[ie/manyvids] Fix ManyVids extractor after website update

yt_dlp/extractor/manyvids.py

@@ -1,21 +1,20 @@
-import re
-
 from .common import InfoExtractor
+from .. import traverse_obj
 from ..utils import (
     determine_ext,
-    extract_attributes,
     int_or_none,
-    str_to_int,
+    parse_count,
+    parse_duration,
+    parse_iso8601,
     url_or_none,
-    urlencode_postdata,
 )
 
 
 class ManyVidsIE(InfoExtractor):
-    _WORKING = False
     _VALID_URL = r'(?i)https?://(?:www\.)?manyvids\.com/video/(?P<id>\d+)'
     _TESTS = [{
-        # preview video
+        # Dead preview video
+        'skip': True,
         'url': 'https://www.manyvids.com/Video/133957/everthing-about-me/',
         'md5': '03f11bb21c52dd12a05be21a5c7dcc97',
         'info_dict': {
@@ -26,6 +25,24 @@ class ManyVidsIE(InfoExtractor):
             'view_count': int,
             'like_count': int,
         },
+    }, {
+        # preview video
+        'url': 'https://www.manyvids.com/Video/530341/mv-tips-tricks',
+        'md5': '738dc723f7735ee9602f7ea352a6d058',
+        'info_dict': {
+            'id': '530341',
+            'ext': 'mp4',
+            'title': 'MV Tips &amp; Tricks (Preview)',
+            'description': 'md5:c3bae98c0f9453237c28b0f8795d9f83',
+            'thumbnail': 'https://cdn5.manyvids.com/php_uploads/video_images/DestinyDiaz/thumbs/thumb_Hs26ATOO7fcZaI9sx3XT_screenshot_001.jpg',
+            'uploader': 'DestinyDiaz',
+            'view_count': int,
+            'like_count': int,
+            'release_timestamp': 1508419904,
+            'tags': ['AdultSchool', 'BBW', 'SFW', 'TeacherFetish'],
+            'release_date': '20171019',
+            'duration': 3167.0,
+        },
     }, {
         # full video
         'url': 'https://www.manyvids.com/Video/935718/MY-FACE-REVEAL/',
@@ -35,79 +52,41 @@ class ManyVidsIE(InfoExtractor):
             'ext': 'mp4',
             'title': 'MY FACE REVEAL',
             'description': 'md5:ec5901d41808b3746fed90face161612',
+            'thumbnail': 'https://ods.manyvids.com/1001061960/3aa5397f2a723ec4597e344df66ab845/screenshots/thumbs/custom_1_180_5be09c1dcce03.jpg',
             'uploader': 'Sarah Calanthe',
             'view_count': int,
             'like_count': int,
+            'release_date': '20181110',
+            'tags': ['EyeContact', 'Interviews', 'MaskFetish', 'MouthFetish', 'Redhead'],
+            'release_timestamp': 1541851200,
+            'duration': 224.0,
         },
     }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        real_url = f'https://www.manyvids.com/video/{video_id}/gtm.js'
+        info = traverse_obj(
-        try:
+            self._download_json(f'https://www.manyvids.com/bff/store/video/{video_id}', video_id),
-            webpage = self._download_webpage(real_url, video_id)
+            ('data', {dict})) or {}
-        except Exception:
-            # probably useless fallback
-            webpage = self._download_webpage(url, video_id)
 
-        info = self._search_regex(
+        video_urls = self._download_json(f'https://www.manyvids.com/bff/store/video/{video_id}/private', video_id)[
-            r'''(<div\b[^>]*\bid\s*=\s*(['"])pageMetaDetails\2[^>]*>)''',
+            'data']
-            webpage, 'meta details', default='')
-        info = extract_attributes(info)
-
-        player = self._search_regex(
-            r'''(<div\b[^>]*\bid\s*=\s*(['"])rmpPlayerStream\2[^>]*>)''',
-            webpage, 'player details', default='')
-        player = extract_attributes(player)
 
         video_urls_and_ids = (
-            (info.get('data-meta-video'), 'video'),
+            (traverse_obj(video_urls, ('teaser', 'filepath')), 'preview'),
-            (player.get('data-video-transcoded'), 'transcoded'),
+            (video_urls.get('transcodedFilepath'), 'transcoded'),
-            (player.get('data-video-filepath'), 'filepath'),
+            (video_urls.get('filepath'), 'filepath'),
-            (self._og_search_video_url(webpage, secure=False, default=None), 'og_video'),
         )
 
-        def txt_or_none(s, default=None):
+        title = traverse_obj(info, 'title')
-            return (s.strip() or default) if isinstance(s, str) else default
 
-        uploader = txt_or_none(info.get('data-meta-author'))
+        # If the video formats JSON only contains a teaser object, then it is a preview
-
+        if video_urls.get('teaser') and not video_urls.get('filepath'):
-        def mung_title(s):
-            if uploader:
-                s = re.sub(rf'^\s*{re.escape(uploader)}\s+[|-]', '', s)
-            return txt_or_none(s)
-
-        title = (
-            mung_title(info.get('data-meta-title'))
-            or self._html_search_regex(
-                (r'<span[^>]+class=["\']item-title[^>]+>([^<]+)',
-                 r'<h2[^>]+class=["\']h2 m-0["\'][^>]*>([^<]+)'),
-                webpage, 'title', default=None)
-            or self._html_search_meta(
-                'twitter:title', webpage, 'title', fatal=True))
-
-        title = re.sub(r'\s*[|-]\s+ManyVids\s*$', '', title) or title
-
-        if any(p in webpage for p in ('preview_videos', '_preview.mp4')):
             title += ' (Preview)'
-
+            video_id += '-preview'
-        mv_token = self._search_regex(
+            self.report_warning(
-            r'data-mvtoken=(["\'])(?P<value>(?:(?!\1).)+)\1', webpage,
+                f'Only extracting preview. Video may be paid or subscription only. {self._login_hint()}')
-            'mv token', default=None, group='value')
-
-        if mv_token:
-            # Sets some cookies
-            self._download_webpage(
-                'https://www.manyvids.com/includes/ajax_repository/you_had_me_at_hello.php',
-                video_id, note='Setting format cookies', fatal=False,
-                data=urlencode_postdata({
-                    'mvtoken': mv_token,
-                    'vid': video_id,
-                }), headers={
-                    'Referer': url,
-                    'X-Requested-With': 'XMLHttpRequest',
-                })
 
         formats = []
         for v_url, fmt in video_urls_and_ids:
@@ -130,33 +109,22 @@ def mung_title(s):
             if f.get('height') is None:
                 f['height'] = int_or_none(
                     self._search_regex(r'_(\d{2,3}[02468])_', f['url'], 'video height', default=None))
-            if '/preview/' in f['url']:
+            if 'preview' in f['format_id']:
-                f['format_id'] = '_'.join(filter(None, (f.get('format_id'), 'preview')))
                 f['preference'] = -10
             if 'transcoded' in f['format_id']:
                 f['preference'] = f.get('preference', -1) - 1
 
-        def get_likes():
-            likes = self._search_regex(
-                rf'''(<a\b[^>]*\bdata-id\s*=\s*(['"]){video_id}\2[^>]*>)''',
-                webpage, 'likes', default='')
-            likes = extract_attributes(likes)
-            return int_or_none(likes.get('data-likes'))
-
-        def get_views():
-            return str_to_int(self._html_search_regex(
-                r'''(?s)<span\b[^>]*\bclass\s*=["']views-wrapper\b[^>]+>.+?<span\b[^>]+>\s*(\d[\d,.]*)\s*</span>''',
-                webpage, 'view count', default=None))
-
         return {
             'id': video_id,
             'title': title,
             'formats': formats,
-            'description': txt_or_none(info.get('data-meta-description')),
+            'description': (traverse_obj(info, 'description')),
-            'uploader': txt_or_none(info.get('data-meta-author')),
+            'uploader': (traverse_obj(info, ('model', 'displayName'))),
             'thumbnail': (
-                url_or_none(info.get('data-meta-image'))
+                url_or_none(traverse_obj(info, 'screenshot')) or url_or_none(traverse_obj(info, 'thumbnail'))),
-                or url_or_none(player.get('data-video-screenshot'))),
+            'view_count': (parse_count(traverse_obj(info, 'views'))),
-            'view_count': get_views(),
+            'like_count': (parse_count(traverse_obj(info, 'likes'))),
-            'like_count': get_likes(),
+            'release_timestamp': (parse_iso8601(traverse_obj(info, 'launchDate'))),
+            'duration': (parse_duration(traverse_obj(info, 'videoDuration'))),
+            'tags': [t.get('label') for t in traverse_obj(info, 'tagList')],
         }