PredEq is a utility for testing objects using an equivalence predicate. It facilitates writing declarative-style
tests for complex requirements, and for cases where the default __eq__ is inapplicable or missing.
At its core, predeq(predicate) is an object which tests equal to X if predicate(X) returns True.
from predeq import predeq, instanceof
def test_tuple_of_int_and_str():
"""Test that value is a tuple of an int and a string."""
value = (123, 'abc')
# without predeq
assert isinstance(value, tuple) and len(value) == 2
assert isinstance(value[0], int)
assert isinstance(value[1], str)
# with predeq
assert value == (
predeq(lambda x: isinstance(x, int)), # provide your own predicate
instanceof(str), # or use one of the recipes
)Since PredEq is primarily designed for tests, it also tries to be helpful when they fail.
If lambda is used as a predicate, predeq attempts to locate its source code and show it in pytest output:
==================================== FAILURES ====================================
_______________________ test_tuple_of_int_and_str_issue_1 ________________________
def test_tuple_of_int_and_str_issue_1():
value = (123.0, 'abc')
> assert value == (
predeq(lambda x: isinstance(x, int)),
instanceof(str),
)
E AssertionError: assert (123.0, 'abc') == (<predeq to m...stanceof(str))
E
E At index 0 diff: 123.0 != <predeq to meet lambda x: isinstance(x, int)>
E Use -v to get more diff
Each built-in recipe also provides a readable representation, which typically looks like the corresponding code:
==================================== FAILURES ====================================
_______________________ test_tuple_of_int_and_str_issue_2 ________________________
def test_tuple_of_int_and_str_issue_2():
value = (123, b'abc')
> assert value == (
predeq(lambda x: isinstance(x, int)),
instanceof(str),
)
E AssertionError: assert (123, b'abc') == (<predeq to m...stanceof(str))
E
E At index 1 diff: b'abc' != instanceof(str)
E Use -v to get more diff