I have this function:

def chunk_load(filename: str, chunksize=1000, max_chunks:Optional[int]=None, expected_chunks=Optional[int]) -> pd.DataFrame:
    display_string = f'Loading {os.path.basename(filename)}|chunksize={chunksize}'
    if max_chunks is None:
        df = pd.concat([chunk for chunk in tqdm(pd.read_csv(filename, chunksize=chunksize), desc=display_string, total=expected_chunks)]) # pylance says that `total` is not compatible
    else:
        chunks = []
        for _ in tqdm(range(max_chunks), desc=display_string):
            try:
                chunk = next(pd.read_csv(filename, chunksize=chunksize))
                chunks.append(chunk)
            except StopIteration:
                break
        df = pd.concat(chunks)
    return df

Pylance is complaining that "UnionType" is incompatible with type "float | None"

I created a small example that I thought would mimic the same dynamic:

from typing import Optional

def foo(foo_val: Optional[float]) -> None:
    pass

def trouble_foo(trouble_val: float | None) -> None:
    pass

def bar(bar_val: Optional[int]) -> None:
    return foo(bar_val)

def trouble_bar(trouble_bar_val: Optional[int]) -> None:
    return trouble_foo(trouble_bar_val)

bar(35)
trouble_bar(42)

Pylance shows no errors for any of this.

Why is it that for the chunk_load function, passing in an Optional[int] for the total parameter is not okay because it expects a float | None, but, in the simple example, doing this is totally fine?