Question

I have a problem with polish national characters. All that characters are converted into html entities that look good on the screen but are very difficult to read in edits.

Witam w systemie informacyjnym sieci komputerowej w Laskach. Jego celem jest informowanie użytkowników o zmianach i nowościach oraz przedstawienie możliwości efektywnego wykorzystania Internetu w edukacji, życiu codziennym i zawodowym.
=Administrator:<br>Andrzej Cuber=
Sieciowa *Biblioteka* - to system multimedialnych publikacji elektronicznych gromadz&#261;cych teksty, d&#378;wi&#281;ki, filmy ró&#380;nego rodzaju publikacji.

The problem is with screen readers and braille dispalys used by the blind. They don't get clear polish when they try to edit something, the words are being mixed with some kind of numbers etc.

I have done some sets:

Set HTTP_EQUIV_ON_VIEW = <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-2">
<meta http-equiv="Content-Language" content="pl">
Set HTTP_EQUIV_ON_EDIT = <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-2"> <meta http-equiv="Content-Language" content="pl">

Is there any way to leave the characters on the page an they were typped? Do they need to be converted? Another problem which I encounted is that no matter what page code is set it is always iso-8859-1. It sometimes turns my screen readers into something strange.

• TWiki version: dec 2001
• Web server: Apache 1.3.26
• Server OS: RedHat Linux 7.2
• Web browser: IE 5.5/6.0, Lynx, Links
• Client OS: Win 9x/Me

-- AndrzejCuber - 30 Jun 2002

Answer

Not sure why this is happening - I have been able to use ISO-8859/1 characters in a TWiki page, and they just appear as normal characters when editing. This is using the TWikiAlphaRelease.

You probably need to edit the templates to read 8859-2 near the top, instead of 8859-1 - I don't think the embedded HTML text has much effect since this is already set in the templates. See NoWarningWhenEnteringNonStdChars.

By the way, it's best to use a WikiWord for new pages, and to create them through the AskedQuestions page so that the SupportStatus is set correctly - ensures more people see your question.

-- RichardDonkin - 01 Jul 2002

There are quite a few InternationalisationEnhancements in the TWikiAlphaRelease of 30 Nov 2002, and this is tested with Lynx as well as graphical browsers, in IS0-8859-1 at least. I'm also going to set the template charset based on the site locale setting - not done yet but is on the way.

-- RichardDonkin - 30 Nov 2002

This should now be fixed in the TWikiAlphaRelease of 8 Dec - by setting the $siteLocale in TWiki.cfg you can select a suitable charset used by all templates. If you can test the latest alpha (now available via HTTP, see TWikiAlphaRelease) that would help clarify whether this fix actually solves your problem.

UPDATE: CyrillicSupport is working, using a non-ISO-8859-1 character set, so your character set should be no problem.

-- RichardDonkin - 08 Dec 2002