Re: No declaration for isblank

[Neil Jerram]

address@hidden (Ludovic Courtès) writes:

> Hi,
>
> I'm slowly getting back to Guile, and the first thing I came across was
> this old Gnulib-related bootstrap issue (HEAD).
>
> address@hidden (Ludovic Courtès) writes:
>
>> My `gnulib-cache.m4' is the same as the one in CVS HEAD:
>>
>>   
>> http://cvs.savannah.gnu.org/viewvc/guile/guile-core/m4/gnulib-cache.m4?revision=1.1&root=guile&view=markup
>>
>> It contains the line "gl_MODULES([alloca strcase])" so `gnulib-tool'
>> must at least pull the `alloca' and `strcase' modules.  It should also
>> pull their dependencies; `strcase' happens to depend on `string', which
>> in turn depends on `extensions', the module that ensures `_GNU_SOURCE'
>> is defined when available.
>>
>> (Actually, we should be importing `extensions' explicitly rather than
>> rely on this side effect.)
>
> I followed my own suggestion and added `extensions' to the list of
> required modules, since it turns out that it was no longer pulled by the
> other modules.  Hopefully, that fixes this very problem once for all.

Sounds good.  I think most of this problem was actually me not
understanding how gnulib works, but I agree that it makes sense to
import extensions explicitly.

> On a related note, when bootstrapping, `autoreconf' complains that we're
> using `AM_GNU_GETTEXT' without having `po' in `SUBDIRS'---which is
> normal since no string is internationalized.  Still, is there something
> we could do about it?

I guess we could start some token internationalization...  Perhaps we
could start with the "ERROR" in display_header(), and who knows where
that might lead?  Translation of Guile's messages can only be a good
thing, n'est-ce pas?

(On the other hand, I haven't noticed anyone clamouring for this.)

> Thanks,
> Ludovic.

Regards,
        Neil