I need to process and close QFileDialog in tests of the application.
The dialog invoked by:
QFileDialog::getOpenFileName( ... );
In the test I 'catch' this dialog by:
QApplication::topLevelWidgets();
Then I choose need file by QFileDialog API.
And now I should close the dialog to get valid answer (filename) from QFileDialog::getOpenFileName();
But QDilaog's slots has no effect (accept(), reject(), close()... ). The dialog stays opened.
The solution described here works, but it is no option in my case. I must to use standard dialogs.
Is there a way to close it properly?
QFileDialog::getOpenFileName is a static method, so you're limited by what you can do with it.
If you want more control, I suggest creating an instance of QFileDialog and using that instead. By calling the instance's close() function, you can programmatically close the dialog.
In response to the comment that this doesn't work, here's example code:-
// Must create the FileDialog on the heap, so we can call close and the dialog is deleted
// Set the Qt::WA_DeleteOnClose flag if the instance is still required
QFileDialog* fileDlg = new QFileDialog(this, QString("Select Config file"), QDir::homePath(), QString("Config (*.xml)"));
// One shot timer to close the dialog programmatically
QTimer *timer = new QTimer(this);
timer->setSingleShot(true);
connect(timer, &QTimer::timeout, [=]()
{
fileDlg->close();
timer->deleteLater();
} );
timer->start(3000);
fileDlg->exec();
In order for to display a native dialog, you have to run exec() or call one of the static functions.
Unfortunately, in Windows, this calls a blocking function in the Windows API making the displayed dialog modal, running it's own event loop. Without returning to the Qt event loop, you cannot execute the close() function using the signals/slots interface.
I tried to bypass this by calling the close() function directly from another thread, but this results in Qt trying to send an event to the underlying dialog. Since sending (as opposed to posting) events across thread boundaries is not allowed in Qt, a fatal error is generated.
So, it seems for Windows at least, this is not possible.
I have not tested on platforms other than Windows. The code I used was:
int main(int argc, char *argv[])
{
QApplication a(argc, argv);
QFileDialog* fileDlg = new QFileDialog(0, QString("Select Config file"), QDir::homePath(), QString("Config (*.xml)"));
// spawn a new thread
QtConcurrent::run([=](){
QTimer timer;
timer.setSingleShot(true);
QEventLoop *loop = new QEventLoop;
QObject::connect(&timer, &QTimer::timeout, [=](){
fileDlg->close();
fileDlg->deleteLater();
loop->quit();
});
timer.start(3000);
loop->exec();
delete loop;
});
fileDlg->exec();
return a.exec();
}
In Windows you can use WinAPI to close the dialog:
#define WAIT(A) while (!(A)) {}
HWND dialogHandle, button;
WAIT(dialogHandle = FindWindow(NULL, L"Open")); //write here title of the dialog
WAIT(button = FindWindowEx(dialogHandle, NULL, L"Button", L"&Open")); //write here title of the button to click
SendMessage(button, BM_CLICK, 0, 0);
Related
I have written a QT code which starts a new process on a button click, that process have to execute a shell script and append the Std output/error on text Browser dynamically based upon the script result. The code gets failed in the custom Slot. This is what my window.h is
class Window : public QWidget
{
Q_OBJECT
public:
explicit Window(QWidget *parent = 0);
QPushButton *goButton;
QTextBrowser *statusWindow;
QProcess *process;
private slots:
void go_Button_Clicked();
void updateOutput();
};
And this is how my window.cpp is
Window::Window(QWidget *parent) : QWidget(parent)
{
// Text Browser
statusWindow = new QTextBrowser(this);
// Create the Go button, make "this" the parent
goButton = new QPushButton(this);
connect(goButton, SIGNAL (clicked()), this, SLOT (go_Button_Clicked()));
}
void Window::go_Button_Clicked()
{
// Step1 - Create a new process
QProcess *process = new QProcess(this);
// Step2 - Connect the Signals and slot
qDebug() << connect(process, SIGNAL(readyReadStandardOutput()), this, SLOT(updateOutput()));
// Step3 - Start the Script process
process->start("/home/root/script.sh");
qDebug() << "Process in main";
}
void Window::updateOutput()
{
qDebug() << "Process in update Output";
//statusWindow->append(process->readAllStandardOutput());
}
So whenever i uncommented lines in update Output the GUI crashes as soon as i hit the button. using qdebug i managed to find that the GUI crashes bcoz of the line "statusWindow->append(process->readAllStandardOutput());" .
If the line is commented the debug message keeps on printing on console, but with the un-commented line i get the debug message once and then GUI crashes.Here is the debug output.
true
true
Process in main
Process in update Output
Process killed by signal
Any idea whats going on here, and i am familiar with Debugging in QT
Replace:
QProcess *process = new QProcess(this);
by
process = new QProcess(this);
You are using uninitialized member variable process, because in Window::go_Button_Clicked() you are creating a local variable.
Edit
Actually, the code is very error prone. What happens, when the user presses the button multiple times. Then in the slot you may read the the output from wrong process.
One workaround could be: You don't define QProcess as member at all, but do it like now - as a local variable. Then in the slot, you can cast the sender to QProcess*, and unless it fails use this instance. It will be always the right one. Afterwards, don't forget to delete the sender. Use deleteLater() for this.
Inside your void Window::go_Button_Clicked() slote you hide QProcess *process declared at .h file with new variable
QProcess *process = new QProcess(this);
Replace this line with
process = new QProcess(this);
But when you click second time you'll get memory leak, and can not get more data from first process. So you chould change you design somehow.
Not much to add; what is the equivalent of JavaScript's setTimeout on qtScript?
Here's how you can extend your script language, by providing a self-contained C++ method (no need for bookkeeping of timer ids or so). Just create the following slot called "setTimeout":
void ScriptGlobalObject::setTimeout(QScriptValue fn, int milliseconds)
{
if (fn.isFunction())
{
QTimer *timer = new QTimer(0);
qScriptConnect(timer, SIGNAL(timeout()), QScriptValue(), fn);
connect(timer, SIGNAL(timeout()), timer, SLOT(deleteLater()));
timer->setSingleShot(true);
timer->start(milliseconds);
} else
context()->throwError(tr("Passed parameter '%1' is not a function.").arg(fn.toString()));
}
And introduce that slot as function to the global object of your script engine. This can be done in different ways, e.g. just creating a QScriptValue-function via the QScriptEngine instance, and setting an accordingly named property on the engine's existing global object. In my case however, the entire ScriptGlobalObject instance is set as new global object, like this:
mScriptGlobalObject = new ScriptGlobalObject(this);
engine->setGlobalObject(engine->newQObject(mScriptGlobalObject));
Note that if you want to use "context()" as shown in the setTimeout code above, your ScriptGlobalObject should derive also from QScriptable, like this:
class ScriptGlobalObject : public QObject, protected QScriptable
In the script you can now use setTimeout to have a method invoked at a later time (as long as the QScriptEngine instance from which it originates isn't deleted in the meantime):
setTimeout(function() {
// do something in three seconds
}, 3000);
You can expose the QTimer as an instantiable class to the script engine. You can then instantiate it via new QTimer().
This is documented in Making Applications Scriptable.
Below is a complete example. The timer fires a second after the script is evaluated, prints timeout on the console, and exits the application.
// https://github.com/KubaO/stackoverflown/tree/master/questions/script-timer-11236970
#include <QtScript>
template <typename T> void addType(QScriptEngine * engine) {
auto constructor = engine->newFunction([](QScriptContext*, QScriptEngine* engine){
return engine->newQObject(new T());
});
auto value = engine->newQMetaObject(&T::staticMetaObject, constructor);
engine->globalObject().setProperty(T::staticMetaObject.className(), value);
}
int main(int argc, char ** argv) {
QCoreApplication app{argc, argv};
QScriptEngine engine;
addType<QTimer>(&engine);
engine.globalObject().setProperty("qApp", engine.newQObject(&app));
auto script =
"var timer = new QTimer(); \n"
"timer.interval = 1000; \n"
"timer.singleShot = true; \n"
"var conn = timer.timeout.connect(function(){ \n"
" print(\"timeout\"); \n"
" qApp.quit(); \n"
"}); \n"
"timer.start();\n";
engine.evaluate(script);
return app.exec();
}
setTimeout and setInterval are not defined in ECMAScript specification because they are not JavaScript features. These functions are part of browser environments. So, QTscript does not have them.
You can use QTimer to achive this functionality.
Here is a quick code how to use it in QTScript.
var timer = new QTimer();
timer.interval = 100; // set the time in milliseconds
timer.singleShot = true; // in-case if setTimout and false in-case of setInterval
timer.timeout.connect(this, function(){console("in setTimout")});
timer.start();
Watch out for any bugs, I just coded it here.
Qt provides an example in the 'context2d' project a way to access to the setInterval/clearInterval setTimeout/clearTimeout functionalities from the script.
In the 'Environment' class of this project, the startTimer function of a QObject is called each time that the script invokes setInterval/setTimeout.
Then the timer identifier is added in a QHash with a reference on the javascript callback.
When the countdown of the timer (c++ part) is over, the timerEvent of the 'Environment' class is called and then the javascript callback called..
Note that all the timers are killed in the Environment::reset() in order to clean the context.
see: https://doc.qt.io/archives/qt-4.7/script-context2d.html
Here's my submission for "kludge of the year"... but this works AND doesn't require recompiling the C++ back end, in a context where that is way too much work to get involved with! I'm not sure how the rest of you may be using the now defunk Qt Script, but my need is for the Qt Installer Framework, and I want to just use it out of the box - not via a custom fork of the whole tool set, for me to then try to maintain (since QtIFW itself is still actively updated), or have to port around, compile on alternate platforms, share with collaborators...
So what's my solution? Well there's no QTimer exposed the to standard script engine, but you can define your own custom (i.e. "dynamic") installer wizard pages, defining the interface via Qt forms (.ui files). With that, you can drop in any QWidget, and then add connections for the signals and slots on the Qt Script side... And so, I just used the first, and perhaps most simplistic, set of widget signals and slots which I saw had some sort of timer mechanism built in that I might capitalize upon.
In a custom form, drop this in somewhere, adding a hidden QPushButton:
<widget class="QPushButton" name="timerKludgeButton">
<property name="visible">
<bool>false</bool>
</property>
</widget>
Then, when you load the component containing the .ui (form), connect the button's "released" signal to a custom "slot" (i.e. Qt Script function):
Component.prototype.componentLoaded = function(){
var page = gui.pageWidgetByObjectName("DynamicMyTimerKludgePage");
page.timerKludgeButton.released.connect(this, this.onTimeOut);
}
Next, define the slot:
Component.prototype.onTimeOut = function(){
console.log("Asynchronous time out!");
}
Finally, where applicable, start your "timer" with the core trick here, using the QPushButton's "animateClick" function:
var page = gui.pageWidgetByObjectName("DynamicMyTimerKludgePage");
page.timerKludgeButton.animateClick(2000);
I'we started experimenting with unit testing in Qt and would like to hear comments on a scenario that involves unit testing signals and slots.
Here is an example:
The code i would like to test is (m_socket is a pointer to QTcpSocket):
void CommunicationProtocol::connectToCamera()
{
m_socket->connectToHost(m_cameraIp,m_port);
}
Since that is an asynchronous call i can't test a returned value. I would however like to test if the response signal that the socket emits on a successful connection (void connected ()) is in fact emitted.
I've written the test below:
void CommunicationProtocolTest::testConnectToCammera()
{
QSignalSpy spy(communicationProtocol->m_socket, SIGNAL(connected()));
communicationProtocol->connectToCamera();
QTest::qWait(250);
QCOMPARE(spy.count(), 1);
}
My motivation was, if the response doesn't happen in 250ms, something is wrong.
However, the signal is never caught, and I can't say for sure if it's even emitted. But I've noticed that I'm not starting the event loop anywhere in the test project. In the development project, the event loop is started in main with QCoreApplication::exec().
To sum it up, when unit testing a class that depends on signals and slots, where should the
QCoreApplication a(argc, argv);
return a.exec();
be run in the test environment?
I realize this is an old thread but as I hit it and as others will, there is no answer and the answer by peter and other comments still miss the point of using QSignalSpy.
To answer you original question about "where the QCoreApplication exec function is needed", basically the answer is, it isn't. QTest and QSignalSpy already has that built in.
What you really need to do in your test case is "run" the existing event loop.
Assuming you are using Qt 5:
http://doc.qt.io/qt-5/qsignalspy.html#wait
So to modify your example to use the wait function:
void CommunicationProtocolTest::testConnectToCammera()
{
QSignalSpy spy(communicationProtocol->m_socket, SIGNAL(connected()));
communicationProtocol->connectToCamera();
// wait returns true if 1 or more signals was emitted
QCOMPARE(spy.wait(250), true);
// You can be pedantic here and double check if you want
QCOMPARE(spy.count(), 1);
}
That should give you the desired behaviour without having to create another event loop.
Good question. Main issues I've hit are (1) needing to let app do app.exec() yet still close-at-end to not block automated builds and (2) needing to ensure pending events get processed before relying on the result of signal/slot calls.
For (1), you could try commenting out the app.exec() in main(). BUT then if someone has FooWidget.exec() in their class that you're testing, it's going to block/hang. Something like this is handy to force qApp to exit:
int main(int argc, char *argv[]) {
QApplication a( argc, argv );
//prevent hanging if QMenu.exec() got called
smersh().KillAppAfterTimeout(300);
::testing::InitGoogleTest(&argc, argv);
int iReturn = RUN_ALL_TESTS();
qDebug()<<"rcode:"<<iReturn;
smersh().KillAppAfterTimeout(1);
return a.exec();
}
struct smersh {
bool KillAppAfterTimeout(int secs=10) const;
};
bool smersh::KillAppAfterTimeout(int secs) const {
QScopedPointer<QTimer> timer(new QTimer);
timer->setSingleShot(true);
bool ok = timer->connect(timer.data(),SIGNAL(timeout()),qApp,SLOT(quit()),Qt::QueuedConnection);
timer->start(secs * 1000); // N seconds timeout
timer.take()->setParent(qApp);
return ok;
}
For (2), basically you have to coerce QApplication into finishing up the queued events if you're trying to verify things like QEvents from Mouse + Keyboard have expected outcome. This FlushEvents<>() method is helpful:
template <class T=void> struct FlushEvents {
FlushEvents() {
int n = 0;
while(++n<20 && qApp->hasPendingEvents() ) {
QApplication::sendPostedEvents();
QApplication::processEvents(QEventLoop::AllEvents);
YourThread::microsec_wait(100);
}
YourThread::microsec_wait(1*1000);
} };
Usage example below.
"dialog" is instance of MyDialog.
"baz" is instance of Baz.
"dialog" has a member of type Bar.
When a Bar selects a Baz, it emits a signal;
"dialog" is connected to the signal and we need to
make sure the associated slot has gotten the message.
void Bar::select(Baz* baz) {
if( baz->isValid() ) {
m_selected << baz;
emit SelectedBaz();//<- dialog has slot for this
} }
TEST(Dialog,BarBaz) { /*<code>*/
dialog->setGeometry(1,320,400,300);
dialog->repaint();
FlushEvents<>(); // see it on screen (for debugging)
//set state of dialog that has a stacked widget
dialog->setCurrentPage(i);
qDebug()<<"on page: "
<<i; // (we don't see it yet)
FlushEvents<>(); // Now dialog is drawn on page i
dialog->GetBar()->select(baz);
FlushEvents<>(); // *** without this, the next test
// can fail sporadically.
EXPECT_TRUE( dialog->getSelected_Baz_instances()
.contains(baz) );
/*<code>*/
}
I had a similar issue with Qt::QueuedConnection (event is queued automatically if the sender and the receiver belongs to different threads). Without a proper event loop in that situation, the internal state of objects depending on event processing will not be updated. To start an event loop when running QTest, change the macro QTEST_APPLESS_MAIN at the bottom of the file to QTEST_MAIN. Then, calling qApp->processEvents() will actually process events, or you can start another event loop with QEventLoop.
QSignalSpy spy(&foo, SIGNAL(ready()));
connect(&foo, SIGNAL(ready()), &bar, SLOT(work()), Qt::QueuedConnection);
foo.emitReady();
QCOMPARE(spy.count(), 1); // QSignalSpy uses Qt::DirectConnection
QCOMPARE(bar.received, false); // bar did not receive the signal, but that is normal: there is no active event loop
qApp->processEvents(); // Manually trigger event processing ...
QCOMPARE(bar.received, true); // bar receives the signal only if QTEST_MAIN() is used
I have a Qt application on Linux.
I'd like to program custom keyboard shortcuts such as CTRL-Q which will then call a subroutine which quits the program.
How can I do this?
Try this:
new QShortcut(QKeySequence(Qt::CTRL + Qt::Key_Q), this, SLOT(close()));
You can create it in the contructor of your form. This allows to avoid polluting your class with a pointer to access the shortcut. You may still want to add a pointer to the shortcut if you want to access it later on. The shortcut will be deleted when the application exits, since it is parented to it. It automatically does the connection, you don't have to do it manually.
Also note that there is no default Ctrl+Q sequence on Windows, but there is one on Linux and MacOS.
Since CTRL-Q may have a menu item or toolbar entry, too, I think you're looking for QAction.
See this:
http://doc.qt.digia.com/4.6/qaction.html#shortcut-prop
LE:
Example of QAction at work:
QAction *foo = new QAction(this);
foo->setShortcut(Qt::Key_Q | Qt::CTRL);
connect(foo, SIGNAL(triggered()), this, SLOT(close()));
this->addAction(foo);
Just create a new Qt GUI project (I used QtCreator) and add that code to the main window's constructor and it should work as expected.
Please note that there is no need of freeing the memory since the Qt framework will take care of that when the app closes.
For the modern Qt (5.9 as of now):
void MainWindow::bootStrap()
{
// create shortcut
QShortcut *shortcut = new QShortcut(QKeySequence(Qt::CTRL + Qt::Key_Q), this);
// connect its 'activated' signal to your function 'foo'
QObject::connect(shortcut, &QShortcut::activated,
this, &MainWindow::foo);
}
// somewhere in the code, define the function 'foo'
void MainWindow::foo()
{
qDebug() << "Ctrl+Q pressed.";
}
Don't forget to #include <QShortcut>.
Further info: http://doc.qt.io/qt-5/qshortcut.html
this is a sample to create file menu and exit action and connection between signal and slot.
QMenu *fileMenu = new QMenu(trUtf8("&File"));
QAction *actionExit = new QAction(tr("E&xit"));
//set "ctrl+q shortcut for exit action
actionExit->setShortcut(tr("CTRL+Q"));
//connect triggered signal of actionExit to close slot
connect(actionExit, SIGNAL(triggered()), this, SLOT(close()));
//add actionExit into file menu
fileMenu->addAction(actionExit);
I'm new to Qt, and want to simply display a video in Qt GUI. I basically got everything figured out, except for some details handling the QThread, which is really annoying.
I reformulate my question into a simpler program, hope it will explains better
first I define this QObject class
#include <QObject>
#include <QDebug>
class myObject : public QObject
{
Q_OBJECT
public:
explicit myObject(QObject *parent = 0);
bool stop;
signals:
void finishWork();
public slots:
void dowork();
void onfinishThread();
};
myObject::myObject(QObject *parent) :
QObject(parent)
{
stop = true;
}
void myObject::dowork(){
qDebug()<<"start working!";
while(!stop){
qDebug()<<"working...";
}
emit finishWork();
qDebug()<<"finish do work!";
}
void myObject::onfinishThread(){
qDebug()<<"thread is finished!";
}
then the main function
#include <QCoreApplication>
#include <QThread>
#include <iostream>
#include <windows.h>
#include "myobject.h"
using namespace std;
int main(int argc, char *argv[])
{
QCoreApplication a(argc, argv);
myObject* myObj = new myObject();
QThread* myThread = new QThread;
myThread->connect(myObj, SIGNAL(finishWork()), SLOT(quit()));
myObj->connect(myThread, SIGNAL(started()), SLOT(dowork()));
myObj->connect(myThread, SIGNAL(finished()), SLOT(onfinishThread()));
myObj->moveToThread(myThread);
myObj->stop = false;
myThread->start();
cout<<"Press ENTER to continue....."<<endl<<endl;
cin.ignore(1);
myObj->stop = true;
Sleep(10);
if(myThread->isRunning()){
qDebug()<<"The thread is still running?!!!";
}
/*
myObj->stop = false;
Sleep(1000);
myThread->start();
myObj->stop = true;
*/
myObj->deleteLater();
myThread->deleteLater();
return a.exec();
}
As you can see, I even used cin to try let the dowork() run first, but it didn't work at all, the output is
so I'm really confused on how scheduling works for QThread...
Also, if you uncomment the part
/*
myObj->stop = false;
Sleep(1000);
myThread->start();
myObj->stop = true;
*/
the output is exactly the same! only stays a while after printing
The thread is still running?!!!
Would anyone help me with this? Thanks a lot. You may simply copy all the code and test it yourself.
-------------------------Original Question, bad explanation, please ignore....----------------------------------------
I made a videoObj Class with only one function to Query the frames, the function is defined as:
void videoObj::ProcessFrame(){
bool getframe;
qDebug()<<"get in ProcessFrame";
while(!stop_flag){
getframe = capture_.read(imgOriginal_);
if(!getframe){
qDebug()<<"Video End!";
break;
}
cv::cvtColor(imgOriginal_, imgOriginal_, CV_BGR2RGB);
QImage qimgOriginal((uchar*)imgOriginal_.data, imgOriginal_.cols, imgOriginal_.rows, imgOriginal_.step, QImage::Format_RGB888);
emit imgProcessed(qimgOriginal);
this->thread()->msleep(10);
//qDebug()<<"processing frames";
}
emit stopProcess();
qDebug()<<"Stop thread";
}
Basically above code is just query frames and whenever got one emit the
SIGNAL imgProcessed(qimgOriginal)
and whenever the stop_flag is set on, stop the while loop and emit the
SIGNAL stopProcess()
I use this class in the MainWindow Class, here is how I define the connection in the constructor:
MainWindow::MainWindow(QWidget *parent) :
QMainWindow(parent),
ui(new Ui::MainWindow)
{
ui->setupUi(this);
video_obj_ = new videoObj();
video_thread_ = new QThread;
this->connect(video_obj_, SIGNAL(imgProcessed(QImage)), SLOT(onImgProcssed(QImage))); \\this is for displaying the image
video_obj_->connect(video_thread_, SIGNAL(started()), SLOT(ProcessFrame()));
video_obj_->moveToThread(video_thread_);
video_thread_->connect(video_obj_, SIGNAL(stopProcess()), SLOT(quit()));
}
Above code works fine in frame query. The problem I don't understand is, if I set video_obj_->stop_flag on in any MainWiow member function, the ProcessFrame() in videoObj Class should emit stopProcess() signal and trigger quit() of video_thread_, and then the thread should finish, that is video_thread_->finished() is true.
However, if I do something like:
connect(video_thread_, SIGNAL(finished()), this, SLOT(onStopProcess())); //onStopProcess() see below
void MainWindow::on_btnStart_clicked()
{
video_obj_->stop_flag = 1;
this->thread()->msleep(10);
video_obj_->capture_.open(ui->lineEditVidoeAddress->text().toStdString());
//... something about videoCapture setting here, not related
video_obj_->capture_.set(CV_CAP_PROP_POS_FRAMES, 0);
video_obj_->stop_flag = 0;
this->thread()->msleep(10);
if(video_thread_->isRunning()){
qDebug()<<"The thread is still running?!!!";
}
video_thread_->start();
}
void MainWindow::onStopProcess(){
qDebug()<<"thread is finished";
}
It will give me the output:
Stop thread
The thread is still running?!!!
thread is finished
Which means triggering the quit() is not finish the thread, or quit() has not been triggered.
If I use:
video_thread_->wait(); //instead of video_thread_->wait(10);
The program will just freeze.
Is anyone can help me with this, it really confuse me about this quit() and finished()...
Thanks!
Since when you call stop_flag=1, the video_thread_ finishes the current function which is
ProcessFrame
before ProcessFrame finish, quit is called on the video_thread_ through emit stopProcess().
However, quit is different from terminate, terminate can exit the thread any time (but it is not safe), quit works with the event loop, if the thread has no eventloop, quit has no effect.
I guess before qthread execute the next event in the eventloop, it checks some flag, which can be set by quit, it the flag is set by quit, then it won't execute the next event in the eventloop. Or it can also be that a quit event is inserted into the eventloop, and the next event in the eventloop will be quit.
After stop_flag = 1, you called video_thread_->wait, that will block the video_thread_ from executing the next event in the eventloop, thus the quit will not take effect before time out, however, the nextlines which print "not finished?!!!" is executed immediately. Instead of calling video_thread->wait, if you call currentThread()->Sleep(some_enough_time), then there will be time for the video_thread_ to execute the next event and quit.
You can read the Qt documentation of QThread, wait is used with terminate to terminate a thread synchronously.
============================== The new code ================================
When you do:
myObj->connect(myThread, SIGNAL(started()), SLOT(dowork()));
The signal source is myThread, the slot also belongs to myThread, since the object "myThread" is created in the main thread, thus as a object it lives in the main thread. You can call myThread->thread() to see this, it will return the main thread instead of the new thread.
However, the started signal is emitted from the new thread namely the thread that myThread represents, thus the connection is a Qt::QueuedConnection. dowork() is posted in the event queue of main thread, and it'll be executed only after a.exec() which executes the main threads eventloop.
The same thing happens to the other 2 connect calls, all the slots will be executed in the eventloop of the main thread.
First the start is emitted from the new thread when myThread->start is called, dowork is posted in the main thread's event queue.
Nothing really happens before you call a.exec();
So the program will go ahead to cin and then set stop to true, and then print "Thread is still running?!!".
Second When a.exec() is called, the dowork is executed in the main thread and "start work" is printed. Nothing is done because the stop flag is already true, and the finishwork signal is emitted from the main thread, print "finish do work";
Third The last step the finishwork is emitted, and the quit slot is directly called. However, before the new thread can really quit, the main thread has already finished the eventqueue, because no more events are posted to the main thread. The application exits without waiting for quit() to take effect.
To test this is true, do not modify any of your code, just add after
emit finishWork();
currentThread()->sleep(1000);
you will see "thread is finished!" printed, because this make time for the new thread to emit finished(), and the onfinishThread() will be add to the main thread's eventqueue.
BTW, your way of working with thread looks like java style, which is not the standard qt way. You can read this before you work on qt thread.
This is not a scheduling issue.
That you did in your code looks like:
Create a thread.
Then this thread emits a signal that it started, run slot dowork()
start a thread.
Wait for user input
echo about thread is running
execute event loop
At point 3 thread is already running and signalled about that. Because myObj is created in main thread and not moved to any other thread (to process events there), your thread does not do anything else now, but just spins event loop. At the same time event that tells you want to run myObj's dowork() is posted on the main thread. At last then it comes to step 6 you start to execute event loop of it, and first thing it finds is event that it needs to call dowork() on myObj.
To make it clear to you how Qt threads and signal-slot connection works, I recommend you to read this article on Qt blog.
In simple to fix it, you can move your object myObj to the thread that you wan't to run.
But to make this really correct, I bet that you really want is to subclass QRunnable and (re-)implement it's run method to do the stuff you wan't in QThread, so that thread does it's job, than finished correctly, so you can join on it. Or depending on your goal you might be even better with using QtConcurrent::run(aFunction)