[SPARK-19851][SQL] Add support for EVERY and ANY (SOME) aggregates

[dilipbiswal]

What changes were proposed in this pull request?

Implements Every, Some, Any aggregates in SQL. Logically these aggregate expressions are mapped to Min and Max, respectively.

Every(x) => Min(x) where x is boolean.
Some(x) => Max(x) where x is boolean.

Any is a synonym for Some.

How was this patch tested?

Added tests in SQLQueryTestSuite, DataframeAggregateSuite

[SparkQA]

Test build #97876 has finished for PR 22797 at commit 905652a.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds no public classes.

[dilipbiswal]

cc @cloud-fan@gatorsmile

[cloud-fan]

Hi @dilipbiswal sorry for the back and forth, can you try one more approach? Basically we want to evaluate how the simplest logical rewrite looks like.

We can create unevaluatable EVERY and ANY expressions, and create a rule to replace them with MAX and MIN. This rule can be put in the beginning of optimizer like RewriteDistinctAggregates. We don't need to create a completed framework to rewrite agg functions.

[dilipbiswal]

@cloud-fan Sure.. Let me give that a try.
One question : So this would be a very specific rewrite like `RewriteEveryAnySome" as opposed to a generic rule to rewrite any aggregate expressions, right ? The reason i ask is because, i have tried the generic rewrite in https://github.com/dilipbiswal/spark/tree/SPARK-19851-rewrite

[cloud-fan]

yea, just a special rule instead of a general agg function rewrite framework.

[dilipbiswal]

@cloud-fan OK.. thanks a LOT.