Wrap Arc to Statistics for partition_statistics API (#20570)

## 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.
-->

- Part of #20184

## Rationale for this change

<!--
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?

<!--
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?

<!--
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?

<!--
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: xudong963

datafusion-examples/examples/relation_planner/table_sample.rs

@@ -727,8 +727,8 @@ impl ExecutionPlan for SampleExec {
         Some(self.metrics.clone_inner())
     }
 
-    fn partition_statistics(&self, partition: Option<usize>) -> Result<Statistics> {
-        let mut stats = self.input.partition_statistics(partition)?;
+    fn partition_statistics(&self, partition: Option<usize>) -> Result<Arc<Statistics>> {
+        let mut stats = Arc::unwrap_or_clone(self.input.partition_statistics(partition)?);
         let ratio = self.upper_bound - self.lower_bound;
 
         // Scale statistics by sampling ratio (inexact due to randomness)
@@ -741,7 +741,7 @@ impl ExecutionPlan for SampleExec {
             .map(|n| (n as f64 * ratio) as usize)
             .to_inexact();
 
-        Ok(stats)
+        Ok(Arc::new(stats))
     }
 
     fn apply_expressions(

datafusion/core/tests/custom_sources_cases/mod.rs

@@ -184,12 +184,12 @@ impl ExecutionPlan for CustomExecutionPlan {
         Ok(Box::pin(TestCustomRecordBatchStream { nb_batch: 1 }))
     }
 
-    fn partition_statistics(&self, partition: Option<usize>) -> Result<Statistics> {
+    fn partition_statistics(&self, partition: Option<usize>) -> Result<Arc<Statistics>> {
         if partition.is_some() {
-            return Ok(Statistics::new_unknown(&self.schema()));
+            return Ok(Arc::new(Statistics::new_unknown(&self.schema())));
         }
         let batch = TEST_CUSTOM_RECORD_BATCH!().unwrap();
-        Ok(Statistics {
+        Ok(Arc::new(Statistics {
             num_rows: Precision::Exact(batch.num_rows()),
             total_byte_size: Precision::Absent,
             column_statistics: self
@@ -208,7 +208,7 @@ impl ExecutionPlan for CustomExecutionPlan {
                     ..Default::default()
                 })
                 .collect(),
-        })
+        }))
     }
 
     fn apply_expressions(

datafusion/core/tests/custom_sources_cases/statistics.rs

@@ -182,11 +182,11 @@ impl ExecutionPlan for StatisticsValidation {
         unimplemented!("This plan only serves for testing statistics")
     }
 
-    fn partition_statistics(&self, partition: Option<usize>) -> Result<Statistics> {
+    fn partition_statistics(&self, partition: Option<usize>) -> Result<Arc<Statistics>> {
         if partition.is_some() {
-            Ok(Statistics::new_unknown(&self.schema))
+            Ok(Arc::new(Statistics::new_unknown(&self.schema)))
         } else {
-            Ok(self.stats.clone())
+            Ok(Arc::new(self.stats.clone()))
         }
     }
 
@@ -255,7 +255,7 @@ async fn sql_basic() -> Result<()> {
     let physical_plan = df.create_physical_plan().await.unwrap();
 
     // the statistics should be those of the source
-    assert_eq!(stats, physical_plan.partition_statistics(None)?);
+    assert_eq!(stats, *physical_plan.partition_statistics(None)?);
 
     Ok(())
 }
@@ -295,7 +295,7 @@ async fn sql_limit() -> Result<()> {
                 .collect(),
             total_byte_size: Precision::Absent
         },
-        physical_plan.partition_statistics(None)?
+        *physical_plan.partition_statistics(None)?
     );
 
     let df = ctx
@@ -304,7 +304,7 @@ async fn sql_limit() -> Result<()> {
         .unwrap();
     let physical_plan = df.create_physical_plan().await.unwrap();
     // when the limit is larger than the original number of lines, statistics remain unchanged
-    assert_eq!(stats, physical_plan.partition_statistics(None)?);
+    assert_eq!(stats, *physical_plan.partition_statistics(None)?);
 
     Ok(())
 }
@@ -324,7 +324,7 @@ async fn sql_window() -> Result<()> {
     let result = physical_plan.partition_statistics(None)?;
 
     assert_eq!(stats.num_rows, result.num_rows);
-    let col_stats = result.column_statistics;
+    let col_stats = &result.column_statistics;
     assert_eq!(2, col_stats.len());
     assert_eq!(stats.column_statistics[1], col_stats[0]);
 

datafusion/core/tests/physical_optimizer/join_selection.rs

@@ -1191,12 +1191,12 @@ impl ExecutionPlan for StatisticsExec {
         unimplemented!("This plan only serves for testing statistics")
     }
 
-    fn partition_statistics(&self, partition: Option<usize>) -> Result<Statistics> {
-        Ok(if partition.is_some() {
+    fn partition_statistics(&self, partition: Option<usize>) -> Result<Arc<Statistics>> {
+        Ok(Arc::new(if partition.is_some() {
             Statistics::new_unknown(&self.schema)
         } else {
             self.stats.clone()
-        })
+        }))
     }
 
     fn apply_expressions(