fix(python): Allow for IntoExprColumn arguments in Series.str.replace and Series.str.replace_all

[MarcoGorelli]

Non-Python-literal input are already allowed here, I think the type hints just didn't get updated in #4516 ?

I've updated to match how it's typed in extract:

polars/py-polars/polars/series/string.py

Line 824 in 88cd1b0

def extract(self, pattern: IntoExprColumn, group_index: int = 1) -> Series:

polars/py-polars/polars/expr/string.py

Line 1436 in 88cd1b0

def extract(self, pattern: IntoExprColumn, group_index: int = 1) -> Expr:

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.71%. Comparing base (1306208) to head (e490df3).

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #24644      +/-   ##
==========================================
- Coverage   81.71%   81.71%   -0.01%     
==========================================
  Files        1700     1700              
  Lines      233283   233283              
  Branches     2996     2996              
==========================================
- Hits       190633   190617      -16     
- Misses      41888    41904      +16     
  Partials      762      762              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[coastalwhite]

I think IntoExprColumn is the wrong annotation here.

IntoExprColumn means "Inputs that can convert into a col expression". That is not the case here. The str here semantically means a string, not a column identifier.

[MarcoGorelli]

Sure, this is one of the cases with str_as_lit=True

It's the same in offset_by, which uses str | IntoExprColumn

polars/py-polars/polars/series/datetime.py

Line 1753 in 769d990

def offset_by(self, by: str | IntoExprColumn) -> Series:

Would that be acceptable?

[coastalwhite]

I think that indeed makes sense and is acceptable.

[mcrumiller]

It's the same in offset_by, which uses str | IntoExprColumn

I am confused by this, if you have a column named "1y" then what happens when you pass offset_by("1y")?

[MarcoGorelli]

Then it would be interpreted as a string literal "1 year"

[mcrumiller]

Then it would be interpreted as a string literal "1 year"

Is this a priority issue then? It first tries to interpret as literal string, then as an column-convertible expression? Maybe I don't understand IntoExprColumn.