Implement `__len__` on SedonaDB DataFrame for Pythonic `len(df)` support

[prantogg]

It would be great if SedonaDB DataFrames supported Python’s built-in len() by implementing __len__.

In Python data tooling, it’s common (and expected) that len(df) returns the number of rows in a dataframe-like object. For example, DuckDB relations/dataframes support this kind of “Pythonic” ergonomics.

Today, calling len(df) on a SedonaDB DataFrame raises a TypeError because the DataFrame type doesn’t implement __len__.

[p-vdp]

See #176

[paleolimbot]

It's a great point and it's come up before. I'm skeptical to add __len__() because it is also typically expected that calling len(obj) is cheap and in the case of a SedonaDB DataFrame this could be very time consuming. Ibis throws an error here explaining the rationale (ExpressionError: Use .count() instead) and maybe we should do the same.