PREG_OFFSET_CAPTURE returns wrong offset if there is a multibyte character

[moxol]

Function preg_match_all($pattern, $subject, PREG_OFFSET_CAPTURE ) returns wrong offset for matched value if there are multibyte characters in subject.
It seems that PHP counts multibyte character as two characters instead of one.

Is there a solution to this problem?

[requinix]

Bytes and characters are not the same thing. You think it's characters, PHP is telling you bytes.

[moxol]

If I use that offset in substr() then it would give me wrong substring.

How to solve this issue?

Edited July 26, 2022 by moxol

[moxol]

I think I found the soluton:

 

 $offset = mb_strlen(substr($text, 0, $offset_capture))

Is this ok?

Edited July 26, 2022 by moxol

[requinix]

Is that what you want to do? Find out the "character" position for a match's byte position?

You're only giving out small pieces of information at a time. It's hard to give advice on broader problems or situations when we have to work with is your use of mb_strlen and substr...

[moxol]

 

36 minutes ago, requinix said:

Is that what you want to do? Find out the "character" position for a match's byte position?

Yes, I am using regex preg_match_all to find offset of matches so that I could use substr to the position of PREG_OFFSET_CAPTURE.
And then I would use mb_strlen(substr($text, 0, $offset_capture))
to find "real" offset if there are multibytes in the text.