◄About Section►  ◄Function Group►                     ◄Up► ◄Next► ◄Previous►
────────────────────────────────────────────────────────────────────────────
 
                               Using Cursors
 
Typically, applications use cursors to mark the text-insertion point in a
text window or to indicate that a window has the keyboard focus.
 
There can be only one cursor in use by the system at one time, so windows
must create and destroy cursors as they gain and lose the keyboard focus.
The following code fragment shows how an application should respond to a
WM_SETFOCUS message when using a cursor in a particular window:
 
case WM_SETFOCUS:
    if (SHORT1FROMMP(mp2)) {
 
        /* gain focus */
 
        WinCreateCursor(...);
        WinShowCursor(hwnd, TRUE);
    }
    else {
 
        /* lose focus */
 
        WinDestroyCursor(hwnd);
    }
    return 0L;
 
Creating a Cursor
 
An application creates a cursor by calling the WinCreateCursor function.
Generally, this is done when a window gains the keyboard focus. An
application specifies the window in which the cursor will be displayed. This
window may be HWND_DESKTOP, an application window, or a control window.
 
An application specifies the cursor position, in window coordinates, and the
cursor height and width. It also specifies whether the cursor rectangle
should be filled, framed, flashing, or halftone.
 
The cursor width is usually zero (nominal border width is used) for
text-insertion cursors. This is preferable to a value of 1, since such a
fine width is almost invisible on a high-resolution monitor. The cursor
width may also be related to the window size; for example, when a button
control uses a dotted-line cursor around the button text to indicate focus.
 
Finally, an application can specify a clipping rectangle, in window
coordinates, that controls the cursor clipping region. Typically, the most
efficient strategy is to specify NULL, causing the rectangle to clip the
cursor to the window rectangle.
 
Destroying a Cursor
 
Applications should destroy cursors by calling the WinDestroyCursor function
when they lose the keyboard focus. It is important that they be destroyed
when they lose the focus. Manipulating two cursors in MS OS/2 simultaneously
can have unpredictable results and affect other applications.
 
Showing the Cursor
 
MS OS/2 maintains a "show level" for the cursor. The cursor is visible when
the show level is zero. Each time the cursor is hidden, its show level is
incremented. Each time the cursor is shown, its show level is decremented.
Because the show/hide relationship is one for one, the show level can never
go below zero.
 
An application should show the cursor when it is first created since the
cursor is created with a show level of 1.
 
MS OS/2 automatically hides the cursor when the WinBeginPaint function is
called, and shows the cursor when the WinEndPaint function is called.
Therefore, there is no conflict with the cursor during WM_PAINT processing.
 
Positioning the Cursor
 
An application can set the position of an existing cursor by calling the
WinCreateCursor function with the CURSOR_SETPOS flag set. This function can
also be used to move a cursor around a window. Position arguments are given
in window coordinates. To change the cursor size, destroy the current cursor
and then create a new one with the desired size.
 
 
                                      ♦