[SPARK-42052][SQL] Codegen Support for HiveSimpleUDF

[panbingkun]

What changes were proposed in this pull request?

As a subtask of SPARK-42050, this PR adds Codegen Support for HiveSimpleUDF

Why are the changes needed?

Improve codegen coverage and performance

Does this PR introduce any user-facing change?

No.

How was this patch tested?

New UT & Pass GA.

[HyukjinKwon]

Hm, I think we should better go and figure out to resolve this by using Invoke so we don't have to manually implement the codegen logic. I was fine with #39555 as a one time thing but seems like there are some more to fix.

[HyukjinKwon]

cc @cloud-fan @yaooqinn @dongjoon-hyun FYI

[dongjoon-hyun]

I agree with @HyukjinKwon 's comment too.

[panbingkun]

Hm, I think we should better go and figure out to resolve this by using Invoke so we don't have to manually implement the codegen logic. I was fine with #39555 as a one time thing but seems like there are some more to fix.

As the first step, I have submitted a new Pr(#39949) to rewrite HiveGenericUDF with Invoke.
cc @HyukjinKwon @cloud-fan @yaooqinn @dongjoon-hyun @LuciferYang

[DonnyZone]

Any progress on this topic? We find Codegen combined with the expression-reuse feature (#29975) can greatly improve performance. Many users write duplicate hiveUDFs in their queries.