[youtube] Fix extraction.
[youtube-dl] / youtube_dl / extractor / redtube.py
1 from __future__ import unicode_literals
2
3 import re
4
5 from .common import InfoExtractor
6 from ..utils import (
7     determine_ext,
8     ExtractorError,
9     int_or_none,
10     merge_dicts,
11     str_to_int,
12     unified_strdate,
13     url_or_none,
14 )
15
16
17 class RedTubeIE(InfoExtractor):
18     _VALID_URL = r'https?://(?:(?:\w+\.)?redtube\.com/|embed\.redtube\.com/\?.*?\bid=)(?P<id>[0-9]+)'
19     _TESTS = [{
20         'url': 'http://www.redtube.com/66418',
21         'md5': 'fc08071233725f26b8f014dba9590005',
22         'info_dict': {
23             'id': '66418',
24             'ext': 'mp4',
25             'title': 'Sucked on a toilet',
26             'upload_date': '20110811',
27             'duration': 596,
28             'view_count': int,
29             'age_limit': 18,
30         }
31     }, {
32         'url': 'http://embed.redtube.com/?bgcolor=000000&id=1443286',
33         'only_matching': True,
34     }, {
35         'url': 'http://it.redtube.com/66418',
36         'only_matching': True,
37     }]
38
39     @staticmethod
40     def _extract_urls(webpage):
41         return re.findall(
42             r'<iframe[^>]+?src=["\'](?P<url>(?:https?:)?//embed\.redtube\.com/\?.*?\bid=\d+)',
43             webpage)
44
45     def _real_extract(self, url):
46         video_id = self._match_id(url)
47         webpage = self._download_webpage(
48             'http://www.redtube.com/%s' % video_id, video_id)
49
50         ERRORS = (
51             (('video-deleted-info', '>This video has been removed'), 'has been removed'),
52             (('private_video_text', '>This video is private', '>Send a friend request to its owner to be able to view it'), 'is private'),
53         )
54
55         for patterns, message in ERRORS:
56             if any(p in webpage for p in patterns):
57                 raise ExtractorError(
58                     'Video %s %s' % (video_id, message), expected=True)
59
60         info = self._search_json_ld(webpage, video_id, default={})
61
62         if not info.get('title'):
63             info['title'] = self._html_search_regex(
64                 (r'<h(\d)[^>]+class="(?:video_title_text|videoTitle|video_title)[^"]*">(?P<title>(?:(?!\1).)+)</h\1>',
65                  r'(?:videoTitle|title)\s*:\s*(["\'])(?P<title>(?:(?!\1).)+)\1',),
66                 webpage, 'title', group='title',
67                 default=None) or self._og_search_title(webpage)
68
69         formats = []
70         sources = self._parse_json(
71             self._search_regex(
72                 r'sources\s*:\s*({.+?})', webpage, 'source', default='{}'),
73             video_id, fatal=False)
74         if sources and isinstance(sources, dict):
75             for format_id, format_url in sources.items():
76                 if format_url:
77                     formats.append({
78                         'url': format_url,
79                         'format_id': format_id,
80                         'height': int_or_none(format_id),
81                     })
82         medias = self._parse_json(
83             self._search_regex(
84                 r'mediaDefinition["\']?\s*:\s*(\[.+?}\s*\])', webpage,
85                 'media definitions', default='{}'),
86             video_id, fatal=False)
87         if medias and isinstance(medias, list):
88             for media in medias:
89                 format_url = url_or_none(media.get('videoUrl'))
90                 if not format_url:
91                     continue
92                 if media.get('format') == 'hls' or determine_ext(format_url) == 'm3u8':
93                     formats.extend(self._extract_m3u8_formats(
94                         format_url, video_id, 'mp4',
95                         entry_protocol='m3u8_native', m3u8_id='hls',
96                         fatal=False))
97                     continue
98                 format_id = media.get('quality')
99                 formats.append({
100                     'url': format_url,
101                     'format_id': format_id,
102                     'height': int_or_none(format_id),
103                 })
104         if not formats:
105             video_url = self._html_search_regex(
106                 r'<source src="(.+?)" type="video/mp4">', webpage, 'video URL')
107             formats.append({'url': video_url})
108         self._sort_formats(formats)
109
110         thumbnail = self._og_search_thumbnail(webpage)
111         upload_date = unified_strdate(self._search_regex(
112             r'<span[^>]+>(?:ADDED|Published on) ([^<]+)<',
113             webpage, 'upload date', default=None))
114         duration = int_or_none(self._og_search_property(
115             'video:duration', webpage, default=None) or self._search_regex(
116                 r'videoDuration\s*:\s*(\d+)', webpage, 'duration', default=None))
117         view_count = str_to_int(self._search_regex(
118             (r'<div[^>]*>Views</div>\s*<div[^>]*>\s*([\d,.]+)',
119              r'<span[^>]*>VIEWS</span>\s*</td>\s*<td>\s*([\d,.]+)',
120              r'<span[^>]+\bclass=["\']video_view_count[^>]*>\s*([\d,.]+)'),
121             webpage, 'view count', default=None))
122
123         # No self-labeling, but they describe themselves as
124         # "Home of Videos Porno"
125         age_limit = 18
126
127         return merge_dicts(info, {
128             'id': video_id,
129             'ext': 'mp4',
130             'thumbnail': thumbnail,
131             'upload_date': upload_date,
132             'duration': duration,
133             'view_count': view_count,
134             'age_limit': age_limit,
135             'formats': formats,
136         })