[XVideos] Support profiles, searches, channels and favorites

yt_dlp/extractor/_extractors.py

@@ -2200,7 +2200,13 @@
 from .xstream import XstreamIE
 from .xtube import XTubeUserIE, XTubeIE
 from .xuite import XuiteIE
-from .xvideos import XVideosIE
+from .xvideos import (
+    XVideosIE,
+    XVideosChannelIE,
+    XVideosPlaylistIE,
+    XVideosRelatedIE,
+    XVideosSearchIE,
+)
 from .xxxymovies import XXXYMoviesIE
 from .yahoo import (
     YahooIE,

yt_dlp/extractor/xvideos.py

@@ -1,37 +1,49 @@
+from __future__ import unicode_literals
+
 import re
+import itertools
 
 from .common import InfoExtractor
-from ..compat import compat_urllib_parse_unquote
+from ..compat import (
+    compat_parse_qs,
+    compat_str,
+    compat_urlparse,
+    compat_urllib_parse_unquote,
+    compat_urllib_parse_urlencode,
+)
 from ..utils import (
     clean_html,
     determine_ext,
+    extract_attributes,
     ExtractorError,
     int_or_none,
     parse_duration,
+    try_get,
+    url_basename,
+    urljoin,
 )
 
 
 class XVideosIE(InfoExtractor):
     _VALID_URL = r'''(?x)
                     https?://
                         (?:
-                            (?:[^/]+\.)?xvideos2?\.com/video|
+                            (?:[^/]+\.)?xvideos2?\.com/(?:video|prof-video-click/model/[^/]+/)|
                             (?:www\.)?xvideos\.es/video|
                             (?:www|flashservice)\.xvideos\.com/embedframe/|
                             static-hw\.xvideos\.com/swf/xv-player\.swf\?.*?\bid_video=
                         )
-                        (?P<id>[0-9]+)
-                    '''
+                        (?P<id>\d+)
+                 '''
     _TESTS = [{
-        'url': 'https://www.xvideos.com/video4588838/motorcycle_guy_cucks_influencer_steals_his_gf',
+        'url': 'http://www.xvideos.com/video4588838/biker_takes_his_girl',
         'md5': '14cea69fcb84db54293b1e971466c2e1',
         'info_dict': {
             'id': '4588838',
             'ext': 'mp4',
-            'title': 'Motorcycle Guy Cucks Influencer, Steals his GF',
+            'title': 'Biker Takes his Girl',
             'duration': 108,
             'age_limit': 18,
-            'thumbnail': r're:^https://img-hw.xvideos-cdn.com/.+\.jpg',
         }
     }, {
         # Broken HLS formats
@@ -48,9 +60,6 @@ class XVideosIE(InfoExtractor):
     }, {
         'url': 'https://flashservice.xvideos.com/embedframe/4588838',
         'only_matching': True,
-    }, {
-        'url': 'https://www.xvideos.com/embedframe/4588838',
-        'only_matching': True,
     }, {
         'url': 'http://static-hw.xvideos.com/swf/xv-player.swf?id_video=4588838',
         'only_matching': True,
@@ -92,9 +101,17 @@ class XVideosIE(InfoExtractor):
         'only_matching': True
     }]
 
+    @classmethod
+    def suitable(cls, url):
+        EXCLUDE_IE = (XVideosRelatedIE, )
+        return (False if any(ie.suitable(url) for ie in EXCLUDE_IE)
+                else super(XVideosIE, cls).suitable(url))
+
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id)
+
+        webpage = self._download_webpage(
+            'https://www.xvideos.com/video%s/0' % video_id, video_id)
 
         mobj = re.search(r'<h1 class="inlineError">(.+?)</h1>', webpage)
         if mobj:
@@ -159,3 +176,215 @@ def _real_extract(self, url):
             'thumbnails': thumbnails,
             'age_limit': 18,
         }
+
+
+class XVideosPlaylistIE(InfoExtractor):
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:[^/]+\.)?xvideos2?\.com/
+                          (?:c(?:/[sm]:[^/]+)*|
+                             profiles|
+                             favorite)/
+                            (?P<id>[^#?/]+)
+                  '''
+    _TESTS = []
+
+    def _extract_videos_from_json_list(self, json_list, path='video'):
+        return (
+            'https://www.xvideos.com/%s%d' % (path, x.get('id'), )
+            for x in json_list if isinstance(x, dict))
+
+    def _get_playlist_url(self, url, playlist_id):
+        """URL of first playlist page"""
+        id_match = re.match(self._VALID_URL, url).groupdict()
+        video_sort = id_match.get('sort')
+        if video_sort:
+            url, _ = compat_urlparse.urldefrag(url)
+            if url.endswith('/'):
+                url = url[:-1]
+            url = '%s/%s' % (url, video_sort.replace('-', '/'))
+        return url
+
+    def _get_next_page(self, url, num, page):
+        '''URL of num'th continuation page of url'''
+        if page.startswith('{'):
+            url, sub = re.subn(r'(/)(\d{1,7})($|[#?/])', r'\g<1>%d\3' % (num, ), url)
+            if sub == 0:
+                url += '/%d' % (num, )
+            return url
+        next_page = self._search_regex(
+            r'''(?s)(<a\s[^>]*?\bclass\s*=\s*(?P<q>'|").*?\bnext-page\b.*?(?P=q)[^>]*?>)''',
+            page, 'next page', default=None)
+        if next_page:
+            next_page = extract_attributes(next_page)
+            next_page = next_page.get('href')
+            if next_page:
+                return urljoin(url, next_page)
+        return False
+
+    def _extract_videos(self, url, playlist_id, num, page):
+        """Get iterable videos plus stop flag"""
+        return ((
+            'https://www.xvideos.com/video' + x.group('video_id')
+            for x in re.finditer(r'''<div\s[^>]*?id\s*=\s*(\'|")video_(?P<video_id>[0-9]+)\1''', page)),
+            None)
+
+    def _real_extract(self, url):
+        id_match = re.match(self._VALID_URL, url).groupdict()
+
+        playlist_id = id_match['id']
+        if url.endswith(playlist_id):
+            url += '/0'
+        next_page = self._get_playlist_url(url, playlist_id)
+        matches = []
+        for count in itertools.count(0):
+            webpage = self._download_webpage(
+                next_page,
+                '%s (+%d)' % (playlist_id, count) if count > 0 else playlist_id)
+
+            vids, stop = self._extract_videos(next_page, playlist_id, count, webpage)
+
+            if vids:
+                matches.append(vids)
+
+            if stop:
+                break
+            next_page = self._get_next_page(next_page, count + 1, webpage)
+            if not next_page:
+                break
+
+        return self.playlist_from_matches(
+            itertools.chain.from_iterable(matches), playlist_id)
+
+
+class XVideosRelatedIE(XVideosPlaylistIE):
+    _VALID_URL = XVideosIE._VALID_URL + r'(?:/[^/]+)*?\#_related-(?P<related>videos|playlists)'
+
+    _TESTS = []
+
+    def _extract_videos(self, url, playlist_id, num, page):
+        id_match = re.match(self._VALID_URL, url).groupdict()
+        related = id_match.get('related')
+        if not related:
+            return super(XVideosRelatedIE, self)._extract_videos(url, playlist_id, num, page)
+
+        if related == 'videos':
+            related_json = self._search_regex(
+                r'(?s)videos_related\s*=\s*(\[.*?])\s*;',
+                page, 'related', default='[]')
+            related_json = self._parse_json(related_json, playlist_id, fatal=False) or []
+            return (self._extract_videos_from_json_list(related_json), True)
+        # playlists
+        related_json = self._download_json(
+            'https://www.xvideos.com/video-playlists/' + playlist_id, playlist_id, fatal=False)
+        return (
+            self._extract_videos_from_json_list(
+                try_get(related_json, lambda x: x['playlists'], list) or [],
+                path='favorite/'),
+            True)
+
+
+class XVideosChannelIE(XVideosPlaylistIE):
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:[^/]+\.)?xvideos2?\.com/
+                          (?:
+                             (?:amateur-|pornstar-|model-)?channel|profile|
+                             pornstar
+                          )s/
+                            (?P<id>[^#?/]+)
+                              (?:\#_tab(?P<tab>Videos|Favorites|Playlists|AboutMe)(?:,(?P<sort>[^,]+))?)?
+                 '''
+    _TESTS = [{
+        'url': 'https://www.xvideos.com/pornstar-channels/sienna-west',
+        'playlist_mincount': 5,
+    }, ]
+
+    def _get_playlist_url(self, url, playlist_id):
+        webpage = self._download_webpage(url, playlist_id)
+        id_match = re.match(self._VALID_URL, url).groupdict()
+        tab = (id_match.get('tab') or '').lower()
+        if tab:
+            if tab in ('videos', 'favorites'):
+                url, frag = compat_urlparse.urldefrag(url)
+                if not url.endswith('/'):
+                    url += '/'
+                frag = frag.split(',')
+                url += tab
+                if tab == 'videos':
+                    url += '/' + (frag[1] if len(frag) > 1 else 'best')
+                url += '/0'
+            return url
+
+        # activity
+        conf = self._search_regex(
+            r'(?s)\.\s*xv\s*\.\s*conf\s*=\s*(\{.*?})[\s;]*</script',
+            webpage, 'XV conf')
+        conf = self._parse_json(conf, playlist_id)
+        act = try_get(conf,
+                      ((lambda x: x['dyn'][y])
+                       for y in ('page_main_cat', 'user_main_cat')),
+                      compat_str) or 'straight'
+
+        url, _ = compat_urlparse.urldefrag(url)
+        if url.endswith('/'):
+            url = url[:-1]
+
+        return '%s/activity/%s' % (url, act, )
+
+    def _get_next_page(self, url, num, page):
+        if page.startswith('{') or '#_tab' in url:
+            return super(XVideosChannelIE, self)._get_next_page(url, num, page)
+
+        act_time = int_or_none(url_basename(url)) or 0
+        last_act = int(self._search_regex(
+            r'(?s)id\s*=\s*"?activity-event-(\d{10})(?!.*id\s*=\s*"?activity-event-\d+.*).+$',
+            page, 'last activity', default=act_time))
+        if last_act == act_time:
+            return False
+        return (
+            url.replace('/%d' % (act_time, ), '/%d' % (last_act, ))
+            if act_time
+            else url + ('/%d' % (last_act, )))
+
+    def _extract_videos(self, url, playlist_id, num, page):
+        tab = next((x for x in ('videos', 'favorites') if '/%s/' % (x, ) in url), None)
+        if tab == 'videos':
+            tab_json = self._parse_json(page, playlist_id, fatal=False) or {}
+            more = try_get(tab_json, lambda x: x['current_page'] + 1, int)
+            more = int_or_none(more, scale=tab_json.get('nb_videos'), invscale=tab_json.get('nb_per_page'), default=0)
+            return (
+                self._extract_videos_from_json_list(
+                    try_get(tab_json, lambda x: x['videos'], list) or []),
+                more > 0)
+
+        if tab == 'favorites':
+            return ((
+                'https://www.xvideos.com' + x.group('playlist')
+                for x in re.finditer(r'''<a\s[^>]*?href\s*=\s*('|")(?P<playlist>/favorite/\d+/[^#?]+?)\1''', page)),
+                None)
+
+        return super(XVideosChannelIE, self)._extract_videos(url, playlist_id, num, page)
+
+
+class XVideosSearchIE(XVideosPlaylistIE):
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:[^/]+\.)?xvideos2?\.com/
+                          \?k=(?P<id>[^#?/&]+)
+                 '''
+    _TESTS = [{
+        # uninteresting search with probably at least two pages of results,
+        # but not too many more
+        'url': 'http://www.xvideos.com/?k=libya&sort=length',
+        'playlist_mincount': 30,
+    }, ]
+
+    def _get_next_page(self, url, num, page):
+        parsed_url = compat_urlparse.urlparse(url)
+        qs = compat_parse_qs(parsed_url.query)
+        qs['p'] = [num]
+        parsed_url = (
+            list(parsed_url[:4])
+            + [compat_urllib_parse_urlencode(qs, True), None])
+        return compat_urlparse.urlunparse(parsed_url), False