[Minor] Remove/reorder logical plan rules (#15421)

* Remove rules that execute twice

* Move optimization rule around

Author: Dandandan

datafusion/optimizer/src/optimizer.rs

@@ -230,7 +230,6 @@ impl Optimizer {
             Arc::new(EliminateDuplicatedExpr::new()),
             Arc::new(EliminateFilter::new()),
             Arc::new(EliminateCrossJoin::new()),
-            Arc::new(CommonSubexprEliminate::new()),
             Arc::new(EliminateLimit::new()),
             Arc::new(PropagateEmptyRelation::new()),
             // Must be after PropagateEmptyRelation
@@ -243,9 +242,8 @@ impl Optimizer {
             Arc::new(SingleDistinctToGroupBy::new()),
             // The previous optimizations added expressions and projections,
             // that might benefit from the following rules
-            Arc::new(SimplifyExpressions::new()),
-            Arc::new(CommonSubexprEliminate::new()),
             Arc::new(EliminateGroupByConstant::new()),
+            Arc::new(CommonSubexprEliminate::new()),
             Arc::new(OptimizeProjections::new()),
         ];
 

datafusion/optimizer/tests/optimizer_integration.rs

@@ -78,8 +78,8 @@ fn subquery_filter_with_cast() -> Result<()> {
     \n    SubqueryAlias: __scalar_sq_1\
     \n      Aggregate: groupBy=[[]], aggr=[[avg(CAST(test.col_int32 AS Float64))]]\
     \n        Projection: test.col_int32\
-    \n          Filter: __common_expr_5 >= Date32(\"2002-05-08\") AND __common_expr_5 <= Date32(\"2002-05-13\")\
-    \n            Projection: CAST(test.col_utf8 AS Date32) AS __common_expr_5, test.col_int32\
+    \n          Filter: __common_expr_4 >= Date32(\"2002-05-08\") AND __common_expr_4 <= Date32(\"2002-05-13\")\
+    \n            Projection: CAST(test.col_utf8 AS Date32) AS __common_expr_4, test.col_int32\
     \n              TableScan: test projection=[col_int32, col_utf8]";
     assert_eq!(expected, format!("{plan}"));
     Ok(())

datafusion/sqllogictest/test_files/explain.slt

@@ -187,7 +187,6 @@ logical_plan after extract_equijoin_predicate SAME TEXT AS ABOVE
 logical_plan after eliminate_duplicated_expr SAME TEXT AS ABOVE
 logical_plan after eliminate_filter SAME TEXT AS ABOVE
 logical_plan after eliminate_cross_join SAME TEXT AS ABOVE
-logical_plan after common_sub_expression_eliminate SAME TEXT AS ABOVE
 logical_plan after eliminate_limit SAME TEXT AS ABOVE
 logical_plan after propagate_empty_relation SAME TEXT AS ABOVE
 logical_plan after eliminate_one_union SAME TEXT AS ABOVE
@@ -196,9 +195,8 @@ logical_plan after eliminate_outer_join SAME TEXT AS ABOVE
 logical_plan after push_down_limit SAME TEXT AS ABOVE
 logical_plan after push_down_filter SAME TEXT AS ABOVE
 logical_plan after single_distinct_aggregation_to_group_by SAME TEXT AS ABOVE
-logical_plan after simplify_expressions SAME TEXT AS ABOVE
-logical_plan after common_sub_expression_eliminate SAME TEXT AS ABOVE
 logical_plan after eliminate_group_by_constant SAME TEXT AS ABOVE
+logical_plan after common_sub_expression_eliminate SAME TEXT AS ABOVE
 logical_plan after optimize_projections TableScan: simple_explain_test projection=[a, b, c]
 logical_plan after eliminate_nested_union SAME TEXT AS ABOVE
 logical_plan after simplify_expressions SAME TEXT AS ABOVE
@@ -210,7 +208,6 @@ logical_plan after extract_equijoin_predicate SAME TEXT AS ABOVE
 logical_plan after eliminate_duplicated_expr SAME TEXT AS ABOVE
 logical_plan after eliminate_filter SAME TEXT AS ABOVE
 logical_plan after eliminate_cross_join SAME TEXT AS ABOVE
-logical_plan after common_sub_expression_eliminate SAME TEXT AS ABOVE
 logical_plan after eliminate_limit SAME TEXT AS ABOVE
 logical_plan after propagate_empty_relation SAME TEXT AS ABOVE
 logical_plan after eliminate_one_union SAME TEXT AS ABOVE
@@ -219,9 +216,8 @@ logical_plan after eliminate_outer_join SAME TEXT AS ABOVE
 logical_plan after push_down_limit SAME TEXT AS ABOVE
 logical_plan after push_down_filter SAME TEXT AS ABOVE
 logical_plan after single_distinct_aggregation_to_group_by SAME TEXT AS ABOVE
-logical_plan after simplify_expressions SAME TEXT AS ABOVE
-logical_plan after common_sub_expression_eliminate SAME TEXT AS ABOVE
 logical_plan after eliminate_group_by_constant SAME TEXT AS ABOVE
+logical_plan after common_sub_expression_eliminate SAME TEXT AS ABOVE
 logical_plan after optimize_projections SAME TEXT AS ABOVE
 logical_plan TableScan: simple_explain_test projection=[a, b, c]
 initial_physical_plan DataSourceExec: file_groups={1 group: [[WORKSPACE_ROOT/datafusion/core/tests/data/example.csv]]}, projection=[a, b, c], file_type=csv, has_header=true

datafusion/sqllogictest/test_files/group_by.slt

@@ -4188,9 +4188,8 @@ EXPLAIN SELECT SUM(DISTINCT CAST(x AS DOUBLE)), MAX(DISTINCT CAST(x AS DOUBLE))
 logical_plan
 01)Projection: sum(alias1) AS sum(DISTINCT t1.x), max(alias1) AS max(DISTINCT t1.x)
 02)--Aggregate: groupBy=[[t1.y]], aggr=[[sum(alias1), max(alias1)]]
-03)----Aggregate: groupBy=[[t1.y, __common_expr_1 AS t1.x AS alias1]], aggr=[[]]
-04)------Projection: CAST(t1.x AS Float64) AS __common_expr_1, t1.y
-05)--------TableScan: t1 projection=[x, y]
+03)----Aggregate: groupBy=[[t1.y, CAST(t1.x AS Float64) AS alias1]], aggr=[[]]
+04)------TableScan: t1 projection=[x, y]
 physical_plan
 01)ProjectionExec: expr=[sum(alias1)@1 as sum(DISTINCT t1.x), max(alias1)@2 as max(DISTINCT t1.x)]
 02)--AggregateExec: mode=FinalPartitioned, gby=[y@0 as y], aggr=[sum(alias1), max(alias1)]
@@ -4201,9 +4200,8 @@ physical_plan
 07)------------CoalesceBatchesExec: target_batch_size=2
 08)--------------RepartitionExec: partitioning=Hash([y@0, alias1@1], 8), input_partitions=8
 09)----------------RepartitionExec: partitioning=RoundRobinBatch(8), input_partitions=1
-10)------------------AggregateExec: mode=Partial, gby=[y@1 as y, __common_expr_1@0 as alias1], aggr=[]
-11)--------------------ProjectionExec: expr=[CAST(x@0 AS Float64) as __common_expr_1, y@1 as y]
-12)----------------------DataSourceExec: partitions=1, partition_sizes=[1]
+10)------------------AggregateExec: mode=Partial, gby=[y@1 as y, CAST(x@0 AS Float64) as alias1], aggr=[]
+11)--------------------DataSourceExec: partitions=1, partition_sizes=[1]
 
 # create an unbounded table that contains ordered timestamp.
 statement ok