python - Find cases of List["str"] but not List[str] = ["a"]

Question

I'm trying to check for outdated type hints syntax in a Python codebase.

I would like to catch cases such as

List["str"]

which could just be

List[str]

I've written the following regular expression:

List[.*".+".*]

however, it also catches the following string:

List[str] = ["a"]

which I don't want to catch.

Examples of strings I want to catch:

• List["str"]
• Union[int, "np.ndarray"]

How could I modify my regular expression accordingly?

Answer 1

To match the example strings, you could match either List or Union followed by using a negated character class matching from the opening till closing square bracket that can not be followed by an equals sign.

(?:List|Union)[[^][]*](?!s*=)

Explanation

• (?:List|Union) Match either List or union preceded by a word boundary to prevent them being part of a longer word.
• [[^][]*] Match from [ till ] not matching either of them in between using a negated character class
• (?!s*=) Assert no = sign to the right

Regex demo