str_replace this character: ö

Question

I am trying to do a str_replace on the following line:

team 4 -1ööööö57167168

I've been able to do it in the past just by doing this:

str_replace("ööööö", " ",trim($line));

The desired outcome should be team 4 -1 57167168

However, now that I'm trying to do it in a Codeigniter application, it fails to replace it and I'm not sure what to try.

Is this an encoding config issue with Codeigniter? How can I get it to replace correctly?

does using the utf-8 escape work? ie, `str_replace("\xC3\x96", " ",trim($line));` — cegfault, Aug 12 '12 at 22:35
Are you going [utf8 all the way through](http://stackoverflow.com/questions/279170/utf-8-all-the-way-through) ? Did you save your file as utf-8? Does this `echo bin2hex('ööööö');` not give you `c3b6c3b6c3b6c3b6c3b6`? — Esailija, Aug 12 '12 at 22:40
@Esailija I think so, the file is generated by the game itself and I have no control over it's encoding. — Motive, Aug 12 '12 at 22:42
Verify that the file is utf8 (e.g. in mac osx terminal run `file -I yourinputfile`). — mAu, Aug 12 '12 at 22:44
@MotiveKyle ok cool, then your file is saved as utf8. The string you are getting is not, where are you getting it from? From a database? What was the actual outcome, btw. — Esailija, Aug 12 '12 at 22:49
@Esailija I'm just uploading the file using Codeigniter's file uploading class and just using fopen to read it from the folder. The file is a log of player actions in a game (for watching replays). I'm not pulling anything from a database yet. Sorry, what do you mean by the actual outcome? — Motive, Aug 12 '12 at 22:56
I mean what you see as a result? You could also post the `bin2hex` of the file. if you see it givin c3b6 for `ö` then your issue isn't with encoding at all. — Esailija, Aug 12 '12 at 22:58

score 1 · Answer 1 · answered Aug 12 '12 at 22:37

1

Checkout the multi byte string functions here http://php.net/manual/en/ref.mbstring.php and specifically the mb_str_replace.

Note, ö can be represented in multiple character sets so make sure you are matching the correct character set!

answered Aug 12 '12 at 22:37

williamvicary

FWIW, if both strings are utf-8, `str_replace` will work just fine here. – Esailija Aug 12 '12 at 22:48

score -3 · Answer 2 · answered Aug 12 '12 at 22:33

-3

Use utf8_encode() and utf8_decode(), PHP isn't a UTF8 native language !

answered Aug 12 '12 at 22:33

HatchBipper

This is pretty harmful advice because there is no need for these conversions (You can just work with utf8) and you will lose out on 99.99% of the unicode characters because ISO-8859-1 cannot encode them. For example, [`utf8_decode("€")` gives `?`](http://codepad.viper-7.com/CisDx3) and there is no way to know from that question mark (`?`) that the original character was `€` – Esailija Aug 12 '12 at 22:59

2 Answers2