If you want to implement a web server, you need to have a general understanding of the operating principles of the web server. Let’s start with a static text server, taking accessing 1.html of the web server as an example
1. The client sends an http request to the server. If the port number the server listens to is 9002, then the address to test access on the machine itself is http://localhost:9002/1.html.
2. The server listens to port 9002. After receiving the request, it can obtain the location in the web directory of the uri resource that needs to be accessed in the request from the http header.
3. The server reads the resource file that needs to be accessed, then fills it into the http entity and returns it to the client.
The schematic diagram is as follows:
<?php class web_config { // 监听的端口号 const PORT = 9003; // 项目根目录 const WEB_ROOT = "/Users/zhoumengkang/Documents/html"; } class server { private $ip; private $port; public function __construct($ip, $port) { $this->ip = $ip; $this->port = $port; $this->await(); } private function await() { $sock = socket_create(AF_INET, SOCK_STREAM, SOL_TCP); if ($sock < 0) { echo "Error:" . socket_strerror(socket_last_error()) . "\n"; } $ret = socket_bind($sock, $this->ip, $this->port); if (!$ret) { echo "BIND FAILED:" . socket_strerror(socket_last_error()) . "\n"; exit; } echo "OK\n"; $ret = socket_listen($sock); if ($ret < 0) { echo "LISTEN FAILED:" . socket_strerror(socket_last_error()) . "\n"; } do { $new_sock = null; try { $new_sock = socket_accept($sock); } catch (Exception $e) { echo $e->getMessage(); echo "ACCEPT FAILED:" . socket_strerror(socket_last_error()) . "\n"; } try { $request_string = socket_read($new_sock, 1024); $response = $this->output($request_string); socket_write($new_sock, $response); socket_close($new_sock); } catch (Exception $e) { echo $e->getMessage(); echo "READ FAILED:" . socket_strerror(socket_last_error()) . "\n"; } } while (TRUE); } /** * @param $request_string * @return string */ private function output($request_string){ // 静态 GET /1.html HTTP/1.1 ... $request_array = explode(" ",$request_string); if(count($request_array) < 2){ return $this->not_found(); } $uri = $request_array[1]; $filename = web_config::WEB_ROOT . $uri; echo "request:".$filename."\n"; // 静态文件的处理 if (file_exists($filename)) { return $this->add_header(file_get_contents($filename)); } else { return $this->not_found(); } } /** * 404 返回 * @return string */ private function not_found(){ $content = " <h1>File Not Found </h1> "; return "HTTP/1.1 404 File Not Found\r\nContent-Type: text/html\r\nContent-Length: ".strlen($content)."\r\n\r\n".$content; } /** * 加上头信息 * @param $string * @return string */ private function add_header($string){ return "HTTP/1.1 200 OK\r\nContent-Length: ".strlen($string)."\r\nServer: mengkang\r\n\r\n".$string; } } $server = new server("127.0.0.1", web_config::PORT);
As mentioned in the above code, as long as the file is executed in the terminal, a static web server will be started.
The picture below is a screenshot of me accessing the 1.jpg file in my web directory
The simple static web server has been completed. The next question is how to make it support the output of dynamic content. Do we only need to execute a certain program inside the web server and return the result to the client? But in this way, the web server code is coupled with the business code. How to solve a web server that can be used in various business scenarios?
The advent of CGI solved this problem. So what is CGI? The following paragraph is copied:
CGI is an interface standard between external applications (CGI programs) and Web servers. It is a procedure for transferring information between CGI programs and Web servers. The CGI specification allows Web servers to execute external programs and send their output to Web browsers. CGI turns the Web's set of simple static hypermedia documents into a complete new interactive media.
How dizzying, to give a specific example, for example, the PHP global variable $_SERVER['QUERY_STRING'] we are using is passed by the web server through the CGI protocol. For example, in Nginx, maybe you remember this fastcgi configuration
fastcgi_param QUERY_STRING $query_string;
Yes nginx passes its global variable $query_string to the environment variable of fastcgi_param.
Below we also use CGI’s QUERY_STRING as a bridge to pass the information in the uri requested by the client to the cgi program. Store QUERY_STRING in the environment variable of the request through putenv.
We agree that the resources accessed in the web server have the .cgi suffix to indicate dynamic access. This is somewhat similar to configuring location in nginx to find php scripts. It's all a rule to check whether a cgi program should be requested. In order to distinguish it from the Web server, I wrote a cgi program in C to query user information and query user information based on user ID.
The general access logic is as follows
Demo code address: https://github.com/zhoumengkang/php/tree/master/php-webserver/dynamic
If you want to run the demo, you need to do the following
1. Modify the project root directory WEB_ROOT in config.php
2. Compile cgi-demouser.c, compile the command gcc -o user.cgi user.c, and then put the user.cgi file under the root directory of your configured project
3. Execute php start.php in the terminal, so that the web server will be started
4. Through http://localhost:9003/user.cgi?id=1 you can access and see the following effect
In fact, I just did some cgi judgment on the basis of the static server to forward the request, and merged the codes of the three files on github into one file for everyone to watch
<?php class web_config { // 监听的端口号 const PORT = 9003; // 项目根目录 const WEB_ROOT = "/Users/zhoumengkang/Documents/html"; // 系统支持的 cgi 程序的文件扩展名 const CGI_EXTENSION = "cgi"; } class server { private $ip; private $port; public function __construct($ip, $port) { $this->ip = $ip; $this->port = $port; $this->await(); } private function await() { $sock = socket_create(AF_INET, SOCK_STREAM, SOL_TCP); if ($sock < 0) { echo "Error:" . socket_strerror(socket_last_error()) . "\n"; } $ret = socket_bind($sock, $this->ip, $this->port); if (!$ret) { echo "BIND FAILED:" . socket_strerror(socket_last_error()) . "\n"; exit; } echo "OK\n"; $ret = socket_listen($sock); if ($ret < 0) { echo "LISTEN FAILED:" . socket_strerror(socket_last_error()) . "\n"; } do { $new_sock = null; try { $new_sock = socket_accept($sock); } catch (Exception $e) { echo $e->getMessage(); echo "ACCEPT FAILED:" . socket_strerror(socket_last_error()) . "\n"; } try { $request_string = socket_read($new_sock, 1024); $response = $this->output($request_string); socket_write($new_sock, $response); socket_close($new_sock); } catch (Exception $e) { echo $e->getMessage(); echo "READ FAILED:" . socket_strerror(socket_last_error()) . "\n"; } } while (TRUE); } /** * @param $request_string * @return string */ private function output($request_string){ // 静态 GET /1.html HTTP/1.1 ... // 动态 GET /user.cgi?id=1 HTTP/1.1 ... $request_array = explode(" ",$request_string); if(count($request_array) < 2){ return ""; } $uri = $request_array[1]; echo "request:".web_config::WEB_ROOT . $uri."\n"; $query_string = null; if ($uri == "/favicon.ico") { return ""; } if (strpos($uri,"?")) { $uriArr = explode("?", $uri); $uri = $uriArr[0]; $query_string = isset($uriArr[1]) ? $uriArr[1] : null; } $filename = web_config::WEB_ROOT . $uri; if ($this->cgi_check($uri)) { $this->set_env($query_string); $handle = popen(web_config::WEB_ROOT.$uri, "r"); $read = stream_get_contents($handle); pclose($handle); return $this->add_header($read); } // 静态文件的处理 if (file_exists($filename)) { return $this->add_header(file_get_contents($filename)); } else { return $this->not_found(); } } /** * 设置环境变量 给 cgi 程序使用 * @param $query_string * @return bool */ private function set_env($query_string){ if($query_string == null){ return false; } if (strpos($query_string, "=")) { putenv("QUERY_STRING=".$query_string); } } /** * 判断请求的 uri 是否是合法的 cgi 资源 * @param $uri * @return bool */ private function cgi_check($uri){ $info = pathinfo($uri); $extension = isset($info["extension"]) ? $info["extension"] : null; if( $extension && in_array($extension,explode(",",web_config::CGI_EXTENSION))){ return true; } return false; } /** * 404 返回 * @return string */ private function not_found(){ $content = "<h1>File Not Found </h1>"; return "HTTP/1.1 404 File Not Found\r\nContent-Type: text/html\r\nContent-Length: ".strlen($content)."\r\n\r\n".$content; } /** * 加上头信息 * @param $string * @return string */ private function add_header($string){ return "HTTP/1.1 200 OK\r\nContent-Length: ".strlen($string)."\r\nServer: mengkang\r\n\r\n".$string; } } $server = new server("127.0.0.1", web_config::PORT);
The above is the entire implementation process of implementing a dynamic web server in PHP. I hope it will be helpful to everyone's learning.