No error on duplicate imports

[mjbvz]

1. Turn on analysis
2. Create a file:

from os.path import join
from os import path
from os.path import join
from os import abort as join

Bug
There are no errors reported about join being imported multiple times

Maybe not a runtime error but seems like a programing mistake we should catch. Also when you run organize imports, it shifts the import order which changes the behavior of the code :/

[erictraut]

This is legal Python code, and it doesn't violate any static typing rules.

If you'd like to enforce a rule like "no duplicate import symbols", that's something that would be more in the domain of a linter. Popular linters for Python code include pylint and ruff. Pyright (the static analysis tool upon which pylance is built) is a type checker, so it focuses on type analysis and type violations.

[mjbvz]

Ok with pushing this to a linter but running organize imports breaks the code by changing which import is resolved in the body of my program

[erictraut]

With static type checking turned on, you should see errors if your code is broken by changing the order of imports.