[cffi-devel] Changes to the type system, among others.

[Rayiner Hashem]

What are the ramifications of the anonymous type stuff? In particular, does 
this mean that defctype behaves more like C's "typedef" (in that the source 
type doesn't yet need to be defined)? Could any of the new type system 
changes achieve that result? The dependence of defctype on having its source 
type already defined makes handling automatic declaration generation quite 
complicated in a lot of idiomatic C code. Since the dependence graph of C 
type definitions might have cycles, it becomes impossible to impose a total 
ordering on definitions. This isn't a problem when typedefs are fully 
resolved to their target types (as Verrazano does currently), but is a 
problem if you want the C-FFI binding to more closely resemble the C header 
file with regards to type naming. 

On 8/26/05, Luis Oliveira <luismbo at gmail.com> wrote:
> 
> Hello,
> 
> Here's the Changelog for the new changes to my branch since my last
> update:
> 
> Fri Aug 26 15:47:08 WEST 2005 Luis Oliveira <loliveira at common-lisp.net>
> * Type system improvements and more.
> 
> - New tests: tests/enums.lisp (4 tests) and tests/misc-types.lisp (3
> tests).
> - New file: objects.lisp, skeleton of the safe typed pointer
> interface. I
> also added a :boolean type here, it's looking for a better home
> still.
> - (early-types.lisp): type parsers, DEFINE-TYPE-SPEC-PARSER and
> PARSE-TYPE.
> Added abbility to have anonymous types (ie. a name is no longer
> required
> for FOREIGN-TYPE). Moved DEFCTYPE to types.lisp.
> - (enum.lisp): improved DEFCENUM, the values for each keyword are now
> optional, like in C. Added the anonymous type (:enum ...)
> - (types.lisp):
> - Renamed :TO-C-ARG to :TO-C-DYNAMIC.
> - Renamed FOREIGN-AREF to MEM-AREF and implemented compiler macros
> for it.
> - New exported macro: DEFINE-FOREIGN-TYPE for parameterized types.
> - Change DEFINE-TYPE-TRANSLATOR back to the previous interface,
> translators take aow and we have translators
> available at runtime now too.
> - FOREIGN-TYPEDEF's now inherit their parent type's translators.
> - New :wrapper type, for anonymous typedefs with translators.
> - (strings.lisp): Renamed string to :string. Added new type,
> :string+ptr.
> - (utils.lisp): renamed let-if to bif.
> - (tests/callbacks.lisp): two tests were not being run for SBCL
> because I thought it triggered a bug but it's actually only one that
> triggers the bug. Fixed that.
> - Updated examples and tests to use :string instead of string.
> 
> Sun Aug 21 14:35:13 WEST 2005 Luis Oliveira <loliveira at common-lisp.net>
> * A couple of minor changes.
> 
> - Move #-cffi/no-foreign-funcall from cffi-tests.asd to
> tests/funcall.lisp and have the lisps that don't support this
> push the feature in cffi-<lisp>.lisp instead of defpackage.
> - New file: utils.lisp
> - Make Allegro, CMUCL, Corman and OpenMCL use callback-symbol-name
> (from cffi-utils) to safely intern callback symbol.
> - Clisp: consider (pointerp NIL) => T
> 
> Thu Aug 11 05:43:53 WEST 2005 Luis Oliveira <loliveira at common-lisp.net>
> * Clisp: use FOREIGN-ADDRESS-NULL in NULL-PTR-P
> 
> Wed Aug 10 02:00:07 WEST 2005 Luis Oliveira <loliveira at common-lisp.net>
> * Lispworks: use :byte for cffi's :char types
> 
> --
> Luís Oliveira
> http://student.dei.uc.pt/~lmoliv/
> Equipa Portuguesa do Translation Project
> http://www2.iro.umontreal.ca/~pinard/po/registry.cgi?team=pt
> _______________________________________________
> cffi-devel mailing list
> cffi-devel at common-lisp.net
> http://common-lisp.net/cgi-bin/mailman/listinfo/cffi-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.common-lisp.net/pipermail/cffi-devel/attachments/20050826/512c8970/attachment.html>