[pornhub] Extract video URL from tv platform site (#12007, #12129)

[youtube-dl] / youtube_dl / extractor / pornhub.py

diff --git a/youtube_dl/extractor/pornhub.py b/youtube_dl/extractor/pornhub.py
index 0724efc09963d1f84fc7b24ec99d89b58e1bdfb6..7a2737032ff27a73825a7787feece5d52a92507d 100644 (file)
--- a/youtube_dl/extractor/pornhub.py
+++ b/youtube_dl/extractor/pornhub.py
@@ -2,27 +2,27 @@
 from __future__ import unicode_literals
 
 import itertools
-import os
+# import os
 import re
 
 from .common import InfoExtractor
 from ..compat import (
     compat_HTTPError,
-    compat_urllib_parse_unquote,
-    compat_urllib_parse_unquote_plus,
-    compat_urllib_parse_urlparse,
+    # compat_urllib_parse_unquote,
+    # compat_urllib_parse_unquote_plus,
+    # compat_urllib_parse_urlparse,
 )
 from ..utils import (
     ExtractorError,
     int_or_none,
     js_to_json,
     orderedSet,
-    sanitized_Request,
+    # sanitized_Request,
     str_to_int,
 )
-from ..aes import (
-    aes_decrypt_text
-)
+# from ..aes import (
+#     aes_decrypt_text
+# )
 
 
 class PornHubIE(InfoExtractor):
@@ -33,7 +33,7 @@ class PornHubIE(InfoExtractor):
                             (?:[a-z]+\.)?pornhub\.com/(?:view_video\.php\?viewkey=|embed/)|
                             (?:www\.)?thumbzilla\.com/video/
                         )
-                        (?P<id>[0-9a-z]+)
+                        (?P<id>[\da-z]+)
                     '''
     _TESTS = [{
         'url': 'http://www.pornhub.com/view_video.php?viewkey=648719015',
@@ -96,12 +96,11 @@ class PornHubIE(InfoExtractor):
         '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(
@@ -110,10 +109,14 @@ 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=(["\'])(?:(?!\1).)*\b(?:removed|userMessageSection)\b(?:(?!\1).)*\1[^>]*>(?P<error>.+?)</div>',
@@ -124,10 +127,19 @@ class PornHubIE(InfoExtractor):
                 '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 = self._html_search_meta(
+        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',
@@ -157,7 +169,19 @@ 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)))
+        """
+        video_variables = {}
+        for video_variablename, quote, video_variable in re.findall(
+                r'(player_quality_[0-9]{3,4}p\w+)\s*=\s*(["\'])(.+?)\2;', webpage):
+            video_variables[video_variablename] = video_variable
+
+        video_urls = []
+        for encoded_video_url in re.findall(
+                r'player_quality_[0-9]{3,4}p\s*=(.+?);', webpage):
+            for varname, varval in video_variables.items():
+                encoded_video_url = encoded_video_url.replace(varname, varval)
+            video_urls.append(re.sub(r'[\s+]', '', encoded_video_url))
+
         if webpage.find('"encrypted":true') != -1:
             password = compat_urllib_parse_unquote_plus(
                 self._search_regex(r'"video_title":"([^"]+)', webpage, 'password'))
@@ -187,6 +211,7 @@ class PornHubIE(InfoExtractor):
                 'height': height,
             })
         self._sort_formats(formats)
+        """
 
         page_params = self._parse_json(self._search_regex(
             r'page_params\.zoneDetails\[([\'"])[^\'"]+\1\]\s*=\s*(?P<data>{[^}]+})',
@@ -199,6 +224,7 @@ class PornHubIE(InfoExtractor):
 
         return {
             'id': video_id,
+            'url': video_url,
             'uploader': video_uploader,
             'title': title,
             'thumbnail': thumbnail,
@@ -207,7 +233,7 @@ class PornHubIE(InfoExtractor):
             'like_count': like_count,
             'dislike_count': dislike_count,
             'comment_count': comment_count,
-            'formats': formats,
+            # 'formats': formats,
             'age_limit': 18,
             'tags': tags,
             'categories': categories,
@@ -230,7 +256,14 @@ 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(
@@ -244,12 +277,12 @@ 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,
     }]