History

Mikolaj Boc 1f6cac0da9 Make WASM export names different across modules The export name is now ${TARGET_NAME}Entry. This can also be overridden by using QT_WASM_EXPORT_NAME, both in CMake and qmake Change-Id: I59c97ae6e22f0b2720716e9d7eff7b6b13d37ab5 Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>		2023-06-20 09:08:03 +02:00
..
qtwasmtestlib.cpp	wasm: add "skip" support to qtwasmtestlib	2022-08-17 04:52:46 +02:00
qtwasmtestlib.h	wasm: add "skip" support to qtwasmtestlib	2022-08-17 04:52:46 +02:00
qtwasmtestlib.js	Make WASM export names different across modules	2023-06-20 09:08:03 +02:00
README.md	Make WASM export names different across modules	2023-06-20 09:08:03 +02:00

README.md

QtWasmTestLib - async auto tests for WebAssembly

QtWasmTestLib supports auto-test cases in the web browser. Like QTestLib, each test case is defined by a QObject subclass with one or more test functions. The test functions may be asynchronous, where they return early and then complete at some later point.

The test lib is implemented as a C++ and JavaScript library, where the test is written using C++ and a hosting html page calls JavaScript API to run the test.

Implementing a basic test case

In the test cpp file, define the test functions as private slots. All test functions must call completeTestFunction() exactly once, or will time out otherwise. Subsequent calls to completeTestFunction will be disregarded. It is advised to use QWASMSUCCESS/QWASMFAIL for reporting the test execution status and QWASMCOMPARE/QWASMVERIFY to assert on test conditions. The call can be made after the test function itself has returned.

class TestTest: public QObject
{
    Q_OBJECT
private slots:
    void timerTest() {
        QTimer::singleShot(timeout, [](){
            completeTestFunction();
         });
    }
};

Then define a main() function which calls initTestCase(). The main() function is async too, as per Emscripten default. Build the .cpp file as a normal Qt for WebAssembly app.

int main(int argc, char **argv)
{
    auto testObject = std::make_shared<TestTest>();
    initTestCase<QCoreApplication>(argc, argv, testObject);
    return 0;
}

Finally provide an html file which hosts the test runner and calls runTestCase()

<!doctype html>
<script type="text/javascript" src="qtwasmtestlib.js"></script>
<script type="text/javascript" src="test_case.js"></script>
<script>
    window.onload = async () => {
        runTestCase(entryFunction, document.getElementById("log"));
    };
</script>
<p>Running Foo auto test.</p>
<div id="log"></div>

Implementing a GUI test case

This is similar to implementing a basic test case, with the difference that the hosting html file provides container elements which becomes QScreens for the test code.

<!doctype html>
<script type="text/javascript" src="qtwasmtestlib.js"></script>
<script type="text/javascript" src="test_case.js"></script>
<script>
    window.onload = async () => {
        let log = document.getElementById("log")
        let containers = [document.getElementById("container")];
        runTestCase(entryFunction, log, containers);
    };
</script>
<p>Running Foo auto test.</p>
<div id="container"></div>
<div id="log"></div>