[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/tweakers.py b/youtube_dl/extractor/tweakers.py
index 3080f9e467d2cafa7397dddc5ca511db039e09bc..2b10d9bcaec909caa303bb33c1621527a3299797 100644 (file)
--- a/youtube_dl/extractor/tweakers.py
+++ b/youtube_dl/extractor/tweakers.py
@@ -1,41 +1,62 @@
-# coding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
+from ..utils import (
+    int_or_none,
+    determine_ext,
+    mimetype2ext,
+)
 
 
 class TweakersIE(InfoExtractor):
-    _VALID_URL = r'https?://tweakers\.net/video/(?P<id>[0-9]+).*'
+    _VALID_URL = r'https?://tweakers\.net/video/(?P<id>\d+)'
     _TEST = {
         'url': 'https://tweakers.net/video/9926/new-nintendo-3ds-xl-op-alle-fronten-beter.html',
-        'md5': 'f7f7f3027166a7f32f024b4ae6571ced',
+        'md5': 'fe73e417c093a788e0160c4025f88b15',
         'info_dict': {
             'id': '9926',
             'ext': 'mp4',
-            'title': 'New-Nintendo-3Ds-Xl-Op-Alle-Fronten-Beter',
-            # TODO more properties, either as:
-            # * A value
-            # * MD5 checksum; start the string with md5:
-            # * A regular expression; start the string with re:
-            # * Any Python type (for example int or float)
+            'title': 'New Nintendo 3DS XL - Op alle fronten beter',
+            'description': 'md5:3789b21fed9c0219e9bcaacd43fab280',
+            'thumbnail': r're:^https?://.*\.jpe?g$',
+            'duration': 386,
+            'uploader_id': 's7JeEm',
         }
     }
 
     def _real_extract(self, url):
-        splitted_url = re.split('.html|/', url)
-        del splitted_url[-1] # To remove extra '/' at the end
-        video_id = splitted_url[4]
-        title = splitted_url[5].title() # Retrieve title for URL and capitalize
-        splitted_url[3] = splitted_url[3] + '/player' # Add /player to get the player page
-        player_url = '/'.join(splitted_url) + '.html'
-        player_page = self._download_webpage(player_url, video_id)
+        video_id = self._match_id(url)
+        video_data = self._download_json(
+            'https://tweakers.net/video/s1playlist/%s/1920/1080/playlist.json' % video_id,
+            video_id)['items'][0]
+
+        title = video_data['title']
+
+        formats = []
+        for location in video_data.get('locations', {}).get('progressive', []):
+            format_id = location.get('label')
+            width = int_or_none(location.get('width'))
+            height = int_or_none(location.get('height'))
+            for source in location.get('sources', []):
+                source_url = source.get('src')
+                if not source_url:
+                    continue
+                ext = mimetype2ext(source.get('type')) or determine_ext(source_url)
+                formats.append({
+                    'format_id': format_id,
+                    'url': source_url,
+                    'width': width,
+                    'height': height,
+                    'ext': ext,
+                })
+        self._sort_formats(formats)
 
         return {
             'id': video_id,
-            'ext': 'mp4',
             'title': title,
-            'url':  re.findall('http.*mp4', player_page)[0],
-            'player_url': player_url
+            'description': video_data.get('description'),
+            'thumbnail': video_data.get('poster'),
+            'duration': int_or_none(video_data.get('duration')),
+            'uploader_id': video_data.get('account'),
+            'formats': formats,
         }