[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    KOffice and Unicode
From:       "Jacek Konieczny" <jajcus () zeus ! polsl ! gliwice ! pl>
Date:       1998-07-11 15:48:22
[Download RAW message or body]

Hi,

KOffice is now in early developement stage and it is the best time to think
about it internals and future use.
Now it doesn't care about character set used. In fact only ISO-8859-1 is
usable now. (some time ago I could also use ISO-8859-2 in kpresenter).

If nothing is done to make koffice more universal now, probably it will
bring problems later. Because it is to be used worldwide it will
have handle any characters.

There are several ways to achieve such universality.
One is TeX solution - to have some own format for representation of
accented characters and special symbols. But it is not portable.
Another is specyfying charset for document. But this is not good for
multilangual documents and does not help much in case of mathematical
symbols.

So the Unicode seems to be the best solution. In can be used to represend
practically any character - all alphabets and a lot of mathematical and
'dings' symbols. It is very portable, and many packages uses it now.
It seems it will be a standard in a few yers time. 

And I thing unicode should be used in koffice from the very beginning,
because later it will be VERY difficult (most problems with Ms  Office 97
were caused by moving it to Unicode).

KDE even now have some Unicode support, but it was developed for HTML and
e-mail use. Qt doesn't help much in this matter. But KOffice uses STL, so
it is no problem to use basic_string<wchar> or something to represent
strings. And use KDE KCharsets classes to convert it to displayable format.
Documents would be saved using UTF-8 for example so they will be readable
in any ASCII viewer, and they could be exported to any charset plain-text
file. 

Maybe KCharsets* classes would need some polishing, but even now they are
usable. I couldn't improve them much, because nothing, except khtmlw,kmail
and krn uses them now.

I don't know how koffice represents text documents now, but it shouldn't be
difficult to convert it to unicode now. It will be much more difficult
later.


Comments are welcome

BTW. Does XDND provide any character-set information for draged data?

Greets,
    Jacek

--
+---------+--------------------------------------------------------+
!      ,  !            Jacek Konieczny, Gliwice, Poland            !      
! Jajcus  ! email: jajcus@zeus.polsl.gliwice.pl, jajcus@polbox.com !
!         ! ICQ# 7149127                           WWW: none (yet) !
+---------+--------------------------------------powered-by-Linux--+

[prev in list] [next in list] [prev in thread] [next in thread]