[collegehumor] Support multiple formats (Fixes #2092)
authorPhilipp Hagemeister <phihag@phihag.de>
Sun, 5 Jan 2014 01:47:54 +0000 (02:47 +0100)
committerPhilipp Hagemeister <phihag@phihag.de>
Sun, 5 Jan 2014 01:50:10 +0000 (02:50 +0100)
Unfortunately, we lose a part of the description in the new JSON format, but that's still better than a non-functioning URL.

youtube_dl/extractor/collegehumor.py

index b27c1dfc52401f3c148d48d2b2897d2b06db3834..a963f6f79febbfa9fedd88ebf046a8ad840ff14f 100644 (file)
@@ -1,12 +1,7 @@
+import json
 import re
 
 from .common import InfoExtractor
-from ..utils import (
-    compat_urllib_parse_urlparse,
-    determine_ext,
-
-    ExtractorError,
-)
 
 
 class CollegeHumorIE(InfoExtractor):
@@ -15,16 +10,16 @@ class CollegeHumorIE(InfoExtractor):
     _TESTS = [{
         u'url': u'http://www.collegehumor.com/video/6902724/comic-con-cosplay-catastrophe',
         u'file': u'6902724.mp4',
-        u'md5': u'1264c12ad95dca142a9f0bf7968105a0',
+        u'md5': u'dcc0f5c1c8be98dc33889a191f4c26bd',
         u'info_dict': {
             u'title': u'Comic-Con Cosplay Catastrophe',
-            u'description': u'Fans get creative this year at San Diego.  Too creative.  And yes, that\'s really Joss Whedon.',
+            u'description': u'Fans get creative this year at San Diego.  Too',
         },
     },
     {
         u'url': u'http://www.collegehumor.com/video/3505939/font-conference',
         u'file': u'3505939.mp4',
-        u'md5': u'c51ca16b82bb456a4397987791a835f5',
+        u'md5': u'72fa701d8ef38664a4dbb9e2ab721816',
         u'info_dict': {
             u'title': u'Font Conference',
             u'description': u'This video wasn\'t long enough, so we made it double-spaced.',
@@ -33,50 +28,30 @@ class CollegeHumorIE(InfoExtractor):
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
-        if mobj is None:
-            raise ExtractorError(u'Invalid URL: %s' % url)
         video_id = mobj.group('videoid')
 
-        info = {
+        jsonUrl = 'http://www.collegehumor.com/moogaloop/video/' + video_id + '.json'
+        data = json.loads(self._download_webpage(
+            jsonUrl, video_id, u'Downloading info JSON'))
+        vdata = data['video']
+
+        PREFS = {'high_quality': 2, 'low_quality': 0}
+        formats = []
+        for format_key in ('mp4', 'webm'):
+            for qname, qurl in vdata[format_key].items():
+                formats.append({
+                    'format_id': format_key + '_' + qname,
+                    'url': qurl,
+                    'format': format_key,
+                    'preference': PREFS.get(qname),
+                })
+
+        self._sort_formats(formats)
+
+        return {
             'id': video_id,
-            'uploader': None,
-            'upload_date': None,
+            'title': vdata['title'],
+            'description': vdata.get('description'),
+            'thumbnail': vdata.get('thumbnail'),
+            'formats': formats,
         }
-
-        self.report_extraction(video_id)
-        xmlUrl = 'http://www.collegehumor.com/moogaloop/video/' + video_id
-        mdoc = self._download_xml(xmlUrl, video_id,
-                                         u'Downloading info XML',
-                                         u'Unable to download video info XML')
-
-        try:
-            videoNode = mdoc.findall('./video')[0]
-            youtubeIdNode = videoNode.find('./youtubeID')
-            if youtubeIdNode is not None:
-                return self.url_result(youtubeIdNode.text, 'Youtube')
-            info['description'] = videoNode.findall('./description')[0].text
-            info['title'] = videoNode.findall('./caption')[0].text
-            info['thumbnail'] = videoNode.findall('./thumbnail')[0].text
-            next_url = videoNode.findall('./file')[0].text
-        except IndexError:
-            raise ExtractorError(u'Invalid metadata XML file')
-
-        if next_url.endswith(u'manifest.f4m'):
-            manifest_url = next_url + '?hdcore=2.10.3'
-            adoc = self._download_xml(manifest_url, video_id,
-                                         u'Downloading XML manifest',
-                                         u'Unable to download video info XML')
-
-            try:
-                video_id = adoc.findall('./{http://ns.adobe.com/f4m/1.0}id')[0].text
-            except IndexError:
-                raise ExtractorError(u'Invalid manifest file')
-            url_pr = compat_urllib_parse_urlparse(info['thumbnail'])
-            info['url'] = url_pr.scheme + '://' + url_pr.netloc + video_id[:-2].replace('.csmil','').replace(',','')
-            info['ext'] = 'mp4'
-        else:
-            # Old-style direct links
-            info['url'] = next_url
-            info['ext'] = determine_ext(info['url'])
-
-        return info