ScrollWindow... Is a control visible?

[John Thompson]

I've got a window where I use ScrollWindow and I'm curious if there's an easy way to determine if various controls on the window are viewable from the ScrollWindow's current position. My window only scrolls vertically, so I'm guessing I could check the location of the control and see if its currently in the range being displayed, but I thought IsWindowVisible (tried this, didn't work) or something similar existed for this situation (but couldn't find it).

[Michael Mattias]

Try the IntersectRect WinApi function using the client area of the screen and window rect of the control. If any areas are in common (a rect is returned) the control is at least partially visible.

MCM

ADDED:
Not tested but I think it will work.

FUNCTION IsControlInVisibleArea (hWndparent, %ID_CONTROL) AS LONG
 LOCAL rParent AS RECT, rCtrl AS RECT, rDest AS RECT, iRet AS LONG

 ' Get RECT of parent Window
  GetClientRect hwndParent, rHwnd  ' client coodinates
' --------------------------------------------------
' convert client coordinates to screen coordinates
' --------------------------------------------------
  MapWindowPoints hWndParent, %HWND_DESKTOP, BYVAL VARPTR(rParent), BYVAL 2&  'when using RECT cPoints=2
' ------------------------------------------------
' Get screen coordinates of the Control's RECT 
' ------------------------------------------------
 GetWindowRect GetDlgItem (hWndParent, %ID_CONTROL), rCtrl  ' screen coordinates
 ' -------------------------------- 
 ' See if the two RECTs intersect. 
 ' -------------------------------- 
 iRet = IntersectRect (rDest, rParent, rCtrl) 

 FUNCTION = ISTRUE (iret)  ' if true is returned the RECTs intersect

END FUNCTION

[Pierre Bellisle]

Hey John,

for the fun of it, here is a little one using
GetClientRect, GetWindowRect, MapWindowPoints and PtInRect.

#COMPILE EXE '#Win 8.04#
#DIM ALL
#INCLUDE "WIN32API.INC" '#2005-01-27#
 
%LABEL = 200
'______________________________________________________________________________
 
CALLBACK FUNCTION DlgProc() AS LONG
 LOCAL si       AS SCROLLINFO
 LOCAL DlgRect  AS RECT
 LOCAL CtrlRect AS RECT
 LOCAL hControl AS DWORD
 LOCAL oldPos   AS LONG
 LOCAL WmScroll AS LONG
 LOCAL iSeeYou  AS LONG
 
 SELECT CASE CBMSG
  CASE %WM_INITDIALOG
    si.cbSize = LEN(si)
    si.fMask  = %SIF_ALL
    si.nMin   =   0
    si.nMax   = 800
    si.nPage  =  50
    si.nPos   =   0
    SetScrollInfo CBHNDL, %SB_VERT, si, 1
    SetScrollInfo CBHNDL, %SB_HORZ, si, 1 
 
  CASE %WM_HSCROLL, %WM_VSCROLL
    si.cbSize = SIZEOF(si)
    si.fMask  = %SIF_ALL
    IF CBMSG = %WM_HSCROLL THEN
      GetScrollInfo CBHNDL, %SB_HORZ, si
      oldPos = si.nPos
      SELECT CASE LOWRD(CBWPARAM)
        CASE %SB_LINELEFT   : si.nPos = si.nPos - 5
        CASE %SB_PAGELEFT   : si.nPos = si.nPos - si.nPage
        CASE %SB_LINERIGHT  : si.nPos = si.nPos + 5
        CASE %SB_PAGERIGHT  : si.nPos = si.nPos + si.nPage
        CASE %SB_THUMBTRACK : si.nPos = HIWRD(CBWPARAM)
        CASE ELSE           : EXIT FUNCTION
      END SELECT
      si.fMask = %SIF_POS
      SetScrollInfo CBHNDL, %SB_HORZ, si, 1
      ScrollWindow CBHNDL, oldPos - si.nPos, 0, BYVAL %NULL, BYVAL %NULL
    ELSE
      GetScrollInfo CBHNDL, %SB_VERT, si
      oldPos = si.nPos
      SELECT CASE LOWRD(CBWPARAM)
        CASE %SB_LINEUP     : si.nPos = si.nPos - 5
        CASE %SB_PAGEUP     : si.nPos = si.nPos - si.nPage
        CASE %SB_LINEDOWN   : si.nPos = si.nPos + 5
        CASE %SB_PAGEDOWN   : si.nPos = si.nPos + si.nPage
        CASE %SB_THUMBTRACK : si.nPos = HIWRD(CBWPARAM)
        CASE ELSE           : EXIT FUNCTION
      END SELECT
      si.fMask = %SIF_POS
      SetScrollInfo CBHNDL, %SB_VERT, si, 1
      ScrollWindow CBHNDL, 0, oldPos - si.nPos, BYVAL %NULL, BYVAL %NULL
    END IF
    FUNCTION = 1
 
    GetClientRect(CBHNDL, DlgRect)
    hControl = GetDlgItem(CBHNDL, %LABEL)
    GetWindowRect(hControl, CtrlRect)
    MapWindowPoints %HWND_DESKTOP, CBHNDL, CtrlRect, 2
 
    iSeeYou = 0
    IF PtInRect(DlgRect, CtrlRect.nLeft,  CtrlRect.nTop )   THEN INCR iSeeYou
    IF PtInRect(DlgRect, CtrlRect.nRight, CtrlRect.nTop)    THEN INCR iSeeYou
    IF PtInRect(DlgRect, CtrlRect.nLeft,  CtrlRect.nBottom) THEN INCR iSeeYou
    IF PtInRect(DlgRect, CtrlRect.nRight, CtrlRect.nBottom) THEN INCR iSeeYou
    IF iSeeYou = 4 THEN
      SetWindowText CBHNDL, "Control is fully viewable"
    ELSEIF iSeeYou = 0 THEN
      SetWindowText CBHNDL, "Control is not viewable"
    ELSE
      SetWindowText CBHNDL, "Control is partially viewable"
    END IF 
 
 END SELECT
 
END FUNCTION
'______________________________________________________________________________
 
FUNCTION PBMAIN () AS LONG
 LOCAL hDlg AS DWORD
 
 DIALOG NEW %HWND_DESKTOP, "isControlViewable",,, 450, 250, _
   %WS_CAPTION OR %WS_SYSMENU OR %WS_MAXIMIZEBOX OR %WS_MINIMIZEBOX,  TO hDlg
 DIALOG SET COLOR hDlg, -1, RGB(75, 165, 125)
 SetClassLong hDlg, %GCL_HICON, LoadIcon(BYVAL %NULL, BYVAL %IDI_INFORMATION) 'Set a nice dialog icon
 
 CONTROL ADD LABEL, hDlg, %LABEL, "Keep scrolling !", 100, 100, 300, 200, _
   %SS_CENTER OR %SS_CENTERIMAGE, %WS_EX_LEFT
 
 CONTROL SET COLOR hDlg, %LABEL, -1, RGB(25, 125, 255)
 
 DIALOG SHOW MODAL hDlg CALL DlgProc
 
END FUNCTION
'______________________________________________________________________________
'

[John Thompson]

Awesome! Thank you both very much. -John

[Michael Mattias]

GetClientRect(CBHNDL, DlgRect)
...
GetWindowRect(hControl, CtrlRect)

These lines will compile and execute in 7x, but not in anything else. The parens around the parameters are a syntax error when the CALL verb is not used. (That would be a bug in 7x which was fixed in 8x).

MCM

[Dave Biggs]

GetClientRect(CBHNDL, DlgRect)
...
GetWindowRect(hControl, CtrlRect)

These lines will compile and execute in 7x, but not in anything else. The parens around the parameters are a syntax error when the CALL verb is not used. (That would be a bug in 7x which was fixed in 8x).

MCM

Is that the right way around?

Call GetClientRect CBHNDL, DlgRect

compiles in PBWin703 but not in PBWn804 (needs parens)

GetClientRect(CBHNDL, DlgRect)

compiles in both as does

GetClientRect CBHNDL, DlgRect

Seems the PBWin8x compiler requires parens when the Call verb is used. The bug was that the PBWin7x compiler would accept use of the Call verb without parens. Wasn't it?

[Michael Mattias]

Yes, I think I said that correctly.

Without CALL, a DECLARE is required and no parens are permitted.
With CALL, DECLARE is not required (but is permitted) and parens are mandatory.

The above is the "corrrect" syntax. What different versions of the compiler of the compiler do... well that's what defines bugs.

>The bug was that the PBWin7x compiler would accept use of the Call verb without parens. Wasn't it?

Actually two bugs..
- Accepted parens when CALL was not used. (But it worked as though the parens weren't there)
- This also 'worked' :

 Value = FunctionName param, param, param

I know the documented syntax works in all versions; so just use that.

[Dave Biggs]

Ok since PBWin8x parens are mandatory with Call.

Without CALL, a DECLARE is required and no parens are permitted.

Yet

GetClientRect(CBHNDL, DlgRect)

compiles and runs in both PBWin703 and PBWin804.

[Michael Mattias]

GetClientRect(CBHNDL, DlgRect) compiles and runs in both PBWin703 and PBWin804.

It shouldn't compile.

Well, I'll be dipped... this compiles in 8.03, too:

FUNCTION TestParens() AS LONG
    LOCAL hWnd AS LONG, R AS RECT
    
    GetClientRect (hWnd, R)
    
END FUNCTION

I could have sworn this was fixed in 8x.

Oh, wait a minute, the help file changed. These are all shown as valid under CALL:

CALL MyProc (parm1, parm2)
MyProc (parm1, Parm2)
MyProc parm1, parm2

Aha, I see now by doing some more RTFM:

This updated syntax now permits macros to be called using the SUB-style convention if/when the macros expand directly to Function calls. For example:

.. and here it is in the What's New section of 8.03 help file:

PowerBASIC now supports parentheses around parameters in calls to Subs (and functions that are being called as Subs) when the CALL keyword is omitted. For example, the following lines demonstrate valid syntax:

I guess it's true the memory is the second thing to go.

MCM

[Pierre Bellisle]

Welcome in 2008.