[gccsdk] hpp and ipp header files and suffix swapping

Alan Buckley alan_baa at hotmail.com
Mon Mar 7 06:12:07 PST 2011

John Tytgat wrote on Friday, March 04, 2011 8:45 PM 

> In message <SNT136-ds190397CA9CAFC06F123384F0C20 at phx.gbl>
          "Alan Buckley" <alan_baa at hotmail.com> wrote:

> > The boost libraries use suffixes hpp for header files and ipp 
> > which I assume are inline function include files.
> > 
> > Should these be added to the sfix list for gcc and all its
> > tools?
> > 
> > If so, do I just need to change the !GCC !run file or are
> > there other locations that will need changing?

> I don't think there is a real technical requirement to define suffix
> swapping for hpp/ipp files in order to be able to use them under RISC OS
> as an "#incude <boost/scoped_ptr.hpp>" should result in reading
> "boost.scoped_ptr/hpp" on RISC OS.

I can confirm the headers work fine on RISC OS in this format.

> However for consistency reasons with normal .h header files we might do
> this as well.

> I don't have a strong opinion about it

Unfortunately I don’t have strong opinions about it either. Consistency
with existing headers is what prompted me to raise the point in the
first place.

Is there anybody else who has an opinion?

I guess if nobody cares I may as well leave things as they stand, though
it will be a little odd with cpp files using suffix swapping and hpp not.

> , except that if we go for hpp/ipp
> suffix swapping we do the following:

> 1. !GCC's !Run file needs to be adapted (and yes, that should be enough)
> 2. We should do hpp/ipp suffix swapping for all libraries (Boost might the
>    only one, I don't know) packaged for RISC OS users from now on
> 3. We need to provide a ReadMe section in those libraries for !GCC 4.1.1
>    users detailing what needs to be changed in their !Run file
>    (otherwise they won't be able to use these libraries).

> As once we go for it, we'll stick to it (as it is rather disruptive for
> our users to revert this decision).

I completely agree with the steps you’ve outlined here if we did go
ahead with the change.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.riscos.info/pipermail/gcc/attachments/20110307/abf2bb46/attachment.html>

More information about the gcc mailing list