]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/pornhub.py
Merge tag 'upstream/2017.03.07'
[youtubedl] / youtube_dl / extractor / pornhub.py
index 5a55c25e78cf4b5ad95c05e5856f639d9bdd8617..9b413590a4078b9e962edc63912a85fbc8312523 100644 (file)
@@ -1,27 +1,40 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
-import os
+import itertools
+# import os
 import re
 
 from .common import InfoExtractor
 from ..compat import (
-    compat_urllib_parse_unquote,
-    compat_urllib_parse_unquote_plus,
-    compat_urllib_parse_urlparse,
+    compat_HTTPError,
+    # compat_urllib_parse_unquote,
+    # compat_urllib_parse_unquote_plus,
+    # compat_urllib_parse_urlparse,
 )
 from ..utils import (
     ExtractorError,
     int_or_none,
-    sanitized_Request,
+    js_to_json,
+    orderedSet,
+    # sanitized_Request,
     str_to_int,
 )
-from ..aes import (
-    aes_decrypt_text
-)
+from ..aes import (
+    aes_decrypt_text
+)
 
 
 class PornHubIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:[a-z]+\.)?pornhub\.com/(?:view_video\.php\?viewkey=|embed/)(?P<id>[0-9a-z]+)'
+    IE_DESC = 'PornHub and Thumbzilla'
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:
+                            (?:[a-z]+\.)?pornhub\.com/(?:view_video\.php\?viewkey=|embed/)|
+                            (?:www\.)?thumbzilla\.com/video/
+                        )
+                        (?P<id>[\da-z]+)
+                    '''
     _TESTS = [{
         'url': 'http://www.pornhub.com/view_video.php?viewkey=648719015',
         'md5': '1e19b41231a02eba417839222ac9d58e',
@@ -36,21 +49,58 @@ class PornHubIE(InfoExtractor):
             'dislike_count': int,
             'comment_count': int,
             'age_limit': 18,
-        }
+            'tags': list,
+            'categories': list,
+        },
+    }, {
+        # non-ASCII title
+        'url': 'http://www.pornhub.com/view_video.php?viewkey=1331683002',
+        'info_dict': {
+            'id': '1331683002',
+            'ext': 'mp4',
+            'title': '重庆婷婷女王足交',
+            'uploader': 'cj397186295',
+            'duration': 1753,
+            'view_count': int,
+            'like_count': int,
+            'dislike_count': int,
+            'comment_count': int,
+            'age_limit': 18,
+            'tags': list,
+            'categories': list,
+        },
+        'params': {
+            'skip_download': True,
+        },
     }, {
         'url': 'http://www.pornhub.com/view_video.php?viewkey=ph557bbb6676d2d',
         'only_matching': True,
     }, {
+        # removed at the request of cam4.com
         'url': 'http://fr.pornhub.com/view_video.php?viewkey=ph55ca2f9760862',
         'only_matching': True,
+    }, {
+        # removed at the request of the copyright owner
+        'url': 'http://www.pornhub.com/view_video.php?viewkey=788152859',
+        'only_matching': True,
+    }, {
+        # removed by uploader
+        'url': 'http://www.pornhub.com/view_video.php?viewkey=ph572716d15a111',
+        'only_matching': True,
+    }, {
+        # private video
+        'url': 'http://www.pornhub.com/view_video.php?viewkey=ph56fd731fce6b7',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.thumbzilla.com/video/ph56c6114abd99a/horny-girlfriend-sex',
+        'only_matching': True,
     }]
 
-    @classmethod
-    def _extract_url(cls, webpage):
-        mobj = re.search(
-            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?pornhub\.com/embed/\d+)\1', webpage)
-        if mobj:
-            return mobj.group('url')
+    @staticmethod
+    def _extract_urls(webpage):
+        return re.findall(
+            r'<iframe[^>]+?src=["\'](?P<url>(?:https?:)?//(?:www\.)?pornhub\.com/embed/[\da-z]+)',
+            webpage)
 
     def _extract_count(self, pattern, webpage, name):
         return str_to_int(self._search_regex(
@@ -59,33 +109,52 @@ class PornHubIE(InfoExtractor):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        req = sanitized_Request(
-            'http://www.pornhub.com/view_video.php?viewkey=%s' % video_id)
-        req.add_header('Cookie', 'age_verified=1')
-        webpage = self._download_webpage(req, video_id)
+        def dl_webpage(platform):
+            return self._download_webpage(
+                'http://www.pornhub.com/view_video.php?viewkey=%s' % video_id,
+                video_id, headers={
+                    'Cookie': 'age_verified=1; platform=%s' % platform,
+                })
+
+        webpage = dl_webpage('pc')
 
         error_msg = self._html_search_regex(
-            r'(?s)<div class="userMessageSection[^"]*".*?>(.*?)</div>',
-            webpage, 'error message', default=None)
+            r'(?s)<div[^>]+class=(["\'])(?:(?!\1).)*\b(?:removed|userMessageSection)\b(?:(?!\1).)*\1[^>]*>(?P<error>.+?)</div>',
+            webpage, 'error message', default=None, group='error')
         if error_msg:
             error_msg = re.sub(r'\s+', ' ', error_msg)
             raise ExtractorError(
                 'PornHub said: %s' % error_msg,
                 expected=True, video_id=video_id)
 
+        tv_webpage = dl_webpage('tv')
+
+        video_url = self._search_regex(
+            r'<video[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//.+?)\1', tv_webpage,
+            'video url', group='url')
+
+        title = self._search_regex(
+            r'<h1>([^>]+)</h1>', tv_webpage, 'title', default=None)
+
+        # video_title from flashvars contains whitespace instead of non-ASCII (see
+        # http://www.pornhub.com/view_video.php?viewkey=1331683002), not relying
+        # on that anymore.
+        title = title or self._html_search_meta(
+            'twitter:title', webpage, default=None) or self._search_regex(
+            (r'<h1[^>]+class=["\']title["\'][^>]*>(?P<title>[^<]+)',
+             r'<div[^>]+data-video-title=(["\'])(?P<title>.+?)\1',
+             r'shareTitle\s*=\s*(["\'])(?P<title>.+?)\1'),
+            webpage, 'title', group='title')
+
         flashvars = self._parse_json(
             self._search_regex(
-                r'var\s+flashv1ars_\d+\s*=\s*({.+?});', webpage, 'flashvars', default='{}'),
+                r'var\s+flashvars_\d+\s*=\s*({.+?});', webpage, 'flashvars', default='{}'),
             video_id)
         if flashvars:
-            video_title = flashvars.get('video_title')
             thumbnail = flashvars.get('image_url')
             duration = int_or_none(flashvars.get('video_duration'))
         else:
-            video_title, thumbnail, duration = [None] * 3
-
-        if not video_title:
-            video_title = self._html_search_regex(r'<h1 [^>]+>([^<]+)', webpage, 'title')
+            title, thumbnail, duration = [None] * 3
 
         video_uploader = self._html_search_regex(
             r'(?s)From:&nbsp;.+?<(?:a href="/users/|a href="/channels/|span class="username)[^>]+>(.+?)<',
@@ -100,58 +169,42 @@ class PornHubIE(InfoExtractor):
         comment_count = self._extract_count(
             r'All Comments\s*<span>\(([\d,.]+)\)', webpage, 'comment')
 
-        video_urls = list(map(compat_urllib_parse_unquote, re.findall(r"player_quality_[0-9]{3}p\s*=\s*'([^']+)'", webpage)))
-        if webpage.find('"encrypted":true') != -1:
-            password = compat_urllib_parse_unquote_plus(
-                self._search_regex(r'"video_title":"([^"]+)', webpage, 'password'))
-            video_urls = list(map(lambda s: aes_decrypt_text(s, password, 32).decode('utf-8'), video_urls))
-
-        formats = []
-        for video_url in video_urls:
-            path = compat_urllib_parse_urlparse(video_url).path
-            extension = os.path.splitext(path)[1][1:]
-            format = path.split('/')[5].split('_')[:2]
-            format = '-'.join(format)
-
-            m = re.match(r'^(?P<height>[0-9]+)[pP]-(?P<tbr>[0-9]+)[kK]$', format)
-            if m is None:
-                height = None
-                tbr = None
-            else:
-                height = int(m.group('height'))
-                tbr = int(m.group('tbr'))
-
-            formats.append({
-                'url': video_url,
-                'ext': extension,
-                'format': format,
-                'format_id': format,
-                'tbr': tbr,
-                'height': height,
-            })
-        self._sort_formats(formats)
+        page_params = self._parse_json(self._search_regex(
+            r'page_params\.zoneDetails\[([\'"])[^\'"]+\1\]\s*=\s*(?P<data>{[^}]+})',
+            webpage, 'page parameters', group='data', default='{}'),
+            video_id, transform_source=js_to_json, fatal=False)
+        tags = categories = None
+        if page_params:
+            tags = page_params.get('tags', '').split(',')
+            categories = page_params.get('categories', '').split(',')
 
         return {
             'id': video_id,
+            'url': video_url,
             'uploader': video_uploader,
-            'title': video_title,
+            'title': title,
             'thumbnail': thumbnail,
             'duration': duration,
             'view_count': view_count,
             'like_count': like_count,
             'dislike_count': dislike_count,
             'comment_count': comment_count,
-            'formats': formats,
+            'formats': formats,
             'age_limit': 18,
+            'tags': tags,
+            'categories': categories,
         }
 
 
 class PornHubPlaylistBaseIE(InfoExtractor):
     def _extract_entries(self, webpage):
         return [
-            self.url_result('http://www.pornhub.com/%s' % video_url, PornHubIE.ie_key())
-            for video_url in set(re.findall(
-                r'href="/?(view_video\.php\?.*\bviewkey=[\da-z]+[^"]*)"', webpage))
+            self.url_result(
+                'http://www.pornhub.com/%s' % video_url,
+                PornHubIE.ie_key(), video_title=title)
+            for video_url, title in orderedSet(re.findall(
+                r'href="/?(view_video\.php\?.*\bviewkey=[\da-z]+[^"]*)"[^>]*\s+title="([^"]+)"',
+                webpage))
         ]
 
     def _real_extract(self, url):
@@ -159,7 +212,14 @@ class PornHubPlaylistBaseIE(InfoExtractor):
 
         webpage = self._download_webpage(url, playlist_id)
 
-        entries = self._extract_entries(webpage)
+        # Only process container div with main playlist content skipping
+        # drop-down menu that uses similar pattern for videos (see
+        # https://github.com/rg3/youtube-dl/issues/11594).
+        container = self._search_regex(
+            r'(?s)(<div[^>]+class=["\']container.+)', webpage,
+            'container', default=webpage)
+
+        entries = self._extract_entries(container)
 
         playlist = self._parse_json(
             self._search_regex(
@@ -173,28 +233,43 @@ class PornHubPlaylistBaseIE(InfoExtractor):
 class PornHubPlaylistIE(PornHubPlaylistBaseIE):
     _VALID_URL = r'https?://(?:www\.)?pornhub\.com/playlist/(?P<id>\d+)'
     _TESTS = [{
-        'url': 'http://www.pornhub.com/playlist/6201671',
+        'url': 'http://www.pornhub.com/playlist/4667351',
         'info_dict': {
-            'id': '6201671',
-            'title': 'P0p4',
+            'id': '4667351',
+            'title': 'Nataly Hot',
         },
-        'playlist_mincount': 35,
+        'playlist_mincount': 2,
     }]
 
 
 class PornHubUserVideosIE(PornHubPlaylistBaseIE):
     _VALID_URL = r'https?://(?:www\.)?pornhub\.com/users/(?P<id>[^/]+)/videos'
     _TESTS = [{
-        'url': 'http://www.pornhub.com/users/rushandlia/videos',
+        'url': 'http://www.pornhub.com/users/zoe_ph/videos/public',
         'info_dict': {
-            'id': 'rushandlia',
+            'id': 'zoe_ph',
         },
-        'playlist_mincount': 13,
+        'playlist_mincount': 171,
+    }, {
+        'url': 'http://www.pornhub.com/users/rushandlia/videos',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
         user_id = self._match_id(url)
 
-        webpage = self._download_webpage(url, user_id)
-
-        return self.playlist_result(self._extract_entries(webpage), user_id)
+        entries = []
+        for page_num in itertools.count(1):
+            try:
+                webpage = self._download_webpage(
+                    url, user_id, 'Downloading page %d' % page_num,
+                    query={'page': page_num})
+            except ExtractorError as e:
+                if isinstance(e.cause, compat_HTTPError) and e.cause.code == 404:
+                    break
+            page_entries = self._extract_entries(webpage)
+            if not page_entries:
+                break
+            entries.extend(page_entries)
+
+        return self.playlist_result(entries, user_id)