在解析socketserver是如工作之前,我们先看看socektserver类的继承关系图:
请求类继承关系:
server类继承关系:
有了上面的继承关系图后,我们解析socketserver就轻松多了,下面,我们从代码开始,慢慢揭开socketserver面纱:
import socketserver import struct, json, os class FtpServer(socketserver.BaseRequestHandler): coding = 'utf-8' server_dir = 'file_upload' max_packet_size = 1024 BASE_DIR = os.path.dirname(os.path.abspath(__file__)) def handle(self): print(self.request) while True: data = self.request.recv(4) data_len = struct.unpack('i', data)[0] head_json = self.request.recv(data_len).decode(self.coding) head_dic = json.loads(head_json) cmd = head_dic['cmd'] if hasattr(self, cmd): func = getattr(self, cmd) func(head_dic) def put(self): pass def get(self): pass if __name__ == '__main__': HOST, PORT = "localhost", 9999 with socketserver.ThreadingTCPServer((HOST, PORT), FtpServer) as server: server.serve_forever()
我们通过socketserver.ThreadingTCPServer实例化对象server,那么此时应用调用类的__init__方法,前往ThreadingTCPServer类看看:
class ThreadingTCPServer(ThreadingMixIn, UDPServer): pass
发现這个类啥都没写,我们知道,如果一个类什么方法都没有定义,那么它的方法肯定都是从其父类继承而来,接着,先到ThreadingMinIn里面看看,
class ThreadingMixIn: daemon_threads = False def process_request_thread(self, request, client_address): passdef process_request(self, request, client_address): pass
这个类也没有__init__方法,因此,我们应该去右继承的父类TCPserver中找:
class TCPServer(BaseServer): address_family = socket.AF_INET socket_type = socket.SOCK_STREAM request_queue_size = 5 allow_reuse_address = False def __init__(self, server_address, RequestHandlerClass, bind_and_activate=True): BaseServer.__init__(self, server_address, RequestHandlerClass)# self.socket = socket.socket(self.address_family,self.socket_type) # 创建套接字对象 if bind_and_activate: try: self.server_bind() #绑定端口和IP self.server_activate() # 监听端口 except: self.server_close() raise
看到Tcpserver的__init__方法,完成了以下几件事:
创建套接字,绑定端口和IP,并监听
将端口、IP和我们创建类传递到Baseserver类中;
此时,对象的初始化工作并没有完成,接着,我们要进入baseserver类,看看该类下的__init__完成了什么工作:
class BaseServer: timeout = None def __init__(self, server_address, RequestHandlerClass): self.server_address = server_address #将端口和IP暂存 self.RequestHandlerClass = RequestHandlerClass #暂存我们创建的类 self.__is_shut_down = threading.Event() # 创建event对象
到此,对象的初始化工作完成。然后是调用serve_forever()方法,开始不断循环监听。下面,我们来看看,这个server_forever实现
注意:我们要清楚一点,我们在找這个方法在哪里的时候,一定要按照顺序去找,也就是说,我们先得从子类开始找,如果子类不存在,就去其父类找。下面我们就遵循這个原则来找找看。
先来看看左继承的父类ThreadingMixIn中有没有server_forever:
class ThreadingMixIn: daemon_threads = False def process_request_thread(self, request, client_address): try: self.finish_request(request, client_address) except Exception: self.handle_error(request, client_address) finally: self.shutdown_request(request) def process_request(self, request, client_address): t = threading.Thread(target = self.process_request_thread, args = (request, client_address)) t.daemon = self.daemon_threads t.start()
再来看看父类Tcpserver:
class TCPServer(BaseServer):def __init__(self, server_address, RequestHandlerClass, bind_and_activate=True): def server_bind(self): def server_activate(self): def server_close(self): def fileno(self): def get_request(self): def shutdown_request(self, request): def close_request(self, request):
我们发现,没有server_forever方法,好,我去其继承的父类BaseServer类看看:
class BaseServer:def __init__(self, server_address, RequestHandlerClass): def server_activate(self): def serve_forever(self, poll_interval=0.5): def shutdown(self): def service_actions(self): def handle_request(self): def _handle_request_noblock(self): def handle_timeout(self): def verify_request(self, request, client_address): def process_request(self, request, client_address): def server_close(self): def finish_request(self, request, client_address): def shutdown_request(self, request): def close_request(self, request): def handle_error(self, request, client_address): def __enter__(self): def __exit__(self, *args):
我们发现server_forever()果然在這个类中,现在,我们的目标是:找到在什么地方调用我们自己写的handle方法。
在我们找到的server_forever()方法中,
def serve_forever(self, poll_interval=0.5): self.__is_shut_down.clear() try: with _ServerSelector() as selector: selector.register(self, selectors.EVENT_READ)#原来底层是用epoll来实现不断循环监听 while not self.__shutdown_request: ready = selector.select(poll_interval) #有新的链接进来 if ready: self._handle_request_noblock() # 这里应该是处理新的链接 self.service_actions() finally: self.__shutdown_request = False self.__is_shut_down.set()
好,我大致找到了链接的处理入口,我们跟进去,继续寻找:
def _handle_request_noblock(self): try: request, client_address = self.get_request() except OSError: return if self.verify_request(request, client_address): try: self.process_request(request, client_address)#注意这里的process_request() except Exception: self.handle_error(request, client_address) self.shutdown_request(request) except: self.shutdown_request(request) raise else: self.shutdown_request(request)
到源码中,我们找到该函数,现在,只看我划线的部分。其他部分都是针对异常的处理,如果没有异常,其他都是不会执行的,所以,其他的异常处理,我们先暂时不看。
我们发现,如果有链接,最后会交给process_request()(我们会发现,在baseserver类和ThreadingMixIn都有這个方法,这里找类方法,一定要按照类的继承顺序来查找),所以,我们到ThreadingMiXin中去看看processs_request()做了哪些事情:
def process_request(self, request, client_address): t = threading.Thread(target = self.process_request_thread,args = (request, client_address)) # 原来开了一个线程,支持并发 t.daemon = self.daemon_threads # 开启守护线程 t.start()
在线程中执行该类下的process_requsest_thread()方法,
def process_request_thread(self, request, client_address): try: self.finish_request(request, client_address) except Exception: self.handle_error(request, client_address) finally: self.shutdown_request(request)
到此为止,链接建立成功!
下面,我们来看看,当有消息发送,是如何进行处理的。
当有消息发送,selector监听到了,
def serve_forever(self, poll_interval=0.5): self.__is_shut_down.clear() try: with _ServerSelector() as selector: selector.register(self, selectors.EVENT_READ)# 监听了活动链接 while not self.__shutdown_request: ready = selector.select(poll_interval) if ready: # 准备好了 self._handle_request_noblock() # 进入处理 self.service_actions() finally: self.__shutdown_request = False self.__is_shut_down.set()
下面我们跟进_handle_request_noblock(),
def _handle_request_noblock(self): try: request, client_address = self.get_request() except OSError: return if self.verify_request(request, client_address): try: self.process_request(request, client_address) except Exception: self.handle_error(request, client_address) self.shutdown_request(request) except: self.shutdown_request(request) raise else: self.shutdown_request(request)
我们到process_request()看看:
def process_request(self, request, client_address): """Start a new thread to process the request.""" t = threading.Thread(target = self.process_request_thread, # start a threading to handle the request args = (request, client_address)) t.daemon = self.daemon_threads t.start()
然后开启线程执行,process_request_thread()方法,
def process_request_thread(self, request, client_address): try: self.finish_request(request, client_address) # -----> to Baseserver find except Exception: self.handle_error(request, client_address) finally: self.shutdown_request(request)
然后调用finish_request()方法,现在我们跟进看看,
def finish_request(self, request, client_address): self.RequestHandlerClass(request, client_address, self)
执行了RequestHandlerClass(request, client_address, self),這个是啥??还记得最开始我们传进来的类保存在哪呢?没错,就是RequestHandlerClass里面,现在这里才开始实例化這个类,也就是说,在这里开始调用我们自己的类了。既然是调用我们自己的类,那么必然要实例化,我们先回到自己创建的类,找找__init__方法。
class MyTCPHandler(socketserver.BaseRequestHandler): def handle(self): self.data = self.request.recv(1024).strip() print("{} wrote:".format(self.client_address[0])) print(self.data) self.request.sendall(self.data.upper())
自己类没有写__init__方法,那么我去它继承的BaseRequestHandler()下面找找看:
class BaseRequestHandler: def __init__(self, request, client_address, server): self.request = request # 接受传进来的请求链接 self.client_address = client_address # 客户端的ip/端口 self.server = server # self.setup() try: self.handle() finally: self.finish() def setup(self): pass def handle(self): pass def finish(self): pass
我们来看看,它继承类实例化完成了哪些操作:
调用handle()方法,我们发现,在这个类中也有一个handle()方法,那么这里调用时调用自己写的还是這个类中的呢?
当然是调用我们自己写!
至此,我们完成了一次通信的完整过程!
总结sockerserver整个流程:
1.开启了线程,支持并发操作
2.I/O多路复用,监听多个文件描述符!