curl_getinfo

nikita dot bulatenko at gmail dot com (2013-03-15 14:11:33)

CURLINFO_SSL_VERIFYRESULT error codes:
0: ok the operation was successful.
2 : unable to get issuer certificate
3: unable to get certificate CRL
4: unable to decrypt certificate's signature
5: unable to decrypt CRL's signature
6: unable to decode issuer public key
7: certificate signature failure
8: CRL signature failure
9: certificate is not yet valid
10: certificate has expired
11: CRL is not yet valid
12:CRL has expired
13: format error in certificate's notBefore field
14: format error in certificate's notAfter field
15: format error in CRL's lastUpdate field
16: format error in CRL's nextUpdate field
17: out of memory
18: self signed certificate
19: self signed certificate in certificate chain
20: unable to get local issuer certificate
21:unable to verify the first certificate
22: certificate chain too long
23: certificate revoked
24: invalid CA certificate
25: path length constraint exceeded
26: unsupported certificate purpose
27: certificate not trusted
28: certificate rejected
29: subject issuer mismatch
30: authority and subject key identifier mismatch
31: authority and issuer serial number mismatch
32: key usage does not include certificate signing
50: application verification failure
details at http://www.openssl.org/docs/apps/verify.html#VERIFY_OPERATION

vince (2010-10-22 16:56:49)

CURLINFO_HTTP_CODE does not return a string, as the docs say, but rather an integer.

<?php
    $c = curl_init('http://www.example.com/');
    if(curl_getinfo($c, CURLINFO_HTTP_CODE) === '200') echo "CURLINFO_HTTP_CODE returns a string.";
    if(curl_getinfo($c, CURLINFO_HTTP_CODE) === 200) echo "CURLINFO_HTTP_CODE returns an integer.";
    curl_close($c);
?>

returns

"CURLINFO_HTTP_CODE returns an integer."

Anonymous (2010-07-20 14:53:05)

The main doc neglects to mention that when the CURLINFO_HEADER_OUT option is set the array returned by this function will included a new property, request_header, that is a string of the headers sent in the request.

raghuveer at spiderace dot com (2008-10-23 05:47:11)

A code snippet that I had attempted to enumerate the Remote File Size and to post Amount of Time taken for that particular Curl Request.

<?php

/* borrowed the Byte Conversion code (byte_convert function) posted by "olafurw at gmail.com" and modified by "d.abromeit" in http://in.php.net/filesize. I did make changes in the notation of expression of multiples of bytes from KiB => KB, MiB => MB etc all symbols in this Byte Code*/

  function byte_convert($bytes)
  {
    $symbol = array('B', 'KB', 'MB', 'GB', 'TB', 'PB', 'EB', 'ZB', 'YB');

    $exp = 0;
    $converted_value = 0;
    if( $bytes > 0 )
    {
      $exp = floor( log($bytes)/log(1024) );
      $converted_value = ( $bytes/pow(1024,floor($exp)) );
    }

    return sprintf( '%.2f '.$symbol[$exp], $converted_value );
  }

   $userAgent = 'Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; .NET CLR 1.1.4322)';
   $url = 'http://www.spiderace.com/samplefiles/testfile.zip';

    // initialize curl with given url
    $ch = curl_init($url);
    // make sure we get the header
    curl_setopt($ch, CURLOPT_HEADER, 1);
    // make it a http HEAD request
    curl_setopt($ch, CURLOPT_NOBODY, 1);
    // add useragent
    curl_setopt($ch, CURLOPT_USERAGENT, $userAgent);
    //Tell curl to write the response to a variable
    curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
    // The maximum number of seconds to allow cURL functions to execute. 
    curl_setopt($ch,CURLOPT_CONNECTTIMEOUT,60);
    // Tell curl to stop when it encounters an error 
    curl_setopt($ch, CURLOPT_FAILONERROR, 1);
    
    $execute = curl_exec($ch);
    
    // Check if any error occured
    if(!curl_errno($ch))
    {
        $bytes = curl_getinfo($ch, CURLINFO_CONTENT_LENGTH_DOWNLOAD);
        // Display the File Size
        echo byte_convert($bytes);
        echo "<br><br>";
        $total_time = curl_getinfo($ch, CURLINFO_TOTAL_TIME);
        echo 'Took ' . $total_time . ' seconds to send a request to ' . $url;
        clearstatcache();
    }
    curl_close($ch);

?>

NOTE: This code worked on Both Linux (tried on CentOS) & Windows Servers (tried on XAMPP For Windows on Windows XP PC), thank you. Hope this will be helpful to others as well.

master-No-sp-aM at io dot ua (2008-07-17 00:40:59)

CURLINFO_CONTENT_LENGTH_DOWNLOAD contains zero in FTP request although Content-Length field is present.

Following workaround may helps:

<?php
$ch = curl_init($url);
... CURL OPTIONS ...
$result=curl_exec($ch);
$pu=parse_url($url);
if($pu['scheme']=='ftp'){
    $regexp="/Content-Length: (\d*)/";
    preg_match($regexp,$result,$matches);
    $con_len=$matches[1];
}
else $con_len = curl_getinfo($ch, CURLINFO_CONTENT_LENGTH_DOWNLOAD);
?>

Its worth to keep using CURLINFO_CONTENT_LENGTH_DOWNLOAD in http requests, becouse there could be several headers pack if CURLOPT_FOLLOWLOCATION is fired, and regular expression for this case is a little bit complicated.

nemetral (2008-06-17 07:57:16)

Just a quick note: if you want to use curl_getinfo() with option CURLINFO_HEADER_OUT in order to debug your cURL request, you must add curl_setopt($handle, CURLINFO_HEADER_OUT, true); first while specifying the options.

luiheidsgoeroe at hotmail dot com (2008-02-04 15:02:09)

Keep in mind that for CURLOPT_RETURNTRANSFER it has to be set with curl_setopt() before execution:

This doesn't work:
<?php
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, "http://www.example.com/");
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_exec($ch);
var_dump(curl_getinfo($ch,CURLINFO_HEADER_OUT));
?>

This works:
<?php
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, "http://www.example.com/");
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLINFO_HEADER_OUT, true);
curl_exec($ch);
var_dump(curl_getinfo($ch,CURLINFO_HEADER_OUT));
?>

Mark Evers (2008-01-16 22:14:10)

There is a constant missing from that list. CURLINFO_REDIRECT_COUNT will give you the number of redirects it went through if CURLOPT_FOLLOWLOCATION was set.

ssttoo at hotmail dot com (2004-04-06 12:13:09)

Here are the response codes ready for pasting in an ini-style file. Can be used to provide more descriptive message, corresponding to 'http_code' index of the arrray returned by curl_getinfo(). 
These are taken from the W3 consortium HTTP/1.1: Status Code Definitions, found at
http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html

[Informational 1xx]
100="Continue"
101="Switching Protocols"

[Successful 2xx]
200="OK"
201="Created"
202="Accepted"
203="Non-Authoritative Information"
204="No Content"
205="Reset Content"
206="Partial Content"

[Redirection 3xx]
300="Multiple Choices"
301="Moved Permanently"
302="Found"
303="See Other"
304="Not Modified"
305="Use Proxy"
306="(Unused)"
307="Temporary Redirect"

[Client Error 4xx]
400="Bad Request"
401="Unauthorized"
402="Payment Required"
403="Forbidden"
404="Not Found"
405="Method Not Allowed"
406="Not Acceptable"
407="Proxy Authentication Required"
408="Request Timeout"
409="Conflict"
410="Gone"
411="Length Required"
412="Precondition Failed"
413="Request Entity Too Large"
414="Request-URI Too Long"
415="Unsupported Media Type"
416="Requested Range Not Satisfiable"
417="Expectation Failed"

[Server Error 5xx]
500="Internal Server Error"
501="Not Implemented"
502="Bad Gateway"
503="Service Unavailable"
504="Gateway Timeout"
505="HTTP Version Not Supported"

And an example usage:
<?php
$ch = curl_init(); // create cURL handle (ch)
if (!$ch) {
    die("Couldn't initialize a cURL handle");
}
// set some cURL options
$ret = curl_setopt($ch, CURLOPT_URL,            "http://mail.yahoo.com");
$ret = curl_setopt($ch, CURLOPT_HEADER,         1);
$ret = curl_setopt($ch, CURLOPT_FOLLOWLOCATION, 1);
$ret = curl_setopt($ch, CURLOPT_RETURNTRANSFER, 0);
$ret = curl_setopt($ch, CURLOPT_TIMEOUT,        30);

// execute
$ret = curl_exec($ch);

if (empty($ret)) {
    // some kind of an error happened
    die(curl_error($ch));
    curl_close($ch); // close cURL handler
} else {
    $info = curl_getinfo($ch);
    curl_close($ch); // close cURL handler

    if (empty($info['http_code'])) {
            die("No HTTP code was returned"); 
    } else {
        // load the HTTP codes
        $http_codes = parse_ini_file("path/to/the/ini/file/I/pasted/above");
        
        // echo results
        echo "The server responded: <br />";
        echo $info['http_code'] . " " . $http_codes[$info['http_code']];
    }

}
?>