Question

I am sorry for the question that seems to be pretty popular however, the sheer number of informations on this has left me confused. In the Twiki Release of Sept 1 2004, is there UTF8 support? For instance, can I save Japanese, Chinese, German, and English text on the same page, etc?

Thanks!

Environment

TWiki version:	TWikiRelease01Sep2004
TWiki plugins:	DefaultPlugin, EmptyPlugin, InterwikiPlugin
Server OS:	Windows 2000
Web server:	Apache 3.0.99-563
Perl version:	Perl 5.8.5
Client OS:	Windows 2000/XP
Web Browser:	Netscape 7.1

-- ChrisSeroty - 15 Oct 2004

Answer

Yes, I am using Twiki with hybrid English and Chinese in a same page. Where did you get the sheer information for this? I only followed the pre-installed Installation guide to make the current TWiki work for Chinese.

• This is probably using gb2312 or similar character set, which includes Roman letters along with Chinese characters. --RD

-- ChunhuaLiao - 16 Oct 2004

There's always Codev - try Googling this site...

-- MartinCleaver - 16 Oct 2004

I'm just starting to work on this myself so thanks for the question. I've got Russian, English and Japanese working on the same page (both in view and edit mode). Working slowly through the hints and instructions on the TWiki.cfg file re internationalisation was helpful. In the end though (and I probably did it the wrong way) I ended up entering codes in the cfg file such as euc-jp in the override line. I'm still attempting to find codes for Korean, Hebrew and Greek.

Just found these sites - may be of interest - http://www.opensource.apple.com/darwinsource/DevToolsDec2003/gnutar-421/tar/intl/locale.alias ; http://docsun.cites.uiuc.edu/sun_docs/C/solaris_9/SUNWdev/I18NDG/p22.html#BASESOLARIS-5

-- SueLocke - 18 Oct 2004

Thanks for the feedback! Interesting - so Chunhua, did you add to the Twiki.cfg file like Sue did? For instance, If I have a WikiWord like this ÉlectricalData, if I click on the link, it takes me to either lectricalData or AlectricalData - depending on my browser. I was wondering which setting is off, so any input is greatly appreciated.

-- ChrisSeroty - 19 Oct 2004

Your best starting point is probably ProposedUTF8SupportForI18N followed by international search of Codev web. RichardDonkin wrote >90% of the internationalisation code but he's not active (on twiki.org) all the time. If he doesn't chime in on this topic in a few days you might want to send him an email. I'm sure he'd love to hear from people who are actually using his code.

-- MattWilkie - 19 Oct 2004

ProposedUTF8SupportForI18N is the best view of what's planned and what's done - only Phase 1 (i.e. EncodeURLsWithUTF8) is in the TWikiRelease01Sep2004, and there are unfortunately some minor patches needed to that release for I18N to work (see the known issues page or InternationalisationIssues for details).

However... You may find you can use the Sept release OK with UTF-8, without support for international-character WikiWords - this may be OK for Japanese, Chinese, and so on, where pages with Japanese/Chinese names aren't WikiWords anyway. This would require lots of testing, and quite a few things would not work.

I did get quite a lot of Phase 2 working on my own private copy of TWiki, but it was quite hairy due to bugs in Perl 5.8 (mixing UTF-8 and locales caused big problems that were hard to track down). Since the current I18N is based on locales, there's a fair amount of work required to refactor the code to either UTF-8 mode or locale mode. If you are interested, have a look at my UTF-8 mode TWiki - partially working only.

There is also a fairly big (2-5 times) performance hit to using Perl's UTF-8 mode (see perldoc perluniintro), in my experience with TWiki, so that also needs to be addressed. Overall, the full UTF-8 Phase 2 is a lot of work, and I don't have much spare time at the moment, unfortunately.

-- RichardDonkin - 20 Oct 2004