python / python/mypy

Can't check only submodules (except if installed with `pip install -e .`)

Open
#15,655 1 comment 3 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

bug
Dominant language
Python
Stars
20.6k
Forks
3.3k
PR merge metrics
PR metrics pending

Description

Bug Report

Checking a submodule only of a project seems impossible, when not installed using pip install -e .

Usecase:
I have project that contains auto-generated API code which is not correctly typed.
I only want to check the controllers that are located in a submodule.

To Reproduce

  1. Create a example project with the following structure
.
├── setup.py
└──src
    └── example_project
        ├── __init__.py
        ├── calc.py
        ├── py.typed
        └── sub
            ├── __init__.py
            └── foo.py

src/example_project/calc.py :

def int_add(a: int, b: int) -> int:
    """
    Simply add two numbers
    """
    return a + b


def float_diff(a: float, b: float) -> float:
    """
    Diff two numbers
    """
    return a / b


if __name__ == "__main__":
    int_add("a", "b")

src/ems_example_project/sub/foo.py:

import example_project.calc

example_project.calc.int_add("foo", "bar")

a: int = 1
a = "2"
  1. Install the package with pip install . OR do not install it at all.

  2. Run mypy for the given submodule

mypy src/example_project/sub

Expected Behavior

Only type errors for the submodule are found:

src/example_project/sub/foo.py:3: error: Argument 1 to "int_add" has incompatible type "str"; expected "int"  [arg-type]
src/example_project/sub/foo.py:3: error: Argument 2 to "int_add" has incompatible type "str"; expected "int"  [arg-type]
src/example_project/sub/foo.py:6: error: Incompatible types in assignment (expression has type "str", variable has type "int")  [assignment]
Found 3 errors in 1 file (checked 2 source files)

This actually works, if I install the package with pip install -e ..
But I expect this to work no matter if I install the package normally or not installing it at all.

Actual Behavior

The type errors for the top level modules are also listed.

src/example_project/calc.py:21: error: Argument 1 to "int_add" has incompatible type "str"; expected "int"  [arg-type]
src/example_project/calc.py:21: error: Argument 2 to "int_add" has incompatible type "str"; expected "int"  [arg-type]
src/example_project/sub/foo.py:3: error: Argument 1 to "int_add" has incompatible type "str"; expected "int"  [arg-type]
src/example_project/sub/foo.py:3: error: Argument 2 to "int_add" has incompatible type "str"; expected "int"  [arg-type]
src/example_project/sub/foo.py:6: error: Incompatible types in assignment (expression has type "str", variable has type "int")  [assignment]
Found 5 errors in 2 files (checked 2 source files)

If I use the option -p example_project.sub instead of the path it works as expected. But in my environment I am only able to add specific files (think pre-commit or similar).
It also works if I install the package using pip install -e.
Still I would not expect a such a difference just depending on how I installed the package.

Your Environment

  • Mypy version used: 1.4.1
  • Mypy command-line flags: none besides the path to the file
  • Mypy configuration options from mypy.ini (and other config files): none
  • Python version used: 3.9.16

This might be related to:

  • #10956
  • #10260
  • #6861
  • #10428

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 reported setup.py layout and reproduce the difference between mypy src/example_project/sub, -p example_project.sub, and an editable versus normal install. Compare diagnostics from calc.py and sub/foo.py, then trace the path-based module discovery behavior. Done means checking only the requested submodule produces the three expected errors without also checking calc.py.

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
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.