[pornhd] Fix formats extraction (Closes #3898)

youtube_dl/extractor/pornhd.py

@@ -4,19 +4,27 @@ import re
 import json
 
 from .common import InfoExtractor
-from ..utils import int_or_none
+from ..utils import (
+    int_or_none,
+    js_to_json,
+    qualities,
+    determine_ext,
+)
 
 
 class PornHdIE(InfoExtractor):
-    _VALID_URL = r'http://(?:www\.)?pornhd\.com/(?:[a-z]{2,4}/)?videos/(?P<id>\d+)'
+    _VALID_URL = r'http://(?:www\.)?pornhd\.com/(?:[a-z]{2,4}/)?videos/(?P<id>\d+)(?:/(?P<display_id>.+))?'
     _TEST = {
         'url': 'http://www.pornhd.com/videos/1962/sierra-day-gets-his-cum-all-over-herself-hd-porn-video',
         'md5': '956b8ca569f7f4d8ec563e2c41598441',
         'info_dict': {
             'id': '1962',
+            'display_id': 'sierra-day-gets-his-cum-all-over-herself-hd-porn-video',
             'ext': 'mp4',
             'title': 'Sierra loves doing laundry',
             'description': 'md5:8ff0523848ac2b8f9b065ba781ccf294',
+            'thumbnail': 're:^https?://.*\.jpg',
+            'view_count': int,
             'age_limit': 18,
         }
     }
@@ -24,8 +32,9 @@ class PornHdIE(InfoExtractor):
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
+        display_id = mobj.group('display_id')
 
-        webpage = self._download_webpage(url, video_id)
+        webpage = self._download_webpage(url, display_id or video_id)
 
         title = self._html_search_regex(
             r'<title>(.+) porn HD.+?</title>', webpage, 'title')
@@ -33,38 +42,21 @@ class PornHdIE(InfoExtractor):
             r'<div class="description">([^<]+)</div>', webpage, 'description', fatal=False)
         view_count = int_or_none(self._html_search_regex(
             r'(\d+) views\s*</span>', webpage, 'view count', fatal=False))
+        thumbnail = self._search_regex(
+            r"'poster'\s*:\s*'([^']+)'", webpage, 'thumbnail', fatal=False)
 
-        videos = re.findall(
-            r'var __video([\da-zA-Z]+?)(Low|High)StreamUrl = \'(http://.+?)\?noProxy=1\'', webpage)
-
-        mobj = re.search(r'flashVars = (?P<flashvars>{.+?});', webpage)
-        if mobj:
-            flashvars = json.loads(mobj.group('flashvars'))
-            for key, quality in [('hashlink', 'low'), ('hd', 'high')]:
-                redirect_url = flashvars.get(key)
-                if redirect_url:
-                    videos.append(('flv', quality, redirect_url))
-            thumbnail = flashvars['urlWallpaper']
-        else:
-            thumbnail = self._og_search_thumbnail(webpage)
-
-        formats = []
-        for format_, quality, redirect_url in videos:
-            format_id = '%s-%s' % (format_.lower(), quality.lower())
-            video_url = self._download_webpage(
-                redirect_url, video_id, 'Downloading %s video link' % format_id, fatal=False)
-            if not video_url:
-                continue
-            formats.append({
-                'url': video_url,
-                'ext': format_.lower(),
-                'format_id': format_id,
-                'quality': 1 if quality.lower() == 'high' else 0,
-            })
+        quality = qualities(['SD', 'HD'])
+        formats = [{
+            'url': source['file'],
+            'format_id': '%s-%s' % (source['label'], determine_ext(source['file'])),
+            'quality': quality(source['label']),
+        } for source in json.loads(js_to_json(self._search_regex(
+            r"(?s)'sources'\s*:\s*(\[.+?\])", webpage, 'sources')))]
         self._sort_formats(formats)
 
         return {
             'id': video_id,
+            'display_id': display_id,
             'title': title,
             'description': description,
             'thumbnail': thumbnail,