[Prev][Next][Index][Thread]
Casual pidgin hacking
-
To: info-dylan@ai.mit.edu
-
Subject: Casual pidgin hacking
-
From: Nolan Darilek <nolan@ethereal.dhis.org>
-
Date: Tue, 2 May 2000 16:30:05 -0400 (EDT)
-
Organization: The University of Texas at Austin; Austin, Texas
-
Xref: traf.lcs.mit.edu comp.lang.dylan:11950
A few weeks ago I was working on a GD port of the C-FFI libraries, but
gave that up temporarily. I then realized that, since Pidgin is
basically a translator, it doesn't need the FFI libraries to be
finished; of course, that means that I can't test it, but I *might* be
able to do some of the grunt work.
I've started implementing a c-ffi-output module which can translate
the structures created by the Pidgin parser into FFI declarations. I'm
designing it to be a recursive function; calling translate() on a
declaration spits out a C-FFI description, while calling translate()
on a <C-int> spits out 'C-int' for the purposes of the macros.
I'm a bit confused about how this works with pointers, though. I
understand that int * is <C-int*>; does that also hold true in the FFI
description, allowing you to use C-int*? Or is there more to it than
that?
I also have one final question about FFI. Was there an implementation
of the C-FFI library in the old 1.2 Harlequi distribution? If so, is
it available under an open license? The documentation which I have is
from HD 1.2, yet I cannot seem to find it in the tarball of sources
available on ftp.gwydiondylan.org. Was the FFI source not released
under the open source license like the rest of the libraries have
been?
Follow-Ups: