I install a Zenphoto on my server. And I saw that in themes don't have a encoding declaration:
` (my server send all pages inISO-8859-2`). Polish chars are broken.

And I saw that Ajax script to add a titles and description save a polish chars (?, ?, ?, ?, ?, ó, ?, ?...) bad.
Example:

Quote:?a?ó?? ?ó?t? ja??
changing to:

Quote:Za%u017Có%u0142%u0107 %u017Có%u0142t%u0105 ja%u017A%u0144

i've sent patches to fix this to tristan some time ago but still no reply..

pronvit: write it in comment.

add to beginning of template-functions.php
`

/**

• Function converts an Javascript escaped string back into a string with specified charset (default is UTF-8).

• Modified function from http://pure-essence.net/stuff/code/utf8RawUrlDecode.phps

• @param string $source escaped with Javascript's escape() function

• @param string $iconv_to destination character set will be used as second paramether in the iconv function. Default is UTF-8.

• @return string

*/

function unescape($source, $iconv_to = 'UTF-8') {

$decodedStr = '';

$pos = 0;

$len = strlen ($source);

while (&#36;pos < &#36;len) {

   $charAt = substr ($source, $pos, 1);

   if ($charAt == '%') {

       $pos++;

       $charAt = substr ($source, $pos, 1);

       if ($charAt == 'u') {

           // we got a unicode character

           $pos++;

           $unicodeHexVal = substr ($source, $pos, 4);

           $unicode = hexdec ($unicodeHexVal);

           $decodedStr .= code2utf($unicode);

           $pos += 4;

       }

       else {

           // we have an escaped ascii character

           $hexVal = substr ($source, $pos, 2);

           $decodedStr .= chr (hexdec ($hexVal));

           $pos += 2;

       }

   }

   else {

       $decodedStr .= $charAt;

       $pos++;

   }

}

if (&#36;iconv_to != "UTF-8") {

   $decodedStr = iconv("UTF-8", $iconv_to, $decodedStr);

}

return &#36;decodedStr;

}

/**

• Function coverts number of utf char into that character.

• Function taken from: http://sk2.php.net/manual/en/function.utf8-encode.php#49336

• @param int &#36;num

• @return utf8char

*/

function code2utf(&#36;num){

if(&#36;num6)+192).chr((&#36;num&63)+128);

if(&#36;num>12)+224).chr(((&#36;num>>6)&63)+128).chr((&#36;num&63)+128);

if(&#36;num>18)+240).chr(((&#36;num>>12)&63)+128).chr(((&#36;num>>6)&63)+128) .chr((&#36;num&63)+128);

return '';

}

`

then change stripslashes(&#36;newtitle) to stripslashes(unescape(&#36;newtitle))
and stripslashes(&#36;newdesc) to stripslashes(unescape(&#36;newdesc)).

and don't forget to add meta content-type = utf-8

Thanks. But one polish letter (ó) doesn't work now fine.

I'm sure there are better ways to do this. I have pronvit's mods, but I'll look into better encoding options.

Thanks for the report.

why this way isn't good enough?

It might be, I just always look for the best way... is there something wrong with that? ;-)

It probably is fine. You're welcome to hack ZP to use it if need be.

nothing wrong.
but I think that these encoding problems are quite critical for non-english users. and why to search for something more if solution is available

Quote:encoding problems are quite critical for non-english users
It's a very big problem.

Easy solution for non-english users.

add
header("Content-type: text/html; charset=utf-8");
into zp/index.php in the first line to override the default charset.
in Sajax.php, replace "escape" with "encodeURIComponent".
It works very well for Chinese characters.
http://blog.dukechina.org/gallery/

WeiChen: thanks

I think, that this line:
header("Content-type: text/html; charset=utf-8");
is good thing to adding into admin.php

Shouldn't a simple htmlentities() PHP function change any special character to a valid html entity? I mean it converts all applicable characters to HTML entities. http://fi.php.net/htmlentities . I just don't know where to put this so that it would convert the text to html entities BEFORE it's saved to the database. I need Finnish/Scandinavian characters and they're coming out as just blocks or question marks. Help, please. Where could I put this?

WeiChen: seems to work ok for German, too! Many thanks!

Perhaps this change should be added in the next version?

Oh, but this doesn't work for Russian language... shit =)

Spacemonkey: I used WeiChen method but with a charset ISO-8859-15 to make Finnish letters work.

However, if the image has Finnish letters, it won't display properly.

In latest version this bug is not resolved...

It is now! Version to be released today with support for any desired character-set, including UTF-8 by default.

It still shows all scandinavian letters (ä,å,ö) as question marks in File Info and Tags. Does anyone know what to do for this?