The Open IoT SDK Unit Tests Application executes all supported unit tests on the target.
The Matter unit tests are included in a set of libraries and allow to validate most of the components used by Matter examples applications. The main goal of this application is to run registered tests on Open IoT SDK target and check the results. The final result is the number of tests that failed.
The required environment is the same as for the Matter examples. For information on how to setup it see Open IoT SDK examples.
The build process means creating separate executable file for each Matter tested component. It assumes the use of all supported test libraries and creating independent applications from them.
You build using a vscode task or call the script directly from the command line.
Command Palette (F1) => Run Task... => Build Open IoT SDK unit-tests => (debug on/off)
This will call the scripts with the selected parameters.
You can call the script directly yourself.
${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh unit-tests
Use --help to get more information about the script options.
Unit-tests applications can be run independently or as an entire set. It runs in the background and opens a telnet session. The script will open telnet for you and connect to the port used by the FVP. When the telnet process is terminated it will also terminate the FVP instance.
You can run the application script from a vscode task or call the script directly.
Expected output of each executed test:
[ATM] Open IoT SDK unit-tests start [ATM] Open IoT SDK unit-tests run... ... [ATM] Test status: 0
Command Palette (F1) => Run Task... => Run Open IoT SDK unit-tests => <test name> or all (to run all tests)
You can call the script directly yourself.
${MATTER_ROOT}/scripts/examples/openiotsdk_example.sh -C run unit-tests <test name> (optional to run specific test)
Debugging can be started using a VS code launch task:
Run and Debug (Ctrl+Shift+D) => Debug Open IoT SDK unit-tests application => Start Debugging (F5) => <test name> => <test name>
As you can see above, you will need to select the name of the test twice. This is because the debug task needs to launch the run task and currently VS code has no way of passing parameters between tasks.