PHP + curl, HTTP POST sample code?

<?php
//
// A very simple PHP example that sends a HTTP POST to a remote site
//

$ch = curl_init();

curl_setopt($ch, CURLOPT_URL,"http://www.example.com/tester.phtml");
curl_setopt($ch, CURLOPT_POST, 1);
curl_setopt($ch, CURLOPT_POSTFIELDS,
            "postvar1=value1&postvar2=value2&postvar3=value3");

// In real life you should use something like:
// curl_setopt($ch, CURLOPT_POSTFIELDS, 
//          http_build_query(array('postvar1' => 'value1')));

// Receive server response ...
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);

$server_output = curl_exec($ch);

curl_close ($ch);

// Further processing ...
if ($server_output == "OK") { ... } else { ... }
?>

Procedural

// set post fields
$post = [
    'username' => 'user1',
    'password' => 'passuser1',
    'gender'   => 1,
];

$ch = curl_init('http://www.example.com');
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_POSTFIELDS, $post);

// execute!
$response = curl_exec($ch);

// close the connection, release resources used
curl_close($ch);

// do anything you want with your response
var_dump($response);

Object oriented

<?php

// mutatis mutandis
namespace MyApp\Http;

class CurlPost
{
    private $url;
    private $options;
           
    /**
     * @param string $url     Request URL
     * @param array  $options cURL options
     */
    public function __construct($url, array $options = [])
    {
        $this->url = $url;
        $this->options = $options;
    }

    /**
     * Get the response
     * @return string
     * @throws \RuntimeException On cURL error
     */
    public function __invoke(array $post)
    {
        $ch = \curl_init($this->url);
        
        foreach ($this->options as $key => $val) {
            \curl_setopt($ch, $key, $val);
        }

        \curl_setopt($ch, \CURLOPT_RETURNTRANSFER, true);
        \curl_setopt($ch, \CURLOPT_POSTFIELDS, $post);

        $response = \curl_exec($ch);
        $error    = \curl_error($ch);
        $errno    = \curl_errno($ch);
        
        if (\is_resource($ch)) {
            \curl_close($ch);
        }

        if (0 !== $errno) {
            throw new \RuntimeException($error, $errno);
        }
        
        return $response;
    }
}

Usage

// create curl object
$curl = new \MyApp\Http\CurlPost('http://www.example.com');

try {
    // execute the request
    echo $curl([
        'username' => 'user1',
        'password' => 'passuser1',
        'gender'   => 1,
    ]);
} catch (\RuntimeException $ex) {
    // catch errors
    die(sprintf('Http error %s with code %d', $ex->getMessage(), $ex->getCode()));
}

Side note here: it would be best to create some kind of interface called AdapterInterface for example with getResponse() method and let the class above implement it. Then you can always swap this implementation with another adapter of your like, without any side effects to your application.

Using HTTPS / encrypting traffic

Usually there's a problem with cURL in PHP under the Windows operating system. While trying to connect to a https protected endpoint, you will get an error telling you that certificate verify failed.

What most people do here is to tell the cURL library to simply ignore certificate errors and continue (curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, false);). As this will make your code work, you introduce huge security hole and enable malicious users to perform various attacks on your app like Man In The Middle attack or such.

Never, ever do that. Instead, you simply need to modify your php.ini and tell PHP where your CA Certificate file is to let it verify certificates correctly:

; modify the absolute path to the cacert.pem file
curl.cainfo=c:\php\cacert.pem

The latest cacert.pem can be downloaded from the Internet or extracted from your favorite browser. When changing any php.ini related settings remember to restart your webserver.


A live example of using php curl_exec to do an HTTP post:

Put this in a file called foobar.php:

<?php
  $ch = curl_init();
  $skipper = "luxury assault recreational vehicle";
  $fields = array( 'penguins'=>$skipper, 'bestpony'=>'rainbowdash');
  $postvars = '';
  foreach($fields as $key=>$value) {
    $postvars .= $key . "=" . $value . "&";
  }
  $url = "http://www.google.com";
  curl_setopt($ch,CURLOPT_URL,$url);
  curl_setopt($ch,CURLOPT_POST, 1);                //0 for a get request
  curl_setopt($ch,CURLOPT_POSTFIELDS,$postvars);
  curl_setopt($ch,CURLOPT_RETURNTRANSFER, true);
  curl_setopt($ch,CURLOPT_CONNECTTIMEOUT ,3);
  curl_setopt($ch,CURLOPT_TIMEOUT, 20);
  $response = curl_exec($ch);
  print "curl response is:" . $response;
  curl_close ($ch);
?>

Then run it with the command php foobar.php, it dumps this kind of output to screen:

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" 
"http://www.w3.org/TR/html4/strict.dtd">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<title>Title</title>

<meta http-equiv="Pragma" content="no-cache">
<meta http-equiv="Expires" content="0">
<body>
  A mountain of content...
</body>
</html>

So you did a PHP POST to www.google.com and sent it some data.

Had the server been programmed to read in the post variables, it could decide to do something different based upon that.