Having trouble with non-latin characters

Kleidi

Hello there!
I'm having some trouble with my page. I have moved my site from a host provider to another one, and now i'm having some problems with non-latin characters, for text that comes from db query and for text that is in html/php file.
For text that comes from db someone suggested me to apply this after db connection:

mysql_query("SET CHARACTER SET utf8");
mysql_query("SET NAMES utf8");

and it did the trick, but, now i'm having the same problem for texts that comes from html/php files: instead of ë or ç appears �
I'm sure that should be e trick someone on the server or somewhere else on the configuration.
Hope that you can help me.
Thank you.

P.S.
I have used at header
<meta http-equiv="content-type" content="text/html; charset=UTF-8" />
and
<meta http-equiv="content-type" content="text/html; charset=iso-8859-1" />

NogDog

Use UTF-8 in your meta tag since that's what you're using for data, and also set it in a header() to ensure that's what the browser sees (in case your web server sends a different header by default):

header('Content-Type: text/html; charset="UTF-8"';);

Also, if using htmlentities(), add the optional character set parameter:

echo htmlentities($text, ENT_QUOTES, 'UTF-8');

Kleidi

I tried the method you suggest NogDog but i had no success. Seems that was a problem on downloading files from the old server to my pc since many of those files now are encoded in ANSI and not in UTF-8. When i try to upload files on server again, files with ANSI encode create the problem.
Can you help me on prevent this error to appear again and, to re-encode all the files in bulk on utf-8 again?
And, i have a question, should i encode to utf-8 or utf-8 with dom?

Thank you!

johanafm

It's not dom, it's BOM - acronym for byte order mark, which is usually not recommend to be used with UTF-8 since it has no use and isn't recommended, which leads to some problems. However, some programs will actually require that it is there, so it depends on where your files will be used. And take care with Microsoft's notepad (and perhaps others as well) since it saves utf-8 BOM by default.

However, if you encode your PHP files as UTF-8 with BOM, and your file starts with

<?php

Then the actual start contents of such a file will be

ï»¿<?php

which, apart from sending annoying characters as output, will also prevent you from sending custom php headers.

Another issue would be two identical utf-8 strings that are not binary-comparable due to one containing BOM while the other doesn't. String concatenation would be a third issue.