[pornhub] Fix extraction (closes #12007)
[youtube-dl] / youtube_dl / extractor / pornhub.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import itertools
5 import os
6 import re
7
8 from .common import InfoExtractor
9 from ..compat import (
10     compat_HTTPError,
11     compat_urllib_parse_unquote,
12     compat_urllib_parse_unquote_plus,
13     compat_urllib_parse_urlparse,
14 )
15 from ..utils import (
16     ExtractorError,
17     int_or_none,
18     js_to_json,
19     orderedSet,
20     sanitized_Request,
21     str_to_int,
22 )
23 from ..aes import (
24     aes_decrypt_text
25 )
26
27
28 class PornHubIE(InfoExtractor):
29     IE_DESC = 'PornHub and Thumbzilla'
30     _VALID_URL = r'''(?x)
31                     https?://
32                         (?:
33                             (?:[a-z]+\.)?pornhub\.com/(?:view_video\.php\?viewkey=|embed/)|
34                             (?:www\.)?thumbzilla\.com/video/
35                         )
36                         (?P<id>[\da-z]+)
37                     '''
38     _TESTS = [{
39         'url': 'http://www.pornhub.com/view_video.php?viewkey=648719015',
40         'md5': '1e19b41231a02eba417839222ac9d58e',
41         'info_dict': {
42             'id': '648719015',
43             'ext': 'mp4',
44             'title': 'Seductive Indian beauty strips down and fingers her pink pussy',
45             'uploader': 'Babes',
46             'duration': 361,
47             'view_count': int,
48             'like_count': int,
49             'dislike_count': int,
50             'comment_count': int,
51             'age_limit': 18,
52             'tags': list,
53             'categories': list,
54         },
55     }, {
56         # non-ASCII title
57         'url': 'http://www.pornhub.com/view_video.php?viewkey=1331683002',
58         'info_dict': {
59             'id': '1331683002',
60             'ext': 'mp4',
61             'title': '重庆婷婷女王足交',
62             'uploader': 'cj397186295',
63             'duration': 1753,
64             'view_count': int,
65             'like_count': int,
66             'dislike_count': int,
67             'comment_count': int,
68             'age_limit': 18,
69             'tags': list,
70             'categories': list,
71         },
72         'params': {
73             'skip_download': True,
74         },
75     }, {
76         'url': 'http://www.pornhub.com/view_video.php?viewkey=ph557bbb6676d2d',
77         'only_matching': True,
78     }, {
79         # removed at the request of cam4.com
80         'url': 'http://fr.pornhub.com/view_video.php?viewkey=ph55ca2f9760862',
81         'only_matching': True,
82     }, {
83         # removed at the request of the copyright owner
84         'url': 'http://www.pornhub.com/view_video.php?viewkey=788152859',
85         'only_matching': True,
86     }, {
87         # removed by uploader
88         'url': 'http://www.pornhub.com/view_video.php?viewkey=ph572716d15a111',
89         'only_matching': True,
90     }, {
91         # private video
92         'url': 'http://www.pornhub.com/view_video.php?viewkey=ph56fd731fce6b7',
93         'only_matching': True,
94     }, {
95         'url': 'https://www.thumbzilla.com/video/ph56c6114abd99a/horny-girlfriend-sex',
96         'only_matching': True,
97     }]
98
99     @staticmethod
100     def _extract_urls(webpage):
101         return re.findall(
102             r'<iframe[^>]+?src=["\'](?P<url>(?:https?:)?//(?:www\.)?pornhub\.com/embed/[\da-z]+)',
103             webpage)
104
105     def _extract_count(self, pattern, webpage, name):
106         return str_to_int(self._search_regex(
107             pattern, webpage, '%s count' % name, fatal=False))
108
109     def _real_extract(self, url):
110         video_id = self._match_id(url)
111
112         req = sanitized_Request(
113             'http://www.pornhub.com/view_video.php?viewkey=%s' % video_id)
114         req.add_header('Cookie', 'age_verified=1')
115         webpage = self._download_webpage(req, video_id)
116
117         error_msg = self._html_search_regex(
118             r'(?s)<div[^>]+class=(["\'])(?:(?!\1).)*\b(?:removed|userMessageSection)\b(?:(?!\1).)*\1[^>]*>(?P<error>.+?)</div>',
119             webpage, 'error message', default=None, group='error')
120         if error_msg:
121             error_msg = re.sub(r'\s+', ' ', error_msg)
122             raise ExtractorError(
123                 'PornHub said: %s' % error_msg,
124                 expected=True, video_id=video_id)
125
126         # video_title from flashvars contains whitespace instead of non-ASCII (see
127         # http://www.pornhub.com/view_video.php?viewkey=1331683002), not relying
128         # on that anymore.
129         title = self._html_search_meta(
130             'twitter:title', webpage, default=None) or self._search_regex(
131             (r'<h1[^>]+class=["\']title["\'][^>]*>(?P<title>[^<]+)',
132              r'<div[^>]+data-video-title=(["\'])(?P<title>.+?)\1',
133              r'shareTitle\s*=\s*(["\'])(?P<title>.+?)\1'),
134             webpage, 'title', group='title')
135
136         flashvars = self._parse_json(
137             self._search_regex(
138                 r'var\s+flashvars_\d+\s*=\s*({.+?});', webpage, 'flashvars', default='{}'),
139             video_id)
140         if flashvars:
141             thumbnail = flashvars.get('image_url')
142             duration = int_or_none(flashvars.get('video_duration'))
143         else:
144             title, thumbnail, duration = [None] * 3
145
146         video_uploader = self._html_search_regex(
147             r'(?s)From:&nbsp;.+?<(?:a href="/users/|a href="/channels/|span class="username)[^>]+>(.+?)<',
148             webpage, 'uploader', fatal=False)
149
150         view_count = self._extract_count(
151             r'<span class="count">([\d,\.]+)</span> views', webpage, 'view')
152         like_count = self._extract_count(
153             r'<span class="votesUp">([\d,\.]+)</span>', webpage, 'like')
154         dislike_count = self._extract_count(
155             r'<span class="votesDown">([\d,\.]+)</span>', webpage, 'dislike')
156         comment_count = self._extract_count(
157             r'All Comments\s*<span>\(([\d,.]+)\)', webpage, 'comment')
158
159         video_variables = {}
160         for video_variablename, quote, video_variable in re.findall(
161                 r'(player_quality_[0-9]{3,4}p[0-9a-z]+?)=\s*(["\'])(.*?)\2;', webpage):
162             video_variables[video_variablename] = video_variable
163
164         encoded_video_urls = []
165         for encoded_video_url in re.findall(
166                 r'player_quality_[0-9]{3,4}p\s*=(.*?);', webpage):
167             encoded_video_urls.append(encoded_video_url)
168
169         # Decode the URLs 
170         video_urls = []
171         for url in encoded_video_urls:
172             for varname, varval in video_variables.items():
173                 url = url.replace(varname, varval)
174             url = url.replace('+', '')
175             url = url.replace(' ', '')
176             video_urls.append(url)
177
178         if webpage.find('"encrypted":true') != -1:
179             password = compat_urllib_parse_unquote_plus(
180                 self._search_regex(r'"video_title":"([^"]+)', webpage, 'password'))
181             video_urls = list(map(lambda s: aes_decrypt_text(s, password, 32).decode('utf-8'), video_urls))
182
183         formats = []
184         for video_url in video_urls:
185             path = compat_urllib_parse_urlparse(video_url).path
186             extension = os.path.splitext(path)[1][1:]
187             format = path.split('/')[5].split('_')[:2]
188             format = '-'.join(format)
189
190             m = re.match(r'^(?P<height>[0-9]+)[pP]-(?P<tbr>[0-9]+)[kK]$', format)
191             if m is None:
192                 height = None
193                 tbr = None
194             else:
195                 height = int(m.group('height'))
196                 tbr = int(m.group('tbr'))
197
198             formats.append({
199                 'url': video_url,
200                 'ext': extension,
201                 'format': format,
202                 'format_id': format,
203                 'tbr': tbr,
204                 'height': height,
205             })
206         self._sort_formats(formats)
207
208         page_params = self._parse_json(self._search_regex(
209             r'page_params\.zoneDetails\[([\'"])[^\'"]+\1\]\s*=\s*(?P<data>{[^}]+})',
210             webpage, 'page parameters', group='data', default='{}'),
211             video_id, transform_source=js_to_json, fatal=False)
212         tags = categories = None
213         if page_params:
214             tags = page_params.get('tags', '').split(',')
215             categories = page_params.get('categories', '').split(',')
216
217         return {
218             'id': video_id,
219             'uploader': video_uploader,
220             'title': title,
221             'thumbnail': thumbnail,
222             'duration': duration,
223             'view_count': view_count,
224             'like_count': like_count,
225             'dislike_count': dislike_count,
226             'comment_count': comment_count,
227             'formats': formats,
228             'age_limit': 18,
229             'tags': tags,
230             'categories': categories,
231         }
232
233
234 class PornHubPlaylistBaseIE(InfoExtractor):
235     def _extract_entries(self, webpage):
236         return [
237             self.url_result(
238                 'http://www.pornhub.com/%s' % video_url,
239                 PornHubIE.ie_key(), video_title=title)
240             for video_url, title in orderedSet(re.findall(
241                 r'href="/?(view_video\.php\?.*\bviewkey=[\da-z]+[^"]*)"[^>]*\s+title="([^"]+)"',
242                 webpage))
243         ]
244
245     def _real_extract(self, url):
246         playlist_id = self._match_id(url)
247
248         webpage = self._download_webpage(url, playlist_id)
249
250         # Only process container div with main playlist content skipping
251         # drop-down menu that uses similar pattern for videos (see
252         # https://github.com/rg3/youtube-dl/issues/11594).
253         container = self._search_regex(
254             r'(?s)(<div[^>]+class=["\']container.+)', webpage,
255             'container', default=webpage)
256
257         entries = self._extract_entries(container)
258
259         playlist = self._parse_json(
260             self._search_regex(
261                 r'playlistObject\s*=\s*({.+?});', webpage, 'playlist'),
262             playlist_id)
263
264         return self.playlist_result(
265             entries, playlist_id, playlist.get('title'), playlist.get('description'))
266
267
268 class PornHubPlaylistIE(PornHubPlaylistBaseIE):
269     _VALID_URL = r'https?://(?:www\.)?pornhub\.com/playlist/(?P<id>\d+)'
270     _TESTS = [{
271         'url': 'http://www.pornhub.com/playlist/4667351',
272         'info_dict': {
273             'id': '4667351',
274             'title': 'Nataly Hot',
275         },
276         'playlist_mincount': 2,
277     }]
278
279
280 class PornHubUserVideosIE(PornHubPlaylistBaseIE):
281     _VALID_URL = r'https?://(?:www\.)?pornhub\.com/users/(?P<id>[^/]+)/videos'
282     _TESTS = [{
283         'url': 'http://www.pornhub.com/users/zoe_ph/videos/public',
284         'info_dict': {
285             'id': 'zoe_ph',
286         },
287         'playlist_mincount': 171,
288     }, {
289         'url': 'http://www.pornhub.com/users/rushandlia/videos',
290         'only_matching': True,
291     }]
292
293     def _real_extract(self, url):
294         user_id = self._match_id(url)
295
296         entries = []
297         for page_num in itertools.count(1):
298             try:
299                 webpage = self._download_webpage(
300                     url, user_id, 'Downloading page %d' % page_num,
301                     query={'page': page_num})
302             except ExtractorError as e:
303                 if isinstance(e.cause, compat_HTTPError) and e.cause.code == 404:
304                     break
305             page_entries = self._extract_entries(webpage)
306             if not page_entries:
307                 break
308             entries.extend(page_entries)
309
310         return self.playlist_result(entries, user_id)