http協議、web服務器、並發服務器（上）

目錄

• 1. HTTP格式
  • 1.1 HTTP GET請求的格式：
  • 1.2 HTTP POST請求的格式：
  • 1.3 HTTP響應的格式：
• 2. Web靜態服務器-顯示固定的頁面
• 3. Web靜態服務器-顯示需要的頁面
• 4. Web靜態服務器-多進程版
• 5. Web靜態服務器-多線程版
• 6. Web靜態服務器-gevent版

1. HTTP格式

每個HTTP請求和響應都遵循相同的格式，一個HTTP包含Header和Body兩部分，其中Body是可選的。HTTP協議是一種文本協議，所以，它的格式也非常簡單。

1.1 HTTP GET請求的格式：

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

1.2 HTTP POST請求的格式：

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

body data goes here...

當遇到連續兩個\r\n時，Header部分結束，后面的數據全部是Body。

1.3 HTTP響應的格式：

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

body data goes here...

HTTP響應如果包含body，也是通過\r\n\r\n來分隔的。

請再次注意，Body的數據類型由Content-Type頭來確定，如果是網頁，Body就是文本，如果是圖片，Body就是圖片的二進制數據。

當存在Content-Encoding時，Body數據是被壓縮的，最常見的壓縮方式是gzip，所以，看到Content-Encoding: gzip時，需要將Body數據先解壓縮，才能得到真正的數據。壓縮的目的在於減少Body的大小，加快網絡傳輸。

2. Web靜態服務器-顯示固定的頁面

import socket


def handle_client(new_client):
        """處理客戶端請求"""

        recv_data = new_client.recv(1024)

        print(recv_data)
    
        # 組裝響應的內容
        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():
    
        # 創建套接字
        tcp_socket_server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
    
        # 綁定本地消息
        tcp_socket_server.bind(("", 8090))

        # 套接字由主動變被動
        tcp_socket_server.listen(128)
    
        while True:
    
                # 接收新的請求
                new_client, client_addr = tcp_socket_server.accept()

                handle_client(new_client)

        # 關閉套接字
        tcp_socket_server.close()


if __name__ == "__main__":

        main()

如上的代碼僅僅只是向瀏覽器發送了簡單的文本內容：6666

在瀏覽器中訪問：

3. Web靜態服務器-顯示需要的頁面

import socket
import re


def handle_client(new_client):
	"""處理客戶端請求"""

	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:	
	
		# 組裝響應的內容
		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():
	
	# 創建套接字
	tcp_socket_server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
	tcp_socket_server.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)	
	
	# 綁定本地消息
	tcp_socket_server.bind(("", 8090))

	# 套接字由主動變被動
	tcp_socket_server.listen(128)
	
	while True:
		
		# 接收新的請求
		new_client, client_addr = tcp_socket_server.accept()

		handle_client(new_client)

	# 關閉套接字
	tcp_socket_server.close()


if __name__ == "__main__":

	main()

在瀏覽器中訪問：

4. Web靜態服務器-多進程版

import socket
import re
import multiprocessing


def handle_client(new_client):
	"""處理客戶端請求"""

	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:	
	
		# 組裝響應的內容
		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():
	
	# 創建套接字
	tcp_socket_server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
	tcp_socket_server.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)	
	
	# 綁定本地消息
	tcp_socket_server.bind(("", 8090))

	# 套接字由主動變被動
	tcp_socket_server.listen(128)
	
	while True:
		
		# 接收新的請求
		new_client, client_addr = tcp_socket_server.accept()
		
		p = multiprocessing.Process(target=handle_client, args=(new_client,))
		p.start()
		
		new_client.close()
		
	# 關閉套接字
	tcp_socket_server.close()


if __name__ == "__main__":

	main()

5. Web靜態服務器-多線程版

import socket
import re
import threading


def handle_client(new_client):
	"""處理客戶端請求"""

	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:	
	
		# 組裝響應的內容
		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():
	
	# 創建套接字
	tcp_socket_server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
	tcp_socket_server.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)	
	
	# 綁定本地消息
	tcp_socket_server.bind(("", 8090))

	# 套接字由主動變被動
	tcp_socket_server.listen(128)
	
	while True:
		
		# 接收新的請求
		new_client, client_addr = tcp_socket_server.accept()
		
		t = threading.Thread(target=handle_client, args=(new_client,))
		t.start()
		
	# 關閉套接字
	tcp_socket_server.close()


if __name__ == "__main__":

	main()

6. Web靜態服務器-gevent版

import socket
import re
import gevent
from gevent import monkey


monkey.patch_all()


def handle_client(new_client):
	"""處理客戶端請求"""

	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:	
	
		# 組裝響應的內容
		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():
	
	# 創建套接字
	tcp_socket_server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
	tcp_socket_server.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)	
	
	# 綁定本地消息
	tcp_socket_server.bind(("", 8090))

	# 套接字由主動變被動
	tcp_socket_server.listen(128)
	
	while True:
		
		# 接收新的請求
		new_client, client_addr = tcp_socket_server.accept()
		
		g = gevent.spawn(handle_client, new_client)
		g.join()
		
	# 關閉套接字
	tcp_socket_server.close()


if __name__ == "__main__":

	main()