Update to_date udf function to support a consistent set of argument types

[Omega359]

Is your feature request related to a problem or challenge?

The docs for to_date say "Supports strings, integer and double types as input." however that is not actually the case - the currently supported types is Int32, Int64, Null, Float64, Date32, Date64 and Utf8, Utf8View and LargeUtf8.

The supported data types should be expanded to support all int types, all uint types, all float types, all Utf8 types, and date types.

Describe the solution you'd like

No response

Describe alternatives you've considered

No response

Additional context

No response

[manishkr]

It should work for all numeric types in principle. However, the implementation relies on Arrow’s casting logic, and that layer doesn’t support this case. We have two option here, either extend the logic here or extend it in arrow. Here’s the relevant reference: https://github.com/apache/arrow-rs/blob/main/arrow-cast/src/cast/mod.rs#L312

[Omega359]

Normally you would be correct however in this case you are not and it's because of the signature of the function signature: Signature::variadic_any(Volatility::Immutable),. The function accepts all types essentially and directly handles each type.

You can see this yourself with a quick test:

$ datafusion-cli
DataFusion CLI v50.3.0
> select to_date(arrow_cast(22222, 'UInt32'));
Execution error: Unsupported data type UInt32 for function to_date

That error message comes from the function itself.

[Omega359]

take