Merge d7c2b5dac8 into 668332b973

* YouPornCategoryIE
* YouPornChannelIE
* YouPornCollectionIE
* YouPornStarIE
* YouPornTagIE
* YouPornVideosIE,

test/test_download.py

@@ -10,6 +10,7 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 from test.helper import (
     assertGreaterEqual,
+    assertLessEqual,
     expect_warnings,
     get_params,
     gettestcases,
@@ -122,7 +123,10 @@ def generator(test_case, tname):
         params['outtmpl'] = tname + '_' + params['outtmpl']
         if is_playlist and 'playlist' not in test_case:
             params.setdefault('extract_flat', 'in_playlist')
-            params.setdefault('playlistend', test_case.get('playlist_mincount'))
+            params.setdefault('playlistend',
+                              test_case['playlist_maxcount'] + 1
+                              if test_case.get('playlist_maxcount')
+                              else test_case.get('playlist_mincount'))
             params.setdefault('skip_download', True)
 
         ydl = YoutubeDL(params, auto_init=False)
@@ -190,6 +194,14 @@ def generator(test_case, tname):
                     'Expected at least %d in playlist %s, but got only %d' % (
                         test_case['playlist_mincount'], test_case['url'],
                         len(res_dict['entries'])))
+            if 'playlist_maxcount' in test_case:
+                assertLessEqual(
+                    self,
+                    len(res_dict['entries']),
+                    test_case['playlist_maxcount'],
+                    'Expected at most %d in playlist %s, but got %d' % (
+                        test_case['playlist_maxcount'], test_case['url'],
+                        len(res_dict['entries'])))
             if 'playlist_count' in test_case:
                 self.assertEqual(
                     len(res_dict['entries']),

youtube_dl/extractor/ardaudiothek.py

@@ -0,0 +1,439 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+try:
+    from urllib.parse import unquote as _unquote_compat
+except ImportError:
+    from urllib import unquote
+
+    def _unquote_compat(str):
+        return unquote(str.encode('utf-8')).decode('utf-8')
+
+from .common import InfoExtractor
+from ..utils import (
+    compat_str,
+    ExtractorError,
+    int_or_none,
+    parse_duration,
+    str_or_none,
+    try_get,
+    unified_strdate,
+    unified_timestamp,
+)
+
+
+class ARDAudiothekBaseIE(InfoExtractor):
+
+    def _extract_episode_info(self, title):
+        """Try to extract episode data from the title."""
+        res = {}
+        if not title:
+            return res
+
+        for pattern in [
+            r'.*(?P<ep_info> \(S(?P<season_number>\d+)/E(?P<episode_number>\d+)\)).*',
+            r'.*(?P<ep_info> \((?:Folge |Teil )?(?P<episode_number>\d+)(?:/\d+)?\)).*',
+            r'.*(?P<ep_info>Folge (?P<episode_number>\d+)(?:\:| -|) )\"(?P<episode>.+)\".*',
+            r'.*(?P<ep_info>Folge (?P<episode_number>\d+)(?:/\d+)?(?:\:| -|) ).*',
+        ]:
+            m = re.match(pattern, title)
+            if m:
+                groupdict = m.groupdict()
+                for int_entry in ['season_number', 'episode_number']:
+                    res[int_entry] = int_or_none(groupdict.get(int_entry))
+
+                for str_entry in ['episode']:
+                    res[str_entry] = str_or_none(groupdict.get(str_entry))
+
+                # Build the episode title by removing numeric episode
+                # information.
+                if groupdict.get('ep_info') and not res['episode']:
+                    res['episode'] = str_or_none(
+                        title.replace(groupdict.get('ep_info'), ''))
+
+                if res['episode']:
+                    res['episode'] = res['episode'].strip()
+
+                break
+
+        # As a fallback use the whole title as the episode name
+        if not res.get('episode'):
+            res['episode'] = title.strip()
+
+        return res
+
+    def _extract_id_title_desc(self, json_data):
+        res = {
+            'id': try_get(json_data, lambda x: x['id'], compat_str),
+            'display_id': try_get(json_data, lambda x: x['slug'], compat_str),
+        }
+        res['title'] = try_get(
+            json_data, lambda x: x['title'], compat_str)
+        res['description'] = try_get(
+            json_data, lambda x: x['summary'], compat_str)
+        return res
+
+    def _extract_episode(self, ep_data):
+        res = self._extract_id_title_desc(ep_data)
+
+        res['url'] = try_get(ep_data, [
+            lambda x: x['enclosure']['download_url'],
+            lambda x: x['enclosure']['playback_url'],
+            lambda x: x['guid'],
+        ], compat_str)
+        if not res['url']:
+            raise ExtractorError(msg='Could not find a URL to download',
+                                 expected=True)
+
+        res['format_note'] = try_get(
+            ep_data, lambda x: x['enclosure']['type'], compat_str)
+        res['duration'] = parse_duration(
+            try_get(ep_data, lambda x: x['duration'], compat_str))
+        res['release_date'] = unified_strdate(
+            try_get(ep_data, lambda x: x['publication_date'], compat_str))
+        res['timestamp'] = unified_timestamp(
+            try_get(ep_data, lambda x: x['publication_date'], compat_str))
+        res['channel'] = try_get(ep_data, [
+            lambda x: x['podcast']['station'],
+            lambda x: x['podcast']['organization_name'],
+        ], compat_str)
+
+        # 'sharing_url' might be a redirecting URL. The generic extractor will
+        # handle the redirection just fine, so that this extractor here will
+        # be used.
+        res['webpage_url'] = try_get(
+            ep_data, lambda x: x['sharing_url'], compat_str)
+
+        res['categories'] = [
+            try_get(ep_data, lambda x: x['podcast']['category'], compat_str),
+        ]
+
+        res['is_live'] = False
+
+        res['series'] = try_get(ep_data,
+                                lambda x: x['podcast']['title'],
+                                compat_str)
+
+        def make_thumbnail(url, id, preference):
+            # Note that the images don't necessarily have the advertised
+            # aspect ratio! So don't set the height based on the aspect
+            # ratio.
+            # Also note that the server will not return an image of any given
+            # width. Most multiples of 32 (or of 64 for higher numbers) seem to
+            # work. When requesting a width of 1080, the server returns an
+            # image with a width of 1024, for instance. Requesting 1400 gives
+            # us 1344, and so on. So a width of 1920 works best for both 1x1
+            # and 16x9 images.
+            thumb_width = 1920
+            return {
+                'id': id,
+                # Only set the width if we actually replace the {width}
+                # placeholder in the URL.
+                'width': thumb_width if '{width}' in url else None,
+                'url': url.replace('{width}', str(thumb_width)),
+                'preference': preference,
+            }
+
+        # We prefer 1x1 images and we prefer episode images. But still provide
+        # all available images so that the user can choose. We use the
+        # thumbnail's 'preference' entry to sort them (the higher the better).
+        # The preferred thumbnail order is:
+        #     (0) podcast-16x9 < (1) episode-16x9
+        #   < (2) podcast-1x1 < (3) episode-1x1
+        thumbnails = []
+        for ar_index, aspect_ratio in enumerate(['16x9', '1x1']):
+            image_key = 'image_%s' % aspect_ratio
+            image_sources = [
+                {'name': 'podcast',
+                 'access': lambda x: x['podcast'][image_key]},
+                {'name': 'episode',
+                 'access': lambda x: x[image_key]},
+            ]
+            for src_index, src in enumerate(image_sources):
+                thumb_url = try_get(ep_data, src['access'], compat_str)
+
+                if thumb_url:
+                    thumbnails.append(make_thumbnail(
+                        thumb_url,
+                        src['name'] + '-' + aspect_ratio,
+                        ar_index * len(image_sources) + src_index))
+        res['thumbnails'] = thumbnails
+
+        res.update(self._extract_episode_info(res.get('title')))
+
+        return res
+
+
+class ARDAudiothekIE(ARDAudiothekBaseIE):
+    _VALID_URL = r'https?://(?:www\.|beta\.)?ardaudiothek\.de/(?:[^/]+)/(?:[^/]+)/(?P<id>[0-9]+)(?:/.*)?'
+    _TESTS = [{
+        'url': 'https://www.ardaudiothek.de/hoerspiel-pool/virginia-woolf-zum-leuchtturm-1-3-die-tuer-aus-glas/53728640',
+        'md5': 'dc12a86bb46faadbdba7a8c9b5a24246',
+        'info_dict': {
+            'id': '53728640',
+            'ext': 'mp3',
+            'title': 'Virginia Woolf: Zum Leuchtturm (1/3) - Die Tür aus Glas',
+            'description': r're:^Am Anfang steht die Frage.*',
+            'thumbnail': compat_str,
+            'timestamp': 1478818860,
+            'upload_date': '20161110',
+        }
+    }, {
+        'url': 'https://www.ardaudiothek.de/eine-stunde-talk/soziologe-matthias-quent-nicht-neutral-gegenueber-rechtsradikalismus/65904422',
+        'md5': '326065e45e8172124165c3b0addd4553',
+        'info_dict': {
+            'id': '65904422',
+            'ext': 'mp3',
+            'title': 'Soziologe Matthias Quent - Nicht neutral gegenüber Rechtsradikalismus',
+            'description': r're:^Matthias Quent erforscht die Ziele.*',
+            'thumbnail': compat_str,
+            'timestamp': 1565809200,
+            'upload_date': '20190814',
+        }
+    }]
+
+    def _real_extract(self, url):
+        episode_id = self._match_id(url)
+
+        api_url = 'https://www.ardaudiothek.de/api/episodes/%s' % episode_id
+        result_data = self._download_json(api_url, episode_id, fatal=False)
+        ep_data = try_get(result_data, lambda x: x['result']['episode'], dict)
+
+        if not ep_data:
+            raise ExtractorError(msg="Could not find any episode data",
+                                 expected=True)
+
+        return self._extract_episode(ep_data)
+
+
+class ARDAudiothekPlaylistIE(ARDAudiothekBaseIE):
+    _VALID_URL = r'https?://(?:www\.|beta\.)?ardaudiothek\.de/(?!kategorie)(?:[^/]+)/(?P<id>[0-9]+)(?:/.*)?'
+    _TESTS = [{
+        'url': 'https://www.ardaudiothek.de/wirtschaft/62037362',
+        'info_dict': {
+            'id': '62037362',
+            'title': 'Wirtschaft',
+            'description': compat_str,
+        },
+        'playlist_mincount': 5,
+    }, {
+        'url': 'https://www.ardaudiothek.de/redezeit/7852070',
+        'info_dict': {
+            'id': '7852070',
+            'title': 'Redezeit',
+            'description': compat_str,
+        },
+        'playlist_mincount': 5,
+    }, {
+        'url': 'https://www.ardaudiothek.de/nur-fuer-starke-nerven-krimis-und-thriller/51581650/alle',
+        'info_dict': {
+            'id': '51581650',
+            'title': r're:^Nur für starke Nerven',
+            'description': compat_str,
+        },
+        'playlist_mincount': 5,
+    }]
+
+    def _get_page_str(self, page):
+        # The API sometimes returns 404s for page=1. So only add that
+        # parameter if we actually are past the first page
+        return '&page=' + compat_str(page) if page > 1 else ''
+
+    def _get_episode_from_array_entry(self, array_entry):
+        # The array entry already is a an 'episode' dict.
+        return array_entry
+
+    def _extract_episodes(
+            self, display_id, api_url_template, default_items_per_page):
+        """
+        Extract episodes by calling a web API end point.
+
+        Sometimes the server does not respond properly when requesting a page.
+        This also happens on the website. It sometimes hangs when trying to
+        load more search results, for instance. Thus the number of entries
+        reported by the API is often wrong and we do not solely rely on that
+        number to stop reading episodes.
+
+        This function handles paginated content in a robust way by skipping
+        over faulty server responses. In this case it reduces the page size to
+        get as many episodes as possible. It also removes duplicate entries
+        from the result.
+
+        Args:
+            display_id: Only used for user feedback.
+            api_url_template: This is the URL of the API to download JSON data
+                from. It is a format string expected to have the following
+                fields:
+                    - {items_per_page}
+                    - {page_str}
+            default_items_per_page: The number of items to fetch per page.
+                It is best to set this to the same value that is used by the
+                website when accessing the API. This function automatically
+                reduces the number of items per page when the server responds
+                with errors or missing data.
+
+        Returns:
+            A list of extracted episode dicts to be used as playlist entries.
+
+        Raises:
+            ExtractorError: Might be raised when extracting episode data.
+
+        """
+        items_per_page = default_items_per_page
+        page = 1
+
+        entries = []
+
+        # The number of entries as reported by the API
+        n_entries = None
+
+        # The API sometimes returns an empty page without any episodes. In this
+        # case the next page often has episodes. This, however, throws off
+        # the total number of entries and it no longer becomes a reliable
+        # stopping condition when comparing it with the number of entries
+        # reported by the API. So we deal with this by not stopping at the
+        # first occurance of an empty page. We skip over a certain number of
+        # empty pages before giving up.
+        max_n_skipped_pages = default_items_per_page + 3
+        n_skipped_pages = 0
+
+        while True:
+            # We need this to check if we actually added any entries
+            n_entries_before_this_page = len(entries)
+
+            # Fetch data
+            api_url = api_url_template.format(
+                page_str=self._get_page_str(page),
+                items_per_page=items_per_page)
+            result_data = self._download_json(api_url, display_id, fatal=False)
+            episodes = try_get(result_data,
+                               lambda x: x['result']['episodes'],
+                               list)
+
+            # Add entries
+            for episode in episodes or []:
+                entry = self._extract_episode(
+                    self._get_episode_from_array_entry(episode))
+                if entry not in entries:
+                    entries.append(entry)
+
+            # Fetch how many episodes the API says it has (it's enough to
+            # read it once)
+            n_entries = n_entries if n_entries is not None else try_get(
+                result_data,
+                lambda x: x['result']['meta']['episodes']['total'],
+                int)
+
+            # Check if we have read the reported number of episodes
+            if n_entries is not None and len(entries) >= n_entries:
+                break
+
+            # Check if we actually added any entries
+            if n_entries_before_this_page == len(entries):
+                # This was an empty page so we have to skip it
+                n_skipped_pages += 1
+                if n_skipped_pages >= max_n_skipped_pages:
+                    # Enough skipping, give up
+                    break
+
+                # Throttle by reading only half as many entries as before
+                if items_per_page > 1:
+                    new_items_per_page = int(max(1, items_per_page / 2))
+                    page = int((page - 1) * items_per_page /
+                               new_items_per_page)
+                    items_per_page = new_items_per_page
+            else:
+                # This page had episodes, so we're no longer skipping
+                n_skipped_pages = 0
+
+                # Try to go back to full speed by going back to the default
+                # items_per_page value if possible.
+                if items_per_page * page % default_items_per_page == 0:
+                    page = int(page * items_per_page /
+                               default_items_per_page)
+                    items_per_page = default_items_per_page
+
+            page += 1
+
+        # Tell the user if we received less entries than the API reported
+        if n_entries is not None and len(entries) < n_entries:
+            self.to_screen('Received {} of {} reported episodes'.format(
+                len(entries), n_entries))
+
+        return entries
+
+    def _real_extract(self, url):
+        podcast_id = self._match_id(url)
+
+        api_url = 'https://www.ardaudiothek.de/api/podcasts/%s' % podcast_id
+        result_data = self._download_json(api_url, podcast_id, fatal=False)
+        pc_data = try_get(result_data, lambda x: x['result']['podcast'], dict)
+
+        if not pc_data:
+            raise ExtractorError(msg="Could not find any playlist data",
+                                 expected=True)
+
+        res = self._extract_id_title_desc(pc_data)
+        res['_type'] = 'playlist'
+
+        # items_per_page works from 1 up to 2147483647 (2^31 - 1).
+        # The website calls the API with items_per_page set to 24. Setting it
+        # to 500 or 1000 would download the data of all episodes in one or two
+        # pages. Increasing this value might however trigger server errors in
+        # the future. So to avoid any problems we will keep using the default
+        # value and just download a few more pages.
+        res['entries'] = self._extract_episodes(
+            podcast_id,
+            'https://www.ardaudiothek.de/api/podcasts/%s/episodes?items_per_page={items_per_page}{page_str}' % podcast_id,
+            24)
+
+        return res
+
+
+class ARDAudiothekSearchIE(ARDAudiothekPlaylistIE):
+    _VALID_URL = r'https?://(?:www\.|beta\.)?ardaudiothek\.de/suche\?(?:(?!q=).*&)?q=(?P<id>[^&]+)(?:&.*)?'
+    _TESTS = [{
+        'url': 'https://www.ardaudiothek.de/suche?q=Sommer',
+        'info_dict': {
+            'id': 'Sommer',
+            'title': 'Sommer',
+            'description': compat_str,
+        },
+        'playlist_mincount': 5,
+    }, {
+        'url': 'https://www.ardaudiothek.de/suche?q=Angela%20Merkel',
+        'info_dict': {
+            'id': 'Angela%20Merkel',
+            'title': 'Angela Merkel',
+            'description': compat_str,
+        },
+        'playlist_mincount': 5,
+    }]
+
+    def _get_page_str(self, page):
+        # The search API always works with a page number
+        return '&page=' + compat_str(page)
+
+    def _get_episode_from_array_entry(self, array_entry):
+        # The array entry is a dict with an 'episode' and a 'search_meta' entry
+        return try_get(array_entry, lambda x: x['episode'], dict)
+
+    def _real_extract(self, url):
+        search_str = self._match_id(url)
+        display_str = _unquote_compat(search_str)
+
+        return {
+            '_type': 'playlist',
+            'id': search_str,
+            'display_id': display_str,
+            'title': display_str,
+            'description': 'ARD Audiothek-Suche nach "' + display_str + '"',
+            # Searching on the website calls the API with items_per_page set
+            # to 8. Other values sometimes cause server errors.
+            'entries': self._extract_episodes(
+                display_str,
+                'https://www.ardaudiothek.de/api/search/%s?focus=episodes{page_str}&items_per_page={items_per_page}' % search_str,
+                8),
+        }

youtube_dl/extractor/extractors.py

@@ -71,6 +71,11 @@ from .ard import (
     ARDIE,
     ARDMediathekIE,
 )
+from .ardaudiothek import (
+    ARDAudiothekIE,
+    ARDAudiothekPlaylistIE,
+    ARDAudiothekSearchIE,
+)
 from .arte import (
     ArteTVIE,
     ArteTVEmbedIE,
@@ -1653,7 +1658,15 @@ from .younow import (
     YouNowChannelIE,
     YouNowMomentIE,
 )
-from .youporn import YouPornIE
+from .youporn import (
+    YouPornIE,
+    YouPornCategoryIE,
+    YouPornChannelIE,
+    YouPornCollectionIE,
+    YouPornStarIE,
+    YouPornTagIE,
+    YouPornVideosIE,
+)
 from .yourporn import YourPornIE
 from .yourupload import YourUploadIE
 from .youtube import (

youtube_dl/extractor/youporn.py

@@ -1,20 +1,38 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import itertools
 import re
+from time import sleep
 
 from .common import InfoExtractor
 from ..utils import (
+    clean_html,
     extract_attributes,
+    ExtractorError,
+    get_element_by_class,
+    get_element_by_id,
     int_or_none,
-    str_to_int,
+    merge_dicts,
+    parse_count,
+    parse_qs,
+    T,
+    traverse_obj,
     unified_strdate,
     url_or_none,
+    urljoin,
 )
 
 
 class YouPornIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?youporn\.com/(?:watch|embed)/(?P<id>\d+)(?:/(?P<display_id>[^/?#&]+))?'
+    _VALID_URL = (
+        r'youporn:(?P<id>\d+)',
+        r'''(?x)
+            https?://(?:www\.)?youporn\.com/(?:watch|embed)/(?P<id>\d+)
+            (?:/(?:(?P<display_id>[^/?#&]+)/?)?)?(?:[#?]|$)
+    '''
+    )
+    _EMBED_REGEX = [r'<iframe[^>]+\bsrc=["\'](?P<url>(?:https?:)?//(?:www\.)?youporn\.com/embed/\d+)']
     _TESTS = [{
         'url': 'http://www.youporn.com/watch/505835/sex-ed-is-it-safe-to-masturbate-daily/',
         'md5': '3744d24c50438cf5b6f6d59feb5055c2',
@@ -34,7 +52,7 @@ class YouPornIE(InfoExtractor):
             'tags': list,
             'age_limit': 18,
         },
-        'skip': 'This video has been disabled',
+        'skip': 'This video has been deactivated',
     }, {
         # Unknown uploader
         'url': 'http://www.youporn.com/watch/561726/big-tits-awesome-brunette-on-amazing-webcam-show/?from=related3&al=2&from_id=561726&pos=4',
@@ -66,57 +84,104 @@ class YouPornIE(InfoExtractor):
     }, {
         'url': 'https://www.youporn.com/watch/13922959/femdom-principal/',
         'only_matching': True,
+    }, {
+        'url': 'https://www.youporn.com/watch/16290308/tinderspecial-trailer1/',
+        'info_dict': {
+            'id': '16290308',
+            'age_limit': 18,
+            'categories': [],
+            'description': None,  # SEO spam using title removed
+            'display_id': 'tinderspecial-trailer1',
+            'duration': 298.0,
+            'ext': 'mp4',
+            'upload_date': '20201123',
+            'uploader': 'Ersties',
+            'tags': [],
+            'thumbnail': 'https://fi1.ypncdn.com/m=eaSaaTbWx/202011/23/16290308/original/3.jpg',
+            'timestamp': 1606147564,
+            'title': 'Tinder In Real Life',
+            'view_count': int,
+        }
     }]
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return re.findall(
-            r'<iframe[^>]+\bsrc=["\']((?:https?:)?//(?:www\.)?youporn\.com/embed/\d+)',
-            webpage)
+    @classmethod
+    def _extract_urls(cls, webpage):
+        def yield_urls():
+            for p in cls._EMBED_REGEX:
+                for m in re.finditer(p, webpage):
+                    yield m.group('url')
+
+        return list(yield_urls())
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
-        display_id = mobj.group('display_id') or video_id
+        # A different video ID (data-video-id) is hidden in the page but
+        # never seems to be used
+        video_id, display_id = self._match_valid_url(url).group('id', 'display_id')
+        url = 'http://www.youporn.com/watch/%s' % (video_id,)
+        webpage = self._download_webpage(
+            url, video_id, headers={'Cookie': 'age_verified=1'})
 
-        definitions = self._download_json(
-            'https://www.youporn.com/api/video/media_definitions/%s/' % video_id,
-            display_id)
+        watchable = self._search_regex(
+            r'''(<div\s[^>]*\bid\s*=\s*('|")?watch-container(?(2)\2|(?!-)\b)[^>]*>)''',
+            webpage, 'watchability', default=None)
+        if not watchable:
+            msg = re.split(r'\s{4}', clean_html(get_element_by_id(
+                'mainContent', webpage)) or '')[0]
+            raise ExtractorError(
+                ('%s says: %s' % (self.IE_NAME, msg))
+                if msg else 'Video unavailable: no reason found',
+                expected=True)
+        # internal ID ?
+        # video_id = extract_attributes(watchable).get('data-video-id')
+
+        playervars = self._search_json(
+            r'\bplayervars\s*:', webpage, 'playervars', video_id)
+
+        def get_fmt(x):
+            v_url = url_or_none(x.get('videoUrl'))
+            if v_url:
+                x['videoUrl'] = v_url
+                return (x['format'], x)
+
+        defs_by_format = dict(traverse_obj(playervars, (
+            'mediaDefinitions', lambda _, v: v.get('format'), T(get_fmt))))
+
+        def get_format_data(f):
+            if f not in defs_by_format:
+                return []
+            return self._download_json(
+                defs_by_format[f]['videoUrl'], video_id, '{0}-formats'.format(f))
 
         formats = []
-        for definition in definitions:
-            if not isinstance(definition, dict):
-                continue
-            video_url = url_or_none(definition.get('videoUrl'))
-            if not video_url:
-                continue
-            f = {
-                'url': video_url,
-                'filesize': int_or_none(definition.get('videoSize')),
-            }
-            height = int_or_none(definition.get('quality'))
+        # Try to extract only the actual master m3u8 first, avoiding the duplicate single resolution "master" m3u8s
+        for hls_url in traverse_obj(
+                get_format_data('hls'),
+                (lambda _, v: not isinstance(v['defaultQuality'], bool), 'videoUrl'),
+                (Ellipsis, 'videoUrl')):
+            formats.extend(self._extract_m3u8_formats(
+                hls_url, video_id, 'mp4', fatal=False, m3u8_id='hls',
+                entry_protocol='m3u8_native'))
+
+        for f in traverse_obj(get_format_data('mp4'), (
+                lambda _, v: v.get('videoUrl'), {
+                    'url': ('videoUrl', T(url_or_none)),
+                    'filesize': ('videoSize', T(int_or_none)),
+                    'height': ('quality', T(int_or_none)),
+                }, T(lambda x: x.get('videoUrl') and x))):
             # Video URL's path looks like this:
             #  /201012/17/505835/720p_1500k_505835/YouPorn%20-%20Sex%20Ed%20Is%20It%20Safe%20To%20Masturbate%20Daily.mp4
             #  /201012/17/505835/vl_240p_240k_505835/YouPorn%20-%20Sex%20Ed%20Is%20It%20Safe%20To%20Masturbate%20Daily.mp4
             #  /videos/201703/11/109285532/1080P_4000K_109285532.mp4
             # We will benefit from it by extracting some metadata
-            mobj = re.search(r'(?P<height>\d{3,4})[pP]_(?P<bitrate>\d+)[kK]_\d+', video_url)
+            mobj = re.search(r'(?P<height>\d{3,4})[pP]_(?P<bitrate>\d+)[kK]_\d+', f['videoUrl'])
             if mobj:
-                if not height:
-                    height = int(mobj.group('height'))
-                bitrate = int(mobj.group('bitrate'))
-                f.update({
-                    'format_id': '%dp-%dk' % (height, bitrate),
-                    'tbr': bitrate,
-                })
-            f['height'] = height
+                if not f.get('height'):
+                    f['height'] = int(mobj.group('height'))
+                f['tbr'] = int(mobj.group('bitrate'))
+                f['format_id'] = '%dp-%dk' % (f['height'], f['tbr'])
             formats.append(f)
         self._sort_formats(formats)
 
-        webpage = self._download_webpage(
-            'http://www.youporn.com/watch/%s' % video_id, display_id,
-            headers={'Cookie': 'age_verified=1'})
-
         title = self._html_search_regex(
             r'(?s)<div[^>]+class=["\']watchVideoTitle[^>]+>(.+?)</div>',
             webpage, 'title', default=None) or self._og_search_title(
@@ -131,8 +196,10 @@ class YouPornIE(InfoExtractor):
         thumbnail = self._search_regex(
             r'(?:imageurl\s*=|poster\s*:)\s*(["\'])(?P<thumbnail>.+?)\1',
             webpage, 'thumbnail', fatal=False, group='thumbnail')
-        duration = int_or_none(self._html_search_meta(
-            'video:duration', webpage, 'duration', fatal=False))
+        duration = traverse_obj(playervars, ('duration', T(int_or_none)))
+        if duration is None:
+            duration = int_or_none(self._html_search_meta(
+                'video:duration', webpage, 'duration', fatal=False))
 
         uploader = self._html_search_regex(
             r'(?s)<div[^>]+class=["\']submitByLink["\'][^>]*>(.+?)</div>',
@@ -148,11 +215,11 @@ class YouPornIE(InfoExtractor):
 
         view_count = None
         views = self._search_regex(
-            r'(<div[^>]+\bclass=["\']js_videoInfoViews["\']>)', webpage,
-            'views', default=None)
+            r'(<div\s[^>]*\bdata-value\s*=[^>]+>)\s*<label>Views:</label>',
+            webpage, 'views', default=None)
         if views:
-            view_count = str_to_int(extract_attributes(views).get('data-value'))
-        comment_count = str_to_int(self._search_regex(
+            view_count = parse_count(extract_attributes(views).get('data-value'))
+        comment_count = parse_count(self._search_regex(
             r'>All [Cc]omments? \(([\d,.]+)\)',
             webpage, 'comment count', default=None))
 
@@ -168,7 +235,10 @@ class YouPornIE(InfoExtractor):
             r'(?s)Tags:.*?</div>\s*<div[^>]+class=["\']tagBoxContent["\'][^>]*>(.+?)</div>',
             'tags')
 
-        return {
+        data = self._search_json_ld(webpage, video_id, expected_type='VideoObject', fatal=False) or {}
+        data.pop('url', None)
+
+        result = merge_dicts(data, {
             'id': video_id,
             'display_id': display_id,
             'title': title,
@@ -183,4 +253,442 @@ class YouPornIE(InfoExtractor):
             'tags': tags,
             'age_limit': age_limit,
             'formats': formats,
-        }
+        })
+        # Remove promotional non-description
+        if result.get('description', '').startswith(
+                'Watch %s online' % (result['title'],)):
+            del result['description']
+        return result
+
+
+class YouPornListBase(InfoExtractor):
+    # pattern in '.title-text' element of page section containing videos
+    _PLAYLIST_TITLEBAR_RE = r'\s+[Vv]ideos\s*$'
+    _PAGE_RETRY_COUNT = 0  # ie, no retry
+    _PAGE_RETRY_DELAY = 2  # seconds
+
+    def _get_next_url(self, url, pl_id, html):
+        return urljoin(url, self._search_regex(
+            r'''<a\s[^>]*?\bhref\s*=\s*("|')(?P<url>(?:(?!\1)[^>])+)\1''',
+            get_element_by_id('next', html) or '', 'next page',
+            group='url', default=None))
+
+    @classmethod
+    def _get_title_from_slug(cls, title_slug):
+        return re.sub(r'[_-]', ' ', title_slug)
+
+    def _entries(self, url, pl_id, html=None, page_num=None):
+
+        # separates page sections
+        PLAYLIST_SECTION_RE = (
+            r'''<div\s[^>]*\bclass\s*=\s*('|")(?:[\w$-]+\s+|\s)*?title-bar(?:\s+[\w$-]+|\s)*\1[^>]*>'''
+        )
+        # contains video link
+        VIDEO_URL_RE = r'''(?x)
+            <div\s[^>]*\bdata-video-id\s*=\s*('|")\d+\1[^>]*>\s*
+            (?:<div\b[\s\S]+?</div>\s*)*
+            <a\s[^>]*\bhref\s*=\s*('|")(?P<url>(?:(?!\2)[^>])+)\2
+        '''
+
+        def yield_pages(url, html=html, page_num=page_num):
+            fatal = not html
+            for pnum in itertools.count(start=page_num or 1):
+                if not html:
+                    html = self._download_webpage(
+                        url, pl_id, note='Downloading page %d' % pnum,
+                        fatal=fatal)
+                if not html:
+                    break
+                fatal = False
+                yield (url, html, pnum)
+                # explicit page: extract just that page
+                if page_num is not None:
+                    break
+                next_url = self._get_next_url(url, pl_id, html)
+                if not next_url or next_url == url:
+                    break
+                url, html = next_url, None
+
+        def retry_page(msg, tries_left, page_data):
+            if tries_left <= 0:
+                return
+            self.report_warning(msg, pl_id)
+            sleep(self._PAGE_RETRY_DELAY)
+            return next(
+                yield_pages(page_data[0], page_num=page_data[2]), None)
+
+        def yield_entries(html):
+            for frag in re.split(PLAYLIST_SECTION_RE, html):
+                if not frag:
+                    continue
+                t_text = get_element_by_class('title-text', frag or '')
+                if not (t_text and re.search(self._PLAYLIST_TITLEBAR_RE, t_text)):
+                    continue
+                for m in re.finditer(VIDEO_URL_RE, frag):
+                    video_url = urljoin(url, m.group('url'))
+                    if video_url:
+                        yield self.url_result(video_url)
+
+        last_first_url = None
+        for page_data in yield_pages(url, html=html, page_num=page_num):
+            # page_data: url, html, page_num
+            first_url = None
+            tries_left = self._PAGE_RETRY_COUNT + 1
+            while tries_left > 0:
+                tries_left -= 1
+                for from_ in yield_entries(page_data[1]):
+                    # may get the same page twice instead of empty page
+                    # or (site bug) intead of actual next page
+                    if not first_url:
+                        first_url = from_['url']
+                        if first_url == last_first_url:
+                            # sometimes (/porntags/) the site serves the previous page
+                            # instead but may provide the correct page after a delay
+                            page_data = retry_page(
+                                'Retrying duplicate page...', tries_left, page_data)
+                            if page_data:
+                                first_url = None
+                                break
+                            continue
+                    yield from_
+                else:
+                    if not first_url and 'no-result-paragarph1' in page_data[1]:
+                        page_data = retry_page(
+                            'Retrying empty page...', tries_left, page_data)
+                        if page_data:
+                            continue
+                    else:
+                        # success/failure
+                        break
+            # may get an infinite (?) sequence of empty pages
+            if not first_url:
+                break
+            last_first_url = first_url
+
+    def _real_extract(self, url, html=None):
+        # exceptionally, id may be None
+        m_dict = self._match_valid_url(url).groupdict()
+        pl_id, page_type, sort = (m_dict.get(k) for k in ('id', 'type', 'sort'))
+
+        qs = parse_qs(url)
+        for q, v in qs.items():
+            if v:
+                qs[q] = v[-1]
+            else:
+                del qs[q]
+
+        base_id = pl_id or 'YouPorn'
+        title = self._get_title_from_slug(base_id)
+        if page_type:
+            title = '%s %s' % (page_type.capitalize(), title)
+        base_id = [base_id.lower()]
+        if sort is None:
+            title += ' videos'
+        else:
+            title = '%s videos by %s' % (title, re.sub(r'[_-]', ' ', sort))
+            base_id.append(sort)
+        if qs:
+            ps = ['%s=%s' % item for item in sorted(qs.items())]
+            title += ' (%s)' % ','.join(ps)
+            base_id.extend(ps)
+        pl_id = '/'.join(base_id)
+
+        return self.playlist_result(
+            self._entries(url, pl_id, html=html,
+                          page_num=int_or_none(qs.get('page'))),
+            playlist_id=pl_id, playlist_title=title)
+
+
+class YouPornCategoryIE(YouPornListBase):
+    IE_DESC = 'YouPorn category, with sorting, filtering and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>category)/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>popular|views|rating|time|duration))?/?(?:[#?]|$)
+    '''
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/category/lingerie/popular/',
+        'info_dict': {
+            'id': 'lingerie/popular',
+            'title': 'Category lingerie videos by popular',
+        },
+        'playlist_mincount': 39,
+    }, {
+        'note': 'Filtered paginated list with single page result',
+        'url': 'https://www.youporn.com/category/lingerie/duration/?min_minutes=10',
+        'info_dict': {
+            'id': 'lingerie/duration/min_minutes=10',
+            'title': 'Category lingerie videos by duration (min_minutes=10)',
+        },
+        'playlist_maxcount': 30,
+    }, {
+        'note': 'Single page of full list',
+        'url': 'https://www.youporn.com/category/lingerie/popular?page=1',
+        'info_dict': {
+            'id': 'lingerie/popular/page=1',
+            'title': 'Category lingerie videos by popular (page=1)',
+        },
+        'playlist_count': 30,
+    }]
+
+
+class YouPornChannelIE(YouPornListBase):
+    IE_DESC = 'YouPorn channel, with sorting and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>channel)/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>rating|views|duration))?/?(?:[#?]|$)
+    '''
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/channel/x-feeds/',
+        'info_dict': {
+            'id': 'x-feeds',
+            'title': 'Channel X-Feeds videos',
+        },
+        'playlist_mincount': 37,
+    }, {
+        'note': 'Single page of full list (no filters here)',
+        'url': 'https://www.youporn.com/channel/x-feeds/duration?page=1',
+        'info_dict': {
+            'id': 'x-feeds/duration/page=1',
+            'title': 'Channel X-Feeds videos by duration (page=1)',
+        },
+        'playlist_count': 24,
+    }]
+
+    @staticmethod
+    def _get_title_from_slug(title_slug):
+        return re.sub(r'_', ' ', title_slug).title()
+
+
+class YouPornCollectionIE(YouPornListBase):
+    IE_DESC = 'YouPorn collection (user playlist), with sorting and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>collection)s/videos/(?P<id>\d+)
+        (?:/(?P<sort>rating|views|time|duration))?/?(?:[#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'^\s*Videos\s+in\s'
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/collections/videos/33044251/',
+        'info_dict': {
+            'id': '33044251',
+            'title': 'Collection Sexy Lips videos',
+            'uploader': 'ph-littlewillyb',
+        },
+        'playlist_mincount': 50,
+    }, {
+        'note': 'Single page of full list (no filters here)',
+        'url': 'https://www.youporn.com/collections/videos/33044251/time?page=1',
+        'info_dict': {
+            'id': '33044251/time/page=1',
+            'title': 'Collection Sexy Lips videos by time (page=1)',
+            'uploader': 'ph-littlewillyb',
+        },
+        'playlist_count': 20,
+    }]
+
+    def _real_extract(self, url):
+        pl_id = self._match_id(url)
+        html = self._download_webpage(url, pl_id)
+        playlist = super(YouPornCollectionIE, self)._real_extract(url, html=html)
+        infos = re.sub(r'\s+', ' ', clean_html(get_element_by_class(
+            'collection-infos', html)) or '')
+        title, uploader = self._search_regex(
+            r'^\s*Collection: (?P<title>.+?) \d+ VIDEOS \d+ VIEWS \d+ days LAST UPDATED From: (?P<uploader>[\w_-]+)',
+            infos, 'title/uploader', group=('title', 'uploader'), default=(None, None))
+
+        return merge_dicts({
+            'title': playlist['title'].replace(playlist['id'].split('/')[0], title),
+            'uploader': uploader,
+        }, playlist) if title else playlist
+
+
+class YouPornTagIE(YouPornListBase):
+    IE_DESC = 'YouPorn tag (porntags), with sorting, filtering and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        porn(?P<type>tag)s/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>views|rating|time|duration))?/?(?:[#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'^\s*Videos\s+tagged\s'
+    _PAGE_RETRY_COUNT = 1
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/porntags/austrian',
+        'info_dict': {
+            'id': 'austrian',
+            'title': 'Tag austrian videos',
+        },
+        'playlist_mincount': 35,
+        'expected_warnings': ['Retrying duplicate page'],
+    }, {
+        'note': 'Filtered paginated list with single page result',
+        'url': 'https://www.youporn.com/porntags/austrian/duration/?min_minutes=10',
+        'info_dict': {
+            'id': 'austrian/duration/min_minutes=10',
+            'title': 'Tag austrian videos by duration (min_minutes=10)',
+        },
+        # number of videos per page is (row x col) 2x3 + 6x4 + 2, or + 3,
+        # or more, varying with number of ads; let's set max as 9x4
+        # NB col 1 may not be shown in non-JS page with site CSS and zoom 100%
+        'playlist_maxcount': 32,
+        'expected_warnings': ['Retrying duplicate page', 'Retrying empty page'],
+    }, {
+        'note': 'Single page of full list',
+        'url': 'https://www.youporn.com/porntags/austrian/?page=1',
+        'info_dict': {
+            'id': 'austrian/page=1',
+            'title': 'Tag austrian videos (page=1)',
+        },
+        'playlist_mincount': 32,
+        'playlist_maxcount': 34,
+        'expected_warnings': ['Retrying duplicate page', 'Retrying empty page'],
+    }]
+
+    # YP tag navigation is broken, loses sort
+    def _get_next_url(self, url, pl_id, html):
+        next_url = super(YouPornTagIE, self)._get_next_url(url, pl_id, html)
+        if next_url:
+            n = self._match_valid_url(next_url)
+            if n:
+                s = n.groupdict().get('sort')
+            if s:
+                u = self._match_valid_url(url)
+                if u:
+                    u = u.groupdict().get('sort')
+                    if s and not u:
+                        n = n.end('sort')
+                        next_url = next_url[:n] + '/' + u + next_url[n:]
+        return next_url
+
+
+class YouPornStarIE(YouPornListBase):
+    IE_DESC = 'YouPorn Pornstar, with description, sorting and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>pornstar)/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>rating|views|duration))?/?(?:[#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'^\s*Videos\s+[fF]eaturing\s'
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/pornstar/daynia/',
+        'info_dict': {
+            'id': 'daynia',
+            'title': 'Pornstar Daynia videos',
+            'description': r're:Daynia Rank \d+ Videos \d+ Views [\d,.]+ .+ Subscribers \d+',
+        },
+        'playlist_mincount': 45,
+    }, {
+        'note': 'Single page of full list (no filters here)',
+        'url': 'https://www.youporn.com/pornstar/daynia/?page=1',
+        'info_dict': {
+            'id': 'daynia/page=1',
+            'title': 'Pornstar Daynia videos (page=1)',
+            'description': 're:.{180,}',
+        },
+        'playlist_count': 26,
+    }]
+
+    @staticmethod
+    def _get_title_from_slug(title_slug):
+        return re.sub(r'_', ' ', title_slug).title()
+
+    def _real_extract(self, url):
+        pl_id = self._match_id(url)
+        html = self._download_webpage(url, pl_id)
+        playlist = super(YouPornStarIE, self)._real_extract(url, html=html)
+        INFO_ELEMENT_RE = r'''(?x)
+            <div\s[^>]*\bclass\s*=\s*('|")(?:[\w$-]+\s+|\s)*?pornstar-info-wrapper(?:\s+[\w$-]+|\s)*\1[^>]*>
+            (?P<info>[\s\S]+?)(?:</div>\s*){6,}
+        '''
+
+        infos = self._search_regex(INFO_ELEMENT_RE, html, 'infos', group='info', default='')
+        if infos:
+            infos = re.sub(
+                r'(?:\s*nl=nl)+\s*', ' ',
+                re.sub(r'(?u)\s+', ' ', clean_html(
+                    re.sub('\n', 'nl=nl', infos)))).replace('ribe Subsc', '')
+
+        return merge_dicts({
+            'description': infos.strip() or None,
+        }, playlist)
+
+
+class YouPornVideosIE(YouPornListBase):
+    IE_DESC = 'YouPorn video (browse) playlists, with sorting, filtering and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+            (?:(?P<id>browse)/)?
+            (?P<sort>(?(id)
+                (?:duration|rating|time|views)|
+                (?:most_(?:favou?rit|view)ed|recommended|top_rated)?))
+            (?:[/#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'\s+(?:[Vv]ideos|VIDEOS)\s*$'
+    _TESTS = [{
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/',
+        'info_dict': {
+            'id': 'youporn',
+            'title': 'YouPorn videos',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/recommended',
+        'info_dict': {
+            'id': 'youporn/recommended',
+            'title': 'YouPorn videos by recommended',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/top_rated',
+        'info_dict': {
+            'id': 'youporn/top_rated',
+            'title': 'YouPorn videos by top rated',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/browse/time',
+        'info_dict': {
+            'id': 'browse/time',
+            'title': 'YouPorn videos by time',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Filtered paginated list with single page result',
+        'url': 'https://www.youporn.com/most_favorited/?res=VR&max_minutes=2',
+        'info_dict': {
+            'id': 'youporn/most_favorited/max_minutes=2/res=VR',
+            'title': 'YouPorn videos by most favorited (max_minutes=2,res=VR)',
+        },
+        'playlist_mincount': 10,
+        'playlist_maxcount': 28,
+    }, {
+        'note': 'Filtered paginated list with several pages',
+        'url': 'https://www.youporn.com/most_favorited/?res=VR&max_minutes=5',
+        'info_dict': {
+            'id': 'youporn/most_favorited/max_minutes=5/res=VR',
+            'title': 'YouPorn videos by most favorited (max_minutes=5,res=VR)',
+        },
+        'playlist_mincount': 45,
+    }, {
+        'note': 'Single page of full list',
+        'url': 'https://www.youporn.com/browse/time?page=1',
+        'info_dict': {
+            'id': 'browse/time/page=1',
+            'title': 'YouPorn videos by time (page=1)',
+        },
+        'playlist_count': 36,
+    }]
+
+    @staticmethod
+    def _get_title_from_slug(title_slug):
+        return 'YouPorn' if title_slug == 'browse' else title_slug