http协议、web服务器、并发服务器（上）

目录

• 1. HTTP格式
  • 1.1 HTTP GET请求的格式：
  • 1.2 HTTP POST请求的格式：
  • 1.3 HTTP响应的格式：
• 2. Web静态服务器-显示固定的页面
• 3. Web静态服务器-显示需要的页面
• 4. Web静态服务器-多进程版
• 5. Web静态服务器-多线程版
• 6. Web静态服务器-gevent版

1. HTTP格式

每个HTTP请求和响应都遵循相同的格式，一个HTTP包含Header和Body两部分，其中Body是可选的。HTTP协议是一种文本协议，所以，它的格式也非常简单。

1.1 HTTP GET请求的格式：

GET /path HTTP/1.1
Header1: Value1
Header2: Value2
Header3: Value3

1.2 HTTP POST请求的格式：

POST /path HTTP/1.1
Header1: Value1
Header2: Value2
Header3: Value3

body data goes here...

当遇到连续两个\r\n时，Header部分结束，后面的数据全部是Body。

1.3 HTTP响应的格式：

HTTP/1.1 200 OK
Header1: Value1
Header2: Value2
Header3: Value3

body data goes here...

HTTP响应如果包含body，也是通过\r\n\r\n来分隔的。

请再次注意，Body的数据类型由Content-Type头来确定，如果是网页，Body就是文本，如果是图片，Body就是图片的二进制数据。

当存在Content-Encoding时，Body数据是被压缩的，最常见的压缩方式是gzip，所以，看到Content-Encoding: gzip时，需要将Body数据先解压缩，才能得到真正的数据。压缩的目的在于减少Body的大小，加快网络传输。

2. Web静态服务器-显示固定的页面

import socket

def handle_client(new_client):
        """处理客户端请求"""

        recv_data = new_client.recv(1024)

        print(recv_data)

        # 组装响应的内容
        response_headers = "HTTP/1.1 200 OK\r\n"
        response_headers += "\r\n"
        response_body = "6666"

        response = response_headers + response_body
        new_client.send(response.encode("utf-8"))

        new_client.close()

def main():

        # 创建套接字
        tcp_socket_server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)

        # 绑定本地消息
        tcp_socket_server.bind(("", 8090))

        # 套接字由主动变被动
        tcp_socket_server.listen(128)

        while True:

                # 接收新的请求
                new_client, client_addr = tcp_socket_server.accept()

                handle_client(new_client)

        # 关闭套接字
        tcp_socket_server.close()

if __name__ == "__main__":

        main()

如上的代码仅仅只是向浏览器发送了简单的文本内容：6666

在浏览器中访问：

3. Web静态服务器-显示需要的页面

import socket
import re

def handle_client(new_client):
	"""处理客户端请求"""

	recv_data = new_client.recv(1024).decode("utf-8")

	# GET / HTTP/1.1
	request_header_lines = recv_data.splitlines()
	for line in request_header_lines:
		print(line)
	http_request_line = request_header_lines[0];
	get_file_name = re.match("[^/]+(/[^ ]*)", http_request_line).group(1)
	print("file name is -> " + get_file_name)	

	# print(recv_data)
	if get_file_name == "/":
		get_file_name = "/index.html"	

	root_path = "./html"
	get_file_name = root_path + get_file_name;
	try:
		f = open(get_file_name, "rb")
	except:
		response_headers = "HTTP/1.1 404 NOT FOUND\r\n"
		response_headers += "\r\n"
		response_body = b"page not found"
	else:	

		# 组装响应的内容
		response_headers = "HTTP/1.1 200 OK\r\n"
		response_headers += "\r\n"
		response_body = f.read()
		f.close()

	new_client.send(response_headers.encode("utf-8"))
	new_client.send(response_body)

	new_client.close()

def main():

	# 创建套接字
	tcp_socket_server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
	tcp_socket_server.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)	

	# 绑定本地消息
	tcp_socket_server.bind(("", 8090))

	# 套接字由主动变被动
	tcp_socket_server.listen(128)

	while True:

		# 接收新的请求
		new_client, client_addr = tcp_socket_server.accept()

		handle_client(new_client)

	# 关闭套接字
	tcp_socket_server.close()

if __name__ == "__main__":

	main()

在浏览器中访问：

4. Web静态服务器-多进程版

import socket
import re
import multiprocessing

def handle_client(new_client):
	"""处理客户端请求"""

	recv_data = new_client.recv(1024).decode("utf-8")

	# GET / HTTP/1.1
	request_header_lines = recv_data.splitlines()
	for line in request_header_lines:
		print(line)
	http_request_line = request_header_lines[0];
	get_file_name = re.match("[^/]+(/[^ ]*)", http_request_line).group(1)
	print("file name is -> " + get_file_name)	

	# print(recv_data)
	if get_file_name == "/":
		get_file_name = "/index.html"	

	root_path = "./html"
	get_file_name = root_path + get_file_name;
	try:
		f = open(get_file_name, "rb")
	except:
		response_headers = "HTTP/1.1 404 NOT FOUND\r\n"
		response_headers += "\r\n"
		response_body = b"page not found"
	else:	

		# 组装响应的内容
		response_headers = "HTTP/1.1 200 OK\r\n"
		response_headers += "\r\n"
		response_body = f.read()
		f.close()

	new_client.send(response_headers.encode("utf-8"))
	new_client.send(response_body)

	new_client.close()

def main():

	# 创建套接字
	tcp_socket_server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
	tcp_socket_server.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)	

	# 绑定本地消息
	tcp_socket_server.bind(("", 8090))

	# 套接字由主动变被动
	tcp_socket_server.listen(128)

	while True:

		# 接收新的请求
		new_client, client_addr = tcp_socket_server.accept()

		p = multiprocessing.Process(target=handle_client, args=(new_client,))
		p.start()

		new_client.close()

	# 关闭套接字
	tcp_socket_server.close()

if __name__ == "__main__":

	main()

5. Web静态服务器-多线程版

import socket
import re
import threading

def handle_client(new_client):
	"""处理客户端请求"""

	recv_data = new_client.recv(1024).decode("utf-8")

	# GET / HTTP/1.1
	request_header_lines = recv_data.splitlines()
	for line in request_header_lines:
		print(line)
	http_request_line = request_header_lines[0];
	get_file_name = re.match("[^/]+(/[^ ]*)", http_request_line).group(1)
	print("file name is -> " + get_file_name)	

	# print(recv_data)
	if get_file_name == "/":
		get_file_name = "/index.html"	

	root_path = "./html"
	get_file_name = root_path + get_file_name;
	try:
		f = open(get_file_name, "rb")
	except:
		response_headers = "HTTP/1.1 404 NOT FOUND\r\n"
		response_headers += "\r\n"
		response_body = b"page not found"
	else:	

		# 组装响应的内容
		response_headers = "HTTP/1.1 200 OK\r\n"
		response_headers += "\r\n"
		response_body = f.read()
		f.close()

	new_client.send(response_headers.encode("utf-8"))
	new_client.send(response_body)

	new_client.close()

def main():

	# 创建套接字
	tcp_socket_server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
	tcp_socket_server.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)	

	# 绑定本地消息
	tcp_socket_server.bind(("", 8090))

	# 套接字由主动变被动
	tcp_socket_server.listen(128)

	while True:

		# 接收新的请求
		new_client, client_addr = tcp_socket_server.accept()

		t = threading.Thread(target=handle_client, args=(new_client,))
		t.start()

	# 关闭套接字
	tcp_socket_server.close()

if __name__ == "__main__":

	main()

6. Web静态服务器-gevent版

import socket
import re
import gevent
from gevent import monkey

monkey.patch_all()

def handle_client(new_client):
	"""处理客户端请求"""

	recv_data = new_client.recv(1024).decode("utf-8")

	# GET / HTTP/1.1
	request_header_lines = recv_data.splitlines()
	for line in request_header_lines:
		print(line)
	http_request_line = request_header_lines[0];
	get_file_name = re.match("[^/]+(/[^ ]*)", http_request_line).group(1)
	print("file name is -> " + get_file_name)	

	# print(recv_data)
	if get_file_name == "/":
		get_file_name = "/index.html"	

	root_path = "./html"
	get_file_name = root_path + get_file_name;
	try:
		f = open(get_file_name, "rb")
	except:
		response_headers = "HTTP/1.1 404 NOT FOUND\r\n"
		response_headers += "\r\n"
		response_body = b"page not found"
	else:	

		# 组装响应的内容
		response_headers = "HTTP/1.1 200 OK\r\n"
		response_headers += "\r\n"
		response_body = f.read()
		f.close()

	new_client.send(response_headers.encode("utf-8"))
	new_client.send(response_body)

	new_client.close()

def main():

	# 创建套接字
	tcp_socket_server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
	tcp_socket_server.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)	

	# 绑定本地消息
	tcp_socket_server.bind(("", 8090))

	# 套接字由主动变被动
	tcp_socket_server.listen(128)

	while True:

		# 接收新的请求
		new_client, client_addr = tcp_socket_server.accept()

		g = gevent.spawn(handle_client, new_client)
		g.join()

	# 关闭套接字
	tcp_socket_server.close()

if __name__ == "__main__":

	main()