[eporner] Fix extraction (Closes #10139)

youtube_dl/extractor/eporner.py

@@ -4,19 +4,23 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
+from ..compat import compat_str
 from ..utils import (
+    encode_base_n,
+    ExtractorError,
+    int_or_none,
     parse_duration,
     str_to_int,
 )
 
 
 class EpornerIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?eporner\.com/hd-porn/(?P<id>\w+)/(?P<display_id>[\w-]+)'
+    _VALID_URL = r'https?://(?:www\.)?eporner\.com/hd-porn/(?P<id>\w+)(?:/(?P<display_id>[\w-]+))?'
     _TESTS = [{
         'url': 'http://www.eporner.com/hd-porn/95008/Infamous-Tiffany-Teen-Strip-Tease-Video/',
         'md5': '39d486f046212d8e1b911c52ab4691f8',
         'info_dict': {
-            'id': '95008',
+            'id': 'qlDUmNsj6VS',
             'display_id': 'Infamous-Tiffany-Teen-Strip-Tease-Video',
             'ext': 'mp4',
             'title': 'Infamous Tiffany Teen Strip Tease Video',
@@ -28,34 +32,72 @@ class EpornerIE(InfoExtractor):
         # New (May 2016) URL layout
         'url': 'http://www.eporner.com/hd-porn/3YRUtzMcWn0/Star-Wars-XXX-Parody/',
         'only_matching': True,
+    }, {
+        'url': 'http://www.eporner.com/hd-porn/3YRUtzMcWn0',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
-        display_id = mobj.group('display_id')
+        display_id = mobj.group('display_id') or video_id
 
-        webpage = self._download_webpage(url, display_id)
+        webpage, urlh = self._download_webpage_handle(url, display_id)
-        title = self._html_search_regex(
-            r'<title>(.*?) - EPORNER', webpage, 'title')
 
-        redirect_url = 'http://www.eporner.com/config5/%s' % video_id
+        video_id = self._match_id(compat_str(urlh.geturl()))
-        player_code = self._download_webpage(
-            redirect_url, display_id, note='Downloading player config')
 
-        sources = self._search_regex(
+        hash = self._search_regex(
-            r'(?s)sources\s*:\s*\[\s*({.+?})\s*\]', player_code, 'sources')
+            r'hash\s*:\s*["\']([\da-f]{32})', webpage, 'hash')
+
+        title = self._og_search_title(webpage, default=None) or self._html_search_regex(
+            r'<title>(.+?) - EPORNER', webpage, 'title')
+
+        # Reverse engineered from vjs.js
+        def calc_hash(s):
+            return ''.join((encode_base_n(int(s[lb:lb + 8], 16), 36) for lb in range(0, 32, 8)))
+
+        video = self._download_json(
+            'http://www.eporner.com/xhr/video/%s' % video_id,
+            display_id, note='Downloading video JSON',
+            query={
+                'hash': calc_hash(hash),
+                'device': 'generic',
+                'domain': 'www.eporner.com',
+                'fallback': 'false',
+            })
+
+        if video.get('available') is False:
+            raise ExtractorError(
+                '%s said: %s' % (self.IE_NAME, video['message']), expected=True)
+
+        sources = video['sources']
 
         formats = []
-        for video_url, format_id in re.findall(r'file\s*:\s*"([^"]+)",\s*label\s*:\s*"([^"]+)"', sources):
+        for kind, formats_dict in sources.items():
-            fmt = {
+            if not isinstance(formats_dict, dict):
-                'url': video_url,
+                continue
+            for format_id, format_dict in formats_dict.items():
+                if not isinstance(format_dict, dict):
+                    continue
+                src = format_dict.get('src')
+                if not isinstance(src, compat_str) or not src.startswith('http'):
+                    continue
+                if kind == 'hls':
+                    formats.extend(self._extract_m3u8_formats(
+                        src, display_id, 'mp4', entry_protocol='m3u8_native',
+                        m3u8_id=kind, fatal=False))
+                else:
+                    height = int_or_none(self._search_regex(
+                        r'(\d+)[pP]', format_id, 'height', default=None))
+                    fps = int_or_none(self._search_regex(
+                        r'(\d+)fps', format_id, 'fps', default=None))
+
+                    formats.append({
+                        'url': src,
                         'format_id': format_id,
-            }
+                        'height': height,
-            m = re.search(r'^(\d+)', format_id)
+                        'fps': fps,
-            if m:
+                    })
-                fmt['height'] = int(m.group(1))
-            formats.append(fmt)
         self._sort_formats(formats)
 
         duration = parse_duration(self._html_search_meta('duration', webpage))