filetype

fileperms

Last updated: Fri, 20 Nov 2009

filesize

(PHP 4, PHP 5)

filesize — Gets file size

Description

int filesize ( string $filename )

Gets the size for the given file.

Parameters

filename: Path to the file.

Return Values

Returns the size of the file in bytes, or FALSE (and generates an error of level E_WARNING) in case of an error.

Note: Because PHP's integer type is signed and many platforms use 32bit integers, filesize() may return unexpected results for files which are larger than 2GB. For files between 2GB and 4GB in size this can usually be overcome by using sprintf("%u", filesize($file)).

Examples

Example #1 filesize() example


<?php

// outputs e.g.  somefile.txt: 1024 bytes

$filename = 'somefile.txt';
echo $filename . ': ' . filesize($filename) . ' bytes';

?>

Notes

Note: The results of this function are cached. See clearstatcache() for more details.

Tip

As of PHP 5.0.0, this function can also be used with some URL wrappers. Refer to List of Supported Protocols/Wrappers for a listing of which wrappers support stat() family of functionality.


My solution for calculating the directory size:





<?php


/**


 * Get the directory size


 * @param directory $directory


 * @return integer


 */


function dirSize($directory) {


    $size = 0;


    foreach(new RecursiveIteratorIterator(new RecursiveDirectoryIterator($directory)) as $file){


        $size+=$file->getSize();


    }


    return $size;


}


?>

tmont
25-Jul-2009 09:05


Here's the best way (that I've found) to get the size of a remote file. Note that HEAD requests don't get the actual body of the request, they just retrieve the headers. So making a HEAD request to a resource that is 100MB will take the same amount of time as a HEAD request to a resource that is 1KB.



<?php

$remoteFile = 'http://us.php.net/get/php-5.2.10.tar.bz2/from/this/mirror';

$ch = curl_init($remoteFile);

curl_setopt($ch, CURLOPT_NOBODY, true);

curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);

curl_setopt($ch, CURLOPT_HEADER, true);

curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true); //not necessary unless the file redirects (like the PHP example we're using here)

$data = curl_exec($ch);

curl_close($ch);

if ($data === false) {

  echo 'cURL failed';

  exit;

}



$contentLength = 'unknown';

$status = 'unknown';

if (preg_match('/^HTTP\/1\.[01] (\d\d\d)/', $data, $matches)) {

  $status = (int)$matches[1];

}

if (preg_match('/Content-Length: (\d+)/', $data, $matches)) {

  $contentLength = (int)$matches[1];

}



echo 'HTTP Status: ' . $status . "\n";

echo 'Content-Length: ' . $contentLength;

?>



Result:



HTTP Status: 302

Content-Length: 8808759

Svetoslav Marinov
23-Jul-2009 05:24


This is an updated version of my previous filesize2bytes.


The return type now it's really an int.





<?php


/**


 * Converts human readable file size (e.g. 10 MB, 200.20 GB) into bytes.


 *


 * @param string $str


 * @return int the result is in bytes


 * @author Svetoslav Marinov


 * @author http://slavi.biz


 */


function filesize2bytes($str) {


    $bytes = 0;





    $bytes_array = array(


        'B' => 1,


        'KB' => 1024,


        'MB' => 1024 * 1024,


        'GB' => 1024 * 1024 * 1024,


        'TB' => 1024 * 1024 * 1024 * 1024,


        'PB' => 1024 * 1024 * 1024 * 1024 * 1024,


    );





    $bytes = floatval($str);





    if (preg_match('#([KMGTP]?B)$#si', $str, $matches) && !empty($bytes_array[$matches[1]])) {


        $bytes *= $bytes_array[$matches[1]];


    }





    $bytes = intval(round($bytes, 2));





    return $bytes;


}


?>

nak5ive at DONT-SPAM-ME dot gmail dot com
11-Jun-2009 05:59


a quick way to convert bytes to a more readable format can be done using this function:





<?php


function formatBytes($bytes, $precision = 2) {


    $units = array('B', 'KB', 'MB', 'GB', 'TB');


   


    $bytes = max($bytes, 0);


    $pow = floor(($bytes ? log($bytes) : 0) / log(1024));


    $pow = min($pow, count($units) - 1);


   


    $bytes /= pow(1024, $pow);


   


    return round($bytes, $precision) . ' ' . $units[$pow];


}


?>

Anonymous
16-May-2009 09:49


Shortest solution to get filesize of a remote website





<?php





$content = file_get_contents("http://www.example.com");





$handle = fopen("example-com.html", "w+");





fwrite($handle, $content);





fclose($handle);





echo size . ': ' . filesize('example-com.html') . ' bytes';





?>

hedora at web dot de
07-Feb-2009 02:23


reliably, quick & dirty:





<?php


function getdirsize($path)


    {


    $result=explode("\t",exec("du -hs ".$path),2);


    return ($result[1]==$path ? $result[0] : "error");


    }


?>

bs_hck at bk dot ru
27-Jan-2009 07:06


Fix 4Gb limit. ( Now limit 8 Gb ;))



<?php

    function GetRealSize($file) {

        // Return size in Mb

        clearstatcache();

        $INT = 4294967295;//2147483647+2147483647+1;

        $size = filesize($file);

        $fp = fopen($file, 'r');

        fseek($fp, 0, SEEK_END);

        if (ftell($fp)==0) $size += $INT;

        fclose($file);

        if ($size<0) $size += $INT;

        return ceil($size/1024/1024);

    }

?>

7r6ivyeo at mail dot com
22-Nov-2008 01:36


I use the CLI version of PHP on Windows Vista and work with large (~ 10G) Virtual Machine files.  Here's how to get a NTFS file's size without the 4G limit:



<?php

function dos_filesize($fn) {



    if (is_file($fn))

        return exec('FOR %A IN ("'.$fn.'") DO @ECHO %~zA');

    else

        return '0';

}

?>



This should work on any Windows OS that provides DOS shell commands.

revoke
16-Sep-2008 09:23


Another way for remote files -> strlen(join('',file('http://www.example.com/)));

Supermagnus
30-Jun-2008 10:38


<?php


function getSizeFile($url) {


    if (substr($url,0,4)=='http') {


        $x = array_change_key_case(get_headers($url, 1),CASE_LOWER);


        if ( strcasecmp($x[0], 'HTTP/1.1 200 OK') != 0 ) { $x = $x['content-length'][1]; }


        else { $x = $x['content-length']; }


    }


    else { $x = @filesize($url); }





    return $x;


}


?>





In case of you have a redirection in the server (like Redirect Permanent in the .htaccess)





In this case we have for exemple:


    [content-length] => Array





        (





            [0] => 294          // Size requested file





            [1] => 357556     // Real Size redirected file





        )

ronnie at logicalt dot com
03-Mar-2008 02:02


Another workaround for the >2GB "stat failed" in Linux is to call the systems stat as follows:



$size = exec ('stat -c %s '. escapeshellarg ($file_path));

jason dot whitehead dot tas at gmail dot com
08-Feb-2008 11:14


I have created a handy function, using parts of code from kaspernj at gmail dot com and md2perpe at gmail dot com, which should get file sizes > 4GB on Windows, Linux and Mac  (at least).





<?php


    function getSize($file) {


        $size = filesize($file);


        if ($size < 0)


            if (!(strtoupper(substr(PHP_OS, 0, 3)) == 'WIN'))


                $size = trim(`stat -c%s $file`);


            else{


                $fsobj = new COM("Scripting.FileSystemObject");


                $f = $fsobj->GetFile($file);


                $size = $file->Size;


            }


        return $size;


    }


?>

webmaster at eclipse org
07-Nov-2007 03:30


On 64-bit platforms, this seems quite reliable for getting the filesize of files > 4GB





<?php


$a = fopen($filename, 'r');


fseek($a, 0, SEEK_END);


$filesize = ftell($a);


fclose($a);


?>

MagicalTux at gmail dot com
16-Oct-2007 05:58


filesize() acts differently between platforms and distributions.



I tried manually compiling PHP on a 32bit platform. Filesize() would fail on files >2G.



Then I compiled again, adding CFLAGS=`getconf LFS_CFLAGS` in front of configure.



Then filesize() would success, but the result would be converted to a 32bit signed integer...



However on 64bit systems, PHP's integers are 64bit signed, and this would work just well...



So now the question is : should linux distributions (Debian, etc) define LFS_CFLAGS or not ? Doing so makes PHP be able to open/seek such files, but makes its behaviour buggy (stat() is supposed to fail if file size is >32bit and appl does not support such integers)...

md2perpe at gmail dot com
17-Jul-2007 09:29


Another way to get a filesize > 2GB on Linux:

<?php

$size = trim(`stat -c%s $file_path`);

?>

Robin Leffmann
26-Jun-2007 12:49


Here's a dodgy little snippet I made to find out how much disk space (in 1K-blocks) any chosen node - file or directory - occupies in the file system..





<?php


function nodesize( $node )


{


    if( !is_readable($node) ) return false; 


    $blah = exec( "/usr/bin/du -sk $node" );


    return substr( $blah, 0, strpos($blah, 9) );


}


?>





The character I search for with strpos() is a tab, as this is what 'du' on OpenBSD uses to separate size and node in the output - this might be whitespaces in other *nixes/distributions.

Josh Finlay <josh at glamourcastle dot com>
12-Nov-2006 08:22


I know there has been alot of remote filesize snippets posted, but I'll post mine also.



It supports HTTP/HTTPS/FTP/FTPS and detects which type it should use. It needs --enable-ftp for the FTP/FTPS functions.



I hope this works for someone.



<?php

    function remotefsize($url) {

        $sch = parse_url($url, PHP_URL_SCHEME);

        if (($sch != "http") && ($sch != "https") && ($sch != "ftp") && ($sch != "ftps")) {

            return false;

        }

        if (($sch == "http") || ($sch == "https")) {

            $headers = get_headers($url, 1);

            if ((!array_key_exists("Content-Length", $headers))) { return false; }

            return $headers["Content-Length"];

        }

        if (($sch == "ftp") || ($sch == "ftps")) {

            $server = parse_url($url, PHP_URL_HOST);

            $port = parse_url($url, PHP_URL_PORT);

            $path = parse_url($url, PHP_URL_PATH);

            $user = parse_url($url, PHP_URL_USER);

            $pass = parse_url($url, PHP_URL_PASS);

            if ((!$server) || (!$path)) { return false; }

            if (!$port) { $port = 21; }

            if (!$user) { $user = "anonymous"; }

            if (!$pass) { $pass = "phpos@"; }

            switch ($sch) {

                case "ftp":

                    $ftpid = ftp_connect($server, $port);

                    break;

                case "ftps":

                    $ftpid = ftp_ssl_connect($server, $port);

                    break;

            }

            if (!$ftpid) { return false; }

            $login = ftp_login($ftpid, $user, $pass);

            if (!$login) { return false; }

            $ftpsize = ftp_size($ftpid, $path);

            ftp_close($ftpid);

            if ($ftpsize == -1) { return false; }

            return $ftpsize;

        }

    }

?>

rico at ws5 dot org
08-Aug-2006 07:41


To get the size of files above 2GB you can use the linux-command filesize like this:



<?php

    function real_filesize_linux($file) {

        @exec("filesize $file",$out,$ret);

        if ( $ret <> '0' ) return FALSE;

        else return($out[0]); 

    }

?>

kaspernj at gmail dot com
18-Jul-2006 01:43


If you want to get the actual filesize for a size above 2 gb in Windows, you can use the COM-extensions in PHP.





An example is as follows:





<?php


    function knj_filesize($file){


        if (file_exists($file)){


            $fsobj = new COM("Scripting.FileSystemObject");


            $file = $fsobj->GetFile($file);


            $var = ($file->Size) + 1 - 1;


            


            return $var;


        }else{


            echo "File does not exist.\n";


            return false;


        }


    }


?>





This will return the corrent filesize. And it is very useful with PHP-GTK applications, where you want to use the filesize for larger files.





This example also works for files over a Windows-network. Try this example with the function:





<?php


   echo knj_filesize("//mycomputer/music/Track1.mp3");


?>





Happy hacking :)

core58 at mail dot ru
14-Apr-2006 03:21


some notes and modifications to previous post.


refering to RFC, when using HTTP/1.1 your request (either GET or POST or HEAD) must contain Host header string, opposite to HTTP/1.1 where Host ain't required. but there's no sure how your remote server would treat the request so you can add Host anyway (it won't be an error for HTTP/1.0).


host value _must_ be a host name (not CNAME and not IP address).





this function catches response, containing Location header and recursively sends HEAD request to host where we are moved until final response is met.


(you can experience such redirections often when downloading something from php scripts or some hash links that use apache mod_rewrite. most all of dowloading masters handle 302 redirects correctly, so this code does it too (running recursively thru 302 redirections).)





[$counter302] specify how much times your allow this function to jump if redirections are met. If initial limit (5 is default) expired -- it returns 0 (should be modified for your purposes whatever).0


ReadHeader() function is listed in previous post


(param description is placed there too).





<?php


function remote_filesize_thru( $ipAddress, $url, $counter302 = 5 )


{


    $socket = fsockopen( "10.233.225.2", 8080 );


    if( !$socket )


    {


        // failed to open TCP socket connection


        // do something sensible here besides exit();


        echo "<br>failed to open socket for [$ipAddress]";


        exit();


    }


                    


    // just send HEAD request to server


    $head = "HEAD $url HTTP/1.0\r\nConnection: Close\r\n\r\n";


    // you may use HTTP/1.1 instead, then your request head string _must_ contain "Host: " header


    fwrite( $socket, $head );


        


    // read the response header


    $header = ReadHeader( $socket );


    if( !$header )


    {


        // handle empty response here the way you need...


        Header( "HTTP/1.1 404 Not Found" );


        exit();


    }


    


    fclose( $socket );


    


    // check for "Location" header


    $locationMarker = "Location: ";


    $pos = strpos( $header, $locationMarker );


    if( $pos > 0 )


    {


            $counter302--;


            if( $counter302 < 0 )


            {


                     // redirect limit (5 by default) expired -- return some warning or do something sensible here


                    echo "warning: too long redirection sequence";


                    return 0;


            }





            // Location is present -- we should determine target host and move there, like any downloading masters do...


            // no need to use regex here


            $end = strpos( $header, "\n", $pos );


            $location = trim( substr( $header, $pos + strlen( $locationMarker ), $end - $pos - strlen( $locationMarker ) ), "\\r\\n" );


             


             // extract pure host (without "http://")


             $host = explode( "/", $location );


             $ipa = gethostbyname( $host[2] );


             // move to Location


             return remote_filesize_thru( $ipa, $location, $counter302 );


    }


        


    // try to acquire Content-Length within the response


    $regex = '/Content-Length:\s([0-9].+?)\s/';


    $count = preg_match($regex, $header, $matches);


                        


    // if there was a Content-Length field, its value


    // will now be in $matches[1]


    if( isset( $matches[1] ) )


          $size = $matches[1];


    else


          $size = 0;


    


    return $size;


}


?>

core58 at mail dot ru
10-Apr-2006 12:09


this is "raw" version of remote_filesize() function.

according to RFC, HTTP servers MUST implement at least GET, POST and HEAD requests, so the function just opens TCP socket connection, sends HEAD request and receives response, parsing length of the resource.

[$ipAddress] is the ip address of remote server.

[$url] is the name of file which size you want to determine.



the code was tested under Apache 2.0.43 and IIS 6.0 and it works correctly in both cases.

i wish the code can save someone's time :)



example:

$ipa = gethostbyname( "www.someserver.com" );

$url = "/docs/somedocument.pdf";



$fsize = remote_filesize2( $ipa, $url );



==========================



<?php



function ReadHeader( $socket )

{

    $i=0;

    $header = "";

    while( true && $i<20 )

    {

        // counter [$i] is used here to avoid deadlock while reading header string

        // it's limited by [20] here cause i really haven't ever met headers with string counter greater than 20

        // *

        $s = fgets( $socket, 4096 );

        $header .= $s;



        if( strcmp( $s, "\r\n" ) == 0 || strcmp( $s, "\n" ) == 0 )

            break;

        $i++;

    }

    if( $i >= 20 )

    {

        // suspicious header strings count was read

        // *

        return false;

    }



    return $header;

}



function remote_filesize2( $ipAddress, $url )

{

    $socket = fsockopen( $ipAddress, 80 );

    if( !$socket )

    {

        // failed to open TCP socket connection

        // do something sensible here besides exit();

        // ...

        exit();

        

    }



    // just send HEAD request to server

    // *

    fwrite( $socket, "HEAD $url HTTP/1.0\r\nConnection: Close\r\n\r\n" );

    

    // read the response header

    // *

    $header = ReadHeader( $socket );

    if( !$header )

    {

        Header( "HTTP/1.1 404 Not Found" );

        exit();

    }



    // try to acquire Content-Length within the response

    // *

    $regex = '/Content-Length:\s([0-9].+?)\s/';

    $count = preg_match($regex, $header, $matches);

   

    // if there was a Content-Length field, its value

    // will now be in $matches[1]

    if( isset( $matches[1] ) )

    {

        $size = $matches[1];

    }

    else

    {

        $size = 0;

    }

   

    fclose( $socket );

    return $size;



}

?>

mkamerma at science dot uva dot nl
12-Mar-2006 02:12


When read/writing binary files you often cannot rely on the feof() function being of much use, since it doesn't get triggered if the pointer is at the eof but hasn't tried to read one more byte. In this case you instead need to check if the file pointer is at filesize yet, but if you don't have the filename handy, you need to pluck it out fstat all the time. Two simple functions that would be nice to have natively in PHP:



<?php

    function fpfilesize(&$fp) { $stat = fstat($fp); return $stat["size"]; }



    function fpeof(&$fp) { return ftell($fp)==fpfilesize($fp); }

?>

aidan at php dot net
13-Jul-2005 04:01


This function quickly calculates the size of a directory:


http://aidanlister.com/repos/v/function.dirsize.php





You can convert filesizes to a human readable size using:


http://aidanlister.com/repos/v/function.size_readable.php





For a faster (unix only) implementation, see function.disk-total-space, note #34100


http://www.php.net/manual/en/function.disk-total-space.php#34100





Also of interest is this wikipedia article, discussing the difference between a kilobyte (1000) and a kibibyte (1024).


http://en.wikipedia.org/wiki/Bytes

bkimble at ebaseweb dot com
19-Nov-2004 11:33


In addition to the handy function Kris posted, here is an upgraded version that does basic http authentication as well.



<?php

/*

* (mixed)remote_filesize($uri,$user='',$pw='')

* returns the size of a remote stream in bytes or

* the string 'unknown'. Also takes user and pw

* incase the site requires authentication to access

* the uri

*/

function remote_filesize($uri,$user='',$pw='')

{

    // start output buffering

    ob_start();

    // initialize curl with given uri

    $ch = curl_init($uri);

    // make sure we get the header

    curl_setopt($ch, CURLOPT_HEADER, 1);

    // make it a http HEAD request

    curl_setopt($ch, CURLOPT_NOBODY, 1);

    // if auth is needed, do it here

    if (!empty($user) && !empty($pw))

    {

        $headers = array('Authorization: Basic ' .  base64_encode($user.':'.$pw));  

        curl_setopt($ch, CURLOPT_HTTPHEADER, $headers);

    }

    $okay = curl_exec($ch);

    curl_close($ch);

    // get the output buffer

    $head = ob_get_contents();

    // clean the output buffer and return to previous

    // buffer settings

    ob_end_clean();

    

    // gets you the numeric value from the Content-Length

    // field in the http header

    $regex = '/Content-Length:\s([0-9].+?)\s/';

    $count = preg_match($regex, $head, $matches);

    

    // if there was a Content-Length field, its value

    // will now be in $matches[1]

    if (isset($matches[1]))

    {

        $size = $matches[1];

    }

    else

    {

        $size = 'unknown';

    }

    

    return $size;

}

?>

add a note