[internetvideoarchive] Fix extraction and support json URLs

youtube_dl/extractor/internetvideoarchive.py

@@ -1,93 +1,91 @@
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
 from ..compat import (
+    compat_parse_qs,
     compat_urlparse,
-    compat_urllib_parse_urlencode,
 )
 from ..utils import (
-    xpath_with_ns,
+    determine_ext,
+    int_or_none,
+    xpath_text,
 )
 
 
 class InternetVideoArchiveIE(InfoExtractor):
-    _VALID_URL = r'https?://video\.internetvideoarchive\.net/flash/players/.*?\?.*?publishedid.*?'
+    _VALID_URL = r'https?://video\.internetvideoarchive\.net/(?:player|flash/players)/.*?\?.*?publishedid.*?'
 
     _TEST = {
-        'url': 'http://video.internetvideoarchive.net/flash/players/flashconfiguration.aspx?customerid=69249&publishedid=452693&playerid=247',
+        'url': 'http://video.internetvideoarchive.net/player/6/configuration.ashx?customerid=69249&publishedid=194487&reporttag=vdbetatitle&playerid=641&autolist=0&domain=www.videodetective.com&maxrate=high&minrate=low&socialplayer=false',
         'info_dict': {
-            'id': '452693',
+            'id': '194487',
             'ext': 'mp4',
-            'title': 'SKYFALL',
-            'description': 'In SKYFALL, Bond\'s loyalty to M is tested as her past comes back to haunt her. As MI6 comes under attack, 007 must track down and destroy the threat, no matter how personal the cost.',
-            'duration': 152,
+            'title': 'KICK-ASS 2',
+            'description': 'md5:c189d5b7280400630a1d3dd17eaa8d8a',
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
         },
     }
 
     @staticmethod
-    def _build_url(query):
-        return 'http://video.internetvideoarchive.net/flash/players/flashconfiguration.aspx?' + query
+    def _build_json_url(query):
+        return 'http://video.internetvideoarchive.net/player/6/configuration.ashx?' + query
 
     @staticmethod
-    def _clean_query(query):
-        NEEDED_ARGS = ['publishedid', 'customerid']
-        query_dic = compat_urlparse.parse_qs(query)
-        cleaned_dic = dict((k, v[0]) for (k, v) in query_dic.items() if k in NEEDED_ARGS)
-        # Other player ids return m3u8 urls
-        cleaned_dic['playerid'] = '247'
-        cleaned_dic['videokbrate'] = '100000'
-        return compat_urllib_parse_urlencode(cleaned_dic)
+    def _build_xml_url(query):
+        return 'http://video.internetvideoarchive.net/flash/players/flashconfiguration.aspx?' + query
 
     def _real_extract(self, url):
         query = compat_urlparse.urlparse(url).query
-        query_dic = compat_urlparse.parse_qs(query)
+        query_dic = compat_parse_qs(query)
         video_id = query_dic['publishedid'][0]
-        url = self._build_url(query)
 
-        flashconfiguration = self._download_xml(url, video_id,
-                                                'Downloading flash configuration')
-        file_url = flashconfiguration.find('file').text
-        file_url = file_url.replace('/playlist.aspx', '/mrssplaylist.aspx')
-        # Replace some of the parameters in the query to get the best quality
-        # and http links (no m3u8 manifests)
-        file_url = re.sub(r'(?<=\?)(.+)$',
-                          lambda m: self._clean_query(m.group()),
-                          file_url)
-        info = self._download_xml(file_url, video_id,
-                                  'Downloading video info')
-        item = info.find('channel/item')
+        if '/player/' in url:
+            configuration = self._download_json(url, video_id)
+
+            # There are multiple videos in the playlist whlie only the first one
+            # matches the video played in browsers
+            video_info = configuration['playlist'][0]
+
+            formats = []
+            for source in video_info['sources']:
+                file_url = source['file']
+                if determine_ext(file_url) == 'm3u8':
+                    formats.extend(self._extract_m3u8_formats(
+                        file_url, video_id, ext='mp4', m3u8_id='hls'))
+                else:
+                    a_format = {
+                        'url': file_url,
+                    }
+
+                    if source.get('label') and source['label'][-4:] == ' kbs':
+                        tbr = int_or_none(source['label'][:-4])
+                        a_format.update({
+                            'tbr': tbr,
+                            'format_id': 'http-%d' % tbr,
+                        })
+                        formats.append(a_format)
 
-        def _bp(p):
-            return xpath_with_ns(
-                p,
-                {
-                    'media': 'http://search.yahoo.com/mrss/',
-                    'jwplayer': 'http://developer.longtailvideo.com/trac/wiki/FlashFormats',
-                }
-            )
-        formats = []
-        for content in item.findall(_bp('media:group/media:content')):
-            attr = content.attrib
-            f_url = attr['url']
-            width = int(attr['width'])
-            bitrate = int(attr['bitrate'])
-            format_id = '%d-%dk' % (width, bitrate)
-            formats.append({
-                'format_id': format_id,
-                'url': f_url,
-                'width': width,
-                'tbr': bitrate,
-            })
+            self._sort_formats(formats)
 
-        self._sort_formats(formats)
+            title = video_info['title']
+            description = video_info.get('description')
+            thumbnail = video_info.get('image')
+        else:
+            configuration = self._download_xml(url, video_id)
+            formats = [{
+                'url': xpath_text(configuration, './file', 'file URL', fatal=True),
+            }]
+            thumbnail = xpath_text(configuration, './image', 'thumbnail')
+            title = 'InternetVideoArchive video %s' % video_id
+            description = None
 
         return {
             'id': video_id,
-            'title': item.find('title').text,
+            'title': title,
             'formats': formats,
-            'thumbnail': item.find(_bp('media:thumbnail')).attrib['url'],
-            'description': item.find('description').text,
-            'duration': int(attr['duration']),
+            'thumbnail': thumbnail,
+            'description': description,
         }