write some subclasses of
RequestHandler and mapped these with routers, pass these routers to
from tornado.ioloop import IOLoop from tornado.web import RequestHandler, Application, url class HelloHandler(RequestHandler): def get(self): self.write("Hello, world") def make_app(): return Application([ url(r"/", HelloHandler), ]) def main(): app = make_app() app.listen(8888) IOLoop.current().start()
Tornado supports strings, bytes, and dictionaries as response, you can use Template to building strings and bytes, also you can write response by hand. If you pass a dictionary to response, it will be encoded as JSON.
Tornado will not parse other types request arguments, e.g. Tornado will not parse JSON request bodies, you need to parse these by yourself(override
CSRF/XSRF prevent: the session cookie from SiteA is not expired, there is a form in SiteB, the form aims doing something to SiteA, if SiteA don’t support XSRF prevent, it will allow submitting if you constructed valid form arguments. If SiteA supports XSRF prevent, it will set a cookie with token to user, and that token will filled in any action form(POST,PUT,DELETE), if these two values are not equal, that request will be prevented, because SiteB can’t retrive SiteA’s cookie, so that form can’t submit with valid cookie value.
Due to the Python GIL (Global Interpreter Lock), it is necessary to run multiple Python processes to take full advantage of multi-CPU machines. Typically it is best to run one process per CPU.