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

List:       xfree-i18n
Subject:    Re: [I18n]Patch to improve Big5-HKSCS support in XFree86
From:       Anthony Fok <anthony () thizlinux ! com>
Date:       2002-03-26 10:14:04
[Download RAW message or body]

On Mon, Mar 25, 2002 at 07:21:33PM +0800, ha shao wrote:
> On Sat, Mar 23, 2002 at 04:33:06PM +0200, Vasilis Vasaitis wrote:
> > > And yeah, if Xlib already has this function, I don't have to go write one
> > > from scratch blindly (because I am new to hashing... ;-)  Thanks for the
> > > tip!  :-)
> >   You might also want to try GNU gperf, a perfect hash function generator. I
> > haven't used it personally, but AFAIK it's a good tool for this kind of
> > task.
> gperf will generate a hash function for a known set of data.
> It will generate an almost perfect hash with a few collisions. Actually,
> the size of the hash table can be control at command line.
> Collisions are handled by the generated hash functions for you.

Thank you both for your tips on gperf, Vasilis and ha shao.
I tried it a few days ago on Debian (apt-get install gperf does it)
and I was really amazed by it.  :-)

As for BIG5-HKSCS table for glibc, James has come up with an excellent
solution in keeping the size down.  A very neat summary index table. 
:-)  More work (esp. the charmap file) is needed before submitting to
glibc, but yes, I am really impressed by his work.  :-)

I'll play with gperf and hashing after some deadlines at work.  Very
helpful tools indeed.  :-)

Cheers,

Anthony

-- 
Anthony Fok Tung-Ling
ThizLinux Laboratory   <anthony@thizlinux.com> http://www.thizlinux.com/
Debian Chinese Project <foka@debian.org>       http://www.debian.org/intl/zh/
Come visit Our Lady of Victory Camp!           http://www.olvc.ab.ca/
_______________________________________________
I18n mailing list
I18n@XFree86.Org
http://XFree86.Org/mailman/listinfo/i18n
[prev in list] [next in list] [prev in thread] [next in thread] 

Configure | About | News | Add a list | Sponsored by KoreLogic