Anyone know of an easy way to encode all files in an directory to UTF-8? Or can someone provide me with a UTF-8 encoded CI 1.5.2?

Could CI be released in UTF-8 (too(?)) in the future?

Zawk - 22 February 2007 09:51 AM

Could CI be released in UTF-8 (too(?)) in the future?

As far as CI is coded with english characters only, its ASCII encoded, and thus UTF-8 too

Zawk - 22 February 2007 09:51 AM

Anyone know of an easy way to encode all files in an directory to UTF-8?

The problem is more yur editor capability to store a file in UTF-8 I guess. But to answer your question, I’d say : it depends on your OS and your tools

If you’re on GNU/Linux, just a

whould suffice. This will convert all files recursively from the current directory to UTF-8 (in files are already in UTF-8, I don’t know the result ).

On Windows, I guess you should rely on an external tool.

I’m not on GNU/Linux But what do you mean about it being UTF-8 too? I put in a special sign in one of the ASCII files and set the character encoding to UTF-8 in a meta tag, and it showed the question mark. I can though change the files encoding in my editor .. one file at a time .. and it works perfectly, I’m just not in the mood to go all the files through

If you do a search on the forums for UTF-8, you should get multiple threads about CI’s existing support for UTF-8. Jozef wrote up a cheat sheet at one time with guidelines for using CI with UTF-8. It may be posted on the Wiki. As I recall, CI can support UTF-8 with some limitations (i.e., one or more libraries did not provide adequate support (validation comes to mind)). Some CI developers are appear to be using one of two methods for handling multilanguage sites (based on the wiki) with UTF-8 and other encodings.

Riri misread your post. He was stating that CI was written using ASCII Decimal which is the base character set for UTF-8.

There are some tools on SourceForge that language translators use to convert various documents with different encodings to other encoding (maybe the babel toolkit or something like that). You might try doing a separate search there. If you can’t find them, leave me a PM and I’ll try to find them here.

The only framework that I know of that claims to support UTF-8 using PHP native libraries is Akelos (a PHP implementation of Ruby on Rails). I could be wrong but believe that Bernie (the Akelos author) checks phpinfo to determine if mbstring is installed and uses Unicode encodings if true for the current language, but falls back on UTF-8 in conjunction with Harry Fuecks UTF-8 string handling functions for PHP (designed originally for the WACT framework).

Zawk - 22 February 2007 03:14 PM

I’m not on GNU/Linux But what do you mean about it being UTF-8 too? I put in a special sign in one of the ASCII files and set the character encoding to UTF-8 in a meta tag, and it showed the question mark. I can though change the files encoding in my editor .. one file at a time .. and it works perfectly, I’m just not in the mood to go all the files through

That’s the default ‘save’ format of your editor I guess, not a CI problem - but look at the other answer below

esra - 22 February 2007 04:35 PM

If you do a search on the forums for UTF-8, you should get multiple threads about CI’s existing support for UTF-8. Jozef wrote up a cheat sheet at one time with guidelines for using CI with UTF-8. It may be posted on the Wiki. As I recall, CI can support UTF-8 with some limitations (i.e., one or more libraries did not provide adequate support (validation comes to mind)). Some CI developers are appear to be using one of two methods for handling multilanguage sites (based on the wiki) with UTF-8 and other encodings.

Riri misread your post. He was stating that CI was written using ASCII Decimal which is the base character set for UTF-8.

That’s what I stated

I didn’t have in mind some problems with some libraries, but it’s only on character recognition I’m right ? For exemple, the validation could not validate the form because some UTF-8 specific characters won’t be treated as classic ones, even if they’re valid.

I’m sure that’s not what he had in mind (more a file save format one), but sure this’s a real problem I’ll check on the forums to see UTF-8 issues too.

Correct. I believe that it is a string handling problem with the Validator and possibly some helpers not being fully UTF-8 aware. Jozef or someone else identified some of the framework problems with UTF-8 string handling in a post back around August or September of last year. I printed the thread for future reference and will try to get back with a url. If you do searches for internationalization and localization or i18n or l10n, you might find the thread yourself.

A STRICT HTML 4.01 page, which the latest DreamWeaver (CS) validated i.e. with no objections, rendered, perfectly, in both DreamWeaver’s internal rendering engine, as well as in the latest Internet Explorer (7).

When run through CI MVC, the page showed occasional AE (Euro signs).  At first, I thought it might be an escaping issue; but, after a day of reading forum postings, I tried changing the page DocType from UTF-8 to Western European, as follows:

Bingo ... no more weird AE signs !

I don’t know if this constitutes a UTF-8 support bug ... or not?

I’m just posting fyi ...

I’ve read that PHP6 will support UTF-8; maybe the issue will disappear, automagically, with PHP6?