+ 'id': video_id.decode('utf-8'),
+ 'url': video_real_url.decode('utf-8'),
+ 'uploader': video_uploader.decode('utf-8'),
+ 'title': video_title,
+ 'stitle': simple_title,
+ 'ext': video_extension.decode('utf-8'),
+ }]
+
+class MetacafeIE(InfoExtractor):
+ """Information Extractor for metacafe.com."""
+
+ _VALID_URL = r'(?:http://)?(?:www\.)?metacafe\.com/watch/([^/]+)/([^/]+)/.*'
+ _DISCLAIMER = 'http://www.metacafe.com/family_filter/'
+ _youtube_ie = None
+
+ def __init__(self, youtube_ie, downloader=None):
+ InfoExtractor.__init__(self, downloader)
+ self._youtube_ie = youtube_ie
+
+ @staticmethod
+ def suitable(url):
+ return (re.match(MetacafeIE._VALID_URL, url) is not None)
+
+ def report_disclaimer(self):
+ """Report disclaimer retrieval."""
+ self.to_stdout(u'[metacafe] Retrieving disclaimer')
+
+ def report_age_confirmation(self):
+ """Report attempt to confirm age."""
+ self.to_stdout(u'[metacafe] Confirming age')
+
+ def report_download_webpage(self, video_id):
+ """Report webpage download."""
+ self.to_stdout(u'[metacafe] %s: Downloading webpage' % video_id)
+
+ def report_extraction(self, video_id):
+ """Report information extraction."""
+ self.to_stdout(u'[metacafe] %s: Extracting information' % video_id)
+
+ def _real_initialize(self):
+ # Retrieve disclaimer
+ request = urllib2.Request(self._DISCLAIMER, None, std_headers)
+ try:
+ self.report_disclaimer()
+ disclaimer = urllib2.urlopen(request).read()
+ except (urllib2.URLError, httplib.HTTPException, socket.error), err:
+ self.to_stderr(u'ERROR: unable to retrieve disclaimer: %s' % str(err))
+ return
+
+ # Confirm age
+ disclaimer_form = {
+ 'filters': '0',
+ 'submit': "Continue - I'm over 18",
+ }
+ request = urllib2.Request('http://www.metacafe.com/', urllib.urlencode(disclaimer_form), std_headers)
+ try:
+ self.report_age_confirmation()
+ disclaimer = urllib2.urlopen(request).read()
+ except (urllib2.URLError, httplib.HTTPException, socket.error), err:
+ self.to_stderr(u'ERROR: unable to confirm age: %s' % str(err))
+ return
+
+ def _real_extract(self, url):
+ # Extract id and simplified title from URL
+ mobj = re.match(self._VALID_URL, url)
+ if mobj is None:
+ self.to_stderr(u'ERROR: invalid URL: %s' % url)
+ return [None]
+
+ video_id = mobj.group(1)
+
+ # Check if video comes from YouTube
+ mobj2 = re.match(r'^yt-(.*)$', video_id)
+ if mobj2 is not None:
+ return self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % mobj2.group(1))
+
+ simple_title = mobj.group(2).decode('utf-8')
+ video_extension = 'flv'
+
+ # Retrieve video webpage to extract further information
+ request = urllib2.Request('http://www.metacafe.com/watch/%s/' % video_id)
+ try:
+ self.report_download_webpage(video_id)
+ webpage = urllib2.urlopen(request).read()
+ except (urllib2.URLError, httplib.HTTPException, socket.error), err:
+ self.to_stderr(u'ERROR: unable retrieve video webpage: %s' % str(err))
+ return [None]
+
+ # Extract URL, uploader and title from webpage
+ self.report_extraction(video_id)
+ mobj = re.search(r'(?m)"mediaURL":"(http.*?\.flv)"', webpage)
+ if mobj is None:
+ self.to_stderr(u'ERROR: unable to extract media URL')
+ return [None]
+ mediaURL = mobj.group(1).replace('\\', '')
+
+ mobj = re.search(r'(?m)"gdaKey":"(.*?)"', webpage)
+ if mobj is None:
+ self.to_stderr(u'ERROR: unable to extract gdaKey')
+ return [None]
+ gdaKey = mobj.group(1)
+
+ video_url = '%s?__gda__=%s' % (mediaURL, gdaKey)
+
+ mobj = re.search(r'(?im)<title>(.*) - Video</title>', webpage)
+ if mobj is None:
+ self.to_stderr(u'ERROR: unable to extract title')
+ return [None]
+ video_title = mobj.group(1).decode('utf-8')
+
+ mobj = re.search(r'(?m)<li id="ChnlUsr">.*?Submitter:<br />(.*?)</li>', webpage)
+ if mobj is None:
+ self.to_stderr(u'ERROR: unable to extract uploader nickname')
+ return [None]
+ video_uploader = re.sub(r'<.*?>', '', mobj.group(1))
+
+ # Return information
+ return [{
+ 'id': video_id.decode('utf-8'),
+ 'url': video_url.decode('utf-8'),
+ 'uploader': video_uploader.decode('utf-8'),