Remove special characters in regex PHP that allow accented words and chinese language

Question

I need a solution that remove all special characters except alphanumeric and accents. I tryed this solution without success.

preg_replace('/[^a-zA-ZáéíóúÁÉÍÓÚâêîôÂÊÎÔãõÃÕçÇ0-9_ \.&-]/s', '', $string);

Furthermore, its need that regex (or other specific solution) allow chinese and arabian charset.

any help its really apprecieted!

So what exactly *are* "special characters"? *"Remove everything except a few selected ones and this whole giant block of other characters which make up the majority of Unicode"* is a bit vague. *Why* do you need to remove those characters? — deceze, Nov 28 '13 at 14:11
Im building a search text field than I need clean keywords from "/%$@* and more... — Vincenzo Lo Palo, Nov 28 '13 at 14:13
What's so special about "/%$@*, why can't people search for those characters? Why only "alphanumeric, accents, Chinese and Arabian", what about, say, Korean, Japanese and Sanskrit? — deceze, Nov 28 '13 at 14:14
So what exactly don't you allow then? Look at the link above by CBroe. Sounds like `\pL` may be what you want, but that's honestly hard to tell. — deceze, Nov 28 '13 at 14:18
Will there be any results for "aslkgjalkgjaljgaslkjg"? No? Then display 0 results. Will there be any results for Chinese characters? No? Then display 0 results. — maček, Nov 28 '13 at 15:26

score 8 · Accepted Answer · answered Nov 28 '13 at 15:21

8

$string = preg_replace('/\PL/u', '', $string);

L is a character attribute meaning letter
\P means does not match attribute
/u is the Unicode modifier, you need this if you want to handle Unicode characters
make sure $string is encoded in UTF-8

So this matches all non-letters and removes them. I can only guess that this matches what you want. See http://www.php.net/manual/en/regexp.reference.unicode.php for more attributes you could match by, e.g. /[^\pL\pS]/u would match everything except letters and "symbols".

answered Nov 28 '13 at 15:21

deceze

510,633
85
743
889

If you need to play around with what matches what, see here: http://rubular.com/r/WcotUy4ju9. (Note that this regex engine requires you to write `\p{L}` instead of `\pL`.) – deceze Nov 28 '13 at 15:33
excellent! thanks and congratulations! – Vincenzo Lo Palo Nov 28 '13 at 17:17
if Do I need exclude space? – Vincenzo Lo Palo Nov 28 '13 at 17:44
Add it to the group in my second example. Either just `\s` or `\pZ`. – deceze Nov 28 '13 at 17:51
$string = preg_replace('\pL\pS/u', '', $string); is it correct? – Vincenzo Lo Palo Nov 28 '13 at 18:04
No, that matches a Letter *followed by* a Symbol. As I wrote: `/[^\pL\pS]/u`. – deceze Nov 28 '13 at 19:05
I wrote so but now work: preg_replace('/[^\pL\pS]/u', '', $string) – Vincenzo Lo Palo Nov 28 '13 at 19:10

score 0 · Answer 2 · answered Nov 28 '13 at 14:22

0

echo preg_replace('/[^أ-يA-Za-z0-9 ]/ui', '', $string);

answered Nov 28 '13 at 14:22

Moeed Farooqui

3,604
1
18
23

1

not work. its ok for special characters but remove foreign language symbols too. – Vincenzo Lo Palo Nov 28 '13 at 14:26

Remove special characters in regex PHP that allow accented words and chinese language

2 Answers2