python / python/mypy

sorted() loses information about namedtuple elements

Open
#17,602 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

bug topic-join-v-union
Dominant language
Python
Stars
20.6k
Forks
3.3k
PR merge metrics
PR metrics pending

Description

Bug Report

it seems that running sorted() on an union of iterables of named tuple, create some missing information

To Reproduce

playground url

from typing import NamedTuple
from typing import reveal_type


class NamedTupleA(NamedTuple):
    field_a: str


class NamedTupleB(NamedTuple):
    field_b: str


def my_fun(my_list: list[NamedTupleA] | list[NamedTupleB]) -> None:
    reveal_type(my_list)
    # Revealed type is "Union[builtins.list[tuple[builtins.str, fallback=__main__.NamedTupleA]], builtins.list[tuple[builtins.str, fallback=__main__.NamedTupleB]]]"

    for nt in my_list:
        if isinstance(nt, NamedTupleA):
            print(nt.field_a)
        else:
            # here nt is recognized of `NamedTupleB` type, I can access field_b
            print(nt.field_b)

    reveal_type(sorted(my_list))
    # Revealed type is "builtins.list[tuple[builtins.str]]"

    for sorted_nt in sorted(my_list):
        if isinstance(sorted_nt, NamedTupleA):
            print(sorted_nt.field_a)
        else:
            # here I expected nt to be recognized of `NamedTupleB` type
            print(sorted_nt.field_b)

if __name__ == '__main__':
    my_fun([NamedTupleA('value_1')])
    my_fun([NamedTupleB('value_2')])
    # this error is expected: I cannot mix NamedTupleA and NamedTupleB
    my_fun([NamedTupleA('value_1'), NamedTupleB('value_2')])

basically I have 2 list of NamedTuples.
my_fun must accept either one.
Ideally I could write a User-Defined Type Guards, but let's say that I'm just checking (with isinstance) each element type.

If I loop directly over the list, the check works for both the if and the else branch (there shouldn't be any other case, the union is between two types)

If I loop directly over the sorted list, the type of the element is different, so in the else branch mypy gives me an error

Expected Behavior

I expect no errors on print(sorted_nt.field_b)

Actual Behavior

main.py:32: error: "tuple[str]" has no attribute "field_b" [attr-defined]

Your Environment

linux, python3.12, mypy 1.10.1

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the linked mypy playground reproduction using Python 3.12 and mypy 1.10.1, then compare the revealed types before and after sorted() in main.py. The fix is complete when the sorted result preserves the NamedTuple alternatives so the isinstance branches accept sorted_nt.field_a and sorted_nt.field_b without an attr-defined error.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
devtools
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Clearly specified
Newbie friendliness
48/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.