Re: Unicode Prerequeites for dia-0.90 ?
- From: Cyrille Chepelov <cyrille chepelov org>
- To: dia-list gnome org
- Subject: Re: Unicode Prerequeites for dia-0.90 ?
- Date: Mon, 3 Jun 2002 00:32:06 +0200
Le Sun, Jun 02, 2002, à 06:29:51PM -0400, Frederick C. Druseikis a écrit:
I've been compiling with /usr/include/unicode.h which comes from the GNOME
libunicode development package. Version is 0.4.0.
OK. I assume it's similar enough to the one on my system.
typedef unsigned int unicode_char_t;
I put this declaration at the beginning of lib/charconv.h There appear to be several places
where it is needed; and there appears to be an assumption that UNICODE_WORK_IN_PROGRESS is
defined because it is set twice in config.h
Yep.
Whether this succeeds or fails, I would love to know the outcome. Sounds
like one extra configure check may help here.
But I also needed to define unicode_iconv_t too (also by putting it in lib/charconv.h).
With these to changes (--with-iconv-prefix=) and defining both of these types, I can get it
all to compile, *but it can't link*.
The linker is looking for symbols such as unicode_strlen unicode_strchr, etc.
So I see these are defined kind-of in lib/charconv.h, but I don't get how they
are supposed to connect to libunicode.
These symbols are defined in -lunicode. For some reason, this should have
been put in your global LDFLAGS by configure.in, but has not.
Is UNICODE_WORK_IN_PROGRESS supposed to be GLOBALLY DEFINED?
Or is it supposed to be Globally UNDEFINED and locally DEFINED as needed?
Yes, UNICODE_WORK_IN_PROGRESS is supposed to be defined globally. Actually,
we're about to remove the need to define it and make it the only case.
-- Cyrille
--
Grumpf.
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]