Separating Benchmarks for physical sorted union over large columns in SQL planner based on Datatype (apache#18599)

## Which issue does this PR close?

<!--
We generally require a GitHub issue to be filed for all bug fixes and
enhancements and this helps us generate change logs for our releases.
You can link an issue to this PR using the GitHub syntax. For example
`Closes #123` indicates that this PR will close issue #123.
-->
 - as discussed in apache#17261


## Rationale for this change
Logical Plan for datatype Int64 and UInt64 differs, UInt64 Logical
Plan's Union are wrapped up in Projection, and EliminateNestedUnion
OptimezerRule is not applied leading to significantly longer execution
time.
<!--
Why are you proposing this change? If this is already explained clearly
in the issue then this section is not needed.
Explaining clearly why changes are proposed helps reviewers understand
your changes and offer better suggestions for fixes.
-->

## What changes are included in this PR?
Separating Benchmarks based on datatype, converting a datatype specific
function to a generic one.
<!--
There is no need to duplicate the description in the issue here but it
is sometimes worth providing a summary of the individual changes in this
PR.
-->

## Are these changes tested?
Yes.
<!--
We typically require tests for all PRs in order to:
1. Prevent the code from being accidentally broken by subsequent changes
2. Serve as another way to document the expected behavior of the code

If tests are not included in your PR, please explain why (for example,
are they covered by existing tests)?
-->

## Are there any user-facing changes?
No, benchmarks only.
<!--
If there are user-facing changes then we may require documentation to be
updated before approving the PR.
-->

<!--
If there are any breaking changes to public APIs, please add the `api
change` label.
-->

---------

Co-authored-by: Andrew Lamb <andrew@nerdnetworks.org>

Author: logan-keede

datafusion/core/benches/sql_planner.rs

@@ -23,13 +23,17 @@ extern crate datafusion;
 mod data_utils;
 
 use crate::criterion::Criterion;
+use arrow::array::PrimitiveArray;
 use arrow::array::{ArrayRef, RecordBatch};
+use arrow::datatypes::ArrowNativeTypeOp;
+use arrow::datatypes::ArrowPrimitiveType;
 use arrow::datatypes::{DataType, Field, Fields, Schema};
 use criterion::Bencher;
 use datafusion::datasource::MemTable;
 use datafusion::execution::context::SessionContext;
 use datafusion_common::{config::Dialect, ScalarValue};
 use datafusion_expr::col;
+use rand_distr::num_traits::NumCast;
 use std::hint::black_box;
 use std::path::PathBuf;
 use std::sync::Arc;
@@ -155,18 +159,30 @@ fn benchmark_with_param_values_many_columns(
 /// 0,100...9900
 /// 0,200...19800
 /// 0,300...29700
-fn register_union_order_table(ctx: &SessionContext, num_columns: usize, num_rows: usize) {
-    // ("c0", [0, 0, ...])
-    // ("c1": [100, 200, ...])
-    // etc
-    let iter = (0..num_columns).map(|i| i as u64).map(|i| {
-        let array: ArrayRef = Arc::new(arrow::array::UInt64Array::from_iter_values(
-            (0..num_rows)
-                .map(|j| j as u64 * 100 + i)
-                .collect::<Vec<_>>(),
-        ));
+fn register_union_order_table_generic<T>(
+    ctx: &SessionContext,
+    num_columns: usize,
+    num_rows: usize,
+) where
+    T: ArrowPrimitiveType,
+    T::Native: ArrowNativeTypeOp + NumCast,
+{
+    let iter = (0..num_columns).map(|i| {
+        let array_data: Vec<T::Native> = (0..num_rows)
+            .map(|j| {
+                let value = (j as u64) * 100 + (i as u64);
+                <T::Native as NumCast>::from(value).unwrap_or_else(|| {
+                    panic!("Failed to cast numeric value to Native type")
+                })
+            })
+            .collect();
+
+        // Use PrimitiveArray which is generic over the ArrowPrimitiveType T
+        let array: ArrayRef = Arc::new(PrimitiveArray::<T>::from_iter_values(array_data));
+
         (format!("c{i}"), array)
     });
+
     let batch = RecordBatch::try_from_iter(iter).unwrap();
     let schema = batch.schema();
     let partitions = vec![vec![batch]];
@@ -183,7 +199,6 @@ fn register_union_order_table(ctx: &SessionContext, num_columns: usize, num_rows
 
     ctx.register_table("t", Arc::new(table)).unwrap();
 }
-
 /// return a query like
 /// ```sql
 /// select c1, 2 as c2, ... n as cn from t ORDER BY c1
@@ -403,13 +418,40 @@ fn criterion_benchmark(c: &mut Criterion) {
 
     // -- Sorted Queries --
     // 100, 200 && 300 is taking too long - https://github.com/apache/datafusion/issues/18366
+    // Logical Plan for datatype Int64 and UInt64 differs, UInt64 Logical Plan's Union are wrapped
+    // up in Projection, and EliminateNestedUnion OptimezerRule is not applied leading to significantly
+    // longer execution time.
+    // https://github.com/apache/datafusion/issues/17261
+
     for column_count in [10, 50 /* 100, 200, 300 */] {
-        register_union_order_table(&ctx, column_count, 1000);
+        register_union_order_table_generic::<arrow::datatypes::Int64Type>(
+            &ctx,
+            column_count,
+            1000,
+        );
 
         // this query has many expressions in its sort order so stresses
         // order equivalence validation
         c.bench_function(
-            &format!("physical_sorted_union_order_by_{column_count}"),
+            &format!("physical_sorted_union_order_by_{column_count}_int64"),
+            |b| {
+                // SELECT ... UNION ALL ...
+                let query = union_orderby_query(column_count);
+                b.iter(|| physical_plan(&ctx, &rt, &query))
+            },
+        );
+
+        let _ = ctx.deregister_table("t");
+    }
+
+    for column_count in [10, 50 /* 100, 200, 300 */] {
+        register_union_order_table_generic::<arrow::datatypes::UInt64Type>(
+            &ctx,
+            column_count,
+            1000,
+        );
+        c.bench_function(
+            &format!("physical_sorted_union_order_by_{column_count}_uint64"),
             |b| {
                 // SELECT ... UNION ALL ...
                 let query = union_orderby_query(column_count);