path: root/doc/src/declarative/focus.qdoc

/****************************************************************************
**
** Copyright (C) 2010 Nokia Corporation and/or its subsidiary(-ies).
** All rights reserved.
** Contact: Nokia Corporation (qt-info@nokia.com)
**
** This file is part of the test suite of the Qt Toolkit.
**
** $QT_BEGIN_LICENSE:LGPL$
** No Commercial Usage
** This file contains pre-release code and may not be distributed.
** You may use this file in accordance with the terms and conditions
** contained in the Technology Preview License Agreement accompanying
** this package.
**
** GNU Lesser General Public License Usage
** Alternatively, this file may be used under the terms of the GNU Lesser
** General Public License version 2.1 as published by the Free Software
** Foundation and appearing in the file LICENSE.LGPL included in the
** packaging of this file.  Please review the following information to
** ensure the GNU Lesser General Public License version 2.1 requirements
** will be met: http://www.gnu.org/licenses/old-licenses/lgpl-2.1.html.
**
** In addition, as a special exception, Nokia gives you certain additional
** rights.  These rights are described in the Nokia Qt LGPL Exception
** version 1.1, included in the file LGPL_EXCEPTION.txt in this package.
**
** If you have questions regarding the use of this file, please contact
** Nokia at qt-info@nokia.com.
**
**
**
**
**
**
**
**
** $QT_END_LICENSE$
**
****************************************************************************/

/*!
\target qmlfocus
\page qmlfocus.html
\title Keyboard Focus in QML

When a key is pressed or released, a key event is generated and delivered to the
focused QML \l Item.  To facilitate the construction of reusable components
and to address some of the cases unique to fluid user interfaces, the QML items add a
\e scope based extension to Qt's traditional keyboard focus model.

\tableofcontents

\section1 Key Handling Overview

When the user presses or releases a key, the following occurs:
\list 1
\o Qt receives the key action and generates a key event.
\o If the Qt widget containing the \l QmlView has focus, the key event is delivered to it. Otherwise, regular Qt key handling continues.
\o The key event is delivered by the scene to the QML \l Item with \e {active focus}.  If no \l Item has \e {active focus}, the key event is \l {QEvent::ignore()}{ignored} and regular Qt key handling continues.
\o If the QML \l Item with \e {active focus} accepts the key event, propagation stops.  Otherwise the event is "bubbled up", by recursively passing it to each \l Item's parent until either the event is accepted, or the root \l Item is reached.

If the \c {Rectangle} element in the following example has active focus and the \e A key is pressed,
it will bubble up to its parent.  However, pressing the \e B key will bubble up to the root
item and thus subsequently be \l {QEvent::ignore()}{ignored}.

\code
Item {
    Item {
        Keys.onPressed: if (event.key == Qt.Key_A) { console.log('Key A was pressed'); event.accepted = true }
        Rectangle {}
    }
}
\endcode

\o If the root \l Item is reached, the key event is \l {QEvent::ignore()}{ignored} and regular Qt key handling continues.

\endlist

See also the \l {Keys}{Keys attached property} and \l {KeyNavigation}{KeyNavigation attached property}.

\section1 Querying the Active Focus Item

Whether or not an \l Item has \e {active focus} can be queried through the
property \c {Item::focus}.  For example, here we have a \l Text
element whose text is determined by whether or not it has \e {active focus}.

\code
Text {
    text: focus ? "I have active focus!" : "I do not have active focus"
}
\endcode

\section1 Acquiring Focus and Focus Scopes

An \l Item requests focus by setting the \c {Item::focus} property to true.

For very simple cases simply setting the \c {Item::focus} property is sometimes
sufficient.  If we run the following example in the \c qmlviewer, we see that
the \c {keyHandler} element has \e {active focus} and pressing the 'A', 'B'
or 'C' keys modifies the text appropriately.

\table
\row
\o \code
    Rectangle {
        color: "lightsteelblue"; width: 240; height: 25
        Text { id: myText }
        Item {
            id: keyHandler
            focus: true
            Keys.onPressed: {
                if (event.key == Qt.Key_A)
                    myText.text = 'Key A was pressed'
                else if (event.key == Qt.Key_B)
                    myText.text = 'Key B was pressed'
                else if (event.key == Qt.Key_C)
                    myText.text = 'Key C was pressed'
            }
        }
    }
\endcode
\o \image declarative-qmlfocus1.png
\endtable

However, were the above example to be used as a self-contained component, this
simple use of the \c {Item::focus} property is no longer sufficient.  The left
hand side of the following table shows what we would like to be able to write.
Here we create two instances of our previously defined component, and set the
second one to have focus.  The intention is that when the \e A, \e B, or \e C
keys are pressed, the second of the two components receives the event and
reponds accordingly.

\table
\row
\o \code
Rectangle {
    color: "red"; width: 240; height: 55
    MyWidget {}
    MyWidget { y: 30; focus: true }
}
\endcode
\o \code
Rectangle {
    color: "red"; width: 240; height: 55
    Rectangle {
        color: "lightsteelblue"; width: 240; height: 25
        Text { id: myText }
        Item {
            id: keyHandler
            focus: true
            Keys.onPressed: {
                if (event.key == Qt.Key_A)
                    myText.text = 'Key A was pressed'
                else if (event.key == Qt.Key_B)
                    myText.text = 'Key B was pressed'
                else if (event.key == Qt.Key_C)
                    myText.text = 'Key C was pressed'
            }
        }
    }
    Rectangle {
        y: 30; focus: true
        color: "lightsteelblue"; width: 240; height: 25
        Text { id: myText }
        Item {
            id: keyHandler
            focus: true
            Keys.onPressed: {
                if (event.key == Qt.Key_A)
                    myText.text = 'Key A was pressed'
                else if (event.key == Qt.Key_B)
                    myText.text = 'Key B was pressed'
                else if (event.key == Qt.Key_C)
                    myText.text = 'Key C was pressed'
            }
        }
    }
}
\endcode
\endtable

The right hand side of the example shows the expanded code - the equivalent QML
without the use of the component \c {MyWidget}.  From this, the problem is
evident - there are no less than three elements that have the \c {Item::focus}
property set to true.  Ultimately only one element can have focus, and the
system has to decide which on.  In this case the first appearance of the
\c {Item::focus} property being set to true on line 4 is selected, and the value
of \c {Item::focus} in the other two instances is reverted back to false.  This
is exactly the opposite of what was wanted!

This problem is fundamentally one of visibility.  The \c {MyWidget}
components each set their \c {keyHandler} Items as focused as that is all they can
do - they don't know how they are going to be used, but they do know that when
they're in use their \c {keyHandler} element is what needs focus.  Likewise
the code that uses the two \c {MyWidgets} sets the second \c {MyWidget} as
focused. While it doesn't know exactly how the \c {MyWidget} is
implemented, it knows that it wants the second one to be focused.  This allows us
to achieve encapsulation, allowing each widget to focus on it's appropriate behaviour
itself.

To solve this problem - allowing components to care about what they know about
and ignore everything else - the QML items introduce a concept known as a
\e {focus scope}.  For existing Qt users, a \e {focus scope} is like an
automatic focus proxy.  A \e {focus scope} is created using the \l FocusScope
element.

In the next example, a \l FocusScope is added to the component, and the visual
result shown.

\table
\row
\o \code
FocusScope {
    width: 240; height: 25
    Rectangle {
        color: "lightsteelblue"; width: 240; height: 25
        Text { id: myText }
        Item {
            id: keyHandler
            focus: true
            Keys.onPressed: {
                if (event.key == Qt.Key_A)
                    myText.text = 'Key A was pressed'
                else if (event.key == Qt.Key_B)
                    myText.text = 'Key B was pressed'
                else if (event.key == Qt.Key_C)
                    myText.text = 'Key C was pressed'
            }
        }
    }
}
\endcode
\o \image declarative-qmlfocus2.png
\endtable

Conceptually \e {focus scopes} are quite simple.
\list
\o Within each \e {focus scope} one element may have \c {Item::focus} set to true.
If more than one \l Item has the \c {Item::focus} property set, the first is selected
and the others are unset, just like when there are no \e {focus scopes}.
\o When a \e {focus scope} receives \e {active focus}, the contained element with
\c {Item::focus} set (if any) also gets \e {active focus}.  If this element is
also a \l FocusScope, the proxying behaviour continues.  Both the
\e {focus scope} and the sub-focused item will have \c {Item::focus} set.
\endlist

So far the example has the second component statically selected.  It is trivial
now to extend this component to make it clickable, and add it to the original
application.  We still set a one of the widgets as focused by default, but from
then on clicking the either one gives it focus.

\table
\row
\o \code
Rectangle {
    color: "red"; width: 240; height: 55
    MyClickableWidget {}
    MyClickableWidget { y: 30; focus: true }
}
\endcode
\o \code
FocusScope {
    id: page; width: 240; height: 25
    MyWidget { focus: true }
    MouseRegion { anchors.fill: parent; onClicked: { page.focus = true } }
}
\endcode
\endtable

\image declarative-qmlfocus3.png

When a QML item explicitly relinquishes focus (by setting its
\c {Item::focus} property to false while it has \e {active focus}), the system
does not automatically select another element to receive focus.  That is, it
is possible for there to be no currently \e {active focus}.

\section1 Advanced uses of Focus Scopes

Focus scopes allow focus to allocation to be easily partitioned.  Several
QML items use it to this effect.

\l ListView, for example, is itself a focus scope.  Generally this isn't
noticable as \l ListView doesn't usually have manually added visual children.
By being a focus scope, \l ListView can focus the current list item without
worrying about how that will effect the rest of the application.  This allows
the current item delegate to react to key presses.

This contrived example shows how this works.  Pressing the \c Return key will
print the name of the current list item.

\table
\row
\o \code
Rectangle {
    color: "lightsteelblue"; width: 240; height: 320

    ListView {
        id: myView; anchors.fill: parent; focus: true
        model: ListModel {
            ListElement { name: "Bob" }
            ListElement { name: "John" }
            ListElement { name: "Michael" }
        }
        delegate: FocusScope {
            width: contents.width; height: contents.height
            Text {
                focus: true
                text: name
                Keys.onReturnPressed: console.log(name)
            }
        }
    }
}
\endcode
\o \image declarative-qmlfocus4.png
\endtable

While the example is simple, there's a lot going on behind the scenes.  Whenever
the current item changes, the \l ListView sets the delegate's \c {Item::focus}
property.  As the \l ListView is a \e {focus scope}, this doesn't effect the
rest of the application.  However, if the \l ListView itself has
\e {active focus} this causes the delegate itself to receive \e {active focus}.
In this example, the root element of the delegate is also a \e {focus scope},
which in turn gives \e {active focus} to the \c {Text} element that
actually performs the work of handling the \e {Return} key.

All of the QML view classes, such as \l PathView and \l GridView, behave
in a similar manner to allow key handling in their respective delegates.

\section1 Focus Panels

Traditional UIs are composed of many top-level windows.  Windows actually
perform two tasks - they act as the visual bounds for a widget, and they segment
focus.  Each window has a separate focused widget, that becomes (to mix
terminologies) the \e {active focus} widget when the window is the active
window.

### Focus panels do basically the same thing.
*/