PHP Conference Nagoya 2025

strtok

(PHP 4, PHP 5, PHP 7, PHP 8)

strtokTokenize string

Опис

strtok(string $string, string $token): string|false

Alternative signature (not supported with named arguments):

strtok(string $token): string|false

strtok() splits a string (string) into smaller strings (tokens), with each token being delimited by any character from token. That is, if you have a string like "This is an example string" you could tokenize this string into its individual words by using the space character as the token.

Note that only the first call to strtok uses the string argument. Every subsequent call to strtok only needs the token to use, as it keeps track of where it is in the current string. To start over, or to tokenize a new string you simply call strtok with the string argument again to initialize it. Note that you may put multiple tokens in the token parameter. The string will be tokenized when any one of the characters in the token argument is found.

Зауваження:

This function behaves slightly different from what one may expect being familiar with explode(). First, a sequence of two or more contiguous token characters in the parsed string is considered to be a single delimiter. Also, a token situated at the start or end of the string is ignored. For example, if a string ";aaa;;bbb;" is used, successive calls to strtok() with ";" as a token would return strings "aaa" and "bbb", and then false. As a result, the string will be split into only two elements, while explode(";", $string) would return an array of 5 elements.

Параметри

string

The string being split up into smaller strings (tokens).

token

The delimiter used when splitting up string.

Значення, що повертаються

A string token, or false if no more tokens are available.

Журнал змін

Версія Опис
8.3.0 Now emits E_WARNING when token is not provided.

Приклади

Приклад #1 strtok() example

<?php
$string
= "This is\tan example\nstring";
/* Use tab and newline as tokenizing characters as well */
$tok = strtok($string, " \n\t");

while (
$tok !== false) {
echo
"Word=$tok<br />";
$tok = strtok(" \n\t");
}
?>

Приклад #2 strtok() behavior on empty part found

<?php
$first_token
= strtok('/something', '/');
$second_token = strtok('/');
var_dump($first_token, $second_token);
?>

Поданий вище приклад виведе:

string(9) "something"
    bool(false)

Приклад #3 The difference between strtok() and explode()

<?php
$string
= ";aaa;;bbb;";

$parts = [];
$tok = strtok($string, ";");
while (
$tok !== false) {
$parts[] = $tok;
$tok = strtok(";");
}
echo
json_encode($parts),"\n";

$parts = explode(";", $string);
echo
json_encode($parts),"\n";

Поданий вище приклад виведе:

["aaa","bbb"]
["","aaa","","bbb",""]

Примітки

Увага

Ця функція може повертати як логічне false, так і не логічне значення, яке прирівнюється до false. Докладніше про це описано в розділі Логічні типи даних. Для перевірки значення, яке повертає ця функція, використовується оператор ===.

Прогляньте також

  • explode() - Розбиває рядок на підрядки

add a note

User Contributed Notes 20 notes

up
98
eep2004 at ukr dot net
11 years ago
<?php
// strtok example
$str = 'Hello to all of Ukraine';
echo
strtok($str, ' ').' '.strtok(' ').' '.strtok(' ');
?>
Result:
Hello to all
up
19
elarlang at gmail dot com
13 years ago
If you have memory-usage critical solution, you should keep in mind, that strtok function holds input string parameter (or reference to it?) in memory after usage.

<?php
function tokenize($str, $token_symbols) {
$word = strtok($str, $token_symbols);
while (
false !== $word) {
// do something here...
$word = strtok($token_symbols);
}
}
?>
Test-cases with handling ~10MB plain-text file:
Case #1 - unset $str variable
<?php
$token_symbols
= " \t\n";
$str = file_get_contents('10MB.txt'); // mem usage 9.75383758545 MB (memory_get_usage() / 1024 / 1024));
tokenize($str, $token_symbols); // mem usage 9.75400161743 MB
unset($str); // 9.75395584106 MB
?>
Case #1 result: memory is still used

Case #2 - call strtok again
<?php
$token_symbols
= " \t\n";
$str = file_get_contents('10MB.txt'); // 9.75401306152 MB
tokenize($str, $token_symbols); // 9.75417709351
strtok('', ''); // 9.75421524048
?>
Case #2 result: memory is still used

Case #3 - call strtok again AND unset $str variable
<?php
$token_symbols
= " \t\n";
$str = file_get_contents('10MB.txt'); // 9.75410079956 MB
tokenize($str, $token_symbols); // 9.75426483154 MB
unset($str);
strtok('', ''); // 0.0543975830078 MB
?>
Case #3 result: memory is free

So, better solution for tokenize function:
<?php
function tokenize($str, $token_symbols, $token_reset = true) {
$word = strtok($str, $token_symbols);
while (
false !== $word) {
// do something here...
$word = strtok($token_symbols);
}

if(
$token_reset)
strtok('', '');
}
?>
up
25
manicdepressive at mindless dot com
20 years ago
<pre><?php
/** get leading, trailing, and embedded separator tokens that were 'skipped'
if for some ungodly reason you are using php to implement a simple parser that
needs to detect nested clauses as it builds a parse tree */

$str = "(((alpha(beta))(gamma))";

$seps = '()';
$tok = strtok( $str,$seps ); // return false on empty string or null
$cur = 0;
$dumbDone = FALSE;
$done = (FALSE===$tok);
while (!
$done) {
// process skipped tokens (if any at first iteration) (special for last)
$posTok = $dumbDone ? strlen($str) : strpos($str, $tok, $cur );
$skippedMany = substr( $str, $cur, $posTok-$cur ); // false when 0 width
$lenSkipped = strlen($skippedMany); // 0 when false
if (0!==$lenSkipped) {
$last = strlen($skippedMany) -1;
for(
$i=0; $i<=$last; $i++){
$skipped = $skippedMany[$i];
$cur += strlen($skipped);
echo
"skipped: $skipped\n";
}
}
if (
$dumbDone) break; // this is the only place the loop is terminated

// process current tok
echo "curr tok: ".$tok."\n";

// update cursor
$cur += strlen($tok);

// get any next tok
if (!$dumbDone){
$tok = strtok($seps);
$dumbDone = (FALSE===$tok);
// you're not really done till you check for trailing skipped
}
};
?></pre>
up
18
eep2004 at ukr dot net
9 years ago
Remove GET variables from the URL
<?php
echo strtok('http://example.com/index.php?foo=1&bar=2', '?');
?>
Result:
http://example.com/index.php
up
5
benighted at gmail dot com
15 years ago
Simple way to tokenize search parameters, including double or single quoted keys. If only one quote is found, the rest of the string is assumed to be part of that token.

<?php
$token
= strtok($keywords,' ');
while (
$token) {
// find double quoted tokens
if ($token{0}=='"') { $token .= ' '.strtok('"').'"'; }
// find single quoted tokens
if ($token{0}=="'") { $token .= ' '.strtok("'")."'"; }

$tokens[] = $token;
$token = strtok(' ');
}
?>

Use substr(1,strlen($token)) and remove the part that adds the trailing quotes if you want your output without quotes.
up
4
Axeia
10 years ago
Might be pointing out the obvious but if you'd rather use a for loop rather than a while (to keep the token strings on the same line for readability for example), it can be done. Added bonus, it doesn't put a $tok variable outside the loop itself either.
Downside however is that you're not able to manually free up the memory used using the technique mentioned by elarlang.

<?php
for($tok = strtok($str, ' _-.'); $tok!==false; $tok = strtok(' _-.'))
{
echo
"$tok </br>";
}
?>
up
6
info at maisuma dot jp
10 years ago
If you want to tokenize by only one letter, explode() is much faster compared to strtok().

<?php
$str
=str_repeat('foo ',10000);

//explode()
$time=microtime(TRUE);
$arr=explode($str,' ');
$time=microtime(TRUE)-$time;
echo
"explode():$time sec.".PHP_EOL;

//strtok()
$time=microtime(TRUE);
$ret=strtok(' ',$str);
while(
$ret!==FALSE){
$ret=strtok(' ');
}
$time=microtime(TRUE)-$time;
echo
"strtok():$time sec.".PHP_EOL;

?>

The result is : (PHP 5.3.3 on CentOS)

explode():0.001317024230957 sec.
strtok():0.0058917999267578 sec.

explode() is about five times fast in short strings, too.
up
5
Logikos
15 years ago
This looks very simple, but it took me a long time to figure out so I thought I'd share it incase someone else was wanting the same thing:

this should work similar to substr() but with tokens instead!

<?php
/* subtok(string,chr,pos,len)
*
* chr = chr used to seperate tokens
* pos = starting postion
* len = length, if negative count back from right
*
* subtok('a.b.c.d.e','.',0) = 'a.b.c.d.e'
* subtok('a.b.c.d.e','.',0,2) = 'a.b'
* subtok('a.b.c.d.e','.',2,1) = 'c'
* subtok('a.b.c.d.e','.',2,-1) = 'c.d'
* subtok('a.b.c.d.e','.',-4) = 'b.c.d.e'
* subtok('a.b.c.d.e','.',-4,2) = 'b.c'
* subtok('a.b.c.d.e','.',-4,-1) = 'b.c.d'
*/
function subtok($string,$chr,$pos,$len = NULL) {
return
implode($chr,array_slice(explode($chr,$string),$pos,$len));
}
?>

explode breaks the tokens up into an array, array slice alows you to pick then tokens you want, and then implode converts it back to a string

although its far from a clone, this was inspired by mIRC's gettok() function
up
2
gilthans at NOSPAM dot gmail dot com
12 years ago
Note that strtok may receive different tokens each time. Therefore, if, for example, you wish to extract several words and then the rest of the sentence:

<?php
$text
= "13 202 5 This is a long message explaining the error codes.";
$error1 = strtok($text, " "); //13
$error2 = strtok(" "); //202
$error3 = strtok(" "); //5
$error_message = strtok(""); //Notice the different token parameter
echo $error_message; //This is a long message explaining the error codes.
?>
up
3
KrazyBox
15 years ago
As of the change in strtok()'s handling of empty strings, it is now useless for scripts that rely on empty data to function.

Take for instance, a standard header. (with UNIX newlines)

http/1.0 200 OK\n
Content-Type: text/html\n
\n
--HTML BODY HERE---

When parsing this with strtok, one would wait until it found an empty string to signal the end of the header. However, because strtok now skips empty segments, it is impossible to know when the header has ended.
This should not be called `correct' behavior, it certainly is not. It has rendered strtok incapable of (properly) processing a very simple standard.

This new functionality, however, does not affect Windows style headers. You would search for a line that only contains "\r"
This, however, is not a justification for the change.
up
1
azeem
15 years ago
Here is a java like StringTokenizer class using strtok function:

<?php

/**
* The string tokenizer class allows an application to break a string into tokens.
*
* @example The following is one example of the use of the tokenizer. The code:
* <code>
* <?php
* $str = 'this is:@\t\n a test!';
* $delim = ' !@:'\t\n; // remove these chars
* $st = new StringTokenizer($str, $delim);
* while ($st->hasMoreTokens()) {
* echo $st->nextToken() . "\n";
* }
* prints the following output:
* this
* is
* a
* test
* ?>
* </code>
*/
class StringTokenizer {

/**
* @var string
*/
private $token;

/**
* @var string
*/
private $delim;
/**
* Constructs a string tokenizer for the specified string
* @param string $str String to tokenize
* @param string $delim The set of delimiters (the characters that separate tokens)
* specified at creation time, default to ' '
*/
public function __construct(/*string*/ $str, /*string*/ $delim = ' ') {
$this->token = strtok($str, $delim);
$this->delim = $delim;
}

public function
__destruct() {
unset(
$this);
}

/**
* Tests if there are more tokens available from this tokenizer's string. It
* does not move the internal pointer in any way. To move the internal pointer
* to the next element call nextToken()
* @return boolean - true if has more tokens, false otherwise
*/
public function hasMoreTokens() {
return (
$this->token !== false);
}

/**
* Returns the next token from this string tokenizer and advances the internal
* pointer by one.
* @return string - next element in the tokenized string
*/
public function nextToken() {
$current = $this->token;
$this->token = strtok($this->delim);
return
$current;
}
}
?>
up
2
voojj3054 at gmail dot com
1 year ago
Hello, portuguese documentation of strtok is wrong, at this part which the example(2) is wrong.

Exemplo #2 Comportamento antigo da strtok()
<?php
$first_token
= strtok('/something', '/');
$second_token = strtok('/');
var_dump ($first_token, $second_token);
?>

O exemplo acima produzirá:

string(0) ""
string(9) "something"

(this example above, should be inverted as this:)

Correct:
string(9) "something"
string(0) ""

(exemple 3 is correct)
Exemplo #3 Novo comportamento da strtok()
<?php
$first_token
= strtok('/something', '/');
$second_token = strtok('/');
var_dump ($first_token, $second_token);
?>

O exemplo acima produzirá:

string(9) "something"
bool(false)
up
1
pradador at me dot com
13 years ago
Here's a simple class that allows you to iterate through string tokens using a foreach loop.

<?php
/**
* The TokenIterator class allows you to iterate through string tokens using
* the familiar foreach control structure.
*
* Example:
* <code>
* <?php
* $string = 'This is a test.';
* $delimiters = ' ';
* $ti = new TokenIterator($string, $delimiters);
*
* foreach ($ti as $count => $token) {
* echo sprintf("%d, %s\n", $count, $token);
* }
*
* // Prints the following output:
* // 0. This
* // 1. is
* // 2. a
* // 3. test.
* </code>
*/
class TokenIterator implements Iterator
{
/**
* The string to tokenize.
* @var string
*/
protected $_string;

/**
* The token delimiters.
* @var string
*/
protected $_delims;

/**
* Stores the current token.
* @var mixed
*/
protected $_token;

/**
* Internal token counter.
* @var int
*/
protected $_counter = 0;

/**
* Constructor.
*
* @param string $string The string to tokenize.
* @param string $delims The token delimiters.
*/
public function __construct($string, $delims)
{
$this->_string = $string;
$this->_delims = $delims;
$this->_token = strtok($string, $delims);
}

/**
* @see Iterator::current()
*/
public function current()
{
return
$this->_token;
}

/**
* @see Iterator::key()
*/
public function key()
{
return
$this->_counter;
}

/**
* @see Iterator::next()
*/
public function next()
{
$this->_token = strtok($this->_delims);

if (
$this->valid()) {
++
$this->_counter;
}
}

/**
* @see Iterator::rewind()
*/
public function rewind()
{
$this->_counter = 0;
$this->_token = strtok($this->_string, $this->_delims);
}

/**
* @see Iterator::valid()
*/
public function valid()
{
return
$this->_token !== FALSE;
}
}
?>
up
1
bohwaz
1 year ago
Please note that strtok memory is shared between all PHP code currently executed, even included files. This can bite you in unexpected ways if you are not careful.

For example:

<?php

$path
= 'dir/file.ext';
$dir_name = strtok($path, '/');

if (
$dir_name !== (new Module)->getAllowedDirName()) {
throw new
\Exception('Invalid directory name');
}

$file_name = strtok('');

?>

Seems easy enough, but if your Module class is not loaded, this triggers the autoloader. The autoloader *MAY* use strtok inside its loading code.

Or your Module class *MAY* use strtok inside its constructor.

This means you will never get your $file_name correctly.

So: you should *always* group strtok calls, without any external code between two strtok calls.

This would be OK:

<?php

$path
= 'dir/file.ext';
$dir_name = strtok($path, '/');
$file_name = strtok('');

if (
$dir_name !== (new Module)->getAllowedDirName()) {
throw new
\Exception('Invalid directory name');
}

?>

This might cause issues:

<?php

$path
= 'one/two#three';
$a = strtok($path, '/');
$b = strtok(Module::NAME_SEPARATOR);
$c = strtok('');

?>

Because your autoloader might be using strtok.

This would be avoided by fetching all parameters used in strtok *before* the calls:

<?php

$path
= 'one/two#three';
$separator = Module::NAME_SEPARATOR;
$a = strtok($path, '/');
$b = strtok($separator);
$c = strtok('');

?>
up
0
heiangus at hotmail dot com
5 years ago
I found this useful for parsing user entered links in text fields.

e.g. This is a link <http://example.com>

function parselink($link) {
$bit1 = trim(strtok($link, '<'));
$bit2 = trim(strtok('>'));
$html = '<a href="'.$bit2.'">'.$bit1.'</a>';
return $html; // <a href="http://example.com">This is a link</a>
}
up
-1
David Spector
3 years ago
After obtaining zero or more tokens with calls to strtok, you can obtain the remainder of the input string by calling strtok with an empty string as the delimiter.
up
0
charlie dot ded at orange dot fr
8 years ago
@maisuma you invert paramaters of explode() and strtok() functions, your code does not do what you expect.
You expect to read the input string token after token so equivalent code for strtok() is arra_filter(explode()) because explode() return lines of empty string when several delimiters are contiguous in the read string, for example 2 whitespaces between.

In fact strtok() is much faster (x2 at least) than arra_filter(explode()) if the read string contains several contiguous delimiters ,
it is slower if the read string contains one and only one delimiter between tokens.

<?php

$repeat
= 10;
$delimiter = ':';
$str=str_repeat('foo:',$repeat);

$timeStrtok=microtime(TRUE);
$token = strtok($str, $delimiter);
while(
$token!==FALSE){
//echo $token . ',';
$token=strtok($delimiter);
}
$timeStrtok -=microtime(TRUE);

$timeExplo=microtime(TRUE);
$arr = explode($delimiter, $str);
//$arr = array_filter($arr);
$timeExplo -=microtime(TRUE);

//print_r($arr);

$X = 1000000; $unit = 'microsec';

echo
PHP_EOL . ' explode() : ' . -$timeExplo . ' ' .$unit .PHP_EOL . ' strtok() : ' . -$timeStrtok . ' ' . $unit .PHP_EOL;

$timeExplo=round(-$timeExplo*$X);
$timeStrtok=round(-$timeStrtok*$X);

echo
PHP_EOL . ' explode() : ' . $timeExplo . ' ' .$unit .PHP_EOL . ' strtok() : ' . $timeStrtok . ' ' . $unit .PHP_EOL;
echo
' ratio explode / strtok : ' . round($timeExplo / $timeStrtok,1) . PHP_EOL;

?>
up
0
fabiolimasouto at gmail dot com
14 years ago
this example will hopefully help you understand how this function works:

<?php
$selector
= 'div.class#id';
$tagname = strtok($selector,'.#');
echo
$tagname.'<br/>';

while(
$tok = strtok('.#'))
{
echo
$tok.'<br/>';
}

?>

Outputs:
div
class
id
up
-2
eep2004 at ukr dot net
10 years ago
Remove GET variables from the URL
<?php
$url
= strtok('http://php.net/manual/en/ref.strings.php?foo=1&bar=2', '?');
// $url = 'http://php.net/manual/en/ref.strings.php'
up
-2
mac.com@nemo
18 years ago
This function takes a string and returns an array with words (delimited by spaces), also taking into account quotes, doublequotes, backticks and backslashes (for escaping stuff).
So

$string = "cp 'my file' to `Judy's file`";
var_dump(parse_cli($string));

would yield:

array(4) {
[0]=>
string(2) "cp"
[1]=>
string(7) "my file"
[2]=>
string(5) "to"
[3]=>
string(11) "Judy's file"
}

Way it works, runs through the string character by character, for each character looking up the action to take, based on that character and its current $state.
Actions can be (one or more of) adding the character/string to the current word, adding the word to the output array, and changing or (re)storing the state.
For example a space will become part of the current 'word' (or 'token') if $state is 'doublequoted', but it will start a new token if $state was 'unquoted'.
I was later told it's a "tokeniser using a finite state automaton". Who knew :-)

<?php

#_____________________
# parse_cli($string) /
function parse_cli($string) {
$state = 'space';
$previous = ''; // stores current state when encountering a backslash (which changes $state to 'escaped', but has to fall back into the previous $state afterwards)
$out = array(); // the return value
$word = '';
$type = ''; // type of character
// array[states][chartypes] => actions
$chart = array(
'space' => array('space'=>'', 'quote'=>'q', 'doublequote'=>'d', 'backtick'=>'b', 'backslash'=>'ue', 'other'=>'ua'),
'unquoted' => array('space'=>'w ', 'quote'=>'a', 'doublequote'=>'a', 'backtick'=>'a', 'backslash'=>'e', 'other'=>'a'),
'quoted' => array('space'=>'a', 'quote'=>'w ', 'doublequote'=>'a', 'backtick'=>'a', 'backslash'=>'e', 'other'=>'a'),
'doublequoted' => array('space'=>'a', 'quote'=>'a', 'doublequote'=>'w ', 'backtick'=>'a', 'backslash'=>'e', 'other'=>'a'),
'backticked' => array('space'=>'a', 'quote'=>'a', 'doublequote'=>'a', 'backtick'=>'w ', 'backslash'=>'e', 'other'=>'a'),
'escaped' => array('space'=>'ap', 'quote'=>'ap', 'doublequote'=>'ap', 'backtick'=>'ap', 'backslash'=>'ap', 'other'=>'ap'));
for (
$i=0; $i<=strlen($string); $i++) {
$char = substr($string, $i, 1);
$type = array_search($char, array('space'=>' ', 'quote'=>'\'', 'doublequote'=>'"', 'backtick'=>'`', 'backslash'=>'\\'));
if (!
$type) $type = 'other';
if (
$type == 'other') {
// grabs all characters that are also 'other' following the current one in one go
preg_match("/[ \'\"\`\\\]/", $string, $matches, PREG_OFFSET_CAPTURE, $i);
if (
$matches) {
$matches = $matches[0];
$char = substr($string, $i, $matches[1]-$i); // yep, $char length can be > 1
$i = $matches[1] - 1;
}else{
// no more match on special characters, that must mean this is the last word!
// the .= hereunder is because we *might* be in the middle of a word that just contained special chars
$word .= substr($string, $i);
break;
// jumps out of the for() loop
}
}
$actions = $chart[$state][$type];
for(
$j=0; $j<strlen($actions); $j++) {
$act = substr($actions, $j, 1);
if (
$act == ' ') $state = 'space';
if (
$act == 'u') $state = 'unquoted';
if (
$act == 'q') $state = 'quoted';
if (
$act == 'd') $state = 'doublequoted';
if (
$act == 'b') $state = 'backticked';
if (
$act == 'e') { $previous = $state; $state = 'escaped'; }
if (
$act == 'a') $word .= $char;
if (
$act == 'w') { $out[] = $word; $word = ''; }
if (
$act == 'p') $state = $previous;
}
}
if (
strlen($word)) $out[] = $word;
return
$out;
}

?>
To Top