| ## `iequals` incorrectly returns `True` when comparing iterables of different lengths with equality-wildcard objects |
|
|
| ### Description |
|
|
| When using `iequals` to compare two iterables of different lengths, it can incorrectly return `True` if one of the iterables contains an object that compares equal to everything (such as `unittest.mock.ANY`). |
|
|
| Here's a minimal reproduction: |
|
|
| ```python |
| from unittest import mock |
| import more_itertools as mi |
| |
| # An empty list vs a list containing mock.ANY |
| result = mi.iequals([], [mock.ANY]) |
| print(result) # prints True, but should be False |
| ``` |
|
|
| The two iterables clearly have different lengths (0 vs 1), so `iequals` should return `False`. But it returns `True`. |
|
|
| ### Root cause |
|
|
| The current implementation uses `zip_longest` with a private sentinel `fillvalue=object()` to detect length mismatches: |
|
|
| ```python |
| return all(map(all_equal, zip_longest(*iterables, fillvalue=object()))) |
| ``` |
|
|
| The idea is that the sentinel won't compare equal to real values. However, `mock.ANY` (and any object that overrides `__eq__` to always return `True`) compares equal to the sentinel too, so the length mismatch goes undetected and `iequals` returns `True` instead of `False`. |
|
|
| ### Expected behavior |
|
|
| `mi.iequals([], [mock.ANY])` should return `False`, since the two iterables have different lengths and cannot be considered equal regardless of the values they contain. |
|
|
| ### Actual behavior |
|
|
| `mi.iequals([], [mock.ANY])` returns `True`. |
|
|
| Work in `/workspace`. Submit your fix in the existing Python source files under `more_itertools`. Preserve the other public behavior. The environment is offline; dependencies are preinstalled. Grading runs the repository's test suite in a fresh environment, using your submitted source files. |
|
|
