depends_* with variants in Portfiles?

[Ryan Schmidt]

On May 25, 2009, at 14:37, PGNet Dev wrote:

>   "dependencies need to allow inclusion of variant"
>    http://trac.macports.org/ticket/126  ( Opened 7 years ago !? )
>
>          " ... but if you would like to discuss why this should be
> implemented and how it could be done, by all means bring it up on
> macports-dev. When a consensus is reached, more code can be written.
> ..."
>
> it's not yet doable :-/
>
> so, 1st, the "why".
>
> if not 'with variants' for depends_*, how -- e.g., in my particular
> case of "ncurses(w) +abi6" -- else should this be done? not an
> uncommon occurrence, I'd suggest.   there is, of course, the path of
> completely separate ports/Portfiles, but that seems silly ...

In ncurses +abi6 you can check whether a particular file installed by  
ncursesw +abi6 is present. If not, bail and tell the user to first  
install ncursesw +abi6.

You can see an example of this strategy in the pango port, whose  
quartz and no_x11 variants check that cairo has been installed with  
the same variants before proceeding.

If ncursesw +abi6 does not cause any different set of files to be  
installed, you could install a dummy file in the variant, which you  
could then check for.