Thread overview
expat.h --> expat.d - help
Sep 01, 2003
Brad Anderson
Sep 01, 2003
Walter
Sep 03, 2003
Brad Anderson
Sep 03, 2003
Brad Anderson
Sep 04, 2003
Walter
Sep 04, 2003
Walter
Sep 19, 2003
Brad Anderson
Oct 25, 2003
Walter
September 01, 2003
I'm trying to wrap expat.h in D and I'm having an issue with the following snippet:

/+
    // original expat.h

    #ifndef XMLPARSEAPI
    #if defined(_MSC_EXTENSIONS) && !defined(__BEOS__) && !defined(__CYGWIN__)
    #define XMLPARSEAPI(type) __declspec(dllimport) type __cdecl
    #else
    #define XMLPARSEAPI(type) type
    #endif
    #endif  /* not defined XMLPARSEAPI */
+/


/+
    // my attempt: expat.d
+/

    version(XMLPARSEAPI) {} else { // version(!XMLPARSEAPI)
        version(_MSC_EXTENSIONS) {
            version(__BEOS__) {} else { // version(!__BEOS__)
                version(__CYGWIN__) {} else { // version(!__CYGWIN__)

                    // ERROR: semicolon expected following function declaration
                    void XMLPARSEAPI(type) __declspec(dllimport) type __cdecl; 

                }
            }
        } else {    // version(!_MSC_EXTENSIONS)

            // ERROR: semicolon expected following function declaration
            void XMLPARSEAPI(type) type;
        }
    }

It doesn't help that I'm a novice at C as well as D, but any assistance is appreciated.

Cheers,
Brad

P.S.  Benji, I'm still looking forward to your DOM parser.  This has just turned into an exercise of converting a header file.


September 01, 2003
Since those macro names have no meaning in D anyway, I think what would work
best is to replace the shebang
with simply:

export
{
 extern (C)
 {
    ... declarations ...
 }
}

The export gives the __declspec(dllimport) semantics, and the extern (C)
does the __cdecl.

"Brad Anderson" <brad@sankaty.com> wrote in message news:biu6rn$2l53$1@digitaldaemon.com...
> I'm trying to wrap expat.h in D and I'm having an issue with the following
snippet:
>
> /+
>      // original expat.h
>
>      #ifndef XMLPARSEAPI
>      #if defined(_MSC_EXTENSIONS) && !defined(__BEOS__) &&
!defined(__CYGWIN__)
>      #define XMLPARSEAPI(type) __declspec(dllimport) type __cdecl
>      #else
>      #define XMLPARSEAPI(type) type
>      #endif
>      #endif  /* not defined XMLPARSEAPI */
> +/
>
>
> /+
>      // my attempt: expat.d
> +/
>
>      version(XMLPARSEAPI) {} else { // version(!XMLPARSEAPI)
>          version(_MSC_EXTENSIONS) {
>              version(__BEOS__) {} else { // version(!__BEOS__)
>                  version(__CYGWIN__) {} else { // version(!__CYGWIN__)
>
>                      // ERROR: semicolon expected following function
declaration
>                      void XMLPARSEAPI(type) __declspec(dllimport) type
__cdecl;
>
>                  }
>              }
>          } else {    // version(!_MSC_EXTENSIONS)
>
>              // ERROR: semicolon expected following function declaration
>              void XMLPARSEAPI(type) type;
>          }
>      }
>
> It doesn't help that I'm a novice at C as well as D, but any assistance is appreciated.
>
> Cheers,
> Brad
>
> P.S.  Benji, I'm still looking forward to your DOM parser.  This has just
turned
> into an exercise of converting a header file.
>
>


September 03, 2003
Walter wrote:
> Since those macro names have no meaning in D anyway, I think what would work
> best is to replace the shebang
> with simply:
> 
> export
> {
>  extern (C)
>  {
>     ... declarations ...
>  }
> }
> 
> The export gives the __declspec(dllimport) semantics, and the extern (C)
> does the __cdecl.
> 

Okay, I tried this...

export
{
  extern (C)
  {
    #ifndef XMLPARSEAPI
    #if defined(_MSC_EXTENSIONS) && !defined(__BEOS__) && !defined(__CYGWIN__)
    #define XMLPARSEAPI(type) __declspec(dllimport) type __cdecl
    #else
    #define XMLPARSEAPI(type) type
    #endif
    #endif  /* not defined XMLPARSEAPI */
  }
}


results from compiler (DMD 0.69, WINXP):
d:\dmd\bin\..\src\phobos\c\expat.d(72): #line integer ["filespec"]\n expected
d:\dmd\bin\..\src\phobos\c\expat.d(73): #line integer ["filespec"]\n expected
d:\dmd\bin\..\src\phobos\c\expat.d(73): semicolon expected following function declaration
d:\dmd\bin\..\src\phobos\c\expat.d(73): Declaration expected, not '&&'

Line 72 is: #ifndef XMLPARSEAPI

Do I just need to get rid of all the macros?  If so, how do I handle the two options?

---

And throughout the rest of the header, I see declarations using the XMLPARSEAPI(type) from above.  Thoughts on those declarations?

ex:
/* This is called for an element declaration. See above for
   description of the model argument. It's the caller's responsibility
   to free model when finished with it.
*/
typedef void (*XML_ElementDeclHandler) (void *userData,
                                        const XML_Char *name,
                                        XML_Content *model);

XMLPARSEAPI(void)
XML_SetElementDeclHandler(XML_Parser parser,
			  XML_ElementDeclHandler eldecl);


Thanks,
Brad

September 03, 2003
one of Pavel's old posts (9/7/02):

Walter wrote:

> In general, it is not. However, most macros fall into predictable patterns
> like:
>
>     #define ABC    3
>
> and can be handled appropriately. Ones that don't would just be ignored.
> This means that the tool won't be perfect, but could handle 95% of the
> conversion chore.


Problems start when it comes to things like this:

    #define FOO 666
    #define BAR FOO                // macro? constant?
    #define BAZ MACRO(BAR/FOO)        // now what?

I'd say that, unfortunately, WinAPI headers have quite a lot of
such things.

And one still have to distinguish between conditional compilation
flags and macro...


--------------

So is this header just too involved to wrap in D, and a rewrite would be more appropriate?

BA

September 04, 2003
"Brad Anderson" <brad@sankaty.com> wrote in message news:bj3n1p$1a51$1@digitaldaemon.com...
> Walter wrote:
> > Since those macro names have no meaning in D anyway, I think what would
work
> > best is to replace the shebang
> > with simply:
> >
> > export
> > {
> >  extern (C)
> >  {
> >     ... declarations ...
> >  }
> > }
> >
> > The export gives the __declspec(dllimport) semantics, and the extern (C)
> > does the __cdecl.
> >
>
> Okay, I tried this...
>
> export
> {
>    extern (C)
>    {
>      #ifndef XMLPARSEAPI
>      #if defined(_MSC_EXTENSIONS) && !defined(__BEOS__) &&
!defined(__CYGWIN__)
>      #define XMLPARSEAPI(type) __declspec(dllimport) type __cdecl
>      #else
>      #define XMLPARSEAPI(type) type
>      #endif
>      #endif  /* not defined XMLPARSEAPI */
>    }
> }
>
>
> results from compiler (DMD 0.69, WINXP):
> d:\dmd\bin\..\src\phobos\c\expat.d(72): #line integer ["filespec"]\n
expected
> d:\dmd\bin\..\src\phobos\c\expat.d(73): #line integer ["filespec"]\n
expected
> d:\dmd\bin\..\src\phobos\c\expat.d(73): semicolon expected following
function
> declaration
> d:\dmd\bin\..\src\phobos\c\expat.d(73): Declaration expected, not '&&'
>
> Line 72 is: #ifndef XMLPARSEAPI
>
> Do I just need to get rid of all the macros?  If so, how do I handle the
two
> options?

D doesn't have a macro preprocessor, nor does it have __declspec(dllimport).

> And throughout the rest of the header, I see declarations using the XMLPARSEAPI(type) from above.  Thoughts on those declarations?
>
> ex:
> /* This is called for an element declaration. See above for
>     description of the model argument. It's the caller's responsibility
>     to free model when finished with it.
> */
> typedef void (*XML_ElementDeclHandler) (void *userData,
>                                          const XML_Char *name,
>                                          XML_Content *model);
>
> XMLPARSEAPI(void)
> XML_SetElementDeclHandler(XML_Parser parser,
>   XML_ElementDeclHandler eldecl);

alias void function(void *, XML_Char *, XML_Content *)
XML_ElementDeclHandler;

export extern (C) void XML_SetElementDeclHandler(XML_Parser parse,
XML_ElementDeclHandler eldecl);



September 04, 2003
"Brad Anderson" <brad@sankaty.com> wrote in message news:bj3no8$1b31$1@digitaldaemon.com...
> So is this header just too involved to wrap in D, and a rewrite would be
more
> appropriate?

Not at all. I'd start with compiling it with -e -l and expand all the macros.


September 19, 2003
Okay, does anyone have any thoughts about converting this mess:


typedef struct {
  void *(*malloc_fcn)(size_t size);
  void *(*realloc_fcn)(void *ptr, size_t size);
  void (*free_fcn)(void *ptr);
} XML_Memory_Handling_Suite;



Thanks,
Brad

P.S.  Walter, thanks for the page htomodule.html.  Along with Mike Wynn's Deimos page, I'm getting places without having much experience with header files in the past.  Now about all that free time I want to have...


October 25, 2003
"Brad Anderson" <brad@sankaty.com> wrote in message news:bkdlai$1k3g$1@digitaldaemon.com...
> Okay, does anyone have any thoughts about converting this mess:
>
>
> typedef struct {
>    void *(*malloc_fcn)(size_t size);
>    void *(*realloc_fcn)(void *ptr, size_t size);
>    void (*free_fcn)(void *ptr);
> } XML_Memory_Handling_Suite;

struct XML_Memory_Handling_Suite {
   void *(*malloc_fcn)(size_t size);
   void *(*realloc_fcn)(void *ptr, size_t size);
   void (*free_fcn)(void *ptr);
}