Re: How to create Unicode input methods for MacOS? (long)

From: Sebastian Hagedorn (
Date: Mon Apr 17 2000 - 11:01:35 EDT

--On Montag, 17. April 2000 5:16 Uhr -0800 Marco Piovanelli
<> wrote:

> On Mon, 17 Apr 2000 03:48:37 -0800, Sebastian Hagedorn
> ( wrote:
> I don't think you need to create a full-fledged input method for Greek.
> A Unicode keyboard layout should suffice. A Unicode keyboard layout is
> basically a 'uchr' resource that tells the system how to map a sequence
> of keystrokes to one or more Unicode characters. 'uchr' is a pretty
> complicated format explained in the PDF document "Supporting Unicode
> Input", available from Apple's site:
> <
> pdf>

I've just taken a look at it and I have remembered that I've seen it
before. Unfortunately a lot is missing:

"Creating a 'uchr' Resource: This section is forthcoming." (page 32)

But this quote

"The 'uchr' resource contains offsets to tables that may be in
any order. Because of the complexity of this format, Rez
may not readily be used to create 'uchr' resources. A
'uchr' resource may be created in any data-description
language that allows the specification of arbitrary binary

doesn't sound very promising... ;-)

The document is from October 1998. Surely there is a more recent draft
somewhere? If it's not been published, when might it be?

Has anybody on this list ever tried to create a 'uchr' resource? Are there
any tools available that might facilitate the process? Is the 'uchr' format
the same for Mac OS X?

>> It would be perfect if the old, GreekKeys-coded fonts could be used. I
>> guess that would mean letting the Text Encoding Converter know that they
>> are used for ancient Greek by adding some resources to the font.
> If you add a Unicode 'cmap' to your fonts, ATSUI should be able
> to use them.
> <>

Thanks. I'll check that out.

>> However, that would require a MacOS script for ancient Greek, right?
> I don't think that's necessary, if you limit yourself to ATSUI-savvy
> applications. Scripts are only needed for traditional, QuickDraw Text-
> based applications.

Right. That seems to be the difference between partial and full Unicode
characters. My custom 'uchr' should have a negative ID if I don't want to
associate it with a script. Do I have to reserve the ID I want to use like
I have to do with creator codes?

Marco, thanks once again for your kind help. It is very much appreciated.

Cheers, Sebastian

Sebastian Hagedorn

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:02 EDT