unittest_mock

unittest.mock

Replace real dependencies with fakes during tests — verify interactions without hitting DB/API.

Files

File	Description
example.py	MagicMock, patch, side_effect, assertions

---

Descriptive Example

Scenario

Test fetch_user without a real database by injecting a mock.

from unittest.mock import MagicMock

def fetch_user(user_id, db):
    return db.get(user_id)

mock_db = MagicMock()
mock_db.get.return_value = {"id": 1, "name": "Alice"}

user = fetch_user(1, mock_db)
print(user)                              # {'id': 1, 'name': 'Alice'}
mock_db.get.assert_called_once_with(1)   # verify interaction

Patching where name is used

from unittest.mock import patch

with patch("builtins.print") as mock_print:
    print("hello")
    mock_print.assert_called_once_with("hello")

Patch the namespace where the name is looked up, not where it's defined.

---

Interview Q&A

Q1: Mock vs stub vs fake?
A: Mock: verifies interactions (was it called? with what args?). Stub: returns canned data, no verification. Fake: working lightweight implementation (in-memory DB).

Q2: What is MagicMock?
A: Auto-creates attributes and methods on access. Records all calls. Returns another MagicMock for chained access.

Q3: Where should you patch?
A: Where the name is looked up (import site), not where defined. If mymodule does from requests import get, patch mymodule.get.

Q4: What is side_effect?
A: Instead of fixed return_value, can be: an exception to raise, an iterable of return values, or a callable for dynamic behavior.

Q5: @patch as decorator vs context manager?
A: Both work. Decorator injects mock as extra argument. Context manager scopes the patch to a block.

Q6: How to mock async functions?
A: AsyncMock from unittest.mock (Python 3.8+). mock_coro = AsyncMock(return_value=result).

---

Run

python3 example.py