There is data validation in my MS Word user form which returns the focus to the textbox where the user entered something incorrect. Now I am trying to accommodate the user's change of mind: instead of correcting the entry, I want him to be able to exit the form (click the Exit command button), in which case the entry would be discarded. I suppose that a solution would start with not using the text box's exit event. I little help from someone who knows the answer would save me a lot of testing time, perhaps to find out that I can't do it.
Does anyone know?
I understand that you are handling the Exit event of the Textbox, setting the Cancel output parameter if the data is not valid.
There's a tricky but simple solution that permits to keep that working and still have an Exit button. It permits to activate the handler of the Exit button without requiring the focus to leave the Textbox. This way you can unload the Form safely in this handler.
Try this it works pretty smoothly:
1- Set the property TakeFocusOnClick of the Exit command button to False. You can do that at design time in the property-sheet, or at run-time i.e. at UserForm_Activate
2- just unload the form when the Exit button is clicked:
Private Sub ExitButton_Click()
Unload Me
End Sub
#A.S.H provided the key to the solution below. His point is that it is possible to call another event procedure while Cancel is active in the Exit procedure of a control. That other procedure can be used to rectify the condition in the first control which is triggering the Cancel, thereby enabling an orderly exit. The all-enabling condition is that the control on whose click event the rectifying procedure is to run must not take the focus when clicked (meaning it can run without triggering an exit from the control stuck on Cancel). I have added code to the exit procedure to set CmdExit.TakeFocusOnClick = False when a Cancel condition arises there. Now, ...
Private Sub CmdExit_Click()
' 12 May 2017
' if CmdExit can't take the focus it can't be the ActiveControl
If Not ActiveControl Is CmdExit Then
Select Case ActiveControl.Name
Case "Cbx107"
Cbx107.Value = ""
Case "Tbx53"
Tbx53.Value = "0"
End Select
With CmdExit
If Not .TakeFocusOnClick Then
.TakeFocusOnClick = True
.SetFocus
End If
End With
End If
' now CmdExit is the ActiveControl
MsgMe "Cmd Exit: ActiveControl = " & ActiveControl.Name
Me.Hide
End Sub
I'm using SDL 2.0, and decided to try out making multiple windows. Unfortunately, now I can't quit out of my program without going back to the IDE and force closing it.
The event handling is as simple as possible, I am only polling for the quit event, and it worked perfectly fine before I added the second window. Is the Quit Event ignored when using multiple windows? If so, how can I turn it back on?
The Quit Event is only sent when the last open window is trying to close, otherwise a window close event is sent.
I also ran into this problem, and the documentation is a little sparse on the topic so I ended up here.
The summary of the problem is:
If you have only one Window, clicking the X button will trigger an SDL_QUIT event.
If you have two or more windows, clicking the X button will trigger an SDL_WINDOWEVENT event, with an internal type of SDL_WINDOWEVENT_CLOSE.
So if your typical code for single-window quit events might look something like this:
SDL_Event e;
while (SDL_PollEvent(&e))
{
if (e.type == SDL_QUIT)
{
// ... Handle close ...
}
}
The multi-window equivalent would be:
SDL_Event e;
while (SDL_PollEvent(&e))
{
if (e.type == SDL_WINDOWEVENT
&& e.window.event == SDL_WINDOWEVENT_CLOSE)
{
// ... Handle window close for each window ...
// Note, you can also check e.window.windowID to check which
// of your windows the event came from.
// e.g.:
if (SDL_GetWindowID(myWindowA) == e.window.windowID)
{
// ... close window A ...
}
}
}
Note that on the last window, you will again receive SDL_QUIT because it's now the only active window - so best to structure your code in a way that correctly handles both depending on the circumstances.
See docs for more info.
I'm trying to design a GUI in SciLab that updates it's properties depending on a checkmark. For example: A checkbox might enable and change the backrounds of several text boxes during a callback; or a pushbutton may require a certain number of checkboxes to be selected.
My problem is that I can't seem to develop a flow control statement for running instructions depending on the checkboxes state during a callback. My current UIControl element looks like this:
handles.chkS11En=uicontrol(f,'unit','normalized','BackgroundColor',[0.8,0.8,0.8],'Enable','on','FontAngle','normal','FontName','helvetica','FontSize',[12],'FontUnits','points','FontWeight','normal','ForegroundColor',[0,0,0],'HorizontalAlignment','center','ListboxTop',[],'Max',[1],'Min',[0],'Position',[0.02140625,0.791119360625398,0.0803125,0.0369667],'Relief','flat','SliderStep',[0.01,0.1],'String','S11','Style','checkbox','Value',[0],'VerticalAlignment','middle','Visible','on','Tag','chkS11En','Callback','chkS11En_callback(handles)')
And my callback that runs when I check the checkbox is this:
cS11En = findobj('tag', 'chkS11En'); // checkbox option
tS11MagUpperBound = findobj('tag', 'txtS11MagUpperBound'); //edit box that is controlled
mprintf("%d\n",cS11En.Value);
if cS11En.Value == [1] then
mprintf("Checked = on \n");
set(tS11MagUpperBound,'BackgroundColor',[1,1,1]);
set(tS11MagUpperBound,"Enable",'on');
set(cS11Save,"Enable",'on');
elseif cS11En.Value == [0] then
mprintf("Checked = off \n");
set(tS11MagUpperBound,'BackgroundColor',[0.8,0.8,0.8]);
set(tS11MagUpperBound,'Enable','off');
set(cS11Save,"Enable",'off');
end
The problem with this code seems to be that the second path (Value = 1) never seems to run, even when i continually toggle the checkbox. I get an output like so:
0
Checked = off
0
Checked = off
0
Checked = off
0
Checked = off
Is there something I'm doing wrong in order to reload checking the element? I want to be able to run both paths, however I can never seem to get a value of 1 from the checkbox element. Does anyone have a solution to this? Thanks!
IF anyone is wondering and finds this through the googles or something, this is how I fixed it:
It turns out that SciLab sometimes doesn't clear all UI variables when the form is closed and a script is running.
The solution is to add a few lines in the top of each of your program that clears all variables, closes all forms, and initializes your variables.
Basically, add this:
// /////////////
// Lemon Pledge
// /////////////
mprintf("\n!!!!!!!!!!!!!!!!!!!\nCLEARING ALL VARIABLES\n!!!!!!!!!!!!!!!!!!!\n")
xdel(winsid());
clear;
clearglobal;
Another less complex solution would be:
Using the same checkbox I left the last attribute in blank.
handles.chkS11En=uicontrol(f,'unit','normalized','BackgroundColor',[0.8,0.8,0.8],'Enable','on','FontAngle','normal','FontName','helvetica','FontSize',[12],'FontUnits','points','FontWeight','normal','ForegroundColor',[0,0,0],'HorizontalAlignment','center','ListboxTop',[],'Max',[1],'Min',[0],'Position',[0.02140625,0.791119360625398,0.0803125,0.0369667],'Relief','flat','SliderStep',[0.01,0.1],'String','S11','Style','checkbox','Value',[0],'VerticalAlignment','middle','Visible','on','Tag','chkS11En','Callback','')
Then I make the callback
function chkS11En_callback(handles)
if handles.chkS11En.Value == [1] then
mprintf("Checked = on \n");
set(tS11MagUpperBound,'BackgroundColor',[1,1,1]);
set(tS11MagUpperBound,"Enable",'on');
set(cS11Save,"Enable",'on');
else
mprintf("Checked = off \n");
set(tS11MagUpperBound,'BackgroundColor',[0.8,0.8,0.8]);
set(tS11MagUpperBound,'Enable','off');
set(cS11Save,"Enable",'off');
end
And voilà, no need to clear your workspace.
I am Using WaitforComplete() in watiN but it doesnt seems to work well. As it executes the next statement even if you have given longer time to wait. I am using thread.sleep() to stop my application until it gets the desired page or element. But the thing is pages are so much dynamic that sometimes it takes much longer time as specified.
Any better solution. Any thing that will catch the page return dynamically and dont go to execute next statments in application.
Sample of Code
'Show Details page
Assert.AreEqual("Confirmation", _internetExplorer.Title)
If (_internetExplorer.Button(Find.ById(New Regex("btnFinish"))).Exists) Then
_internetExplorer.Button(Find.ById(New Regex("btnFinish"))).Click()
Else
Assert.Fail("Could not Find Finish Booking Button on Confirmation Page")
End If
System.Threading.Thread.Sleep(100000)
'Show Booking Summary page
Assert.AreEqual("Display Booking", _internetExplorer.Title)
I want something that detect the return of page dynamically. instead of giving some constant value.
WaitForComplete only works well if there is a postback after some action. Otherwise you have to find something else to wait for. Following an example on how to wait for the specified title:
_internetExplorer.Element("title", "Confirmation").WaitUntilExists();
I would always prefer to use one of the WaitXXX methods instead of Thread.Sleep cause the WaitXXX methods do only wait until the contraint is met. Where as Sleep waits for the time you specified. If its to long, time is waisted. If its to short, problems arise.
HTH,
Jeroen
The WaitForComplete method esentially moves on once the browser has set it's readystate to comllete and the busy state to false.
What I typically do is to try and access what you need to, then perform a thread.sleep for say half a second, then try again. I also have a global timeout that quits after say 10 seconds.
int timeout = 20;
bool controlFound = false;
for (int i = 0; i < timeout; i++)
{
if (_internetExplorer.Button(Find.ById(New Regex("btnFinish"))).Exists)
{
_internetExplorer.Button(Find.ById(New Regex("btnFinish"))).Click();
controlFound = true;
break;
}
else
{
System.Threading.Thread.Sleep(500);
}
}
if (!controlFound)
{
Assert.Fail("Control not found");
}
If it is executing the next statement, it should be finding the corresponding element. I suggest posting a sample of the code you are trying.
It's been a while since I used GTK+, and the last time I did was in C, not using gtkmm and C++ as I am now. Anyway, I have what I think should be an easy problem to solve:
I have a pop-up menu consisting of a list of radio buttons, and when I click one of them I want some action to occur. The code goes like this:
Gtk::RadioMenuItem::Group group;
for ( size_t i = 1; i < LH_MAX; ++i )
{
Gtk::RadioMenuItem* pItem = new Gtk::RadioMenuItem( group, names[i], names[i] );
pItem->set_name( names[i] );
pItem->signal_activate().connect( sigc::mem_fun(*this, &MyClass::on_item_activated) );
pItem->show();
m_Menu.append( *Gtk::manage(pItem) );
}
The only problem I see is that MyClass::on_item_activated gets called twice when a previously-unselected radio button is chosen from the menu. It's called only once when the already-selected radio button is clicked.
I'm guessing that the first firing is to say "something is no longer activate," and the second is for the new radio button activation. Whether I'm right or wrong, the question is the same: how best can I have my handler only take action once per click? Either I need the handler to get called only once, or I need something to check from inside it to know if the callback is a "duplicate" or not.
You could use sigc::bind to supply the item as a argument to the callback function.
pItem->signal_activate().sigc::bind(sigc::mem_fun(*this,&MyClass::on_item_activated),pItem));
Then you can use item->get_active() in the callback to respond to activations only.
void MyClass::on_item_activated(Gtk::RadioMenuItem* item) {
if (item->get_active()) {
// Do some stuff
}
}
That's what I do too, connect to signal_toggled() and check if get_active() is true.
I don't know exactly what you're trying to accomplish (or what MyClass is and what base classes it inherits from), but connecting to signal_toggled() might be more useful than signal_activate()
/Agree with Johannes. Check if the item is activated when receiving the signal.