Suppose we have the following:
$(TARGETS): %.a_pattern: $(subst something, something_else, %.b_pattern)
The problem is that first the subst function is applied to the .b_pattern
and only then the stem is substituted.
For example:
p/x.class p/y.class: %.class: $(subst p,q,%.java)
has the result
p/x.java p/y.java
although it would have been most useful if the stem were first substituted,
and
only then the function subst applied, therefore obtaining the result
q/x.java q/y.java.
My question is if it is possible to obtain such effect, as if the stem were
first subtituted into the dependency pattern
and after that an arbitrary string function to be applied on it.
Is the current behavior intentional? Why, since it could be simulated with
the more clear construct
%$(subst something, something_else, .b_pattern)
I use gnu make 3.79.1 on NT 4.0 Workstation.
Thank you for your time.
Sincerely yours,
Ovidiu Gheorghies
