Chapter 7: One Thing Leads to Another

This example shows how to create custom widgets with signals and slots, and how to connect them together in more complex ways. For the first time, the source is split among several files.

• lcdrange.h contains the LCDRange class definition
• lcdrange.cpp contains the LCDRange implementation
• main.cpp contains MyWidget and main.
• Makefile contains some rules for generating the meta object information necessary for signal/slot creation.

Line by Line Walk-Through

lcdrange.h

This file is mainly lifted from main.cpp in chapter 6 and only the changes are noted here.

    #ifndef LCDRANGE_H
    #define LCDRANGE_H

This is the classic C construction to avoid errors if a header file happens to be included more than once. If you don't use it already: It is a very good habit. The #ifndef should enclose all of the header file.

    #include <qvbox.h>

qwidget.h is included. LCDRange inherits QWidget, and the header file of a parent class must always be included. Until now, qwidget.h has been included indirectly via other header files like qpushbutton.h.

    class QSlider;

This is another classic trick, but one that's much less used. Since we don't need QSlider in the interface of the class, only in the implementation, we use a forward declaration of the class in the header file, and #include the header file for QSlider in the .cpp file.

This makes compilation of big projects much faster, since when a header file has changed, fewer files need to be recompiled. It can often speed up big compiled by a factor of two or more.

    class LCDRange : public QVBox
    {
        Q_OBJECT
    public:
        LCDRange( QWidget *parent=0, const char *name=0 );

Note the Q_OBJECT. This macro must be included in all classes that contain signals and/or slots. If you are curious, it defines the functions that are implemented in the meta object file.

        int value() const;
    public slots:
        void setValue( int );
    
    signals:
        void valueChanged( int );

These three members make up an interface between this widget and other components in a program. Until now, LCDRange didn't really have an interface at all.

value() is a public function for accessing the value of the LCDRange. setValue() is our first custom slot and valueChanged() is our first custom signal.

Slots must be implemented in the normal way (remember, a slot is also a C++ member function). Signals are automatically implemented in the meta object file. Signals follow the access rules of protected C++ functions, i.e. they can only be emitted by the class they are defined in or by classes inheriting from it.

The signal valueChanged() is used when the LCDRange's value has changed - just as you guessed from the name. This is not the last signal you'll see called somethingChanged().

lcdrange.cpp

This file is mainly lifted from t6/main.cpp and only the changes are noted here.

        connect( slider, SIGNAL(valueChanged(int)), lcd, SLOT(display(int)) );
        connect( slider, SIGNAL(valueChanged(int)), SIGNAL(valueChanged(int)) );
    }
    int LCDRange::value() const

This code is from the LCDRange constructor.

The first connect is the same you've seen in the previous chapter. The second is new: It connects slider's valueChanged() signal to this object's valueChanged signal. connect() with 3 arguments always connect to signals or slots in this object.

Yes, that's right. Signals can be connected to other signals. When the first is emitted, the second signal is also emitted.

Let's look at what happens when the user operates the slider: The slider sees that its value has changed, and emits the valueChanged() signal. That signal is connected both to the display() slot of the QLCDNumber and to the valueChanged() signal of the LCDRange.

Thus, when the signal is emitted, LCDRange emits its own valueChanged() signal. In addition, QLCDNumber::display() is called and shows the new number.

Note that you're not guaranteed any particular order of execution - LCDRange::valueChanged() may be emitted before or after QLCDNumber::display(), entirely arbitrarily.

The implementation of value() is straight forward, it simply returns the slider's value.

The implementation of setValue() is equally straight forward. Note that since the slider and LCD number are connected, setting the slider's value automatically updates the LCD number as well. In addition, the slider will automatically adjust the value if it is outside its legal range.

main.cpp

        LCDRange *previous = 0;
        for( int r = 0 ; r < 4 ; r++ ) {
            for( int c = 0 ; c < 4 ; c++ ) {
                LCDRange* lr = new LCDRange( grid );
                if ( previous )
                    connect( lr, SIGNAL(valueChanged(int)),
                             previous, SLOT(setValue(int)) );
                previous = lr;
            }
        }

All of main.cpp is copied from the previous chapter, except that in the constructor for MyWidget, when we create the 16 LCDRange object we now connect them together using the signal/slot mechanism. Each one has its valueChanged() signal connected to the setValue() slot in the previous one. Since LCDRange emits the signal valueChanged() when its value changes (surprise!), we are here creating a "chain" of signals and slots.

Behavior

On startup, the program's appearance is identical to the previous one. Try operating the slider to the bottom right...

Excercises

Use the bottom right slider to set all LCDs to 30. Then set the top half to 29 by using the rightmost slider on the 2nd row. Now, use the one to the left of the last one operated to set the first seven LCDs back to 30. Click on the left arrow on the bottom right slider. What happens? Why is this the correct behavior?

You may now go on to chapter eight.

[Previous tutorial] [Next tutorial] [Main tutorial page]