cacheapi.py 文件源码

def cache_resource(self, url):
        if self.proxy_url is not None:
            proxy = urllib2.ProxyHandler({'http': self.proxy_url})
            opener = urllib2.build_opener(proxy)
            urllib2.install_opener(opener)
        request = urllib2.Request(url)
        user_agent = 'Mozilla/5.0 (X11; Linux i686) AppleWebKit/535.1 (KHTML, like Gecko) Chrome/14.0.835.35 Safari/535.1'
        request.add_header('User-Agent', user_agent)
        handler = urllib2.urlopen(request, timeout=self.http_timeout)
        try:
            resource_type = MIME_TYPES[handler.headers.get('Content-Type')]
            if not resource_type:
                raise UnsupportedResourceFormat("Resource format not found")
        except KeyError:
            raise UnsupportedResourceFormat("Resource format not supported")
        etag = handler.headers.get('ETag')
        last_modified = handler.headers.get('Last-Modified')
        resource_key = self.get_resource_key(url)
        stream = handler.read()
        self.update_resource_params(resource_key, resource_type, etag, last_modified, stream)
        return stream, resource_type