a little HTTP server in Python
proj1 contains the initial API implementation with a test-case which hosts a simple website with a semi-realtime chat, DM system and user database. The API consists of the following functions:
HttpServer.__init__(root_dir, error_dir, host, port, logger_file=None, logger_folder="")HttpServer.add_route(path, handlers)HttpServer.handle_http_requests(worker_thd, *, backlog=5, proxy_worker_thd=lambda fn, *args, **kwargs: fn(*args, **kwargs), halt_on_ret=False, timeout=2)
proj2 is the refined API, with the following API functions:
HttpServer.__init__(root_dir, host, port, logger_file=None, max_conn=10)@staticmethod HttpServer.parse_http_request(data)HttpServer.add_route(methods_supported, path, handler)HttpServer.redirect_route(src_path, dst_path, *, inherit_methods=False)HttpServer.remove_route(path)HttpServer.get_route(conn, addr, method, path)HttpServer.handle_http_connections()HttpServer.close_connections()
Both variants are based from socket-level, using socket alone with delegating instances of threading.Thread per request, maintaing (probably) a persistent TCP connection, enforcing the keep-alive standard where necessary. However, neither projects strictly abide RFC 2616 or any such semantic definitions of grammars such as the URI, GET/POST parameters, etc..
At the moment (4/12/20), both the proj1 and proj2 variants have test-cases displaying their usage. proj2 in particular may be seen below, running locally with the filestructure currently uploaded in this commit.
