strict_parametrize

[blackboxsw]

description

pytest==9.1.0 strict_parametrization_ids does not consider different data types as unique test cases

pytest is recommending provide unique pytest.param ids for two tests which are validating behavior str values vs int values.

Returning the error:

_________________________________________________________ ERROR collecting test_strict.py __________________________________________________________
Duplicate parametrization IDs detected, but strict_parametrization_ids is set.

Test name:      test_strict.py::test_double_accepts_str_and_int
Parameters:     val, expected
Parameter sets: [1, 2], ['1', 2]
IDs:            1-2, 1-2
Duplicates:     1-2

You can fix this problem using `@pytest.mark.parametrize(..., ids=...)` or `pytest.param(..., id=...)`.

pip list

$ pip list
Package                   Version
------------------------- ---------
attrs                     26.1.0
certifi                   2026.5.20
charset-normalizer        3.4.7
configobj                 5.0.9
coverage                  7.14.1
execnet                   2.1.2
idna                      3.18
iniconfig                 2.3.0
Jinja2                    3.1.6
jsonpatch                 1.33
jsonpointer               3.1.1
jsonschema                4.26.0
jsonschema-specifications 2025.9.1
MarkupSafe                3.0.3
oauthlib                  3.3.1
packaging                 26.2
passlib                   1.7.4
pip                       25.1.1
pluggy                    1.6.0
pyfakefs                  6.2.0
Pygments                  2.20.0
pyserial                  3.5
pytest                    9.1.0
pytest-cov                7.1.0
pytest-mock               3.15.1
pytest-xdist              3.8.0
PyYAML                    6.0.3
referencing               0.37.0
requests                  2.34.2
responses                 0.26.1
rpds-py                   2026.5.1
urllib3                   2.7.0

reproducer:

cat > test_strict.py <<EOF
import pytest


def double(val):
    try:
        return 2 * int(val)
    except ValueError:
        return 2 * val


@pytest.mark.parametrize("val,expected", ((1, 2), ("1", 2)))
def test_double_accepts_str_and_int(val, expected):
    assert expected == double(val)
EOF

pytest test_strict.py
....

[The-Compiler]

This is unfortunate but I don't see a bug here. 1-2 is the correct test ID for both (1, 2) and ("1", 2) (having " for strings in all test IDs strikes me as quite verbose and error-prone when copy-pasting IDs), and indeed that means you end up with two tests with the same ID. The recommendation to override that manually seems nothing but correct to me.

[blackboxsw]

I've updated the unit test to be more explicit about the use-case. We'd like to validate a function which accepts and handles ints or strings as they are unique input params. If pytest will not treat these value as unique due to datatype provided because it is converting all values to strings, then yes I agree, we'd need to provide manual id override for the test case.

[RonnyPfannschmidt]

Final test ids are strings

The simplification means that a string and a number with the same string result end up being the same id

Its highly recommended to use manual ids for putting a logical name into the test id

[The-Compiler]

Closing this as I don't think there's anything pytest can do here if we don't want to change the format of autogenerated string IDs to include quotes.