[SPARK-50789][CONNECT] The inputs for typed aggregations should be analyzed

[ueshin]

What changes were proposed in this pull request?

Fixes SparkConnectPlanner to analyze the inputs for typed aggregations.

Why are the changes needed?

The inputs for typed aggregations should be analyzed.

For example:

val ds = Seq("abc", "xyz", "hello").toDS().select("*").as[String]
ds.groupByKey(_.length).reduceGroups(_ + _).show()

fails with:

org.apache.spark.SparkException: [INTERNAL_ERROR] Invalid call to toAttribute on unresolved object SQLSTATE: XX000
  org.apache.spark.sql.catalyst.analysis.Star.toAttribute(unresolved.scala:439)
  org.apache.spark.sql.catalyst.plans.logical.Project.$anonfun$output$1(basicLogicalOperators.scala:74)
  scala.collection.immutable.List.map(List.scala:247)
  scala.collection.immutable.List.map(List.scala:79)
  org.apache.spark.sql.catalyst.plans.logical.Project.output(basicLogicalOperators.scala:74)
  org.apache.spark.sql.connect.planner.SparkConnectPlanner.transformExpressionWithTypedReduceExpression(SparkConnectPlanner.scala:2340)
  org.apache.spark.sql.connect.planner.SparkConnectPlanner.$anonfun$transformKeyValueGroupedAggregate$1(SparkConnectPlanner.scala:2244)
  scala.collection.immutable.List.map(List.scala:247)
  scala.collection.immutable.List.map(List.scala:79)
  org.apache.spark.sql.connect.planner.SparkConnectPlanner.transformKeyValueGroupedAggregate(SparkConnectPlanner.scala:2244)
  org.apache.spark.sql.connect.planner.SparkConnectPlanner.transformAggregate(SparkConnectPlanner.scala:2232)
...

Does this PR introduce any user-facing change?

The failure will not appear.

How was this patch tested?

Added the related tests.

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

No.

[ueshin]

The remaining test failures are not related to this PR.

[ueshin]

Thanks! merging to master.