[SPARK-54871] [SQL] Trim aliases from grouping and aggregate expressions before handling grouping analytics

[mihailoale-db]

What changes were proposed in this pull request?

In this PR I propose to trim aliases from grouping and aggregate expressions before handling grouping analytics. This is needed for the following query:

SELECT col1 AS k2 FROM values(1) GROUP BY CUBE(k2)

Here we have col1 in the single-pass whereas in the fixed-point we have col1 AS k2 before constructing an Aggregate in ResolveGroupingAnalytics. Change removes the AS k2 part and keeps the compatibility between single-pass and fixed-point analyzers without changing outputs (analyzed plans are different, only names).
The change also inlines the behavior with regular aggregates (without grouping analytics). In other words:

SELECT col1 + col2 AS a FROM VALUES(1,2) GROUP BY a

Here col1 + col2 should be the grouping expression (as it is) and we inline the behaviors.

Why are the changes needed?

To keep the compatibility between fixed-point and single-pass analyzers.

Does this PR introduce any user-facing change?

No.

How was this patch tested?

Changed tests.

Was this patch authored or co-authored using generative AI tooling?

No.

[github-actions]

JIRA Issue Information

=== Improvement SPARK-54871 ===
Summary: Trim aliases from grouping and aggregate expressions before handling grouping analytics
Assignee: None
Status: Open
Affected: ["4.1.0"]

---

This comment was automatically generated by GitHub Actions

[mihailotim-db]

LGTM. Let's also mention in the PR description that this is not only for single-pass analyzer, but would also be consistent with how normal aggregates currently work:

[mihailoale-db]

@mihailotim-db Done! @cloud-fan PTAL when you find time. Thanks!

[cloud-fan]

thanks, merging to master!