Re: Is there any value in a lisp-like $(cond ...) function?

[Eric Melski]

J.T. Conklin wrote:
> While testing my implementation, I realized that using $(or ...) can
> mostly approximate this:
>
>         t:
>                 $(or $(if ${COND1},${RESULTS1}), \
>                      $(if ${COND2},${RESULTS2}), \
>                      $(if ${COND3},${RESULTS3}), \
>                      ${RESULTS4})
>
>
> The significant difference is that if ${CONDN} is true but ${RESULTN}
> expands to false, $(or ...) will continue on where $(cond ...) would 
> have stopped.  Accidently adding an else-part to an $(if ...) would
> cause problems too.
>   

You don't show what CONDN or RESULTSN are in your case, but generally 
speaking at least, you can address your first concern above by appending 
a space character _after_ the reference to each RESULTSN variable.  For 
example:

        t:
                $(or $(if ${COND1},${RESULTS1} ), \
                     $(if ${COND2},${RESULTS2} ), \
                    $(if ${COND3},${RESULTS3} ), \
                    ${RESULTS4})


That's sufficient to make the expression evaluate to "true" for purposes 
of gmake and short-circuit out of the evaluation of the other 
expressions.  You can trace which expressions are evaluated by inserting 
a $(info) call into them, like this:

        COND1=$(info COND1)
        COND2=$(info COND2)
        COND3=$(info COND3)
        RESULTS1=$(info RESULTS1)
        RESULTS2=$(info RESULTS2)
        RESULTS3=$(info RESULTS3)
        RESULTS4=$(info RESULTS4)


With this set of values, you'll see that COND1, COND2, COND3 and 
RESULTS4 are evaluated; change COND2 to "$(info COND2)x" (again, the x 
is enough to make it "true") and instead you get COND1, COND2 and 
RESULTS2 evaluated.

Hope that helps,

Eric Melski
Electric Cloud, Inc.