fix according to "https://github.com/rg3/youtube-dl/pull/15079#discussion_r158688607"
[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     def _extract_timestamp(self, video_data):
22         return int_or_none(xpath_attr(video_data, 'dateCreated', 'uts'))
23
24     def _extract_cvp_info(self, data_src, video_id, path_data={}, ap_data={}):
25         video_data = self._download_xml(data_src, video_id)
26         video_id = video_data.attrib['id']
27         title = xpath_text(video_data, 'headline', fatal=True)
28         content_id = xpath_text(video_data, 'contentId') or video_id
29         # rtmp_src = xpath_text(video_data, 'akamai/src')
30         # if rtmp_src:
31         #     splited_rtmp_src = rtmp_src.split(',')
32         #     if len(splited_rtmp_src) == 2:
33         #         rtmp_src = splited_rtmp_src[1]
34         # aifp = xpath_text(video_data, 'akamai/aifp', default='')
35
36         tokens = {}
37         urls = []
38         formats = []
39         rex = re.compile(
40             r'(?P<width>[0-9]+)x(?P<height>[0-9]+)(?:_(?P<bitrate>[0-9]+))?')
41         # Possible formats locations: files/file, files/groupFiles/files
42         # and maybe others
43         for video_file in video_data.findall('.//file'):
44             video_url = video_file.text.strip()
45             if not video_url:
46                 continue
47             ext = determine_ext(video_url)
48             if video_url.startswith('/mp4:protected/'):
49                 continue
50                 # TODO Correct extraction for these files
51                 # protected_path_data = path_data.get('protected')
52                 # if not protected_path_data or not rtmp_src:
53                 #     continue
54                 # protected_path = self._search_regex(
55                 #     r'/mp4:(.+)\.[a-z0-9]', video_url, 'secure path')
56                 # auth = self._download_webpage(
57                 #     protected_path_data['tokenizer_src'], query={
58                 #         'path': protected_path,
59                 #         'videoId': content_id,
60                 #         'aifp': aifp,
61                 #     })
62                 # token = xpath_text(auth, 'token')
63                 # if not token:
64                 #     continue
65                 # video_url = rtmp_src + video_url + '?' + token
66             elif video_url.startswith('/secure/'):
67                 secure_path_data = path_data.get('secure')
68                 if not secure_path_data:
69                     continue
70                 video_url = secure_path_data['media_src'] + video_url
71                 secure_path = self._search_regex(r'https?://[^/]+(.+/)', video_url, 'secure path') + '*'
72                 token = tokens.get(secure_path)
73                 if not token:
74                     query = {
75                         'path': secure_path,
76                         'videoId': content_id,
77                     }
78                     if ap_data.get('auth_required'):
79                         query['accessToken'] = self._extract_mvpd_auth(ap_data['url'], video_id, ap_data['site_name'], ap_data['site_name'])
80                     auth = self._download_xml(
81                         secure_path_data['tokenizer_src'], video_id, query=query)
82                     error_msg = xpath_text(auth, 'error/msg')
83                     if error_msg:
84                         raise ExtractorError(error_msg, expected=True)
85                     token = xpath_text(auth, 'token')
86                     if not token:
87                         continue
88                     tokens[secure_path] = token
89                 video_url = video_url + '?hdnea=' + token
90             elif not re.match('https?://', video_url):
91                 base_path_data = path_data.get(ext, path_data.get('default', {}))
92                 media_src = base_path_data.get('media_src')
93                 if not media_src:
94                     continue
95                 video_url = media_src + video_url
96             if video_url in urls:
97                 continue
98             urls.append(video_url)
99             format_id = video_file.get('bitrate')
100             if ext == 'smil':
101                 formats.extend(self._extract_smil_formats(
102                     video_url, video_id, fatal=False))
103             elif ext == 'm3u8':
104                 m3u8_formats = self._extract_m3u8_formats(
105                     video_url, video_id, 'mp4',
106                     m3u8_id=format_id or 'hls', fatal=False)
107                 if '/secure/' in video_url and '?hdnea=' in video_url:
108                     for f in m3u8_formats:
109                         f['_seekable'] = False
110                 formats.extend(m3u8_formats)
111             elif ext == 'f4m':
112                 formats.extend(self._extract_f4m_formats(
113                     update_url_query(video_url, {'hdcore': '3.7.0'}),
114                     video_id, f4m_id=format_id or 'hds', fatal=False))
115             else:
116                 f = {
117                     'format_id': format_id,
118                     'url': video_url,
119                     'ext': ext,
120                 }
121                 mobj = rex.search(format_id + video_url)
122                 if mobj:
123                     f.update({
124                         'width': int(mobj.group('width')),
125                         'height': int(mobj.group('height')),
126                         'tbr': int_or_none(mobj.group('bitrate')),
127                     })
128                 elif isinstance(format_id, compat_str):
129                     if format_id.isdigit():
130                         f['tbr'] = int(format_id)
131                     else:
132                         mobj = re.match(r'ios_(audio|[0-9]+)$', format_id)
133                         if mobj:
134                             if mobj.group(1) == 'audio':
135                                 f.update({
136                                     'vcodec': 'none',
137                                     'ext': 'm4a',
138                                 })
139                             else:
140                                 f['tbr'] = int(mobj.group(1))
141                 formats.append(f)
142         self._sort_formats(formats)
143
144         subtitles = {}
145         for source in video_data.findall('closedCaptions/source'):
146             for track in source.findall('track'):
147                 track_url = track.get('url')
148                 if not isinstance(track_url, compat_str) or track_url.endswith('/big'):
149                     continue
150                 lang = track.get('lang') or track.get('label') or 'en'
151                 subtitles.setdefault(lang, []).append({
152                     'url': track_url,
153                     'ext': {
154                         'scc': 'scc',
155                         'webvtt': 'vtt',
156                         'smptett': 'tt',
157                     }.get(source.get('format'))
158                 })
159
160         thumbnails = [{
161             'id': image.get('cut'),
162             'url': image.text,
163             'width': int_or_none(image.get('width')),
164             'height': int_or_none(image.get('height')),
165         } for image in video_data.findall('images/image')]
166
167         is_live = xpath_text(video_data, 'isLive') == 'true'
168
169         return {
170             'id': video_id,
171             'title': self._live_title(title) if is_live else title,
172             'formats': formats,
173             'subtitles': subtitles,
174             'thumbnails': thumbnails,
175             'thumbnail': xpath_text(video_data, 'poster'),
176             'description': strip_or_none(xpath_text(video_data, 'description')),
177             'duration': parse_duration(xpath_text(video_data, 'length') or xpath_text(video_data, 'trt')),
178             'timestamp': self._extract_timestamp(video_data),
179             'upload_date': xpath_attr(video_data, 'metas', 'version'),
180             'series': xpath_text(video_data, 'showTitle'),
181             'season_number': int_or_none(xpath_text(video_data, 'seasonNumber')),
182             'episode_number': int_or_none(xpath_text(video_data, 'episodeNumber')),
183             'is_live': is_live,
184         }