Populate a QTableWidget with QTreeViews

[Patou355]

I have QTreeViews contained in a QVector and I want them to populate a QTableWidget using the method QTableWidget::setCellWidget:

//create the TreeViews and their associate models
int i;
for (i = 0 ; i < treeViewAdresses.size() ; ++i) {
    forest << new QTreeView();
    models << new TreeModel(foo);
    forest[i]->setVisible(true);
    forest[i]->setModel(models.at(i));
    forest[i]->setDragEnabled(true);
    forest[i]->setAcceptDrops(true);
    _manageOrderedTable->setCellWidget(0, treeViewAdresses.at(i), forest[i]);
}

The TreeViews are successfully created but the problem is that they stay outside the table. See below:

treeViewAdresses is a vector of integers containing the position (I checked, they are good)
I tried to:

• replace the forest[i] with forest.at(i).
• tried to insert a simple QWidget in the table instead.
  Same problem in both cases.

[mrjj]

Hi
Sadly the image upload is broken and we cannot see it.
Please use external site and post links here.

[mrjj]

Oh
When they become windows it means you did not give them a parent.
So do that in constructor
or via -> setParent

new QTreeView( PARENTPTR );

[VRonin]

@mrjj Unfortunately setCellWidget takes ownership of the widget so it can't be because of the parent thing.

very strange that even with a base QWidget does not work.

Can you try with a clean project and see if it happens the same?

[mrjj]

@VRonin
" setCellWidget takes ownership of the widget"
Hi, thx for adding :)
yes 100% agree,
but what else would make them windows`?

[VRonin]

@mrjj said in Populate a QTableWidget with QTreeViews:

but what else would make them windows`?

my only guess would be the cell(0, treeViewAdresses.at(i)) is null so since the cell does not exist the widget is not added. I'm just 10% confident on this one though

[mrjj]

@VRonin
I also have a 10% question ;)
The at() returns a const ref
Will it then be able to take ownership ?

[VRonin]

@mrjj I think the actual widget is forest[i] which should return a non-const ref.

Anyway even at() will return a const reference to a non const pointer so it can take ownership as the content of the pointer is non const, just like QWidget* const

[mrjj]

@VRonin
Yeah, just thinking out loud.

Well test is always the best

    for (int i = 0 ; i < 10 ; ++i) {
        QTreeView* forest = new QTreeView(ui->tableWidget);
        forest->setVisible(true);
        ui->tableWidget->setCellWidget(i, 0, forest);
    }

This puts them inside

But if you change
QTreeView* forest = new QTreeView(ui->tableWidget);
-->
QTreeView* forest = new QTreeView();

I get windows too.

So it was the parent !?!?!

runable test project
https://www.dropbox.com/s/8l2an24l7vp0x0l/treesintable.zip?dl=0

[VRonin]

@mrjj Big up then. typical example of

Probably here it's the doc's fault as I suspect "ownership" doesn't mean parent-child ownership in this case

[mrjj]

@VRonin
Hehe. And after that came the rule. If it works dont mess with it :)

Well normally when reparenting a widget, all win flags are stripped so it
can be inside. Maybe in this case that part is missing?
Using a QLabel just works.

[mrjj]

@mrjj

Update:
Suddenly it works without setting parent. ( i rebooted. )

If i move show() to be last, its perfect as it wont show as window for a brief moment

 for (int i = 0 ; i < 10 ; ++i) {
        QTreeView* forest = new QTreeView();
        ui->tableWidget->setCellWidget(i, 0, forest);
        forest->setVisible(true);
    }

This suddenly works.

So my other test must be flawed and the parent is not important as
setCellWidget does as we think.

Meh :)

[Patou355]

Hello guys. Thanks for your answerS.
The "error" was simple, stupid. In fact the _manageOrderedTable is an instance of a class called ManageOrderedTable which inherits from QTableWidget.
In this class, there's a method called fillAll and I called setColumnCount() from there. Bad idea.
When I call this method from upon the for loop, everything is fine. I don't understand why.
I could understand if setColumnCount() were called too late, but in this case I should have a segfault, trying to call a cell that doesn't exist.
Well, I don't really understand why but now it works...

Tank you for your help, Patrick.