Why coverage measures test completeness in PyTest

Introduction

Coverage shows how much of your code is tested. It helps find parts not checked by tests.

You want to know if your tests check all important parts of your program.

You need to improve your tests to catch more bugs.

You want to avoid missing errors in code that is never tested.

You want to show your team or boss how complete your testing is.

You want to focus testing on parts of code that are rarely used.

Syntax

PyTest

pytest --cov=your_package tests/

This command runs tests and measures coverage for 'your_package'.

Replace 'your_package' with the folder or module you want to check.

Examples

Runs tests in 'tests/' and measures coverage for 'myapp' folder.

PyTest

pytest --cov=myapp tests/

Runs a single test file and measures coverage for 'module_name'.

PyTest

pytest --cov=module_name test_module.py

Shows coverage report with lines not covered in the terminal.

PyTest

pytest --cov=src --cov-report=term-missing tests/

Sample Program

This simple code has two functions: add and subtract. The test only checks add.

Running coverage will show subtract is not tested.

PyTest

def add(a, b):
    return a + b

def subtract(a, b):
    return a - b


def test_add():
    assert add(2, 3) == 5

# Note: No test for subtract function

Important Notes

Coverage does not guarantee your tests find all bugs, only that code runs during tests.

High coverage is good, but tests must also check correct behavior.

Use coverage reports to improve tests by adding missing cases.

Summary

Coverage measures how much code your tests run.

It helps find untested parts of your program.

Use coverage to improve test completeness and quality.

Practice

1. What does test coverage measure in pytest?

easy

A. How much of the code is executed by tests

B. How many tests are written

C. How fast the tests run

D. How many errors tests find

Solution

1. Step 1: Understand the meaning of coverage

   Coverage shows which parts of the code are run when tests execute.

2. Step 2: Compare options to coverage definition

   Only How much of the code is executed by tests matches this meaning, others describe different test aspects.

3. Final Answer:

   How much of the code is executed by tests -> Option A

4. Quick Check:

   Coverage = executed code percentage [OK]

Hint: Coverage = code run by tests, not test count [OK]

Common Mistakes:

• Confusing coverage with number of tests
• Thinking coverage measures test speed
• Believing coverage counts errors found

2. Which pytest command correctly runs tests with coverage measurement?

easy

A. pytest --cover

B. pytest --coverage

C. pytest -cov-report

D. pytest --cov

Solution

1. Step 1: Recall pytest coverage plugin syntax

   The correct flag to measure coverage is '--cov'.

2. Step 2: Check options for correctness

   Only pytest --cov uses the exact correct flag '--cov'. Others are invalid or incomplete.

3. Final Answer:

   pytest --cov -> Option D

4. Quick Check:

   Use --cov to enable coverage [OK]

Hint: Use '--cov' flag to measure coverage in pytest [OK]

Common Mistakes:

• Using '--coverage' instead of '--cov'
• Mixing coverage report flags with coverage run flags
• Typing '--cover' which is invalid

3. Given this pytest coverage output:

Name          Stmts   Miss  Cover
my_module.py     10      2    80%

What does the 'Miss' number mean?

medium

A. Number of lines not executed by tests

B. Number of errors in code

C. Number of tests skipped

D. Number of tests that failed

Solution

1. Step 1: Understand coverage report columns

   'Miss' shows how many lines of code were not run by tests.

2. Step 2: Match 'Miss' meaning to options

   Number of lines not executed by tests correctly describes 'Miss' as unexecuted lines; others describe unrelated test results.

3. Final Answer:

   Number of lines not executed by tests -> Option A

4. Quick Check:

   Miss = untested lines count [OK]

Hint: 'Miss' means lines tests did not run [OK]

Common Mistakes:

• Thinking 'Miss' counts failed tests
• Confusing 'Miss' with skipped tests
• Assuming 'Miss' means code errors

4. You run pytest with coverage but get 0% coverage report. What is the most likely cause?

medium

A. Tests passed too quickly

B. Tests did not execute any code

C. Coverage plugin is not installed

D. Code has no functions

Solution

1. Step 1: Analyze 0% coverage meaning

   0% coverage means no code lines were run during tests.

2. Step 2: Evaluate causes

   If the coverage plugin is not installed, pytest will not measure coverage and may silently produce 0% coverage report or no coverage data.

3. Final Answer:

   Coverage plugin is not installed -> Option C

4. Quick Check:

   Missing plugin causes no coverage data [OK]

Hint: 0% coverage often means coverage plugin missing [OK]

Common Mistakes:

• Assuming plugin missing causes 0% without errors
• Thinking fast tests mean low coverage
• Believing code without functions can't be covered

5. You want to improve test completeness using coverage. Which approach is best?

hard

A. Write more tests without checking coverage

B. Add tests targeting uncovered code lines shown by coverage report

C. Ignore coverage and focus on test speed

D. Remove tests that run covered code

Solution

1. Step 1: Understand coverage report use

   Coverage shows which code lines lack tests, guiding where to add tests.

2. Step 2: Evaluate options for improving completeness

   Only Add tests targeting uncovered code lines shown by coverage report uses coverage data to add tests for uncovered code, improving completeness.

3. Final Answer:

   Add tests targeting uncovered code lines shown by coverage report -> Option B

4. Quick Check:

   Use coverage to find and test missing code [OK]

Hint: Add tests where coverage report shows gaps [OK]

Common Mistakes:

• Writing tests blindly without coverage info
• Ignoring coverage to focus on speed
• Removing tests that cover code