Main Content

Collect Code Testing Metrics Programmatically

This example shows how to programmatically assess the status and quality of code testing activities in a project. You can assess the testing status of a unit by using the metric API to collect metric data on the test status and coverage results. After collecting metric results, you can access the results or export them to a file. By running a script that collects these metrics, you can automatically analyze the testing status of your project to, for example, design a continuous integration system. Use the results to monitor testing completeness or to detect downstream testing impacts when you make changes to artifacts in the project.

Run Model and Code Tests

Open a project that contains models and testing artifacts. For this example, in the MATLAB® Command Window, enter:

openExample("slcheck/ExploreTestingMetricDataInModelTestingDashboardExample");
openProject("cc_CruiseControl");
The example project contains models and tests for the models. The example project also has the project setting Track tool outputs to detect outdated results enabled. Before you programmatically collect metrics, make sure that the Track tool outputs to detect outdated results setting is enabled in the Simulink section of your project settings. For information, see Monitor Artifact Traceability and Detect Outdated Results with Digital Thread.

For this example, run a test case in Normal mode and then in Software-in-the-loop (SIL) mode by entering:

cp = currentProject;
rf = cp.RootFolder;
tf = fullfile(rf,"tests","cc_DriverSwRequest_Tests.mldatx");
tfObj = sltest.testmanager.load(tf);
tsObj = getTestSuites(tfObj);
tcObj = getTestCases(tsObj);
tc3 = tcObj(3);
run(tc3,SimulationMode='Normal');
run(tc3,SimulationMode='Software-in-the-loop (SIL)');

Collect Metric Results for Software Units in Project

Create a metric.Engine object for the current project.

metric_engine = metric.Engine();

Update the trace information for metric_engine to reflect pending artifact changes and to track the test results.

updateArtifacts(metric_engine);

Create an array of metric identifiers for the metrics you want to collect. For this example, create a list of the metric identifiers used in the SIL Code Testing dashboard by specifying the Dashboard as "ModelUnitSILTesting". For more information, see getAvailableMetricIds.

metric_Ids = getAvailableMetricIds(metric_engine,...
App = "DashboardApp",...
Dashboard="ModelUnitSILTesting");
For a list of code testing metrics, see Code Testing Metrics.

Execute the metric engine to collect the metric results.

execute(metric_engine, metric_Ids);
By default, the metric engine collects results for each unit in the project. If you only want to collect results for a single unit, specify the unit using the ArtifactScope argument. For more information, see execute.

Access Results

After you collect metric results, you can access the results programmatically or generate a report for offline review.

View Test and Coverage Results Programmatically

To access the results programmatically, use the getMetrics function. The function returns the metric.Result objects that contain the result data for the specified unit and metrics. For this example, store the results for the metrics slcomp.sil.TestStatusDistribution and slcomp.sil.CoverageBreakdown in corresponding arrays.

results_silTests = getMetrics(metric_engine, "slcomp.sil.TestStatusDistribution");
results_silCoverage = getMetrics(metric_engine, "slcomp.sil.CoverageBreakdown");

The metric slcomp.sil.TestStatusDistribution returns a distribution of the number of tests that passed, failed, were disabled, or were untested. For this example, the distribution values for the unit cc_DriverSwRequest were in results_silTests(4).Value. The order of the units in your results might be different on your machine.

results_silTests(4).Value
ans = 

  struct with fields:

       BinCounts: [4×1 double]
        BinEdges: [4×1 double]
    OverallCount: 7
          Ratios: [4×1 double]
Use the disp function to display the bin counts of the distribution, which are fields in the Value field of the metric.Result object.
disp(['Unit:  ', results_silTests(4).Scope(1).Name])
disp([' ', num2str(results_silTests(4).Value.BinCounts(1)), ' SIL test FAILED'])
disp([' ', num2str(results_silTests(4).Value.BinCounts(2)), ' SIL tests PASSED'])
disp([' ', num2str(results_silTests(4).Value.BinCounts(3)), ' SIL tests DISABLED'])
disp([' ', num2str(results_silTests(4).Value.BinCounts(4)), ' SIL tests UNTESTED'])
Unit:  cc_DriverSwRequest
 1 SIL test FAILED
 0 SIL tests PASSED
 0 SIL tests DISABLED
 9 SIL tests UNTESTED
This result shows that for the unit cc_DriverSwRequest, one SIL test failed and six SIL tests remain untested. For code testing results to be compliant, each of the tests should have passed. If one or more tests are untested, disabled, or failed, those issues should be addressed before you analyze the code coverage results. For information on how to address the test failure in this example project, see Identify and Troubleshoot Gaps in Code Testing Results and Coverage.

The metric slcomp.sil.CoverageBreakdown returns the percentage of coverage achieved, justified, completed, or missed for each coverage type. For this example, the coverage results for the unit cc_DriverSwRequest were in results_silCoverage(4):

results_silCoverage(4).Value
ans = 

  struct with fields:

       Statement: [1×1 struct]
        Decision: [1×1 struct]
       Condition: [1×1 struct]
            MCDC: [1×1 struct]
        Function: [1×1 struct]
    FunctionCall: [1×1 struct]
You can access the results for each coverage type using the fields in the Value field of the metric.Result object. For example, to access the SIL decision coverage results for the unit:
decisionCoverage = results_silCoverage(4).Value.Decision
decisionCoverage = 

  struct with fields:

               Achieved: 48.4848
              Justified: 0
                 Missed: 51.5152
    AchievedOrJustified: 48.4848
This result shows that for the unit cc_DriverSwRequest, so far the SIL tests achieved 48% of decision coverage and missed 52% of decision coverage. For code coverage results to be compliant, 100% of test results for each coverage type should either be achieved or justified. When the coverage results for each coverage type have either been achieved or justified, the dashboard considers the coverage to be completed.

Generate Report for Offline Review

Generate a report file that contains the SIL testing results for each of the units in the project. For this example, specify the HTML file format, use pwd to provide the path to the current folder, and name the report "SILResultsReport.html".

reportLocation = fullfile(pwd, "SILResultsReport.html");
generateReport(metric_engine,Type="html-file",Location=reportLocation,...
App = "DashboardApp", Dashboard = "ModelUnitSILTesting");
The generated SIL Code Testing Report opens automatically.

To open the table of contents and navigate to results for each unit, click the menu icon in the top-left corner of the report. For each unit in the report, there is a summary of the SIL test results and coverage results.

SIL test statuses and aggregated coverage results

For more information on the report, see generateReport.

Saving the metric results in a report file allows you to access the results without opening the project and the dashboard. Alternatively, you can open the dashboard to see the results and explore the artifacts. In the MATLAB Command Window, enter:

modelTestingDashboard
In the Add Dashboard section of the toolstrip, click SIL Code Testing to open the SIL Code Testing dashboard.

See Also

| | | |

Related Topics