[RESOLVED] Charsets in HTML + PHP

[Desdinova]

I thought I finally figured out how it worked. Simply use UTF-8 in the charset attribute of the content-type meta of your valid HTML page, and voila. If that doesn't work, add a php content-type header with charset, and then voila.

After doing that, my 'ë' still looks like a question mark when I view the page 🙁 in both the rendered page and source.

My ë is written as 'ë' in the file, not as 'ë', and generally I don't really want that to change. this file gets read by PHP and then echoed.

can someone tell me what to look for?

[nrg_alpha]

The question mark simply signifies that that particular unicode's code point doesn't contain an equivalent in the given chararcter set you are using (in your case, UTF-8)... I don't see the problem with using ë Not hard to implement.

PHP will echo that just fine.

EDIT - Here is a good article on unicodes.

EDIT 2: Perhaps passing the file though something like utf8_encode() if you don't want to go through the file and manually change things...

You can always use str_replace() to change ë to ë throughout the file if you don't want to always have to pass the file through utf8_encode().

[Desdinova]

king! utf8_encode() does the trick.

I considered a str_replace, but I hate having to list alllll possibilities (ë, é, è, etc).

Thanks a bunch.

I do find it fairly odd that reading out a file through file_get_contents() doesn't just read it without altering it.

[nrg_alpha]

Desdinova;10914128 wrote:

I considered a str_replace, but I hate having to list alllll possibilities (ë, é, è, etc).

You could always simply use strtr() with an array consisting of characters you would like to replace:

$str = 'Here are some example characters: ë, é and è';
$trans = array('ë'=>'ë', 'é'=>'é', 'è'=>'è');
$str = strtr($str, $trans);

Sure, you create a list.. but once done.. it's done (then again, if utf8_encode() works for you, perfect). As for file_get_contents(), this just reads an entire file into a string (don't worry, newlines and return carriages are still there - If that's what you're getting at).

[Desdinova]

well yes I understand. but the chance you miss a char is pretty big, and that makes it nasty to manage properly.

utf8_encode works like a charm so far. 🙂 implemented that structurally, so I don't expect anymore of those problems. another victory in the charset saga!

I shall mark this resolved.