PHPerKaigi 2024

mb_strimwidth

(PHP 4 >= 4.0.6, PHP 5, PHP 7, PHP 8)

mb_strimwidthGet truncated string with specified width

Description

mb_strimwidth(
    string $string,
    int $start,
    int $width,
    string $trim_marker = "",
    ?string $encoding = null
): string

Truncates string string to specified width, where halfwidth characters count as 1, and fullwidth characters count as 2. See » http://www.unicode.org/reports/tr11/ for details regarding East Asian character widths.

Parameters

string

The string being decoded.

start

The start position offset. Number of characters from the beginning of string (first character is 0), or if start is negative, number of characters from the end of the string.

width

The width of the desired trim. If negative width is specified, count from the end of the string.

Note:

Passing negative width is deprecated as of PHP 8.3.0.

trim_marker

A string that is added to the end of string when string is truncated.

encoding

The encoding parameter is the character encoding. If it is omitted or null, the internal character encoding value will be used.

Return Values

The truncated string. If trim_marker is set, trim_marker replaces the last chars to match the width.

Changelog

Version Description
8.3.0 Passing a negative width to mb_strimwidth() is now deprecated.
8.0.0 encoding is nullable now.
7.1.0 Support for negative starts and widths has been added.

Examples

Example #1 mb_strimwidth() example

<?php
echo mb_strimwidth("Hello World", 0, 10, "...");
// output: "Hello W..."
?>

See Also

add a note

User Contributed Notes 2 notes

up
1
dregad at NOSPAM-mantisbt dot org
10 months ago
I see a lot of usage in the field where people rely on this function to truncate a string to a given length and append some chars at the end following example #1 in the above documentation.

While this works just fine with Western alphabets, it should be noted that a string's width is NOT necessarily the same as its length.

In Chinese, Japanese and Korean, some characters can be represented as full or half width, which may lead to unexpected results...

<?php
$str
= ['English' => 'Switzerland',
'Half width' => 'スイス',
'Full width' => 'スイス',
];
foreach (
$str as $w => $s) {
printf("%-10s: %s (bytes=%d chars=%d width=%d)\nSubstring : %s\nTrim width: %s\n\n",
$w, $s,
strlen($s), mb_strlen($s), mb_strwidth($s),
mb_substr($s, 0, 3),
mb_strimwidth($s, 0, 3)
);
}
/* Output
# With ASCII, chars == width, so everything works as expected
English : Switzerland (bytes=11 chars=11 width=11)
Substring : Swi
Trim width: Swi

# With half-width katakanas, it works too
Half width: スイス (bytes=9 chars=3 width=3)
Substring : スイス
Trim width: スイス

# Full-width katakanas are twice as wide, so we only get the 1st 'su' !
Full width: スイス (bytes=9 chars=3 width=6)
Substring : スイス
Trim width: ス
*/
>?
up
1
jamesgrimshaw2006 at gmail dot com
4 months ago
mb_strimwidth will only append "trim_marker" if the string is over the width.

*test code*
$HashTags = 'Results';
$socialmediatext = 'abcdefghijklmnopqrstuvwxyz';
for( $i=0; $i<=20; $i++ )
{
$socialmediatext .= '.';
$Twittext = mb_strimwidth($socialmediatext,0,35,$HashTags);
echo "<p>[".strlen($Twittext)."]$Twittext</p>";
}

*output*
[27]abcdefghijklmnopqrstuvwxyz.
[28]abcdefghijklmnopqrstuvwxyz..
[29]abcdefghijklmnopqrstuvwxyz...
[30]abcdefghijklmnopqrstuvwxyz....
[31]abcdefghijklmnopqrstuvwxyz.....
[32]abcdefghijklmnopqrstuvwxyz......
[33]abcdefghijklmnopqrstuvwxyz.......
[34]abcdefghijklmnopqrstuvwxyz........
[35]abcdefghijklmnopqrstuvwxyz.........
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
[35]abcdefghijklmnopqrstuvwxyz..Results
To Top