I’m trying to figure out how to use pytest-randomly properly, but the problem is I always get the same set of random numbers for every test. All I did was install pytest-randomly (3.15.0) and run tests normally.

Here’s an example code:

import random


def generate_random_numbers(size):
    return [random.randint(1, 100) for _ in range(size)]


def test_random_numbers_size_1():
    numbers = generate_random_numbers(10)
    print(f"test_random_numbers_size 1.1: Generated numbers = {numbers}")
    numbers = generate_random_numbers(10)
    print(f"test_random_numbers_size 1.2: Generated numbers = {numbers}")
    assert len(numbers) == 10


def test_random_numbers_size_2():
    numbers = generate_random_numbers(10)
    print(f"test_random_numbers_size 2.1: Generated numbers = {numbers}")
    numbers = generate_random_numbers(10)
    print(f"test_random_numbers_size 2.2: Generated numbers = {numbers}")
    assert len(numbers) == 10

The result:

...
Using --randomly-seed=2523483435

test_example.py::test_random_numbers_size_2 PASSED                       [ 50%]
test_random_numbers_size 2.1: Generated numbers = [26, 52, 25, 95, 2, 69, 11, 94, 74, 48]
test_random_numbers_size 2.2: Generated numbers = [72, 41, 6, 27, 60, 28, 54, 48, 100, 76]

test_example.py::test_random_numbers_size_1 PASSED                       [100%]
test_random_numbers_size 1.1: Generated numbers = [26, 52, 25, 95, 2, 69, 11, 94, 74, 48]
test_random_numbers_size 1.2: Generated numbers = [72, 41, 6, 27, 60, 28, 54, 48, 100, 76]

As you can see the numbers are repeated for each test method. Shouldn’t they be different? Is it possible to change the pytest-randomly seed for each test method?
I’ve tried with some fixtures but nothing worked.

The docs say:

• Resets the global random.seed() at the start of every test case and test to a fixed number – this defaults to time.time() from the
  start of your test run, but you can pass in --randomly-seed to
  repeat a randomness-induced failure.