[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Still not able to get gettext working
From: |
Bruno Haible |
Subject: |
Re: Still not able to get gettext working |
Date: |
Sun, 11 Nov 2007 02:24:34 +0100 |
User-agent: |
KMail/1.5.4 |
David Bruce wrote:
> I've been trying very diligently over the last couple of days to get gettext
> working on my program (tuxmath) - again went through the FAQ and Chapter 13
> of the gettext manual, and AFAICT I've followed everything, but my program
> still doesn't call bindtextdomain, textdomain, etc. as if NLS is disabled.
>
> I am definitely calling AM_GNU_GETTEXT, and ENABLE_NLS does indeed get set to
> 1, but gettext is somehow still not activated. I'm hoping someone can
> suggest other pitfalls I can look into.
When you say, "ENABLE_NLS gets set to 1", you mean it is defined to 1
in config.h, I assume? Then, normally, this definition should cause
gettext.h to not define gettext and textdomain calls away, and these
warnings should not occur:
> ../../trunk/src/tuxmath.c: In function ‘main’:
> ../../trunk/src/tuxmath.c:35: warning: statement with no effect
> ../../trunk/src/tuxmath.c:36: warning: statement with no effect
> ../../trunk/src/tuxmath.c:37: warning: statement with no effect
So, the best tool to use at this point is "gcc -E" and "gcc -E -dM".
"gcc -E" shows the preprocessed output, and "gcc -E -dM" shows the state
of the macros at the end of the compilation unit. It will be interesting
to see whether ENABLE_NLS, gettext, and textdomain are defined as macros
when you look at "gcc -E -dM".
The two pitfalls that I can think of are:
- not including <config.h> as the first header file in every compilation
unit.
- a #undef ENABLE_NLS hidden somewhere.
> Here are what I think are the relevant snippets
> I apologize for the length of this post
On the contrary, the relevant snippets and attachments help to understand
your problem. People who ask questions but do *not* provide enough relevant
information should apologize!
Bruno