| Build | Unit Test | Coverage |
|---|---|---|
 |
 |
 |
The Lab1 subproject includes a simple Makefile that can be used on Linux/WSL:
cd Lab1
make all # build the test binary
./Lab1 # run the unit tests
make coverage # Generate coverage report -> coverage_checksum.txt
cat coverage_checksum.txt # see percentage for checksum.c- Isolation of a section of code and validating its correctness
- Testing each modules behavior
- Involves breaking a program into pieces and subjecting each piece to a series of tests
- To verify the working of code
- Provides documentation of the code
- Easy debugging
- Bugs are found quicker and fixed at the early stages of SDLC process
- Improves design and quality of code
- Saves time and money
- Helps in refactoring the code for performance and resusability
- Simplifies the integration
- Code covergae can be measured
- It is done during the development stage of the project
- Software developers who write the code are resposible for performing Unit testing
- Some safety domain standards like DO178 require independence between those who write code andthose who write unit tests
- By following the AAA method:
- Arrange: Initialize the piece of the application to be tested
- Act: Invoke the units and capture the results
- Assert: Observe and retrospect the behaviour
- Methods of Unit testing:
- Manual
- Automation
- It takes more time to write code with Unit test cases, and we don’t have time for that
- In reality, it would save your development time in the long run
- Unit testing will find all bugs
- It won’t, as the intent of the Unit test is not to find bugs but develop robust software components that will have fewer defects in later stages of SDLC
- 100% code coverage means 100% test coverage
- This does not guarantee that code is error-free