How do I detect non-ASCII characters in a string?

Question

If I have a PHP string, how can I determine if it contains at least one non-ASCII character or not, in an efficient way? And by non-ASCII character, I mean any character that is not part of this table, http://www.asciitable.com/, positions 32 - 126 inclusive.

So not only does it have to be part of the ASCII table, but it also has to be printable. I want to detect a string that contains at least one character that does not meet these specifications (either non-printable ASCII, or a different character altogether, such as a Unicode character that is not part of that table.

So you do not mean Unicode, but non `US-ASCII`? I think this is worth to specify if you're looking for something efficient. — hakre, Jun 27 '11 at 19:14
Can you make any safe assumption about the string, such as encoding? — Álvaro González, Jun 27 '11 at 19:19
All ASCII characters are <= 127, and any UTF-8 character sequence that decodes to a non-ASCII character has at least one byte with the highest bit set. Thus, if you have no byte >127, it's ASCII. Detecting UTF-8 encoding as suggested in the answers below will probably work too, but could possibly be ambiguous (since ASCII characters are incidentially _also_ UTF-8 characters). — Damon, Jun 27 '11 at 19:22
Similar to http://stackoverflow.com/questions/4147646/determine-if-utf-8-text-is-all-ascii — Gras Double, Jan 07 '15 at 17:30

score 76 · Accepted Answer · edited Aug 01 '19 at 16:47

76

I found it more useful to detect if any character falls out of the list

if(preg_match('/[^\x20-\x7e]/', $string))

edited Aug 01 '19 at 16:47

BenMorel

34,448
50
182
322

answered Jun 27 '11 at 19:33

Karolis

9,396
29
38

This answer is good but you can find more solutions in this post http://stackoverflow.com/questions/4147646/determine-if-utf-8-text-is-all-ascii – ElSinus Nov 04 '15 at 08:18
@wheresrhys I think your snippets tests if all characters in string are ascii, for any character code should be `/[^\x20-\x7f]/.test(theString)` – Igor Jerosimić Jul 27 '18 at 13:19

score 44 · Answer 2 · answered Jun 27 '11 at 19:21

44

You can use mb_detect_encoding and check for ASCII:

mb_detect_encoding($str, 'ASCII', true)

This will return false if $str contains at least one non-ASCI character (byte value > 0x7F).

answered Jun 27 '11 at 19:21

Gumbo

643,351
109
780
844

20

[`mb_check_encoding`](http://php.net/manual/en/function.mb-check-encoding.php) would be more appropriate: `mb_check_encoding($str, 'ASCII')` – Gras Double Jan 07 '15 at 17:26

score 4 · Answer 3 · answered Dec 30 '13 at 21:51

4

The function ctype_print returns true iff all characters fall into the ASCII range 32-126 (PHP unit test).

answered Dec 30 '13 at 21:51

Steffen

952
7
8

`php -r 'echo ctype_print("\xa0");'` prints `1` so there's something fishy with this function. – forthrin Aug 29 '16 at 11:47
@forthrin: I cannot confirm that. For me, `php -r 'var_dump(ctype_print("\xa0"));'` returns false (using PHP 7.0.10). – Steffen Aug 30 '16 at 13:00
I'm on PHP 7.0.10 too, Homebrew version (OS X). Could the difference be caused by terminal, locale, php.ini or other environmental factors? – forthrin Aug 30 '16 at 17:49
Doesn't work for me either, PHP 7.0.5 on Windows - no idea why. It does not seem to work anymore. We should probably open a bug report? – mindplay.dk Sep 16 '16 at 13:04

score 4 · Answer 4 · answered Aug 18 '20 at 12:07

I benchmarked the suggested functions as I need this check for batch processing of shorter (1000 characters max) strings. I tested 10k iterations of 30 different strings (empty, short, longer, ascii, accents, japanese, emoji, non-ascii start, non-ascii end etc). Here are the rough results:

mb_check_encoding: 95ms average. Performance degrades way faster than preg_match and ctype as the strings get longer (1MB+).

mb_check_encoding($input, 'ASCII');

preg_match: 85ms average. Decently fast for 1MB+ strings (walks the string, so faster if there are non-ascii characters early in the string).

!preg_match('/[\\x80-\\xff]/', $input);

ctype_print: 83ms average. Decently fast for 1MB+ strings (walks the string, so faster if there are non-ascii characters early in the string). DO NOTE that this is not really an ascii check.

ctype_print($input);

while/ord: 500ms average. I'm still waiting for the 1MB+ strings test to finish.

function is_ascii($input) {
    $num = 0;
    while( isset( $string[$num] ) ) {
        if( ord( $string[$num] ) & 0x80 ) {
            return false;
        }
        $num++;
    }
    return true;
}

score 4 · Answer 5 · edited Aug 29 '23 at 21:32

4

Try (mb_detect_encoding). For example:

mb_check_encoding($identifier, 'ASCII');

edited Aug 29 '23 at 21:32

Dharman

30,962
25
85
135

answered Jun 27 '11 at 19:15

Hans Kerkhof

442
3
7

score 2 · Answer 6 · answered Apr 01 '13 at 02:09

Try: (Source)

function is_ascii( $string = '' ) {
    return ( bool ) ! preg_match( '/[\\x80-\\xff]+/' , $string );
}

Although, all of the above answers are correct, but depending upon the input, these solutions may give wrong answers. See the last section in this ASCII validation post.

score 1 · Answer 7 · answered Jun 27 '11 at 19:12

1

You could use:

mb_detect_encoding

but it will be maybe not as precise as you want it to be.

answered Jun 27 '11 at 19:12

fyr

20,227
7
37
53

score -1 · Answer 8 · answered Jun 27 '11 at 19:26

I suggest you look into utf8_encode or utf8_decode under PHP's manual:

http://www.php.net/manual/en/function.utf8-encode.php

Look into the examples down below as it may have something there that leads you to the right direction if not finding what you are looking for.

How do I detect non-ASCII characters in a string?

8 Answers8

Linked

Related