Diese Seite bei php.net
The PHP manual text and comments are covered by the Creative Commons Attribution 3.0 License © the PHP Documentation Group - Impressum - mail("TO:Reinhard Neidl",...)
(PHP 4, PHP 5)
base64_encode — Kodiert Daten MIME base64
Kodiert den übergebenene Parameter data mittels base64.
Diese Form des Encodings wurde entworfen, um Binärdaten unbeschädigt mittels Technologien wie E-Mail übertragen zu können, die nicht korrekt mit 8-Bit-Zeichen umgehen.
Base64-kodierte Daten benötigen ungefähr 33% mehr Speicher als die Originaldaten.
Die zu kodierenden Daten.
Die kodierten Daten in Stringform.
Beispiel #1 base64_encode()-Beispiel
<?php
$str = 'Dies ist ein zu kodierender String';
echo base64_encode($str);
?>
Das oben gezeigte Beispiel erzeugt folgende Ausgabe:
RGllcyBpc3QgZWluIHp1IGtvZGllcmVuZGVyIFN0cmluZw==
output images into html:
<?php
$imgfile = "test.gif";
$handle = fopen($filename, "r");
$imgbinary = fread(fopen($imgfile, "r"), filesize($imgfile));
echo '<img src="data:image/gif;base64,' . base64_encode($imgbinary) . '" />';
?>
gif - data:image/gif;base64,...
jpg - data:image/jpeg;base64,...
png - data:image/png;base64,...
etc.
Some files can give a real headhache when script file is coded in UTF-8. Change it to ANSI and it'll work fine.
Note that some applications, such as OpenSSL's enc command, require that there be a line break every 64 characters in order for their base64 decode function to work. The following function will take care of this problem:
<?php
function ($encodeMe) {
$data = base64_encode($encodeMe);
$datalb = "";
while (strlen($data) > 64) {
$datalb .= substr($data, 0, 64) . "\n";
$data = substr($data,64);
}
$datalb .= $data;
return $datalb;
}
?>
Unfortunately my "function" for encoding base64 on-the-fly from 2007 [which has been removed from the manual in favor of this post] had 2 errors!
The first led to an endless loop because of a missing "$feof"-check, the second caused the rare mentioned errors when encoding failed for some reason in larger files, especially when
setting fgets($fh, 2) for example. But lower values then 1024 are bad overall because they slow down the whole process, so 4096 will be fine for all purposes, I guess.
The error was caused by the use of "empty()".
Here comes the corrected version which I have tested for all kind of files and length (up to 4,5 Gb!) without any error:
<?php
$fh = fopen('Input-File', 'rb');
//$fh2 = fopen('Output-File', 'wb');
$cache = '';
$eof = false;
while (1) {
if (!$eof) {
if (!feof($fh)) {
$row = fgets($fh, 4096);
} else {
$row = '';
$eof = true;
}
}
if ($cache !== '')
$row = $cache.$row;
elseif ($eof)
break;
$b64 = base64_encode($row);
$put = '';
if (strlen($b64) < 76) {
if ($eof) {
$put = $b64."\n";
$cache = '';
} else {
$cache = $row;
}
} elseif (strlen($b64) > 76) {
do {
$put .= substr($b64, 0, 76)."\n";
$b64 = substr($b64, 76);
} while (strlen($b64) > 76);
$cache = base64_decode($b64);
} else {
if (!$eof && $b64{75} == '=') {
$cache = $row;
} else {
$put = $b64."\n";
$cache = '';
}
}
if ($put !== '') {
echo $put;
//fputs($fh2, $put);
//fputs($fh2, base64_decode($put)); // for comparing
}
}
//fclose($fh2);
fclose($fh);
?>
If you want to send a very long value over URL, you might consider using base64_encode, and discover that IE6 only supports 2000 or so chars.
So, Using a little bit of magic you can do this and be happy:
<?php
$string = 'Blah';
$encoded = strtr(base64_encode(addslashes(gzcompress(serialize($string),9))), '+/=', '-_,');
$string= unserialize(gzuncompress(stripslashes(base64_decode(strtr($encoded, '-_,', '+/=')))));
?>
I omitted the strtr functions in my examples. Here are corrected functions:
<?php
function encode($x) {
return strtr(base64_encode(substr($_SESSION['Cksum'],rand(0,28),4) . $x), '+/=', '-_~');
}
function decode($x) {
$y = base64_decode(strtr($x, '-_~', '+/='));
if (strpos($_SESSION['Cksum'],substr($y,0,4)) === false) return false;
return substr($y,4-strlen($y));
}
?>
I have another solution that is simple and elegant. Create a pseudorandom string of characters. Then, each time you want to obfuscate your key, append a random substring from the pseudorandom string and use base64 encoding. When you want to de-obfuscate, convert back from base64. If the prefix is not in your pseudorandom source, then the value is forged. Otherwise, strip the prefix and recover your original key.
The advantages are that the string will look different even for the same key, and encoding and decoding should be extremely fast.
Here's an example:
<?php
// Call makeCksum once upon landing on the homepage
function makeCksum() {
$str = "";
for ($i=0;$i<32;++$i)
$str .= chr(rand(32,126));
$_SESSION['Cksum'] = $str;
}
function encode($x) {
return base64_encode(substr($_SESSION['Cksum'],rand(0,28),4) . $x);
}
function decode($x) {
$y = base64_decode($x);
if (strpos($_SESSION['Cksum'],substr($y,0,4)) === false) return false;
return substr($y,4-strlen($y));
}
?>
At Andi:
The last 6 chars of the decoded is used as some checksum + cryptographical salt - this prevents only in/decrementation of a primary key (ID) in an URL.
But there is no obfuscation of the ID, it is visible almost immediately: the number before "-". Hackers are thus guided to this internal ID of the database record and can use this knowledge...
Obfuscate better, making details a little less obvious, but not more secure:
* avoid the "-" (the "checksum/salt" is always the same lenght...),
* obfuscate the integer by converting ist to hex, like the hash.
<?php
// unchanged, thanx Tom, Andy, fsx.nr01
function base64_url_encode($input) {
return strtr(base64_encode($input), '+/=', '-_,');
}
function base64_url_decode($input) {
return base64_decode(strtr($input, '-_,', '+/='));
}
// some variables are used for clarity, they can be avoided and lines can be shortened:
function encryptId($int, $TableSalt='') {
global $GlobalSalt; // global secret for salt.
$HashedChecksum = substr(sha1($TableSalt.$int.$GlobalSalt), 0, 6);
// The length of the "HashedChecksum" is another little secret,
// but when the integers are small, it reveals...
$hex = dechex($int);
// The integer is better obfuscated by being HEX like the hash.
return base64url_encode($HashedChecksum.$hex);
// reordered, alternatively use substr() with negative lengths...
}
function decryptId($string, $TableSalt='') {
// checks if the second part of the base64 encoded string is correct.
global $GlobalSalt; // global secret for salt.
$parts = base64url_decode($string);
$hex = substr($parts, 6);
$int = hexdec($hex);
$part1 = substr($parts, 0, 6); // The "checksum/salt" is always the same length
return substr(sha1($TableSalt.$int.$GlobalSalt), 0, 6) === $part1
? $int
: false; // distinguish "0" and "false"
}
?>
There are some questions remaining:
* Why should we use sha1() for such an low-security task, it needs a lot of calculation.
* Would other algorithms be faster?
If your goal is, to confuse ordinary users, crc32() is much faster than sha1(). (But it can reveal the "salt" which then is much more a filler than a secret.)
If you want to secure against hackers, you should use strong encryption methods instead of obfuscation.
Different approach: these calculations for every URL can be avoided:
Take the database-ID, salt it (with a 'secret'), calculate a hash (like sha1), store it to a (unique - if you don't trust the truncated hash-algorithm any more) column in the database in addition to the primary key, index that field, and use that strange string for the URL and for the query. These calculations are done once and never again, but it costs you more memory for the database.
As secure as the other solution, but much faster (at least for php).
Shortened the base64url_encode and decode functions from Andi (http://nl2.php.net/manual/en/function.base64-encode.php#82200)
function base64_url_encode($input)
{
return strtr(base64_encode($input), '+/=', '-_,');
}
function base64_url_decode($input)
{
return base64_decode(strtr($input, '-_,', '+/='));
}
I needed a simple way to obfuscate auto_increment primary keys in databases when they are visible to users in URIs or API calls. The users should not be able to increment the id in the URL and see the next data record in the database table.
My solution (uses modified base64 functions by Tom):
function base64url_encode($plainText) {
$base64 = base64_encode($plainText);
$base64url = strtr($base64, '+/=', '-_,');
return $base64url;
}
function base64url_decode($plainText) {
$base64url = strtr($plainText, '-_,', '+/=');
$base64 = base64_decode($base64url);
return $base64;
}
function encryptId($int, $class='') {
return base64url_encode($int.'-'.substr(sha1($class.$int.encryptionKey), 0, 6));
}
function decryptId($string, $class='') {
$parts = explode('-', base64url_decode($string));
if (count($parts) != 2) {
return 0;
}
$int = $parts[0];
return substr(sha1($class.$int.encryptionKey), 0, 6) === $parts[1]
? (int)$int
: 0;
}
- The optional 2nd argument is the class name, so two equal ids of different tables will not result in two equal obfuscated ids.
- encryptionKey is a global secret key for encryption.
- decryptId() checks if the second part of the base64 encoded string is correct.
a note on URI -safe base64.
Simply replacing + = and / with _ - and . doesn't work as the base64_encode function will insert \r \n chars as well which are not URI-safe. So unless we have a base64encode function that does not insert any newline and padddings, the output can never be URI -safe.
referring to the note posted by " web at pkasperski dot com"
- you might as well use $encoded = strtr ( base64_encode ($data), 'ABCDEFG.....', 'aBcDEfG....' ).
this is more efficient. Also, Is simply changing the casing of the letters more secure? Why don't you try swapping letters around instead
- for your utf8_encode function, the "ord" function returns a number in the range 0-255. Remember that strings in PHP are actually a sequence of bytes rather than chars. So your utf8 encode func may not work properly. and the line "for ($n = 0; $n < strlen($input); $n++) {" is inefficient, you should assign a variable $count to strlen($input) and use it.
referring to
"dlyaza aT yahoo DOT com"
whats the usefulness of encoding your images in a php file??
referring to "php at ianco dot co dot uk"
- what version of php are you using? Your code works fine for me. I see all the output.
I have implemented the base64_encode() function with some custom mapping table so I could encode some binary data more securely without anyone knowing how to decode it
here is the class
<?php
class Base64 {
/**
* I have changed letter placement (P <=> x, S <=> 9) and the cases
* You can completely redo the mapping table
*/
private static $BinaryMap = array(
'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', // 7
'i', 'j', 'k', 'l', 'm', 'n', 'o', 'x', // 15
'q', 'r', '9', 't', 'u', 'v', 'w', 'x', // 23
'y', 'z', 'A', 'B', 'C', 'D', 'E', 'F', // 31
'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', // 39
'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', // 47
'W', 'P', 'Y', 'Z', '0', '1', '2', '3', // 55
'4', '5', '6', '7', '8', 'S', '+', '/', // 63
'=', // padding char
);
public function __construct() {}
public function base64_encode($input) {
$output = "";
$chr1 = $chr2 = $chr3 = $enc1 = $enc2 = $enc3 = $enc4 = null;
$i = 0;
// $input = self::utf8_encode($input);
while($i < strlen($input)) {
$chr1 = ord($input[$i++]);
$chr2 = ord($input[$i++]);
$chr3 = ord($input[$i++]);
$enc1 = $chr1 >> 2;
$enc2 = (($chr1 & 3) << 4) | ($chr2 >> 4);
$enc3 = (($chr2 & 15) << 2) | ($chr3 >> 6);
$enc4 = $chr3 & 63;
if (is_nan($chr2)) {
$enc3 = $enc4 = 64;
} else if (is_nan($chr3)) {
$enc4 = 64;
}
$output .= self::$BinaryMap[$enc1]
. self::$BinaryMap[$enc2]
. self::$BinaryMap[$enc3]
. self::$BinaryMap[$enc4];
}
return $output;
}
public function utf8_encode($input) {
$utftext = null;
for ($n = 0; $n < strlen($input); $n++) {
$c = ord($input[$n]);
if ($c < 128) {
$utftext .= chr($c);
} else if (($c > 128) && ($c < 2048)) {
$utftext .= chr(($c >> 6) | 192);
$utftext .= chr(($c & 63) | 128);
} else {
$utftext .= chr(($c >> 12) | 224);
$utftext .= chr((($c & 6) & 63) | 128);
$utftext .= chr(($c & 63) | 128);
}
}
return $utftext;
}
}
?>
and the usage as follows:
<?php
$string = pack('H*', "31c85c5aaa56c1f0102301ea497d0ab010e4e131af261787"); // HEX to binary
echo Base64::base64_encode($string);
echo "<br />";
echo base64_encode($string);
?>
and the output will be:
mCHCwQPwWFaqiWhQ9x0kSbdK4tgVjHEh // with custom mapping
MchcWqpWwfAQIwHqSX0KsBDk4TGvJheH // the base64_encode()
This function supports "base64url" as described in Section 5 of RFC 4648, "Base 64 Encoding with URL and Filename Safe Alphabet"
<?php
function base64url_encode($plainText)
{
$base64 = base64_encode($plainText);
$base64url = strtr($base64, '+/', '-_');
return ($base64url);
}
?>
You may wish to rtrim (or escape) trailing ='s for use in a URI.
Note that at least some Windows systems will not print a line of characters longer than a certain length unless it has line breaks of some kind. So if you base-64 encode a file, print it back for debugging purposes, and see nothing, don't be alarmed.
Using Function:
Output for HTML Put:
<img src="$self?image=file" border="0" alt="file">
<img src="$self?image=folder" border="0" alt="folder">
function getimage ($image) {
switch ($image) {
case 'file':
return base64_decode('R0lGODlhEQANAJEDAJmZmf///wAAAP///yH5BAHoAwMALAAAA
AARAA0AAAItnIGJxg0B42rsiSvCA/REmXQWhmnih3LUSGaqg35vF
bSXucbSabunjnMohq8CADsA');
case 'folder':
return base64_decode('R0lGODlhEQANAJEDAJmZmf///8zMzP///yH5BAHoAwMALAAAAA
ARAA0AAAIqnI+ZwKwbYgTPtIudlbwLOgCBQJYmCYrn+m3smY5v
Gc+0a7dhjh7ZbygAADsA');
case 'hidden_file':
return base64_decode('R0lGODlhEQANAJEDAMwAAP///5mZmf///yH5BAHoAwMALAAAA
AARAA0AAAItnIGJxg0B42rsiSvCA/REmXQWhmnih3LUSGaqg35vF
bSXucbSabunjnMohq8CADsA');
case 'link':
return base64_decode('R0lGODlhEQANAKIEAJmZmf///wAAAMwAAP///wAAAAAAAAAAA
CH5BAHoAwQALAAAAAARAA0AAAM5SArcrDCCQOuLcIotwgTYUll
NOA0DxXkmhY4shM5zsMUKTY8gNgUvW6cnAaZgxMyIM2zBLCaHlJgAADsA');
case 'smiley':
return base64_decode('R0lGODlhEQANAJECAAAAAP//AP///wAAACH5BAHoAwIALAAAA
AARAA0AAAIslI+pAu2wDAiz0jWD3hqmBzZf1VCleJQch0rkdnppB3
dKZuIygrMRE/oJDwUAOwA=');
case 'arrow':
return base64_decode('R0lGODlhEQANAIABAAAAAP///yH5BAEKAAEALAAAAAARAA0AA
AIdjA9wy6gNQ4pwUmav0yvn+hhJiI3mCJ6otrIkxxQAOw==');
}
}
I am finding a length restriction with base64_encode (or possibly with echo) in PHP 4.3.9.
This works ok for me:
<?php
echo strlen(str_repeat('-', 3273)); // 3273
echo strlen(base64_encode(str_repeat('-', 3273))); // 4364
echo base64_encode(str_repeat('-', 3273)); // LS0t repeated
?>
But change the length to 3274 and the third echo prints nothing.
<?php
echo strlen(str_repeat('-', 3274)); // 3274
echo strlen(base64_encode(str_repeat('-', 3274))); // 4368
echo base64_encode(str_repeat('-', 3274)); // Nothing at all printed
?>
This has obvious implications if you're wanting to encode a fairly large serialized array and echo it to a form field.
If you want to decode base64 encoded data in Javascript, you can use the tool (Webtoolkit.base64) on this website: http://www.webtoolkit.info/
I had massive problems storing a serialized Object which contained UTF-8 parts and some ascii parts (from the serialization i think) into mysql.
So i used base64_encode to get a clean string which could be safely decoded and unserialized.
this is bulletproof - if you ever have trouble use this.
the runtime is imho no problem.
Just a minor tweak of massimo's functions.
<?
$data = str_replace(array('+','/','='),array('-','_','.'),$data);
//replace '=' with '.' instead of with nothing, that way the process is reversible. '.' is uri-safe according to http://www.w3.org/Addressing/URL/5_URI_BNF.html
?>
$data = str_replace(array('+','/','='),array('-','_',),$data); // MIME::Base64::URLSafe implementation
$data = str_replace(array('+','/'),array('-','_'),$data); // Python raise "TypeError: Incorrect padding" if you remove "=" chars when decoding
function urlsafe_b64encode($string) {
$data = base64_encode($string);
$data = str_replace(array('+','/','='),array('-','_',''),$data);
return $data;
}
function urlsafe_b64decode($string) {
$data = str_replace(array('-','_'),array('+','/'),$string);
$mod4 = strlen($data) % 4;
if ($mod4) {
$data .= substr('====', $mod4);
}
return base64_decode($data);
}
Php version of perl's MIME::Base64::URLSafe, that provides an url-safe base64 string encoding/decoding (compatible with python base64's urlsafe methods)
If the function doesn't exist, this is a messy but effective way of doing it:
<?
echo bencode("Gabriel Malca");
// R2FicmllbCBNYWxjYQ==
function bencode($string='') {
$binval = convert_binary_str($string);
$final = "";
$start = 0;
while ($start < strlen($binval)) {
if (strlen(substr($binval,$start)) < 6)
$binval .= str_repeat("0",6-strlen(substr($binval,$start)));
$tmp = bindec(substr($binval,$start,6));
if ($tmp < 26)
$final .= chr($tmp+65);
elseif ($tmp > 25 && $tmp < 52)
$final .= chr($tmp+71);
elseif ($tmp == 62)
$final .= "+";
elseif ($tmp == 63)
$final .= "/";
elseif (!$tmp)
$final .= "A";
else
$final .= chr($tmp-4);
$start += 6;
}
if (strlen($final)%4>0)
$final .= str_repeat("=",4-strlen($final)%4);
return $final;
}
function convert_binary_str($string) {
if (strlen($string)<=0) return;
$tmp = decbin(ord($string[0]));
$tmp = str_repeat("0",8-strlen($tmp)).$tmp;
return $tmp.convert_binary_str(substr($string,1));
}
?>
There is an error on the example of passing an array through an HTML Form.
In the line:
$array = unserialize(base64_decode($coded_array);
There is a ')' missing. it should be:
$array = unserialize(base64_decode($coded_array));
If you encode text that contains symbols like < > and want to send it in GET query, be sure to urlencode the result of base64_encode, as it sometimes adds a + (and it's a special symbol) at the end:
<?php
echo base64_encode('<html>');
?>
returns:
PGh0bWw+
A function like this could also be useful:
<?php
function base64_urlencode($str) {
return urlencode(base64_encode($str));
};
?>
Using base64_encode to produce clean filenames from usernames (e.g. for image upload) is a bad idea if Umlaute (ö,ä,ü) are allowed. Then there is a slash (/) added in the filename, that will lead to a nonexisting directory beeing looked for and your script crashing!
I've used base64_encode and base64_decode for file attachment both in MySQL (using a BLOB field) and MSSQL (using a TEXT field). For MSSQL remember to set in PHP.INI file both mssql.textsize and mssql.textlimit to 2147483647.
Here's the code:
######### MSSQL(mssql_)/MySQL(mysql_) file attach
$val=$HTTP_POST_FILES['lob_upload']['tmp_name'];
$valn=$HTTP_POST_FILES['lob_upload']['name'];
$valt=$HTTP_POST_FILES['lob_upload']['type'];
$data=base64_encode(addslashes(fread(fopen($val, "r"), filesize($val))));
mssql_connect("srv","usr","pass") or die ("");
mssql_select_db("db") or die ("");
$query = "UPDATE $table SET $field='$data', $fieldname='$valn', $fieldtype='$valt' WHERE DocID='$DocID'";
$result = mssql_query($query) or die(mssql_error());
mssql_close();
######### MSSQL(mssql_)/MySQL(mysql_) open file attached
mssql_connect("srv","usr","pass") or die ("");
mssql_select_db("db") or die ("");
$query = "SELECT $field,$fieldtype FROM $table WHERE DocID='$DocID'";
$result = mssql_query($query) or die(mssql_error());
$row = mssql_fetch_array($result);
header("Content-type: $row[1]");
echo stripslashes(base64_decode($row[0]));
This strategy is good for Microsoft Word, Acrobat PDF, JPG image and so on (even zipped files!!!)
Problem: mime attachments sending as blank or almost completely blank documents (all data is lost)
Explanation: After a couple days of trying to mime pdf attachments without losing all data, I finally came across this function in some obsolete obscure post:
set_magic_quotes_runtime()
This is set to on by default in the machine, and it causes fread() and/or base64_encode() (both used in most mime examples I've seen) to read or encrypt binary without slashes for special characters. This causes sent files to process incorrectly, breaking, thus truncating most of the data in the file.
Fix: pass 0 to this function and it will do a one time turn off while your code executes.
example:
<?php
set_magic_quotes_runtime(0);
?>
This can also been turned off in the php.ini file, but I'm not sure what uses that setting or what the consequences might be.
info:
http://us2.php.net/manual/en/function.set-magic-quotes-runtime.php
If you use base64encoded strings as cookie names, make sure you remove '=' characters. At least Internet Explorer refuses cookie names containing '=' characters or urlencoded cookie names containing %xx character replacements. Use the function below to turn base64 encoded strings to bare alphabets (get rid of / and + characters as well)
<?php
function base64clean($base64string)
{
$base64string = str_replace(array('=','+','/'),'',$base64string);
return $base64string;
}
?>
As someone suggested above:
using base64_encode() to encode image data and finally output to browser using "data" scheme of IMG src:
<?
// ...
echo '<img src="data:image/png;base64,'.$encoded.' ">';
?>
Netscape browser supports this... However, Windows' Internet Explorer does not.
To embed binary contents in ascii text based html file for IE, you need use MIME multipart.
Re the message on 10-May-2003 04:02
You'll want to call urlencode on the base_64 encoded data before putting it into a GET. IIUC, base 64 output includes the plus and the slash, both of which will be mungered by browsers.
if you want to insert the base64 encoded image in your html <img src> you need to write 'data:datatype;base64,encodeddata' . For example here's a way to embed an PNG image data:
<?
//get the base64 encoded image
$handle = fopen($tempfile,'rb');
$file_content = fread($handle,filesize($tempfile));
fclose($handle);
$encoded = chunk_split(base64_encode($file_content));
//then echo to browser as:
echo '<img src="data:image/png;base64,'.$encoded.' ">';
?>
If you want attach a binary file into mail, pay attention to use mode with "B" flag into fopen function (This is useful only on systems which differentiate between binary and text files, i.e. Windows) Include the 'b' flag in order to make your scripts more portable.
<?php
$handle = fopen($source_file,'rb');
$file_content = fread($handle,filesize($source_file));
fclose($handle);
$encoded = chunk_split(base64_encode($file_content));
?>
I had to send a php array trough a FORM in HTML, and came up with this solution:
<?
$array[] = array("foo", "bar");
$coded_array = base64_encode(serialize($array));
?>
now u can put the $coded_array into an input field or even a GET link ex:
<a href="some_script.php?coded_array=<?=$coded_array;?>">script link</a>
after receiving it in the script you send it to, do the following:
<?
$coded_array = $_GET["coded_array"] // or $_POST off course
$array = unserialize(base64_decode($coded_array);
?>
You can use base64_encode to transfer image file into string text and then display them. I used this to store my images in a database and display them form there. First I open the files using fread, encoded the result, and stored that result in the database. Useful for creating random images.
image.php:
<?
header(" Content-Type: image/jpeg");
header(" Content-Disposition: inline");
$sql = "SELECT data FROM image where name='".$img."'";
$result = mysql_query($sql);
$row = mysql_fetch_row($result);
$image = $row[0];
echo base64_decode($image);
?>
And in the html file you put:
<img src="image.php?img=test3" border="0" alt="">
Guy Laor
Diese Seite bei php.net