr'] = int(format_id)
137                     else:
138                         mobj = re.match(r'ios_(audio|[0-9]+)$', format_id)
139                         if mobj:
140                             if mobj.group(1) == 'audio':
141                                 f.update({
142                                     'vcodec': 'none',
143                                     'ext': 'm4a',
144                                 })
145                             else:
146                                 f['tbr'] = int(mobj.group(1))
147                 formats.append(f)
148         self._sort_formats(formats)
149
150         subtitles = {}
151         for source in video_data.findall('closedCaptions/source'):
152             for track in source.findall('track'):
153                 track_url = track.get('url')
154                 if not isinstance(track_url, compat_str) or track_url.endswith('/big'):
155                     continue
156                 lang = track.get('lang') or track.get('label') or 'en'
157                 subtitles.setdefault(lang, []).append({
158                     'url': track_url,
159                     'ext': {
160                         'scc': 'scc',
161                         'webvtt': 'vtt',
162                         'smptett': 'tt',
163                     }.get(source.get('format'))
164                 })
165
166         thumbnails = [{
167             'id': image.get('cut'),
168             'url': image.text,
169             'width': int_or_none(image.get('width')),
170             'height': int_or_none(image.get('height')),
171         } for image in video_data.findall('images/image')]
172
173         is_live = xpath_text(video_data, 'isLive') == 'true'
174
175         return {
176             'id': video_id,
177             'title': self._live_title(title) if is_live else title,
178             'formats': formats,
179             'subtitles': subtitles,
180             'thumbnails': thumbnails,
181             'thumbnail': xpath_text(video_data, 'poster'),
182             'description': strip_or_none(xpath_text(video_data, 'description')),
183             'duration': parse_duration(xpath_text(video_data, 'length') or xpath_text(video_data, 'trt')),
184             'timestamp': self._extract_timestamp(video_data),
185             'upload_date': xpath_attr(video_data, 'metas', 'version'),
186             'series': xpath_text(video_data, 'showTitle'),
187             'season_number': int_or_none(xpath_text(video_data, 'seasonNumber')),
188             'episode_number': int_or_none(xpath_text(video_data, 'episodeNumber')),
189             'is_live': is_live,
190         }