[Kea-users] Selection of awk binary in configure script

[Patrik Lundin]

On Tue, Jun 14, 2016 at 08:35:38AM -0500, Jeremy C. Reed wrote:
> > While looking at this I also noticed something I did not understand. The
> > configure script in the kea-1.0.0 tar-ball does this:
> > ===
> > for ac_prog in gawk mawk nawk awk
> > ===
> >
> > However, the configure.ac file in that same tar-ball only contains these
> > awk-related settings:
> > ===
> > AC_PATH_PROGS(AWK, gawk awk)
> 
> Looks like that is redundant.
> 

It turns out the configure script actually checks for an
awk tool twice:
===
$ grep gawk configure
for ac_prog in gawk mawk nawk awk
for ac_prog in gawk awk
===

The first loop works with ac_cv_prog_AWK while the later one works with
ac_cv_path_AWK. It appears the first variable is already handled on
OpenBSD, it is only the ac_cv_path_AWK variable that is not handled.

Is there a reason for having both? Should the later one be removed? I looked
for clues in the git log, but the commit that introduced those lines in
configure.ac did not mention that specifically (f1a85d0c).

> > I find no trace of mawk or nawk in any files other than the supposedely
> > generated configure script. Where does these come from?
> 
> It comes from autoconf's programs.m4.
> 

Ah, right, thanks for the hint :).

>
> Some (automated)
> build systems create wrappers in the execute path for gawk to awk and
> maybe you can consider that.
>

I can override the variable via the port Makefile. At this point I am trying to
figure out if the proper change is instead to remove that extra awk check from
configure.ac.

-- 
Patrik Lundin