Fix extraction from youtube.

[youtubedl] / youtube_dl / extractor / youjizz.py

diff --git a/youtube_dl/extractor/youjizz.py b/youtube_dl/extractor/youjizz.py
index c642075dcfabbfb025d64b92e392d614578f42b1..dff69fcb7aca250373fc0e70b2f8278ed2661755 100644 (file)
--- a/youtube_dl/extractor/youjizz.py
+++ b/youtube_dl/extractor/youjizz.py
@@ -4,58 +4,92 @@ import re
 
 from .common import InfoExtractor
 from ..utils import (
-    ExtractorError,
+    determine_ext,
+    int_or_none,
+    parse_duration,
+    url_or_none,
 )
 
 
 class YouJizzIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:\w+\.)?youjizz\.com/videos/[^/#?]+-(?P<id>[0-9]+)\.html(?:$|[?#])'
-    _TEST = {
+    _VALID_URL = r'https?://(?:\w+\.)?youjizz\.com/videos/(?:[^/#?]*-(?P<id>\d+)\.html|embed/(?P<embed_id>\d+))'
+    _TESTS = [{
         'url': 'http://www.youjizz.com/videos/zeichentrick-1-2189178.html',
-        'md5': '07e15fa469ba384c7693fd246905547c',
+        'md5': 'b1e1dfaa8bb9537d8b84eeda9cf4acf4',
         'info_dict': {
             'id': '2189178',
-            'ext': 'flv',
-            "title": "Zeichentrick 1",
-            "age_limit": 18,
+            'ext': 'mp4',
+            'title': 'Zeichentrick 1',
+            'age_limit': 18,
+            'duration': 2874,
         }
-    }
+    }, {
+        'url': 'http://www.youjizz.com/videos/-2189178.html',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.youjizz.com/videos/embed/31991001',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id') or mobj.group('embed_id')
+
         webpage = self._download_webpage(url, video_id)
-        age_limit = self._rta_search(webpage)
-        video_title = self._html_search_regex(
-            r'<title>\s*(.*)\s*</title>', webpage, 'title')
-
-        embed_page_url = self._search_regex(
-            r'(https?://www.youjizz.com/videos/embed/[0-9]+)',
-            webpage, 'embed page')
-        webpage = self._download_webpage(
-            embed_page_url, video_id, note='downloading embed page')
-
-        # Get the video URL
-        m_playlist = re.search(r'so.addVariable\("playlist", ?"(?P<playlist>.+?)"\);', webpage)
-        if m_playlist is not None:
-            playlist_url = m_playlist.group('playlist')
-            playlist_page = self._download_webpage(playlist_url, video_id,
-                                                   'Downloading playlist page')
-            m_levels = list(re.finditer(r'<level bitrate="(\d+?)" file="(.*?)"', playlist_page))
-            if len(m_levels) == 0:
-                raise ExtractorError('Unable to extract video url')
-            videos = [(int(m.group(1)), m.group(2)) for m in m_levels]
-            (_, video_url) = sorted(videos)[0]
-            video_url = video_url.replace('%252F', '%2F')
+
+        title = self._html_search_regex(
+            r'<title>(.+?)</title>', webpage, 'title')
+
+        formats = []
+
+        encodings = self._parse_json(
+            self._search_regex(
+                r'encodings\s*=\s*(\[.+?\]);\n', webpage, 'encodings',
+                default='[]'),
+            video_id, fatal=False)
+        for encoding in encodings:
+            if not isinstance(encoding, dict):
+                continue
+            format_url = url_or_none(encoding.get('filename'))
+            if not format_url:
+                continue
+            if determine_ext(format_url) == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    format_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    m3u8_id='hls', fatal=False))
+            else:
+                format_id = encoding.get('name') or encoding.get('quality')
+                height = int_or_none(self._search_regex(
+                    r'^(\d+)[pP]', format_id, 'height', default=None))
+                formats.append({
+                    'url': format_url,
+                    'format_id': format_id,
+                    'height': height,
+                })
+
+        if formats:
+            info_dict = {
+                'formats': formats,
+            }
         else:
-            video_url = self._search_regex(r'so.addVariable\("file",encodeURIComponent\("(?P<source>[^"]+)"\)\);',
-                                           webpage, 'video URL')
+            # YouJizz's HTML5 player has invalid HTML
+            webpage = webpage.replace('"controls', '" controls')
+            info_dict = self._parse_html5_media_entries(
+                url, webpage, video_id)[0]
+
+        duration = parse_duration(self._search_regex(
+            r'<strong>Runtime:</strong>([^<]+)', webpage, 'duration',
+            default=None))
+        uploader = self._search_regex(
+            r'<strong>Uploaded By:.*?<a[^>]*>([^<]+)', webpage, 'uploader',
+            default=None)
 
-        return {
+        info_dict.update({
             'id': video_id,
-            'url': video_url,
-            'title': video_title,
-            'ext': 'flv',
-            'format': 'flv',
-            'player_url': embed_page_url,
-            'age_limit': age_limit,
-        }
+            'title': title,
+            'age_limit': self._rta_search(webpage),
+            'duration': duration,
+            'uploader': uploader,
+        })
+
+        return info_dict