Switch Vimeo to scraping: fixes #285
authorFilippo Valsorda <filippo.valsorda@gmail.com>
Mon, 5 Mar 2012 16:51:16 +0000 (17:51 +0100)
committerFilippo Valsorda <filippo.valsorda@gmail.com>
Mon, 5 Mar 2012 16:51:16 +0000 (17:51 +0100)
youtube-dl

index 77723155d3a65d586af8e6c31d5c3edeedcb3c04..6497490a0015ded39794b03ed3e6510b16185ad6 100755 (executable)
@@ -2058,7 +2058,7 @@ class VimeoIE(InfoExtractor):
                video_id = mobj.group(1)
 
                # Retrieve video webpage to extract further information
-               request = urllib2.Request("http://vimeo.com/moogaloop/load/clip:%s" % video_id, None, std_headers)
+               request = urllib2.Request(url, None, std_headers)
                try:
                        self.report_download_webpage(video_id)
                        webpage = urllib2.urlopen(request).read()
@@ -2071,77 +2071,66 @@ class VimeoIE(InfoExtractor):
                # and latter we extract those that are Vimeo specific.
                self.report_extraction(video_id)
 
-               # Extract title
-               mobj = re.search(r'<caption>(.*?)</caption>', webpage)
-               if mobj is None:
-                       self._downloader.trouble(u'ERROR: unable to extract video title')
+               # Extract the config JSON
+               config = webpage.split(' = {config:')[1].split(',assets:')[0]
+               try:
+                       config = json.loads(config)
+               except:
+                       self._downloader.trouble(u'ERROR: unable to extract info section')
                        return
-               video_title = mobj.group(1).decode('utf-8')
+               
+               # Extract title
+               video_title = config["video"]["title"]
                simple_title = _simplify_title(video_title)
 
                # Extract uploader
-               mobj = re.search(r'<uploader_url>http://vimeo.com/(.*?)</uploader_url>', webpage)
-               if mobj is None:
-                       self._downloader.trouble(u'ERROR: unable to extract video uploader')
-                       return
-               video_uploader = mobj.group(1).decode('utf-8')
+               video_uploader = config["video"]["owner"]["name"]
 
                # Extract video thumbnail
-               mobj = re.search(r'<thumbnail>(.*?)</thumbnail>', webpage)
-               if mobj is None:
-                       self._downloader.trouble(u'ERROR: unable to extract video thumbnail')
-                       return
-               video_thumbnail = mobj.group(1).decode('utf-8')
+               video_thumbnail = config["video"]["thumbnail"]
 
-               # # Extract video description
-               # mobj = re.search(r'<meta property="og:description" content="(.*)" />', webpage)
-               # if mobj is None:
-               #       self._downloader.trouble(u'ERROR: unable to extract video description')
-               #       return
-               # video_description = mobj.group(1).decode('utf-8')
-               # if not video_description: video_description = 'No description available.'
-               video_description = 'Foo.'
-
-               # Vimeo specific: extract request signature
-               mobj = re.search(r'<request_signature>(.*?)</request_signature>', webpage)
-               if mobj is None:
-                       self._downloader.trouble(u'ERROR: unable to extract request signature')
-                       return
-               sig = mobj.group(1).decode('utf-8')
+               # Extract video description
+               try:
+                       lxml.etree
+               except NameError:
+                       video_description = u'No description available.'
+                       mobj = re.search(r'<meta name="description" content="(.*?)" />', webpage, re.MULTILINE)
+                       if mobj is not None:
+                               video_description = mobj.group(1)
+               else:
+                       html_parser = lxml.etree.HTMLParser()
+                       vwebpage_doc = lxml.etree.parse(StringIO.StringIO(webpage), html_parser)
+                       video_description = u''.join(vwebpage_doc.xpath('id("description")//text()')).strip()
+                       # TODO use another parser
 
-               # Vimeo specific: extract video quality information
-               mobj = re.search(r'<isHD>(\d+)</isHD>', webpage)
-               if mobj is None:
-                       self._downloader.trouble(u'ERROR: unable to extract video quality information')
-                       return
-               quality = mobj.group(1).decode('utf-8')
+               # Extract upload date
+               video_upload_date = u'NA'
+               mobj = re.search(r'<span id="clip-date" style="display:none">[^:]*: (.*?)( \([^\(]*\))?</span>', webpage)
+               if mobj is not None:
+                       video_upload_date = mobj.group(1)
 
-               if int(quality) == 1:
-                       quality = 'hd'
-               else:
-                       quality = 'sd'
+               # Vimeo specific: extract request signature and timestamp
+               sig = config['request']['signature']
+               timestamp = config['request']['timestamp']
 
-               # Vimeo specific: Extract request signature expiration
-               mobj = re.search(r'<request_signature_expires>(.*?)</request_signature_expires>', webpage)
-               if mobj is None:
-                       self._downloader.trouble(u'ERROR: unable to extract request signature expiration')
-                       return
-               sig_exp = mobj.group(1).decode('utf-8')
+               # Vimeo specific: extract video quality information
+               # TODO bind to format param
+               if 'hd' in config["video"]["files"]["h264"]: quality = 'hd'
+               else: quality = 'sd'
 
-               video_url = "http://vimeo.com/moogaloop/play/clip:%s/%s/%s/?q=%s" % (video_id, sig, sig_exp, quality)
+               video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=H264&type=moogaloop_local&embed_location=" \
+                                       %(video_id, sig, timestamp, quality)
 
                try:
                        # Process video information
                        self._downloader.process_info({
-                               'id':           video_id.decode('utf-8'),
+                               'id':           video_id,
                                'url':          video_url,
                                'uploader':     video_uploader,
-                               'upload_date':  u'NA',
+                               'upload_date':  video_upload_date,
                                'title':        video_title,
                                'stitle':       simple_title,
                                'ext':          u'mp4',
-                               'thumbnail':    video_thumbnail.decode('utf-8'),
-                               'description':  video_description,
                                'thumbnail':    video_thumbnail,
                                'description':  video_description,
                                'player_url':   None,