[wistia] reduce embed extraction false positives and support inline embeds(closes...
[youtube-dl] / youtube_dl / extractor / wistia.py
1 from __future__ import unicode_literals
2
3 import re
4
5 from .common import InfoExtractor
6 from ..utils import (
7     ExtractorError,
8     int_or_none,
9     float_or_none,
10     unescapeHTML,
11 )
12
13
14 class WistiaIE(InfoExtractor):
15     _VALID_URL = r'(?:wistia:|https?://(?:fast\.)?wistia\.(?:net|com)/embed/(?:iframe|medias)/)(?P<id>[a-z0-9]{10})'
16     _API_URL = 'http://fast.wistia.com/embed/medias/%s.json'
17     _IFRAME_URL = 'http://fast.wistia.net/embed/iframe/%s'
18
19     _TESTS = [{
20         'url': 'http://fast.wistia.net/embed/iframe/sh7fpupwlt',
21         'md5': 'cafeb56ec0c53c18c97405eecb3133df',
22         'info_dict': {
23             'id': 'sh7fpupwlt',
24             'ext': 'mov',
25             'title': 'Being Resourceful',
26             'description': 'a Clients From Hell Video Series video from worldwidewebhosting',
27             'upload_date': '20131204',
28             'timestamp': 1386185018,
29             'duration': 117,
30         },
31     }, {
32         'url': 'wistia:sh7fpupwlt',
33         'only_matching': True,
34     }, {
35         # with hls video
36         'url': 'wistia:807fafadvk',
37         'only_matching': True,
38     }, {
39         'url': 'http://fast.wistia.com/embed/iframe/sh7fpupwlt',
40         'only_matching': True,
41     }, {
42         'url': 'http://fast.wistia.net/embed/medias/sh7fpupwlt.json',
43         'only_matching': True,
44     }]
45
46     # https://wistia.com/support/embed-and-share/video-on-your-website
47     @staticmethod
48     def _extract_url(webpage):
49         match = re.search(
50             r'<(?:meta[^>]+?content|(?:iframe|script)[^>]+?src)=["\'](?P<url>(?:https?:)?//(?:fast\.)?wistia\.(?:net|com)/embed/(?:iframe|medias)/[a-z0-9]{10})', webpage)
51         if match:
52             return unescapeHTML(match.group('url'))
53
54         match = re.search(
55             r'''(?sx)
56                 <script[^>]+src=(["'])(?:https?:)?//fast\.wistia\.com/assets/external/E-v1\.js\1[^>]*>.*?
57                 <div[^>]+class=(["']).*?\bwistia_async_(?P<id>[a-z0-9]{10})\b.*?\2
58             ''', webpage)
59         if match:
60             return 'wistia:%s' % match.group('id')
61
62         match = re.search(r'(?:data-wistia-?id=["\']|Wistia\.embed\(["\']|id=["\']wistia_)(?P<id>[a-z0-9]{10})', webpage)
63         if match:
64             return 'wistia:%s' % match.group('id')
65
66     def _real_extract(self, url):
67         video_id = self._match_id(url)
68
69         data_json = self._download_json(
70             self._API_URL % video_id, video_id,
71             # Some videos require this.
72             headers={
73                 'Referer': url if url.startswith('http') else self._IFRAME_URL % video_id,
74             })
75
76         if data_json.get('error'):
77             raise ExtractorError(
78                 'Error while getting the playlist', expected=True)
79
80         data = data_json['media']
81         title = data['name']
82
83         formats = []
84         thumbnails = []
85         for a in data['assets']:
86             aurl = a.get('url')
87             if not aurl:
88                 continue
89             astatus = a.get('status')
90             atype = a.get('type')
91             if (astatus is not None and astatus != 2) or atype in ('preview', 'storyboard'):
92                 continue
93             elif atype in ('still', 'still_image'):
94                 thumbnails.append({
95                     'url': aurl,
96                     'width': int_or_none(a.get('width')),
97                     'height': int_or_none(a.get('height')),
98                 })
99             else:
100                 aext = a.get('ext')
101                 is_m3u8 = a.get('container') == 'm3u8' or aext == 'm3u8'
102                 formats.append({
103                     'format_id': atype,
104                     'url': aurl,
105                     'tbr': int_or_none(a.get('bitrate')),
106                     'vbr': int_or_none(a.get('opt_vbitrate')),
107                     'width': int_or_none(a.get('width')),
108                     'height': int_or_none(a.get('height')),
109                     'filesize': int_or_none(a.get('size')),
110                     'vcodec': a.get('codec'),
111                     'container': a.get('container'),
112                     'ext': 'mp4' if is_m3u8 else aext,
113                     'protocol': 'm3u8' if is_m3u8 else None,
114                     'preference': 1 if atype == 'original' else None,
115                 })
116
117         self._sort_formats(formats)
118
119         return {
120             'id': video_id,
121             'title': title,
122             'description': data.get('seoDescription'),
123             'formats': formats,
124             'thumbnails': thumbnails,
125             'duration': float_or_none(data.get('duration')),
126             'timestamp': int_or_none(data.get('createdAt')),
127         }