>I noticed in Y.T's lib everything is defaulted to Unicode, im going to wrap them in version statements and update soon.

Wait...make me do it !

There are some symbols not only "UNICODE".
I'm looking, they are entangled intricately in #if.
I selected to resolve #if on stage of Perl instead of version statement.
(I've tried using version statement for all symbols, but it's more complex than
I thought.)
If you wanted the ANSI-version, I'll re-run the script.

YT

>Are you familiar with Mike Wynn's Win32 work at converting headers (http://www.geocities.com/one_mad_alien/dcom_not_dcom.html)?

It's more beautiful than mine :-)

But I think covering all API is more important than beautiful.
I'm using Perl. Handmade will make troubles and have limit.
For example, Mike's doesn't have "AllocConsole" etc.
Of course, if Mike can cover them, I'll stop.

YT

Yes, I think the automation is most important.  On that fact I vote we keep win32.lib as the standard, it can easily be updated as new SDK's come along

I am currently trying to modify Mike's DFC work to use win32.lib headers, its nothing major really.

C

"Y.Tomino" <Y.Tomino_member@pathlink.com> wrote in message news:bnin5v$11j9$1@digitaldaemon.com...
> >Are you familiar with Mike Wynn's Win32 work at converting headers (http://www.geocities.com/one_mad_alien/dcom_not_dcom.html)?
>
> It's more beautiful than mine :-)
>
> But I think covering all API is more important than beautiful.
> I'm using Perl. Handmade will make troubles and have limit.
> For example, Mike's doesn't have "AllocConsole" etc.
> Of course, if Mike can cover them, I'll stop.
>
> YT
>
>

Could you re-run it ?

Is it possible to have the perl script , once it sees a unicode statement, wrap it in a version statement ?

C


<demoonlit@inter7.jp> wrote in message news:bnijpc$rjt$1@digitaldaemon.com...
> >I noticed in Y.T's lib everything is defaulted to Unicode, im going to
wrap
> >them in version statements and update soon.
>
> Wait...make me do it !
>
> There are some symbols not only "UNICODE".
> I'm looking, they are entangled intricately in #if.
> I selected to resolve #if on stage of Perl instead of version statement.
> (I've tried using version statement for all symbols, but it's more complex
than
> I thought.)
> If you wanted the ANSI-version, I'll re-run the script.
>
> YT
>
>

>Could you re-run it ?
Yes.
I tried to make some A-versions today. win32\ansi\winuser.d, etc.
(http://hp.vector.co.jp/authors/VA028375/contents/D_windows.h.html)

>Is it possible to have the perl script , once it sees a unicode statement,
wrap it in a version statement ?
No.

Version statement can't have complex expression.
For example, windef.h has this #if.

#if (!defined(_MAC)) && ((_MSC_VER >= 800) || defined(_STDCALL_SUPPORTED))

it shold be converted ...

version(_MAC)
else{
  version(_MSC_VER >= 800 /*invalid*/ ){
    version = MAC_and_MSC_VER_ge800_or_STDCALL_SUPPORTED;
  }else version(_STDCALL_SUPPORTED){
    version = MAC_and_MSC_VER_ge800_or_STDCALL_SUPPORTED;
  }
}
if(MAC_and_MSC_VER_ge800_or_STDCALL_SUPPORTED){
  ...

I gave up use of a version statement.

Please think this pattern.
It's impossible to give special treatment only to "UNICODE".

#if defined(UNICODE)
#define XXX
#endif

#if EXPRESSION-USING-XXX
...
#endif

YT

> #if defined(UNICODE)
> #define XXX
> #endif
>
> #if EXPRESSION-USING-XXX
> ...
> #endif

Ahh good point.  Hmm , i guess two seperate libs : win32u.lib, win32a.lib , will be ok ?

C


"Y.Tomino" <demoonlit@inter7.jp> wrote in message news:bnk066$ao2$1@digitaldaemon.com...
> >Could you re-run it ?
> Yes.
> I tried to make some A-versions today. win32\ansi\winuser.d, etc.
> (http://hp.vector.co.jp/authors/VA028375/contents/D_windows.h.html)
>
> >Is it possible to have the perl script , once it sees a unicode
statement,
> wrap it in a version statement ?
> No.
>
> Version statement can't have complex expression.
> For example, windef.h has this #if.
>
> #if (!defined(_MAC)) && ((_MSC_VER >= 800) || defined(_STDCALL_SUPPORTED))
>
> it shold be converted ...
>
> version(_MAC)
> else{
>   version(_MSC_VER >= 800 /*invalid*/ ){
>     version = MAC_and_MSC_VER_ge800_or_STDCALL_SUPPORTED;
>   }else version(_STDCALL_SUPPORTED){
>     version = MAC_and_MSC_VER_ge800_or_STDCALL_SUPPORTED;
>   }
> }
> if(MAC_and_MSC_VER_ge800_or_STDCALL_SUPPORTED){
>   ...
>
> I gave up use of a version statement.
>
> Please think this pattern.
> It's impossible to give special treatment only to "UNICODE".
>
> #if defined(UNICODE)
> #define XXX
> #endif
>
> #if EXPRESSION-USING-XXX
> ...
> #endif
>
> YT
>

Charles Sanders wrote:
> Wow, lots of good stuff!  I hadnt realized so much had been done, is Mike
> still actively working on these ?  This does awnser alot of my questions (
> you <i>can</i> have a static method used for WndProc :D ) Ill use this as a
> base.

I suspect Mike's busy with real life (his most recent post was apparently at the end of September).

> 
> So on the win32 stuff, whose are we going to use ?   How are we going to
> decide ?  I think we ( we as the community of developers! ) need to decide
> this ASAP.
> 
> C