Creating Proper Test Cases

[chamilbuddhima]

No description provided.

[chamilbuddhima]

Framework for Test Cases

Objectives

• Ensure all parts of the pipeline function correctly under various conditions.
• Handle unexpected issues ( e.g. - sensor failures, missing data) without crashes, by recovering if able or giving error messages.

Key areas for testing

• Data synchronization- Verify camera and LiDAR data align correctly.
• Realtime functionality - Make sure that system can handle live data without delays.
• Data export - Check whether results are saved in the correct formats (e.g. - depth maps, point clouds) ,Check whether the data is exported to the intended save location(File) with proper naming.
• Edge cases – Test how the system recover from problems like sensor dropouts or corrupted data.

Implementation plan

• Design test cases - Write tests for normal, stress, and failure scenarios.
• Automate testing - Use scripts to run tests and log results.
• Record test details and results in a shared repository on GitHub.

[chamilbuddhima]

Choosing a Testing Framework

Decided to go with PyTest Framework after comparing it with Unittest.

Reasons :
• PyTest can discover tests automatically based on the filenames while unittest requires creation of test cases and proper naming conventions.
• When a test fails on PyTest it gives a more detailed/informative report why the test failed.
• Pytest’s built-in support for parameterized testing allows efficient testing of multiple input combinations, significantly reducing redundancy. This feature is not natively available in unittest and requires additional setup.
• PyTest includes numerous built-in conveniences like monkeypatching (modifying code at runtime for testing), temporary directory handling, and more, often reducing the need for external libraries.

[chamilbuddhima]

PyTest examples

1.Data export
Verify that results are saved in the correct format and location.

• Normal case-Test if the function saves a file in the correct format and to the specified directory with the expected name.
• Failure case- Test if the function raises an error when trying to save in an unsupported file format.

2.Real-Time data handling
Test the ability to process live data streams without delays.

• Normal case -Simulate live data with acceptable processing delays. The function should process data without raising an error.
• Failure case - Simulate live data with excessive delays . The function should raise a TimeoutError or handle the issue.

[chamilbuddhima]

Example 1
Code to test

PyTest code

Results after running

[chamilbuddhima]

Test case example for MVP

Verify Image Conversion
Ensure the RGB to BGR conversion logic works as expected.
•Send a sample Image message with known RGB values. Verify the output Image message has correctly converted BGR values.

Code to be tested - imageconversion.py