Parameterized Tests¶
Exercise 1: Sets of example data¶
The tests in test_parameterized.py check a list of pairs (word, count) that apply to the text file mobydick_summary.txt:
```python
PAIRS = [
('whale', 5),
('goldfish', 0),
('captain', 4),
('white', 2),
('jellyfish', 99),
('harpoon', 1),
]
```
Run the tests and see what happens.
Exercise 2: Fixing the test¶
Fix the test by using the count.get(word) method instead of [word].
Run the test again. What happens?
Exercise 3: Parameterize¶
There is a deeper issue here. We want all test examples to be independent. We will create six tests from the example data.
Use the test parametrization in pytest. Change the test function by adding the following decorator:
import pytest @pytest.mark.parametrize('word, number', PAIRS) def test_count_words_dict(word, number): ...
The two arguments will be filled in automatically.
Now remove the for loop.
Run the test. You should see six tests instead of one. Make sure all six pass.
Exercise 2: Write another parameterized test¶
Use test parametrization to test the count_words() function on the following test cases.
| text | expected result |
|---|---|
| Call me Ishmael | 3 |
| . | 0 |
| 1234 | 1 |
| x | 1 |
| We are the Borg - Resistance is futile | 7 |
Write one parameterized test that checks these five examples.