Question 1

What is the difference between unit tests and integration tests?

Accepted Answer

Unit tests exercise a small unit (a function or class) in isolation from external dependencies and run fast, while integration tests exercise several real parts working together, such as code with a database or HTTP.

Question 2

What is the role of fixtures in pytest?

Accepted Answer

A fixture prepares a dependency or data a test needs, which the test requests by name as a parameter, and it controls its own lifetime (scope) and cleanup after yield.

Question 3

How do you test many cases without repeating code?

Accepted Answer

Use pytest.mark.parametrize to run the same test over several inputs and expected results, so each case appears as an independent test in the report.

Question 4

How do you use mock and patch correctly?

Accepted Answer

Apply patch to the name where the code under test uses it, not where it is defined, and use autospec so the mock fails if it is called with the wrong signature.

Question 5

How do you test asynchronous (async) code?

Accepted Answer

Use IsolatedAsyncioTestCase in unittest, or a plugin such as pytest-asyncio or anyio in pytest, so the test runs inside an event loop and the coroutine is actually awaited.

Question 6

How do you verify that a function raises the correct exception?

Accepted Answer

Use pytest.raises (or assertRaises in unittest) as a context manager around only the line expected to fail, and check the type and the important fields or part of the message through match.

Question 7

How do you keep time and randomness from making tests flaky?

Accepted Answer

Pass the clock and the source of randomness as replaceable dependencies instead of calling datetime.now() or random directly inside domain logic, then pin them in the test.

Question 8

How do you test HTTP code without relying on a real network?

Accepted Answer

Test the client through a fake transport or fake server, simulating failure cases (timeouts, 5xx, invalid payloads), then add a limited number of contract or integration tests against the real service.

Question 9

Does a high coverage percentage mean the code is well tested?

Accepted Answer

No; coverage measures the lines or branches that ran during the tests, not the quality of the assertions or the coverage of risky cases. It is a useful signal but not a goal in itself.

Question 10

What is a contract test and when is it useful?

Accepted Answer

A contract test verifies that two parties (such as a service and its client, or an adapter and its interface) agree on the shape and semantics of the data, without running the whole system end to end.