[youporn] Fix formats and view count extraction (closes #29216)

youtube_dl/extractor/youporn.py

@@ -4,13 +4,12 @@ import re
 
 from .common import InfoExtractor
 from ..utils import (
+    extract_attributes,
     int_or_none,
     str_to_int,
-    unescapeHTML,
     unified_strdate,
     url_or_none,
 )
-from ..aes import aes_decrypt_text
 
 
 class YouPornIE(InfoExtractor):
@@ -34,6 +33,7 @@ class YouPornIE(InfoExtractor):
             'tags': list,
             'age_limit': 18,
         },
+        'skip': 'This video has been disabled',
     }, {
         # 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',
@@ -78,6 +78,40 @@ class YouPornIE(InfoExtractor):
         video_id = mobj.group('id')
         display_id = mobj.group('display_id') or video_id
 
+        definitions = self._download_json(
+            'https://www.youporn.com/api/video/media_definitions/%s/' % video_id,
+            display_id)
+
+        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'))
+            # 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)
+            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
+            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'})
@@ -88,65 +122,6 @@ class YouPornIE(InfoExtractor):
             webpage, default=None) or self._html_search_meta(
             'title', webpage, fatal=True)
 
-        links = []
-
-        # Main source
-        definitions = self._parse_json(
-            self._search_regex(
-                r'mediaDefinition\s*[=:]\s*(\[.+?\])\s*[;,]', webpage,
-                'media definitions', default='[]'),
-            video_id, fatal=False)
-        if definitions:
-            for definition in definitions:
-                if not isinstance(definition, dict):
-                    continue
-                video_url = url_or_none(definition.get('videoUrl'))
-                if video_url:
-                    links.append(video_url)
-
-        # Fallback #1, this also contains extra low quality 180p format
-        for _, link in re.findall(r'<a[^>]+href=(["\'])(http(?:(?!\1).)+\.mp4(?:(?!\1).)*)\1[^>]+title=["\']Download [Vv]ideo', webpage):
-            links.append(link)
-
-        # Fallback #2 (unavailable as at 22.06.2017)
-        sources = self._search_regex(
-            r'(?s)sources\s*:\s*({.+?})', webpage, 'sources', default=None)
-        if sources:
-            for _, link in re.findall(r'[^:]+\s*:\s*(["\'])(http.+?)\1', sources):
-                links.append(link)
-
-        # Fallback #3 (unavailable as at 22.06.2017)
-        for _, link in re.findall(
-                r'(?:videoSrc|videoIpadUrl|html5PlayerSrc)\s*[:=]\s*(["\'])(http.+?)\1', webpage):
-            links.append(link)
-
-        # Fallback #4, encrypted links (unavailable as at 22.06.2017)
-        for _, encrypted_link in re.findall(
-                r'encryptedQuality\d{3,4}URL\s*=\s*(["\'])([\da-zA-Z+/=]+)\1', webpage):
-            links.append(aes_decrypt_text(encrypted_link, title, 32).decode('utf-8'))
-
-        formats = []
-        for video_url in set(unescapeHTML(link) for link in links):
-            f = {
-                'url': video_url,
-            }
-            # 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)
-            if mobj:
-                height = int(mobj.group('height'))
-                bitrate = int(mobj.group('bitrate'))
-                f.update({
-                    'format_id': '%dp-%dk' % (height, bitrate),
-                    'height': height,
-                    'tbr': bitrate,
-                })
-            formats.append(f)
-        self._sort_formats(formats)
-
         description = self._html_search_regex(
             r'(?s)<div[^>]+\bid=["\']description["\'][^>]*>(.+?)</div>',
             webpage, 'description',
@@ -169,13 +144,12 @@ class YouPornIE(InfoExtractor):
 
         age_limit = self._rta_search(webpage)
 
-        average_rating = int_or_none(self._search_regex(
+        view_count = None
-            r'<div[^>]+class=["\']videoRatingPercentage["\'][^>]*>(\d+)%</div>',
+        views = self._search_regex(
-            webpage, 'average rating', fatal=False))
+            r'(<div[^>]+\bclass=["\']js_videoInfoViews["\']>)', webpage,
-
+            'views', default=None)
-        view_count = str_to_int(self._search_regex(
+        if views:
-            r'(?s)<div[^>]+class=(["\']).*?\bvideoInfoViews\b.*?\1[^>]*>.*?(?P<count>[\d,.]+)<',
+            view_count = str_to_int(extract_attributes(views).get('data-value'))
-            webpage, 'view count', fatal=False, group='count'))
         comment_count = str_to_int(self._search_regex(
             r'>All [Cc]omments? \(([\d,.]+)\)',
             webpage, 'comment count', default=None))
@@ -201,7 +175,6 @@ class YouPornIE(InfoExtractor):
             'duration': duration,
             'uploader': uploader,
             'upload_date': upload_date,
-            'average_rating': average_rating,
             'view_count': view_count,
             'comment_count': comment_count,
             'categories': categories,