StacklessHTTPServer page

# # An example which adapts the standard library module BaseHTTPServer to # handle concurrent requests each on their own tasklet (as opposed to # using the ThreadingMixIn from the SocketServer module). # # Author: Richard Tew # # This code was written to serve as an example of Stackless Python usage. # Feel free to email me with any questions, comments, or suggestions for # improvement. # # But a better place to discuss Stackless Python related matters is the # mailing list: # # http://www.tismer.com/mailman/listinfo/stackless # # Monkeypatch in the stacklesssocket module. import sys, time import stackless import stacklesssocket #sys.modules["socket"] = stacklesssocket stacklesssocket.install() from BaseHTTPServer import BaseHTTPRequestHandler, HTTPServer body = """ StacklessHTTPServer page Nothing to see here, move along.. """ delays = [ 20, 5 ] delayIdx = 0 class RequestHandler(BaseHTTPRequestHandler): # Respect keep alive requests. protocol_version = "HTTP/1.1" def do_GET(self): global delayIdx self.send_response(200) self.send_header("Content-type", "text/html") self.send_header("Content-Length", len(body)) self.end_headers() t = time.time() delay = delays[delayIdx % len(delays)] delayIdx += 1 print stackless.getcurrent(), "blocked", delayIdx, "delay", delay while time.time() < t + delay: stackless.schedule() print stackless.getcurrent(), "done" self.wfile.write(body) class StacklessHTTPServer(HTTPServer): def handle_request(self): try: request, client_address = self.get_request() except socket.error: return stackless.tasklet(self.handle_request_tasklet)(request, client_address) def handle_request_tasklet(self, request, client_address): if self.verify_request(request, client_address): try: self.process_request(request, client_address) except: self.handle_error(request, client_address) self.close_request(request) def Run(): server = StacklessHTTPServer(('', 80), RequestHandler) server.serve_forever() if __name__ == "__main__": stackless.tasklet(Run)() stackless.run()