Re: A simplification of _AC_CANONICAL_SPLIT

[Stepan Kasal]

Hello,

On Thu, Aug 18, 2005 at 01:22:10PM -0700, Paul Eggert wrote:
> > +[case $ac_cv_$1 in
> > +  [[!-]]*-*-*) ;;
> > +  *) AC_MSG_ERROR([non-canonical value of ac_cv_$1: $ac_cv_$1]);;
> > +esac
> 
> Sorry, I don't understand what that pattern is trying to do.  Does it
> really catch all non-canonical values?

The pattern catches values which start with "-" or which contain less
than two "-".
This ensures that the code below:
        IFS=-; set $ac_cv_$1; shift; shift
won't fail.

> Or even common problems?  If not, then I wouldn't bother.

The main goal is not to catch non-canonical values.  The goal is to
catch evil values which would break the code below.

> Also, the [!-] doesn't look portable to me.

I thought that it was a portable way to match any character other than "-".
If I were wrong, we can use:

ac_good=:
case $ac_cv_$1 in
  -*) ac_good=false;;
  *-?*-?*) ;;
  *) ac_good=false;;
esac
ac_good || AC_MSG_ERROR([non-canonical value of ac_cv_$1: $ac_cv_$1])

Have a nice day,
        Stepan Kasal