[youtube] Fix extraction.

[youtube-dl] / youtube_dl / extractor / izlesene.py

diff --git a/youtube_dl/extractor/izlesene.py b/youtube_dl/extractor/izlesene.py
index d16d483eeb0d533debe041b7cd6c7b4826d41dde..f8fca6c8f47c9978e78d1c2c76d0dcb84a8d7f9d 100644 (file)
--- a/youtube_dl/extractor/izlesene.py
+++ b/youtube_dl/extractor/izlesene.py
@@ -1,9 +1,11 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
+from ..compat import (
+    compat_str,
+    compat_urllib_parse_unquote,
+)
 from ..utils import (
     determine_ext,
     float_or_none,
@@ -28,9 +30,9 @@ class IzleseneIE(InfoExtractor):
                 'ext': 'mp4',
                 'title': 'Sevinçten Çıldırtan Doğum Günü Hediyesi',
                 'description': 'md5:253753e2655dde93f59f74b572454f6d',
-                'thumbnail': 're:^http://.*\.jpg',
+                'thumbnail': r're:^https?://.*\.jpg',
                 'uploader_id': 'pelikzzle',
-                'timestamp': 1404302298,
+                'timestamp': int,
                 'upload_date': '20140702',
                 'duration': 95.395,
                 'age_limit': 0,
@@ -43,10 +45,9 @@ class IzleseneIE(InfoExtractor):
                 'id': '17997',
                 'ext': 'mp4',
                 'title': 'Tarkan Dortmund 2006 Konseri',
-                'description': 'Tarkan Dortmund 2006 Konseri',
-                'thumbnail': 're:^http://.*\.jpg',
+                'thumbnail': r're:^https://.*\.jpg',
                 'uploader_id': 'parlayankiz',
-                'timestamp': 1163322193,
+                'timestamp': int,
                 'upload_date': '20061112',
                 'duration': 253.666,
                 'age_limit': 0,
@@ -57,59 +58,50 @@ class IzleseneIE(InfoExtractor):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        url = 'http://www.izlesene.com/video/%s' % video_id
-        webpage = self._download_webpage(url, video_id)
+        webpage = self._download_webpage('http://www.izlesene.com/video/%s' % video_id, video_id)
+
+        video = self._parse_json(
+            self._search_regex(
+                r'videoObj\s*=\s*({.+?})\s*;\s*\n', webpage, 'streams'),
+            video_id)
+
+        title = video.get('videoTitle') or self._og_search_title(webpage)
+
+        formats = []
+        for stream in video['media']['level']:
+            source_url = stream.get('source')
+            if not source_url or not isinstance(source_url, compat_str):
+                continue
+            ext = determine_ext(url, 'mp4')
+            quality = stream.get('value')
+            height = int_or_none(quality)
+            formats.append({
+                'format_id': '%sp' % quality if quality else 'sd',
+                'url': compat_urllib_parse_unquote(source_url),
+                'ext': ext,
+                'height': height,
+            })
+        self._sort_formats(formats)
 
-        title = self._og_search_title(webpage)
-        description = self._og_search_description(webpage)
-        thumbnail = self._proto_relative_url(
+        description = self._og_search_description(webpage, default=None)
+        thumbnail = video.get('posterURL') or self._proto_relative_url(
             self._og_search_thumbnail(webpage), scheme='http:')
 
         uploader = self._html_search_regex(
             r"adduserUsername\s*=\s*'([^']+)';",
-            webpage, 'uploader', fatal=False, default='')
+            webpage, 'uploader', fatal=False)
         timestamp = parse_iso8601(self._html_search_meta(
-            'uploadDate', webpage, 'upload date', fatal=False))
+            'uploadDate', webpage, 'upload date'))
 
-        duration = float_or_none(self._html_search_regex(
-            r'"videoduration"\s*:\s*"([^"]+)"',
-            webpage, 'duration', fatal=False), scale=1000)
+        duration = float_or_none(video.get('duration') or self._html_search_regex(
+            r'videoduration["\']?\s*=\s*(["\'])(?P<value>(?:(?!\1).)+)\1',
+            webpage, 'duration', fatal=False, group='value'), scale=1000)
 
         view_count = str_to_int(get_element_by_id('videoViewCount', webpage))
         comment_count = self._html_search_regex(
             r'comment_count\s*=\s*\'([^\']+)\';',
             webpage, 'comment_count', fatal=False)
 
-        family_friendly = self._html_search_meta(
-            'isFamilyFriendly', webpage, 'age limit', fatal=False)
-
-        content_url = self._html_search_meta(
-            'contentURL', webpage, 'content URL', fatal=False)
-        ext = determine_ext(content_url, 'mp4')
-
-        # Might be empty for some videos.
-        streams = self._html_search_regex(
-            r'"qualitylevel"\s*:\s*"([^"]+)"',
-            webpage, 'streams', fatal=False, default='')
-
-        formats = []
-        if streams:
-            for stream in streams.split('|'):
-                quality, url = re.search(r'\[(\w+)\](.+)', stream).groups()
-                formats.append({
-                    'format_id': '%sp' % quality if quality else 'sd',
-                    'url': url,
-                    'ext': ext,
-                })
-        else:
-            stream_url = self._search_regex(
-                r'"streamurl"\s?:\s?"([^"]+)"', webpage, 'stream URL')
-            formats.append({
-                'format_id': 'sd',
-                'url': stream_url,
-                'ext': ext,
-            })
-
         return {
             'id': video_id,
             'title': title,
@@ -120,6 +112,6 @@ class IzleseneIE(InfoExtractor):
             'duration': duration,
             'view_count': int_or_none(view_count),
             'comment_count': int_or_none(comment_count),
-            'age_limit': 18 if family_friendly == 'False' else 0,
+            'age_limit': self._family_friendly_search(webpage),
             'formats': formats,
         }