Re: RFE: support per-program noinst_HEADERS

[Chris Pickett]

Ralf Wildenhues wrote:
> * Chris Pickett wrote on Wed, Mar 12, 2008 at 08:10:41PM CET:
> > Ralf Wildenhues wrote:
> > > * Chris Pickett wrote on Wed, Mar 12, 2008 at 07:50:46PM CET:
> > > > I would like to be able to use noinst_prog_HEADERS, but with 1.10.1 I 
> > > >  get a complaint about noinst_progdir not being defined:
> > > You can just put all your not-to-be installed headers into the *_SOURCES
> > > alongside the .c/.cc/... files.  *_HEADERS really is necessary only for
> > > headers that are to be installed.
> > That doesn't work for me.  I am #include'ing .c files, and I do not want  
> > objects built for them.  This works fine with noinst_HEADERS for a  
> > single library or program:
>
> So why not this?
>
> noinst_HEADERS =
> if PROG1
> noinst_HEADERS += ...
> else
> noinst_HEADERS += ...
> endif
>
> if PROG2
> noinst_HEADERS += ...
> endif

I'm sorry, I don't really understand... how do I set PROG1 and PROG2? 
What I want is this: per-test program sets of dependencies so that if a 
file in the library is touched, only those test programs that are 
possibly dependent upon it need to be recompiled.  I don't want every 
test program dependent upon the same noinst_HEADERS.  Can you elaborate?

My main test Makefile.am is here:

http://svn.sable.mcgill.ca/sable/spmt/libspmt/tests/Makefile.am

and you can see my (awful) generated code here:

http://svn.sable.mcgill.ca/sable/spmt/libspmt/tests/Makefile.sources.am

> Really adding files to noinst_HEADERS isn't much different from just
> adding them to EXTRA_DIST.
>
> I agree that the warning 
>
> > > > `noinst_prog_HEADERS' is used but `noinst_progdir' is undefined
>
> is weird.  I'm mostly arguing to avoid having to do the work of fixing
> the warning.  ;-)

Yeah, if it's a lot of work, forget it.  I do have a workaround...

Chris