Is message forwarding good enough ?

[Ralph Berger]

Semen;

i think you're right. Looking a the c code preparsed from C++ is
like a view under the hood. Ensures me that procedural compilers
like PB will be alive for a long time.

Ralph

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

[Semen Matusovski]

From one side, C-code is more clever - one API call less.
But when I see that typically my CPU is busy 2-3%, I don't think that it's very actual.
Anyway, typically, processing each message includes some API calls -
GetMessage - Is DialogMessage - TranslateMessage - DispatchMessage - DefWindowProc and so on.



------------------
E-MAIL: [email protected]

[Ralph Berger]

Hi there,

just had a look on some C code generated by winCFront from CPP
sources. Most interesting is the implementation of message mapping.

I guess with a procedural lang. like PB most users will reflect msgs
by forwarding them to the windows it belongs.

iE. %WM_CTLCOLOREDIT handeld by MainWndProc like this :

CALLBACK FUNCTION MainWndProc
      SELECT CASE CBMSG
         CASE %WM_COMMAND
         CASE %WM_CTLCOLOREDIT  : FUNCTION = sendmessage ( CBLPARAM, %WM_CTLCOLOR_REFLECT, CBWPARAM, 0 )
      END SELECT
END FUNCTION

refelected by control's wndProc

FUNCTION SuperEditProc( BYVAL hWnd AS LONG, BYVAL wMsg AS LONG, BYVAL wParam AS LONG, BYVAL lParam AS LONG ) AS LONG
      SELECT CASE wMsg
         CASE %WM_CREATE  : edit_create  hWnd, OffsetWndExtra
         CASE %WM_DESTROY : edit_destroy hWnd, OffsetWndExtra
         CASE %WM_CTLCOLOR_REFLECT
              SetBkColor   WPARAM, GetSysColor( %COLOR_HIGHLIGHT )
              SetTextColor WPARAM, %CYAN
              FUNCTION = gETsTOCKoBJECT(%GRAY_BRUSH)
              EXIT FUNCTION
      END SELECT
      FUNCTION = CallWindowProc(BaseWndProc, hWnd, wMsg, WPARAM, LPARAM)
END FUNCTION

The c code doesn't use sendmessage. It call's SuperEditProc direct,
after retrieiving the wndProc CodePtr via GetWindowLong.

Do you see any benefits calling the child wndProc directly ?

Ralph


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