This is an explanation of curl in PHP. Simply put, curl is a library that allows you to hook up, talk to, and communicate in depth with many different types of servers through URLs, and it also supports many protocols. And people also said that curl can support https authentication, http post, ftp upload, proxy, cookies, simple password authentication and other functions.
Having said so much, I actually don’t feel much about it. I can only feel it in the application. At first, I needed to initiate a POST request to another server on the server side before I started to get in touch with curl, and then I felt it.
Before I officially talk about how to use it, let me mention that you must first install and enable the curl module in your PHP environment. I won’t go into the specific method. Different systems have different installation methods. You can check it on Google. It's quite simple to check it out, or check the official PHP documentation.
After you get the tool, you must first play with it and see if it is comfortable for you. Otherwise, use it as soon as you get it and make your own code better. How can we flirt with the server in such a mess?
For example, let’s take Baidu, the famous “test network connection” website, as an example to try curl
<?php
// create curl resource
$ch = curl_init();
// set url
curl_setopt($ch, CURLOPT_URL, "baidu.com");
//return the transfer as a string
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
// $output contains the output string
$output = curl_exec($ch);
//echo output
echo $output; // close curl resource to free up system resources
curl_close($ch);
?>When you open this php file in the local environment browser When the page appears, it is Baidu’s homepage. What about the “localhost” I just entered?
The above code and comments have fully explained what this code is doing.
$ch = curl_init(), creates a curl session resource, and returns a handle successfully;
curl_setopt($ch, CURLOPT_URL, "baidu.com"), sets the URL, needless to say;
The above two sentences can be combined into one sentence $ch = curl_init("baidu.com");
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 0) This is to set whether to store the response result in a variable. 1 is stored, 0 is echoed out directly;
$output = curl_exec($ch) is executed, and then the response result is stored in the $output variable for the following echo;
curl_close($ ch) Close this curl session resource.
The use of curl in PHP is roughly in this form. The second step, setting parameters through the curl_setopt method is the most complicated and important. If you are interested, you can read the official detailed reference on settable parameters. It's long enough to make you want to vomit, but practice makes perfect as needed.
To summarize, the usage of curl in PHP is: create curl session -> Configuration parameters -> Execute -> Close session.
Let’s take a look at some common scenarios, how we need to “dress ourselves” (configuration parameters) in order to correctly “pick up girls” (correctly pick up the server).
Say hello to the server first, send a Hello to the server and see how she responds. Here is the best The convenient way is to send a GET request to the server. Of course, a small note like POST is also OK.
We take "searching for keywords in a famous gay dating website github" as an example
//通过curl进行GET请求的案例<?php
// create curl resource
$ch = curl_init();
// set url
curl_setopt($ch, CURLOPT_URL, "https://github.com/search?q=react");
//return the transfer as a string
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
// $output contains the output string
$output = curl_exec($ch);
//echo output
echo $output; // close curl resource to free up system resources
curl_close($ch);
?>It seems to be no different from the previous example , but here are two points that can be mentioned:
1. The default request method is GET, so there is no need to explicitly specify the GET method;
2. https requests, non-http requests, some people may see them in various places When making an HTTPS request, you need to add a few lines of code to bypass the SSL certificate check to successfully request the resource, but it doesn't seem to be needed here. What's the reason?
The two Curl options are defined as:
CURLOPT_SSL_VERIFYPEER - verify the peer's SSL certificate CURLOPT_SSL_VERIFYHOST - verify the certificate's name against host
They both default to true in Curl, and shouldn't be disabled unless you've got a good reason. Disabling them is generally only needed if you're sending requests to servers with invalid or self-signed certificates, which is only usually an issue in development. Any publicly-facing site should be presenting a valid certificate, and by disabling these options you're potentially opening yourself up to security issues.
That is, unless you are using an illegal or homemade certificate, which mostly occurs in a development environment, you should set these two lines to false To avoid SSL certificate checking, otherwise there is no need to do this, which is unsafe.
How to make a POST request? For testing, first pass a script to receive POST to a test server:
//testRespond.php<?php
$phpInput=file_get_contents('php://input'); echo urldecode($phpInput);?>Then write a request locally:
<?php
$data=array( "name" => "Lei", "msg" => "Are you OK?"
);
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, "http://测试服务器的IP马赛克/testRespond.php");
curl_setopt($ch, CURLOPT_POST, 1); //The number of seconds to wait while trying to connect. Use 0 to wait indefinitely.
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 60);
curl_setopt($ch, CURLOPT_POSTFIELDS , http_build_query($data));
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
$output = curl_exec($ch);
echo $output;
curl_close($ch);
?>Browser run The result is:
name=Lei&msg=Are you OK?
Here we construct an array and pass it to the server as POST data:
curl_setopt($ch, CURLOPT_POST, 1) indicates that it is a POST request;
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 60) sets the longest tolerable connection time in seconds. You can’t wait forever and become a mummy;
curl_setopt($ch, CURLOPT_POSTFIELDS , http_build_query($data))设置POST的数据域，因为这里是数组数据形式的（等会来讲json格式），所以用http_build_query处理一下。
<?php
$data='{"name":"Lei","msg":"Are you OK?"}';
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, "http://测试服务器的IP马赛克/testRespond.php");
curl_setopt($ch, CURLOPT_POST, 1);
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 60);
curl_setopt($ch, CURLOPT_HTTPHEADER, array('Content-Type: application/json', 'Content-Length:' . strlen($data)));
curl_setopt($ch, CURLOPT_POSTFIELDS , $data);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
$output = curl_exec($ch);
echo $output;
curl_close($ch);
?>浏览器执行，显示：
{"name":"Lei","msg":"Are you OK?"}已经和服务器勾搭上了，这时候得要个照片来看一看了吧，你也得把自己的照片发上去让人看一看了，虽然两个人在一起外貌不重要，但是男俊女靓总是最棒的。
同样远程服务器端我们先传好一个接收脚本,接收图片并且保存到本地，注意文件和文件夹权限问题，需要有写入权限：
<?php
if($_FILES){
$filename = $_FILES['upload']['name'];
$tmpname = $_FILES['upload']['tmp_name']; //保存图片到当前脚本所在目录
if(move_uploaded_file($tmpname,dirname(__FILE__).'/'.$filename)){ echo ('上传成功');
}
}?>然后我们再来写我们本地服务器的php curl部分：
<?php
$data = array('name'=>'boy', "upload"=>"@boy.png");
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, "http://远程服务器地址马赛克/testRespond.php");
curl_setopt($ch, CURLOPT_POST, 1);
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 60);
curl_setopt($ch, CURLOPT_POSTFIELDS , $data);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
$output = curl_exec($ch);
echo $output;
curl_close($ch);
?>浏览器中运行一下，什么都米有，去看一眼远程的服务器，还是什么都没有，并没有上传成功。
为什么会这样呢？上面的代码应该是大家搜索curl php POST图片最常见的代码，这是因为我现在用的是PHP5.6以上版本，@符号在PHP5.6之后就弃用了，PHP5.3依旧可以用，所以有些同学发现能执行啊，有些发现不能执行，大抵是因为PHP版本的不同，而且curl在这两版本中实现是不兼容的，上面是PHP5.3的实现。
下面来讲PHP5.6及以后的实现，：
<?php
$data = array('name'=>'boy', "upload"=>"");
$ch = curl_init();
$data['upload']=new CURLFile(realpath(getcwd().'/boy.png'));
curl_setopt($ch, CURLOPT_URL, "http://115.29.247.189/test/testRespond.php");
curl_setopt($ch, CURLOPT_POST, 1);
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 60);
curl_setopt($ch, CURLOPT_POSTFIELDS , $data);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
$output = curl_exec($ch);
echo $output;
curl_close($ch);
?>这里引入了一个CURLFile对象进行实现，关于此的具体可查阅文档进行了解。这时候再去远程服务器目录下看看，发现有了一张图片了，而且确实是我们刚才上传的图片。
服务器妹子也挺实诚的，看了照骗觉得我长得挺慈眉善目的，就大方得拿出了她自己的照片，但是有点害羞的是，她不愿意主动拿过来，得我们自己去取。
远程服务器在她自己的目录下存放了一个图片叫girl.jpg，地址是她的web服务器根目录/girl.jpg，现在我要去获取这张照片。
<?php
$ch = curl_init();
$fp=fopen('./girl.jpg', 'w');
curl_setopt($ch, CURLOPT_URL, "http://远程服务器地址马赛克/girl.jpg");
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 60);
curl_setopt($ch, CURLOPT_FILE, $fp);
$output = curl_exec($ch);
$info = curl_getinfo($ch);
fclose($fp);
$size = filesize("./girl.jpg"); if ($size != $info['size_download']) { echo "下载的数据不完整，请重新下载";
} else { echo "下载数据完整";
}
curl_close($ch);
?>现在，在我们当前目录下就有了一张刚拿到的照片啦，是不是很激动呢！
这里值得一说的是curl_getinfo方法，这是一个获取本次请求相关信息的方法，对于调试很有帮助，要善用。
这个时候呢，服务器的家长说这个我们女儿还太小，不能找对象，就将她女儿关了起来，并且上了一个密码锁，所谓的HTTP认证，服务器呢偷偷托信鸽将HTTP认证的用户名和密码给了你，要你去见她，带她私奔。
那么拿到了用户名和密码，我们怎么通过PHP CURL搞定HTTP认证呢？
PS:这里偷懒就不去搭HTTP认证去试了，直接放一段代码，我们分析下。
function curl_auth($url,$user,$passwd){
$ch = curl_init();
curl_setopt_array($ch, [
CURLOPT_USERPWD => $user.':'.$passwd,
CURLOPT_URL => $url,
CURLOPT_RETURNTRANSFER => true
]);
$result = curl_exec($ch);
curl_close($ch); return $result;
}
$authurl = 'http://要请求HTTP认证的地址';echo curl_auth($authurl,'vace','passwd');这里有一个地方比较有意思：
curl_setopt_array 这个方法可以通过数组一次性地设置多个参数，防止有些需要多处设置的出现密密麻麻的curl_setopt方法。
这时你成功见到了服务器妹子，想带她私奔，但是无奈没有盘缠走不远，服务器妹子说，她妈服务器上有金库，可以登陆上去搞一点下来。
首先我们先来分析一下，这个事情分两步，一是去登陆界面通过账号密码登陆，然后获取cookie，二是去利用cookie模拟登陆到信息页面获取信息，大致的框架是这样的。
<?php
//设置post的数据
$post = array (
'email' => '账户',
'pwd' => '密码'
);
//登录地址
$url = "登陆地址";
//设置cookie保存路径
$cookie = dirname(__FILE__) . '/cookie.txt';
//登录后要获取信息的地址
$url2 = "登陆后要获取信息的地址";
//模拟登录
login_post($url, $cookie, $post);
//获取登录页的信息
$content = get_content($url2, $cookie);
//删除cookie文件
@ unlink($cookie);
var_dump($content);
?>然后我们思考下下面两个方法的实现：
login_post($url, $cookie, $post)
get_content($url2, $cookie)
//模拟登录 function login_post($url, $cookie, $post) {
$curl = curl_init();
curl_setopt($curl, CURLOPT_URL, $url);
curl_setopt($curl, CURLOPT_RETURNTRANSFER, 0);
curl_setopt($curl, CURLOPT_COOKIEJAR, $cookie);
curl_setopt($curl, CURLOPT_POST, 1);
curl_setopt($curl, CURLOPT_POSTFIELDS, http_build_query($post));
curl_exec($curl);
curl_close($curl);
}//登录成功后获取数据 function get_content($url, $cookie) {
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, $url);
curl_setopt($ch, CURLOPT_HEADER, 0);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
curl_setopt($ch, CURLOPT_COOKIEFILE, $cookie);
$rs = curl_exec($ch);
curl_close($ch);
return $rs;
}至此，总算是模拟登陆成功，一切顺利啦，通过php CURL“撩”服务器就是这么简单。
当然，CURL的能力远不止于此，本文仅希望就后端PHP开发中最常用的几种场景做一个整理和归纳。最后一句话，具体问题具体分析。
相关推荐：
The above is the detailed content of How to use CURL in PHP. For more information, please follow other related articles on the PHP Chinese website!
![[Web front-end] Node.js quick start](https://img.php.cn/upload/course/000/000/067/662b5d34ba7c0227.png)
