This is an automated email from the ASF dual-hosted git repository.
englefly pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/doris.git
The following commit(s) were added to refs/heads/master by this push:
new bc400256317 [opt](Nereids)Join cluster connectivity (#27833)
bc400256317 is described below
commit bc4002563175d1a24f3a1b5ac8ae7ddd4eb4df49
Author: minghong <engle...@gmail.com>
AuthorDate: Fri Dec 8 14:55:10 2023 +0800
[opt](Nereids)Join cluster connectivity (#27833)
* estimation join stats by connectivity
---
.../java/org/apache/doris/nereids/PlanContext.java | 4 ++
.../org/apache/doris/nereids/StatementContext.java | 12 ++++++
.../org/apache/doris/nereids/cost/CostModelV1.java | 28 ++++++++++++
.../rules/exploration/join/JoinCommute.java | 2 +-
.../doris/nereids/rules/rewrite/ReorderJoin.java | 1 +
.../org/apache/doris/nereids/memo/RankTest.java | 2 +-
.../predicate_infer/infer_predicate.out | 29 +++++++------
.../noStatsRfPrune/query14.out | 8 ++--
.../noStatsRfPrune/query16.out | 8 ++--
.../noStatsRfPrune/query31.out | 8 ++--
.../noStatsRfPrune/query58.out | 24 +++++------
.../noStatsRfPrune/query59.out | 35 ++++++++-------
.../noStatsRfPrune/query94.out | 8 ++--
.../noStatsRfPrune/query95.out | 50 +++++++++++-----------
.../no_stats_shape/query14.out | 8 ++--
.../no_stats_shape/query16.out | 10 ++---
.../no_stats_shape/query31.out | 8 ++--
.../no_stats_shape/query58.out | 30 ++++++-------
.../no_stats_shape/query59.out | 35 ++++++++-------
.../no_stats_shape/query94.out | 10 ++---
.../no_stats_shape/query95.out | 48 ++++++++++-----------
.../nostats_rf_prune/q21.out | 22 +++++-----
.../shape_no_stats/q21.out | 22 +++++-----
23 files changed, 228 insertions(+), 184 deletions(-)
diff --git a/fe/fe-core/src/main/java/org/apache/doris/nereids/PlanContext.java
b/fe/fe-core/src/main/java/org/apache/doris/nereids/PlanContext.java
index f560dabf8b3..3ab95423e24 100644
--- a/fe/fe-core/src/main/java/org/apache/doris/nereids/PlanContext.java
+++ b/fe/fe-core/src/main/java/org/apache/doris/nereids/PlanContext.java
@@ -88,4 +88,8 @@ public class PlanContext {
public List<Statistics> getChildrenStatistics() {
return childrenStats;
}
+
+ public StatementContext getStatementContext() {
+ return connectContext.getStatementContext();
+ }
}
diff --git
a/fe/fe-core/src/main/java/org/apache/doris/nereids/StatementContext.java
b/fe/fe-core/src/main/java/org/apache/doris/nereids/StatementContext.java
index da81698a8b3..33ef6997835 100644
--- a/fe/fe-core/src/main/java/org/apache/doris/nereids/StatementContext.java
+++ b/fe/fe-core/src/main/java/org/apache/doris/nereids/StatementContext.java
@@ -43,6 +43,7 @@ import com.google.common.collect.Maps;
import com.google.common.collect.Sets;
import java.util.ArrayList;
+import java.util.Collection;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
@@ -88,6 +89,9 @@ public class StatementContext {
private final Map<CTEId, LogicalPlan> rewrittenCteConsumer = new
HashMap<>();
private final Set<String> viewDdlSqlSet = Sets.newHashSet();
+ // collect all hash join conditions to compute node connectivity in join
graph
+ private final List<Expression> joinFilters = new ArrayList<>();
+
private final List<Hint> hints = new ArrayList<>();
public StatementContext() {
@@ -242,4 +246,12 @@ public class StatementContext {
public List<Hint> getHints() {
return ImmutableList.copyOf(hints);
}
+
+ public List<Expression> getJoinFilters() {
+ return joinFilters;
+ }
+
+ public void addJoinFilters(Collection<Expression> newJoinFilters) {
+ this.joinFilters.addAll(newJoinFilters);
+ }
}
diff --git
a/fe/fe-core/src/main/java/org/apache/doris/nereids/cost/CostModelV1.java
b/fe/fe-core/src/main/java/org/apache/doris/nereids/cost/CostModelV1.java
index 1ff398cce09..7704c9630fd 100644
--- a/fe/fe-core/src/main/java/org/apache/doris/nereids/cost/CostModelV1.java
+++ b/fe/fe-core/src/main/java/org/apache/doris/nereids/cost/CostModelV1.java
@@ -22,6 +22,7 @@ import org.apache.doris.nereids.properties.DistributionSpec;
import org.apache.doris.nereids.properties.DistributionSpecGather;
import org.apache.doris.nereids.properties.DistributionSpecHash;
import org.apache.doris.nereids.properties.DistributionSpecReplicated;
+import org.apache.doris.nereids.trees.expressions.Expression;
import org.apache.doris.nereids.trees.plans.Plan;
import org.apache.doris.nereids.trees.plans.physical.PhysicalAssertNumRows;
import
org.apache.doris.nereids.trees.plans.physical.PhysicalDeferMaterializeOlapScan;
@@ -48,6 +49,8 @@ import org.apache.doris.statistics.Statistics;
import com.google.common.base.Preconditions;
+import java.util.Collections;
+
class CostModelV1 extends PlanVisitor<Cost, PlanContext> {
// for a join, skew = leftRowCount/rightRowCount
@@ -262,6 +265,17 @@ class CostModelV1 extends PlanVisitor<Cost, PlanContext> {
double leftRowCount = probeStats.getRowCount();
double rightRowCount = buildStats.getRowCount();
+ if (leftRowCount == rightRowCount
+ && physicalHashJoin.getGroupExpression().isPresent()
+ && physicalHashJoin.getGroupExpression().get().getOwnerGroup()
!= null
+ &&
!physicalHashJoin.getGroupExpression().get().getOwnerGroup().isStatsReliable())
{
+ int leftConnectivity =
computeConnectivity(physicalHashJoin.left(), context);
+ int rightConnectivity =
computeConnectivity(physicalHashJoin.right(), context);
+ if (rightConnectivity < leftConnectivity) {
+ leftRowCount += 1;
+ }
+ }
+
/*
pattern1: L join1 (Agg1() join2 Agg2())
result number of join2 may much less than Agg1.
@@ -310,6 +324,20 @@ class CostModelV1 extends PlanVisitor<Cost, PlanContext> {
);
}
+ /*
+ in a join cluster graph, if a node has higher connectivity, it is more
likely to be reduced
+ by runtime filters, and it is also more likely to produce effective
runtime filters.
+ Thus, we prefer to put the node with higher connectivity on the join right
side.
+ */
+ private int computeConnectivity(
+ Plan plan, PlanContext context) {
+ int connectCount = 0;
+ for (Expression expr : context.getStatementContext().getJoinFilters())
{
+ connectCount += Collections.disjoint(expr.getInputSlots(),
plan.getOutputSet()) ? 0 : 1;
+ }
+ return connectCount;
+ }
+
@Override
public Cost visitPhysicalNestedLoopJoin(
PhysicalNestedLoopJoin<? extends Plan, ? extends Plan>
nestedLoopJoin,
diff --git
a/fe/fe-core/src/main/java/org/apache/doris/nereids/rules/exploration/join/JoinCommute.java
b/fe/fe-core/src/main/java/org/apache/doris/nereids/rules/exploration/join/JoinCommute.java
index d6df03e1c0b..efdd46f821f 100644
---
a/fe/fe-core/src/main/java/org/apache/doris/nereids/rules/exploration/join/JoinCommute.java
+++
b/fe/fe-core/src/main/java/org/apache/doris/nereids/rules/exploration/join/JoinCommute.java
@@ -93,7 +93,7 @@ public class JoinCommute extends OneExplorationRuleFactory {
if (swapType == SwapType.LEFT_ZIG_ZAG) {
double leftRows =
join.left().getGroup().getStatistics().getRowCount();
double rightRows =
join.right().getGroup().getStatistics().getRowCount();
- return leftRows < rightRows && isZigZagJoin(join);
+ return leftRows <= rightRows && isZigZagJoin(join);
}
return true;
diff --git
a/fe/fe-core/src/main/java/org/apache/doris/nereids/rules/rewrite/ReorderJoin.java
b/fe/fe-core/src/main/java/org/apache/doris/nereids/rules/rewrite/ReorderJoin.java
index a9e685b1c89..31b0dc904ab 100644
---
a/fe/fe-core/src/main/java/org/apache/doris/nereids/rules/rewrite/ReorderJoin.java
+++
b/fe/fe-core/src/main/java/org/apache/doris/nereids/rules/rewrite/ReorderJoin.java
@@ -87,6 +87,7 @@ public class ReorderJoin extends OneRewriteRuleFactory {
Plan plan = joinToMultiJoin(filter, planToHintType);
Preconditions.checkState(plan instanceof MultiJoin);
MultiJoin multiJoin = (MultiJoin) plan;
+ ctx.statementContext.addJoinFilters(multiJoin.getJoinFilter());
ctx.statementContext.setMaxNAryInnerJoin(multiJoin.children().size());
Plan after = multiJoinToJoin(multiJoin, planToHintType);
return after;
diff --git
a/fe/fe-core/src/test/java/org/apache/doris/nereids/memo/RankTest.java
b/fe/fe-core/src/test/java/org/apache/doris/nereids/memo/RankTest.java
index e3571395e47..57773bb0c4f 100644
--- a/fe/fe-core/src/test/java/org/apache/doris/nereids/memo/RankTest.java
+++ b/fe/fe-core/src/test/java/org/apache/doris/nereids/memo/RankTest.java
@@ -55,7 +55,7 @@ class RankTest extends TestWithFeService {
shape.add(memo.unrank(memo.rank(i + 1).first).shape(""));
}
System.out.println(shape);
- Assertions.assertEquals(1, shape.size());
+ Assertions.assertEquals(2, shape.size());
Assertions.assertEquals(bestPlan.shape(""),
memo.unrank(memo.rank(1).first).shape(""));
}
}
diff --git
a/regression-test/data/nereids_rules_p0/predicate_infer/infer_predicate.out
b/regression-test/data/nereids_rules_p0/predicate_infer/infer_predicate.out
index c106bb9407a..bed06fa4848 100644
--- a/regression-test/data/nereids_rules_p0/predicate_infer/infer_predicate.out
+++ b/regression-test/data/nereids_rules_p0/predicate_infer/infer_predicate.out
@@ -465,21 +465,22 @@ PhysicalResultSink
-- !infer5 --
PhysicalResultSink
---PhysicalProject
-----hashJoin[INNER_JOIN] hashCondition=((t1.id = t3.id)) otherCondition=()
build RFs:RF1 id->[id]
-------hashJoin[INNER_JOIN] hashCondition=((t1.id = t2.id)) otherCondition=()
build RFs:RF0 id->[id]
---------filter((t1.id = 1))
-----------PhysicalLimit[GLOBAL]
-------------PhysicalDistribute
---------------PhysicalLimit[LOCAL]
-----------------PhysicalProject
-------------------PhysicalOlapScan[t1] apply RFs: RF0 RF1
---------PhysicalDistribute
+--PhysicalDistribute
+----PhysicalProject
+------hashJoin[INNER_JOIN] hashCondition=((t1.id = t3.id)) otherCondition=()
build RFs:RF1 id->[id]
+--------hashJoin[INNER_JOIN] hashCondition=((t1.id = t2.id)) otherCondition=()
build RFs:RF0 id->[id]
----------filter((t2.id = 1))
-------------PhysicalOlapScan[t2]
-------PhysicalDistribute
---------filter((t3.id = 1))
-----------PhysicalOlapScan[t]
+------------PhysicalOlapScan[t2] apply RFs: RF0
+----------PhysicalDistribute
+------------filter((t1.id = 1))
+--------------PhysicalLimit[GLOBAL]
+----------------PhysicalDistribute
+------------------PhysicalLimit[LOCAL]
+--------------------PhysicalProject
+----------------------PhysicalOlapScan[t1] apply RFs: RF1
+--------PhysicalDistribute
+----------filter((t3.id = 1))
+------------PhysicalOlapScan[t]
-- !infer6 --
PhysicalResultSink
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query14.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query14.out
index 619c5f69f02..ad7cd03a059 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query14.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query14.out
@@ -157,12 +157,12 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
--------------------------------------hashJoin[INNER_JOIN]
hashCondition=((web_sales.ws_sold_date_sk = date_dim.d_date_sk))
otherCondition=() build RFs:RF20 d_date_sk->[ws_sold_date_sk]
----------------------------------------PhysicalProject
------------------------------------------hashJoin[INNER_JOIN]
hashCondition=((web_sales.ws_item_sk = item.i_item_sk)) otherCondition=() build
RFs:RF19 i_item_sk->[ws_item_sk]
---------------------------------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((web_sales.ws_item_sk = cross_items.ss_item_sk))
otherCondition=() build RFs:RF18 ss_item_sk->[ws_item_sk]
-----------------------------------------------PhysicalDistribute
-------------------------------------------------PhysicalProject
---------------------------------------------------PhysicalOlapScan[web_sales]
apply RFs: RF18 RF19 RF20
+--------------------------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((web_sales.ws_item_sk = cross_items.ss_item_sk))
otherCondition=() build RFs:RF18 ws_item_sk->[ss_item_sk]
----------------------------------------------PhysicalDistribute
------------------------------------------------PhysicalCteConsumer (
cteId=CTEId#0 )
+----------------------------------------------PhysicalDistribute
+------------------------------------------------PhysicalProject
+--------------------------------------------------PhysicalOlapScan[web_sales]
apply RFs: RF19 RF20
--------------------------------------------PhysicalDistribute
----------------------------------------------PhysicalProject
------------------------------------------------PhysicalOlapScan[item]
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query16.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query16.out
index 2fac57e7525..b08f9d62730 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query16.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query16.out
@@ -11,7 +11,10 @@ PhysicalResultSink
----------------PhysicalProject
------------------hashJoin[LEFT_ANTI_JOIN] hashCondition=((cs1.cs_order_number
= cr1.cr_order_number)) otherCondition=()
--------------------hashJoin[INNER_JOIN] hashCondition=((cs1.cs_call_center_sk
= call_center.cc_call_center_sk)) otherCondition=() build RFs:RF3
cc_call_center_sk->[cs_call_center_sk]
-----------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((cs1.cs_order_number = cs2.cs_order_number)) otherCondition=((
not (cs_warehouse_sk = cs_warehouse_sk)))
+----------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((cs1.cs_order_number = cs2.cs_order_number)) otherCondition=((
not (cs_warehouse_sk = cs_warehouse_sk))) build RFs:RF2
cs_order_number->[cs_order_number]
+------------------------PhysicalDistribute
+--------------------------PhysicalProject
+----------------------------PhysicalOlapScan[catalog_sales] apply RFs: RF2
------------------------PhysicalDistribute
--------------------------hashJoin[INNER_JOIN]
hashCondition=((cs1.cs_ship_addr_sk = customer_address.ca_address_sk))
otherCondition=() build RFs:RF1 ca_address_sk->[cs_ship_addr_sk]
----------------------------hashJoin[INNER_JOIN]
hashCondition=((cs1.cs_ship_date_sk = date_dim.d_date_sk)) otherCondition=()
build RFs:RF0 d_date_sk->[cs_ship_date_sk]
@@ -25,9 +28,6 @@ PhysicalResultSink
------------------------------PhysicalProject
--------------------------------filter((customer_address.ca_state = 'WV'))
----------------------------------PhysicalOlapScan[customer_address]
-------------------------PhysicalDistribute
---------------------------PhysicalProject
-----------------------------PhysicalOlapScan[catalog_sales]
----------------------PhysicalDistribute
------------------------PhysicalProject
--------------------------filter(cc_county IN ('Barrow County', 'Daviess
County', 'Luce County', 'Richland County', 'Ziebach County'))
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query31.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query31.out
index 67bf996d44b..b99861d990f 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query31.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query31.out
@@ -46,6 +46,10 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
--------------------hashJoin[INNER_JOIN] hashCondition=((ss1.ca_county =
ws1.ca_county)) otherCondition=()
----------------------PhysicalProject
------------------------hashJoin[INNER_JOIN] hashCondition=((ss2.ca_county =
ss3.ca_county)) otherCondition=()
+--------------------------PhysicalDistribute
+----------------------------PhysicalProject
+------------------------------filter((ss3.d_qoy = 3) and (ss3.d_year = 2000))
+--------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
--------------------------hashJoin[INNER_JOIN] hashCondition=((ss1.ca_county =
ss2.ca_county)) otherCondition=()
----------------------------PhysicalDistribute
------------------------------PhysicalProject
@@ -55,10 +59,6 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
------------------------------PhysicalProject
--------------------------------filter((ss2.d_qoy = 2) and (ss2.d_year = 2000))
----------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
---------------------------PhysicalDistribute
-----------------------------PhysicalProject
-------------------------------filter((ss3.d_qoy = 3) and (ss3.d_year = 2000))
---------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
----------------------PhysicalDistribute
------------------------PhysicalProject
--------------------------filter((ws1.d_qoy = 1) and (ws1.d_year = 2000))
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query58.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query58.out
index 080d5b2afea..1389ea6996c 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query58.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query58.out
@@ -24,10 +24,7 @@ PhysicalResultSink
--------------------------------------PhysicalOlapScan[item] apply RFs: RF12
RF13
------------------------------PhysicalDistribute
--------------------------------PhysicalProject
-----------------------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=() build
RFs:RF9 d_date->[d_date]
-------------------------------------PhysicalDistribute
---------------------------------------PhysicalProject
-----------------------------------------PhysicalOlapScan[date_dim] apply RFs:
RF9
+----------------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=()
------------------------------------PhysicalDistribute
--------------------------------------PhysicalProject
----------------------------------------hashJoin[INNER_JOIN]
hashCondition=((date_dim.d_week_seq = date_dim.d_week_seq)) otherCondition=()
build RFs:RF8 d_week_seq->[d_week_seq]
@@ -40,6 +37,9 @@ PhysicalResultSink
------------------------------------------------PhysicalProject
--------------------------------------------------filter((date_dim.d_date =
'2001-03-24'))
----------------------------------------------------PhysicalOlapScan[date_dim]
+------------------------------------PhysicalDistribute
+--------------------------------------PhysicalProject
+----------------------------------------PhysicalOlapScan[date_dim]
------------------PhysicalProject
--------------------hashAgg[GLOBAL]
----------------------PhysicalDistribute
@@ -55,10 +55,7 @@ PhysicalResultSink
--------------------------------------PhysicalOlapScan[item]
------------------------------PhysicalDistribute
--------------------------------PhysicalProject
-----------------------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=() build
RFs:RF5 d_date->[d_date]
-------------------------------------PhysicalDistribute
---------------------------------------PhysicalProject
-----------------------------------------PhysicalOlapScan[date_dim] apply RFs:
RF5
+----------------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=()
------------------------------------PhysicalDistribute
--------------------------------------PhysicalProject
----------------------------------------hashJoin[INNER_JOIN]
hashCondition=((date_dim.d_week_seq = date_dim.d_week_seq)) otherCondition=()
build RFs:RF4 d_week_seq->[d_week_seq]
@@ -71,6 +68,9 @@ PhysicalResultSink
------------------------------------------------PhysicalProject
--------------------------------------------------filter((date_dim.d_date =
'2001-03-24'))
----------------------------------------------------PhysicalOlapScan[date_dim]
+------------------------------------PhysicalDistribute
+--------------------------------------PhysicalProject
+----------------------------------------PhysicalOlapScan[date_dim]
--------------PhysicalProject
----------------hashAgg[GLOBAL]
------------------PhysicalDistribute
@@ -86,10 +86,7 @@ PhysicalResultSink
----------------------------------PhysicalOlapScan[item]
--------------------------PhysicalDistribute
----------------------------PhysicalProject
-------------------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=() build
RFs:RF1 d_date->[d_date]
---------------------------------PhysicalDistribute
-----------------------------------PhysicalProject
-------------------------------------PhysicalOlapScan[date_dim] apply RFs: RF1
+------------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=()
--------------------------------PhysicalDistribute
----------------------------------PhysicalProject
------------------------------------hashJoin[INNER_JOIN]
hashCondition=((date_dim.d_week_seq = date_dim.d_week_seq)) otherCondition=()
build RFs:RF0 d_week_seq->[d_week_seq]
@@ -102,4 +99,7 @@ PhysicalResultSink
--------------------------------------------PhysicalProject
----------------------------------------------filter((date_dim.d_date =
'2001-03-24'))
------------------------------------------------PhysicalOlapScan[date_dim]
+--------------------------------PhysicalDistribute
+----------------------------------PhysicalProject
+------------------------------------PhysicalOlapScan[date_dim]
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query59.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query59.out
index 5c2a0e7a195..d731b71a900 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query59.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query59.out
@@ -18,31 +18,30 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
--------PhysicalDistribute
----------PhysicalTopN[LOCAL_SORT]
------------PhysicalProject
---------------hashJoin[INNER_JOIN] hashCondition=((d.d_week_seq =
d_week_seq1)) otherCondition=() build RFs:RF5 d_week_seq->[d_week_seq]
+--------------hashJoin[INNER_JOIN] hashCondition=((wss.ss_store_sk =
store.s_store_sk) and (y.s_store_id1 = x.s_store_id2)) otherCondition=() build
RFs:RF4 s_store_id2->[s_store_id];RF5 s_store_sk->[ss_store_sk]
----------------PhysicalProject
-------------------hashJoin[INNER_JOIN] hashCondition=((wss.ss_store_sk =
store.s_store_sk) and (y.s_store_id1 = x.s_store_id2)) otherCondition=() build
RFs:RF3 s_store_id1->[s_store_id];RF4 s_store_sk->[ss_store_sk]
---------------------PhysicalProject
-----------------------hashJoin[INNER_JOIN] hashCondition=((wss.ss_store_sk =
store.s_store_sk)) otherCondition=() build RFs:RF2 s_store_sk->[ss_store_sk]
-------------------------hashJoin[INNER_JOIN] hashCondition=((d.d_week_seq =
d_week_seq2)) otherCondition=() build RFs:RF1 d_week_seq->[d_week_seq]
---------------------------hashJoin[INNER_JOIN]
hashCondition=((expr_cast(d_week_seq1 as BIGINT) = expr_(d_week_seq2 - 52)))
otherCondition=()
-----------------------------PhysicalDistribute
-------------------------------PhysicalProject
---------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
-----------------------------PhysicalDistribute
-------------------------------PhysicalProject
---------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
+------------------hashJoin[INNER_JOIN] hashCondition=((wss.ss_store_sk =
store.s_store_sk)) otherCondition=() build RFs:RF3 s_store_sk->[ss_store_sk]
+--------------------hashJoin[INNER_JOIN] hashCondition=((d.d_week_seq =
d_week_seq1)) otherCondition=() build RFs:RF2 d_week_seq->[d_week_seq]
+----------------------hashJoin[INNER_JOIN] hashCondition=((d.d_week_seq =
d_week_seq2)) otherCondition=() build RFs:RF1 d_week_seq->[d_week_seq]
+------------------------hashJoin[INNER_JOIN]
hashCondition=((expr_cast(d_week_seq1 as BIGINT) = expr_(d_week_seq2 - 52)))
otherCondition=()
--------------------------PhysicalDistribute
----------------------------PhysicalProject
-------------------------------filter((d.d_month_seq <= 1219) and
(d.d_month_seq >= 1208))
---------------------------------PhysicalOlapScan[date_dim]
+------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
+--------------------------PhysicalDistribute
+----------------------------PhysicalProject
+------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
------------------------PhysicalDistribute
--------------------------PhysicalProject
-----------------------------PhysicalOlapScan[store] apply RFs: RF3
+----------------------------filter((d.d_month_seq <= 1219) and (d.d_month_seq
>= 1208))
+------------------------------PhysicalOlapScan[date_dim]
+----------------------PhysicalDistribute
+------------------------PhysicalProject
+--------------------------filter((d.d_month_seq <= 1207) and (d.d_month_seq >=
1196))
+----------------------------PhysicalOlapScan[date_dim]
--------------------PhysicalDistribute
----------------------PhysicalProject
-------------------------PhysicalOlapScan[store]
+------------------------PhysicalOlapScan[store] apply RFs: RF4
----------------PhysicalDistribute
------------------PhysicalProject
---------------------filter((d.d_month_seq <= 1207) and (d.d_month_seq >= 1196))
-----------------------PhysicalOlapScan[date_dim]
+--------------------PhysicalOlapScan[store]
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query94.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query94.out
index 43482826444..a7f921c597b 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query94.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query94.out
@@ -11,7 +11,10 @@ PhysicalResultSink
----------------PhysicalProject
------------------hashJoin[LEFT_ANTI_JOIN] hashCondition=((ws1.ws_order_number
= wr1.wr_order_number)) otherCondition=()
--------------------hashJoin[INNER_JOIN] hashCondition=((ws1.ws_web_site_sk =
web_site.web_site_sk)) otherCondition=() build RFs:RF3
web_site_sk->[ws_web_site_sk]
-----------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = ws2.ws_order_number)) otherCondition=((
not (ws_warehouse_sk = ws_warehouse_sk)))
+----------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = ws2.ws_order_number)) otherCondition=((
not (ws_warehouse_sk = ws_warehouse_sk))) build RFs:RF2
ws_order_number->[ws_order_number]
+------------------------PhysicalDistribute
+--------------------------PhysicalProject
+----------------------------PhysicalOlapScan[web_sales] apply RFs: RF2
------------------------PhysicalDistribute
--------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_addr_sk = customer_address.ca_address_sk))
otherCondition=() build RFs:RF1 ca_address_sk->[ws_ship_addr_sk]
----------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_date_sk = date_dim.d_date_sk)) otherCondition=()
build RFs:RF0 d_date_sk->[ws_ship_date_sk]
@@ -25,9 +28,6 @@ PhysicalResultSink
------------------------------PhysicalProject
--------------------------------filter((customer_address.ca_state = 'OK'))
----------------------------------PhysicalOlapScan[customer_address]
-------------------------PhysicalDistribute
---------------------------PhysicalProject
-----------------------------PhysicalOlapScan[web_sales]
----------------------PhysicalDistribute
------------------------PhysicalProject
--------------------------filter((web_site.web_company_name = 'pri'))
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query95.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query95.out
index 47884ed767d..62afa403979 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query95.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/noStatsRfPrune/query95.out
@@ -3,13 +3,13 @@
PhysicalCteAnchor ( cteId=CTEId#0 )
--PhysicalCteProducer ( cteId=CTEId#0 )
----PhysicalProject
-------hashJoin[INNER_JOIN] hashCondition=((ws1.ws_order_number =
ws2.ws_order_number)) otherCondition=(( not (ws_warehouse_sk =
ws_warehouse_sk))) build RFs:RF0 ws_order_number->[ws_order_number]
+------hashJoin[INNER_JOIN] hashCondition=((ws1.ws_order_number =
ws2.ws_order_number)) otherCondition=(( not (ws_warehouse_sk =
ws_warehouse_sk)))
--------PhysicalDistribute
----------PhysicalProject
-------------PhysicalOlapScan[web_sales] apply RFs: RF0 RF6
+------------PhysicalOlapScan[web_sales]
--------PhysicalDistribute
----------PhysicalProject
-------------PhysicalOlapScan[web_sales] apply RFs: RF6
+------------PhysicalOlapScan[web_sales]
--PhysicalResultSink
----PhysicalTopN[MERGE_SORT]
------PhysicalTopN[LOCAL_SORT]
@@ -19,14 +19,25 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
--------------hashAgg[GLOBAL]
----------------hashAgg[LOCAL]
------------------PhysicalProject
---------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = web_returns.wr_order_number))
otherCondition=()
-----------------------hashJoin[INNER_JOIN] hashCondition=((ws1.ws_web_site_sk
= web_site.web_site_sk)) otherCondition=() build RFs:RF5
web_site_sk->[ws_web_site_sk]
-------------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = ws_wh.ws_order_number)) otherCondition=()
build RFs:RF4 ws_order_number->[ws_order_number]
---------------------------PhysicalDistribute
-----------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_addr_sk = customer_address.ca_address_sk))
otherCondition=() build RFs:RF3 ca_address_sk->[ws_ship_addr_sk]
-------------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_date_sk = date_dim.d_date_sk)) otherCondition=()
build RFs:RF2 d_date_sk->[ws_ship_date_sk]
+--------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = ws_wh.ws_order_number)) otherCondition=()
+----------------------PhysicalDistribute
+------------------------PhysicalProject
+--------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
+----------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = web_returns.wr_order_number))
otherCondition=() build RFs:RF5 ws_order_number->[wr_order_number]
+------------------------PhysicalProject
+--------------------------hashJoin[INNER_JOIN]
hashCondition=((web_returns.wr_order_number = ws_wh.ws_order_number))
otherCondition=() build RFs:RF4 wr_order_number->[ws_order_number]
+----------------------------PhysicalDistribute
+------------------------------PhysicalProject
+--------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
+----------------------------PhysicalDistribute
+------------------------------PhysicalProject
+--------------------------------PhysicalOlapScan[web_returns] apply RFs: RF5
+------------------------PhysicalDistribute
+--------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_web_site_sk = web_site.web_site_sk)) otherCondition=()
build RFs:RF3 web_site_sk->[ws_web_site_sk]
+----------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_addr_sk = customer_address.ca_address_sk))
otherCondition=() build RFs:RF2 ca_address_sk->[ws_ship_addr_sk]
+------------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_date_sk = date_dim.d_date_sk)) otherCondition=()
build RFs:RF1 d_date_sk->[ws_ship_date_sk]
--------------------------------PhysicalProject
-----------------------------------PhysicalOlapScan[web_sales] apply RFs: RF2
RF3 RF4 RF5
+----------------------------------PhysicalOlapScan[web_sales] apply RFs: RF1
RF2 RF3
--------------------------------PhysicalDistribute
----------------------------------PhysicalProject
------------------------------------filter((date_dim.d_date <= '1999-04-02')
and (date_dim.d_date >= '1999-02-01'))
@@ -35,19 +46,8 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
--------------------------------PhysicalProject
----------------------------------filter((customer_address.ca_state = 'NC'))
------------------------------------PhysicalOlapScan[customer_address]
---------------------------PhysicalDistribute
-----------------------------PhysicalProject
-------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
-------------------------PhysicalDistribute
---------------------------PhysicalProject
-----------------------------filter((web_site.web_company_name = 'pri'))
-------------------------------PhysicalOlapScan[web_site]
-----------------------PhysicalProject
-------------------------hashJoin[INNER_JOIN]
hashCondition=((web_returns.wr_order_number = ws_wh.ws_order_number))
otherCondition=() build RFs:RF1 wr_order_number->[ws_order_number];RF6
wr_order_number->[ws_order_number,ws_order_number]
---------------------------PhysicalDistribute
-----------------------------PhysicalProject
-------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
---------------------------PhysicalDistribute
-----------------------------PhysicalProject
-------------------------------PhysicalOlapScan[web_returns]
+----------------------------PhysicalDistribute
+------------------------------PhysicalProject
+--------------------------------filter((web_site.web_company_name = 'pri'))
+----------------------------------PhysicalOlapScan[web_site]
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query14.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query14.out
index ce01e32c216..5aebc906be6 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query14.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query14.out
@@ -157,12 +157,12 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
--------------------------------------hashJoin[INNER_JOIN]
hashCondition=((web_sales.ws_sold_date_sk = date_dim.d_date_sk))
otherCondition=() build RFs:RF20 d_date_sk->[ws_sold_date_sk]
----------------------------------------PhysicalProject
------------------------------------------hashJoin[INNER_JOIN]
hashCondition=((web_sales.ws_item_sk = item.i_item_sk)) otherCondition=() build
RFs:RF19 i_item_sk->[ws_item_sk]
---------------------------------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((web_sales.ws_item_sk = cross_items.ss_item_sk))
otherCondition=() build RFs:RF18 ss_item_sk->[ws_item_sk]
-----------------------------------------------PhysicalDistribute
-------------------------------------------------PhysicalProject
---------------------------------------------------PhysicalOlapScan[web_sales]
apply RFs: RF18 RF19 RF20
+--------------------------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((web_sales.ws_item_sk = cross_items.ss_item_sk))
otherCondition=() build RFs:RF18 ws_item_sk->[ss_item_sk]
----------------------------------------------PhysicalDistribute
------------------------------------------------PhysicalCteConsumer (
cteId=CTEId#0 )
+----------------------------------------------PhysicalDistribute
+------------------------------------------------PhysicalProject
+--------------------------------------------------PhysicalOlapScan[web_sales]
apply RFs: RF19 RF20
--------------------------------------------PhysicalDistribute
----------------------------------------------PhysicalProject
------------------------------------------------PhysicalOlapScan[item]
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query16.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query16.out
index d6737a1cd78..b08f9d62730 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query16.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query16.out
@@ -11,12 +11,15 @@ PhysicalResultSink
----------------PhysicalProject
------------------hashJoin[LEFT_ANTI_JOIN] hashCondition=((cs1.cs_order_number
= cr1.cr_order_number)) otherCondition=()
--------------------hashJoin[INNER_JOIN] hashCondition=((cs1.cs_call_center_sk
= call_center.cc_call_center_sk)) otherCondition=() build RFs:RF3
cc_call_center_sk->[cs_call_center_sk]
-----------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((cs1.cs_order_number = cs2.cs_order_number)) otherCondition=((
not (cs_warehouse_sk = cs_warehouse_sk))) build RFs:RF2
cs_order_number->[cs_order_number]
+----------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((cs1.cs_order_number = cs2.cs_order_number)) otherCondition=((
not (cs_warehouse_sk = cs_warehouse_sk))) build RFs:RF2
cs_order_number->[cs_order_number]
+------------------------PhysicalDistribute
+--------------------------PhysicalProject
+----------------------------PhysicalOlapScan[catalog_sales] apply RFs: RF2
------------------------PhysicalDistribute
--------------------------hashJoin[INNER_JOIN]
hashCondition=((cs1.cs_ship_addr_sk = customer_address.ca_address_sk))
otherCondition=() build RFs:RF1 ca_address_sk->[cs_ship_addr_sk]
----------------------------hashJoin[INNER_JOIN]
hashCondition=((cs1.cs_ship_date_sk = date_dim.d_date_sk)) otherCondition=()
build RFs:RF0 d_date_sk->[cs_ship_date_sk]
------------------------------PhysicalProject
---------------------------------PhysicalOlapScan[catalog_sales] apply RFs: RF0
RF1 RF2 RF3
+--------------------------------PhysicalOlapScan[catalog_sales] apply RFs: RF0
RF1 RF3
------------------------------PhysicalDistribute
--------------------------------PhysicalProject
----------------------------------filter((date_dim.d_date <= '2002-05-31') and
(date_dim.d_date >= '2002-04-01'))
@@ -25,9 +28,6 @@ PhysicalResultSink
------------------------------PhysicalProject
--------------------------------filter((customer_address.ca_state = 'WV'))
----------------------------------PhysicalOlapScan[customer_address]
-------------------------PhysicalDistribute
---------------------------PhysicalProject
-----------------------------PhysicalOlapScan[catalog_sales]
----------------------PhysicalDistribute
------------------------PhysicalProject
--------------------------filter(cc_county IN ('Barrow County', 'Daviess
County', 'Luce County', 'Richland County', 'Ziebach County'))
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query31.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query31.out
index 299fb8974d7..281bc604c6b 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query31.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query31.out
@@ -46,6 +46,10 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
--------------------hashJoin[INNER_JOIN] hashCondition=((ss1.ca_county =
ws1.ca_county)) otherCondition=()
----------------------PhysicalProject
------------------------hashJoin[INNER_JOIN] hashCondition=((ss2.ca_county =
ss3.ca_county)) otherCondition=()
+--------------------------PhysicalDistribute
+----------------------------PhysicalProject
+------------------------------filter((ss3.d_qoy = 3) and (ss3.d_year = 2000))
+--------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
--------------------------hashJoin[INNER_JOIN] hashCondition=((ss1.ca_county =
ss2.ca_county)) otherCondition=()
----------------------------PhysicalDistribute
------------------------------PhysicalProject
@@ -55,10 +59,6 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
------------------------------PhysicalProject
--------------------------------filter((ss2.d_qoy = 2) and (ss2.d_year = 2000))
----------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
---------------------------PhysicalDistribute
-----------------------------PhysicalProject
-------------------------------filter((ss3.d_qoy = 3) and (ss3.d_year = 2000))
---------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
----------------------PhysicalDistribute
------------------------PhysicalProject
--------------------------filter((ws1.d_qoy = 1) and (ws1.d_year = 2000))
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query58.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query58.out
index 0e27975e186..e54081b840d 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query58.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query58.out
@@ -24,22 +24,22 @@ PhysicalResultSink
--------------------------------------PhysicalOlapScan[item] apply RFs: RF12
RF13
------------------------------PhysicalDistribute
--------------------------------PhysicalProject
-----------------------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=() build
RFs:RF9 d_date->[d_date]
-------------------------------------PhysicalDistribute
---------------------------------------PhysicalProject
-----------------------------------------PhysicalOlapScan[date_dim] apply RFs:
RF9
+----------------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=() build
RFs:RF9 d_date->[d_date]
------------------------------------PhysicalDistribute
--------------------------------------PhysicalProject
----------------------------------------hashJoin[INNER_JOIN]
hashCondition=((date_dim.d_week_seq = date_dim.d_week_seq)) otherCondition=()
build RFs:RF8 d_week_seq->[d_week_seq]
------------------------------------------PhysicalDistribute
--------------------------------------------PhysicalProject
-----------------------------------------------PhysicalOlapScan[date_dim] apply
RFs: RF8
+----------------------------------------------PhysicalOlapScan[date_dim] apply
RFs: RF8 RF9
------------------------------------------PhysicalDistribute
--------------------------------------------PhysicalAssertNumRows
----------------------------------------------PhysicalDistribute
------------------------------------------------PhysicalProject
--------------------------------------------------filter((date_dim.d_date =
'2001-03-24'))
----------------------------------------------------PhysicalOlapScan[date_dim]
+------------------------------------PhysicalDistribute
+--------------------------------------PhysicalProject
+----------------------------------------PhysicalOlapScan[date_dim]
------------------PhysicalProject
--------------------hashAgg[GLOBAL]
----------------------PhysicalDistribute
@@ -55,22 +55,22 @@ PhysicalResultSink
--------------------------------------PhysicalOlapScan[item]
------------------------------PhysicalDistribute
--------------------------------PhysicalProject
-----------------------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=() build
RFs:RF5 d_date->[d_date]
-------------------------------------PhysicalDistribute
---------------------------------------PhysicalProject
-----------------------------------------PhysicalOlapScan[date_dim] apply RFs:
RF5
+----------------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=() build
RFs:RF5 d_date->[d_date]
------------------------------------PhysicalDistribute
--------------------------------------PhysicalProject
----------------------------------------hashJoin[INNER_JOIN]
hashCondition=((date_dim.d_week_seq = date_dim.d_week_seq)) otherCondition=()
build RFs:RF4 d_week_seq->[d_week_seq]
------------------------------------------PhysicalDistribute
--------------------------------------------PhysicalProject
-----------------------------------------------PhysicalOlapScan[date_dim] apply
RFs: RF4
+----------------------------------------------PhysicalOlapScan[date_dim] apply
RFs: RF4 RF5
------------------------------------------PhysicalDistribute
--------------------------------------------PhysicalAssertNumRows
----------------------------------------------PhysicalDistribute
------------------------------------------------PhysicalProject
--------------------------------------------------filter((date_dim.d_date =
'2001-03-24'))
----------------------------------------------------PhysicalOlapScan[date_dim]
+------------------------------------PhysicalDistribute
+--------------------------------------PhysicalProject
+----------------------------------------PhysicalOlapScan[date_dim]
--------------PhysicalProject
----------------hashAgg[GLOBAL]
------------------PhysicalDistribute
@@ -86,20 +86,20 @@ PhysicalResultSink
----------------------------------PhysicalOlapScan[item]
--------------------------PhysicalDistribute
----------------------------PhysicalProject
-------------------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=() build
RFs:RF1 d_date->[d_date]
---------------------------------PhysicalDistribute
-----------------------------------PhysicalProject
-------------------------------------PhysicalOlapScan[date_dim] apply RFs: RF1
+------------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((date_dim.d_date = date_dim.d_date)) otherCondition=() build
RFs:RF1 d_date->[d_date]
--------------------------------PhysicalDistribute
----------------------------------PhysicalProject
------------------------------------hashJoin[INNER_JOIN]
hashCondition=((date_dim.d_week_seq = date_dim.d_week_seq)) otherCondition=()
build RFs:RF0 d_week_seq->[d_week_seq]
--------------------------------------PhysicalDistribute
----------------------------------------PhysicalProject
-------------------------------------------PhysicalOlapScan[date_dim] apply
RFs: RF0
+------------------------------------------PhysicalOlapScan[date_dim] apply
RFs: RF0 RF1
--------------------------------------PhysicalDistribute
----------------------------------------PhysicalAssertNumRows
------------------------------------------PhysicalDistribute
--------------------------------------------PhysicalProject
----------------------------------------------filter((date_dim.d_date =
'2001-03-24'))
------------------------------------------------PhysicalOlapScan[date_dim]
+--------------------------------PhysicalDistribute
+----------------------------------PhysicalProject
+------------------------------------PhysicalOlapScan[date_dim]
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query59.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query59.out
index ebc8191498f..65458d6bf5d 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query59.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query59.out
@@ -18,31 +18,30 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
--------PhysicalDistribute
----------PhysicalTopN[LOCAL_SORT]
------------PhysicalProject
---------------hashJoin[INNER_JOIN] hashCondition=((d.d_week_seq =
d_week_seq1)) otherCondition=() build RFs:RF5 d_week_seq->[d_week_seq]
+--------------hashJoin[INNER_JOIN] hashCondition=((wss.ss_store_sk =
store.s_store_sk) and (y.s_store_id1 = x.s_store_id2)) otherCondition=() build
RFs:RF4 s_store_id2->[s_store_id];RF5 s_store_sk->[ss_store_sk]
----------------PhysicalProject
-------------------hashJoin[INNER_JOIN] hashCondition=((wss.ss_store_sk =
store.s_store_sk) and (y.s_store_id1 = x.s_store_id2)) otherCondition=() build
RFs:RF3 s_store_id1->[s_store_id];RF4 s_store_sk->[ss_store_sk]
---------------------PhysicalProject
-----------------------hashJoin[INNER_JOIN] hashCondition=((wss.ss_store_sk =
store.s_store_sk)) otherCondition=() build RFs:RF2 s_store_sk->[ss_store_sk]
-------------------------hashJoin[INNER_JOIN] hashCondition=((d.d_week_seq =
d_week_seq2)) otherCondition=() build RFs:RF1 d_week_seq->[d_week_seq]
---------------------------hashJoin[INNER_JOIN]
hashCondition=((expr_cast(d_week_seq1 as BIGINT) = expr_(d_week_seq2 - 52)))
otherCondition=()
-----------------------------PhysicalDistribute
-------------------------------PhysicalProject
---------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
-----------------------------PhysicalDistribute
-------------------------------PhysicalProject
---------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
+------------------hashJoin[INNER_JOIN] hashCondition=((wss.ss_store_sk =
store.s_store_sk)) otherCondition=() build RFs:RF3 s_store_sk->[ss_store_sk]
+--------------------hashJoin[INNER_JOIN] hashCondition=((d.d_week_seq =
d_week_seq1)) otherCondition=() build RFs:RF2 d_week_seq->[d_week_seq]
+----------------------hashJoin[INNER_JOIN] hashCondition=((d.d_week_seq =
d_week_seq2)) otherCondition=() build RFs:RF1 d_week_seq->[d_week_seq]
+------------------------hashJoin[INNER_JOIN]
hashCondition=((expr_cast(d_week_seq1 as BIGINT) = expr_(d_week_seq2 - 52)))
otherCondition=()
--------------------------PhysicalDistribute
----------------------------PhysicalProject
-------------------------------filter((d.d_month_seq <= 1219) and
(d.d_month_seq >= 1208))
---------------------------------PhysicalOlapScan[date_dim]
+------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
+--------------------------PhysicalDistribute
+----------------------------PhysicalProject
+------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
------------------------PhysicalDistribute
--------------------------PhysicalProject
-----------------------------PhysicalOlapScan[store] apply RFs: RF3
+----------------------------filter((d.d_month_seq <= 1219) and (d.d_month_seq
>= 1208))
+------------------------------PhysicalOlapScan[date_dim]
+----------------------PhysicalDistribute
+------------------------PhysicalProject
+--------------------------filter((d.d_month_seq <= 1207) and (d.d_month_seq >=
1196))
+----------------------------PhysicalOlapScan[date_dim]
--------------------PhysicalDistribute
----------------------PhysicalProject
-------------------------PhysicalOlapScan[store]
+------------------------PhysicalOlapScan[store] apply RFs: RF4
----------------PhysicalDistribute
------------------PhysicalProject
---------------------filter((d.d_month_seq <= 1207) and (d.d_month_seq >= 1196))
-----------------------PhysicalOlapScan[date_dim]
+--------------------PhysicalOlapScan[store]
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query94.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query94.out
index 107f8f94a90..a7f921c597b 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query94.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query94.out
@@ -11,12 +11,15 @@ PhysicalResultSink
----------------PhysicalProject
------------------hashJoin[LEFT_ANTI_JOIN] hashCondition=((ws1.ws_order_number
= wr1.wr_order_number)) otherCondition=()
--------------------hashJoin[INNER_JOIN] hashCondition=((ws1.ws_web_site_sk =
web_site.web_site_sk)) otherCondition=() build RFs:RF3
web_site_sk->[ws_web_site_sk]
-----------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = ws2.ws_order_number)) otherCondition=((
not (ws_warehouse_sk = ws_warehouse_sk))) build RFs:RF2
ws_order_number->[ws_order_number]
+----------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = ws2.ws_order_number)) otherCondition=((
not (ws_warehouse_sk = ws_warehouse_sk))) build RFs:RF2
ws_order_number->[ws_order_number]
+------------------------PhysicalDistribute
+--------------------------PhysicalProject
+----------------------------PhysicalOlapScan[web_sales] apply RFs: RF2
------------------------PhysicalDistribute
--------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_addr_sk = customer_address.ca_address_sk))
otherCondition=() build RFs:RF1 ca_address_sk->[ws_ship_addr_sk]
----------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_date_sk = date_dim.d_date_sk)) otherCondition=()
build RFs:RF0 d_date_sk->[ws_ship_date_sk]
------------------------------PhysicalProject
---------------------------------PhysicalOlapScan[web_sales] apply RFs: RF0 RF1
RF2 RF3
+--------------------------------PhysicalOlapScan[web_sales] apply RFs: RF0 RF1
RF3
------------------------------PhysicalDistribute
--------------------------------PhysicalProject
----------------------------------filter((date_dim.d_date <= '2000-04-01') and
(date_dim.d_date >= '2000-02-01'))
@@ -25,9 +28,6 @@ PhysicalResultSink
------------------------------PhysicalProject
--------------------------------filter((customer_address.ca_state = 'OK'))
----------------------------------PhysicalOlapScan[customer_address]
-------------------------PhysicalDistribute
---------------------------PhysicalProject
-----------------------------PhysicalOlapScan[web_sales]
----------------------PhysicalDistribute
------------------------PhysicalProject
--------------------------filter((web_site.web_company_name = 'pri'))
diff --git
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query95.out
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query95.out
index 47884ed767d..41041719b24 100644
---
a/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query95.out
+++
b/regression-test/data/nereids_tpcds_shape_sf100_p0/no_stats_shape/query95.out
@@ -6,10 +6,10 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
------hashJoin[INNER_JOIN] hashCondition=((ws1.ws_order_number =
ws2.ws_order_number)) otherCondition=(( not (ws_warehouse_sk =
ws_warehouse_sk))) build RFs:RF0 ws_order_number->[ws_order_number]
--------PhysicalDistribute
----------PhysicalProject
-------------PhysicalOlapScan[web_sales] apply RFs: RF0 RF6
+------------PhysicalOlapScan[web_sales] apply RFs: RF0
--------PhysicalDistribute
----------PhysicalProject
-------------PhysicalOlapScan[web_sales] apply RFs: RF6
+------------PhysicalOlapScan[web_sales]
--PhysicalResultSink
----PhysicalTopN[MERGE_SORT]
------PhysicalTopN[LOCAL_SORT]
@@ -19,14 +19,25 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
--------------hashAgg[GLOBAL]
----------------hashAgg[LOCAL]
------------------PhysicalProject
---------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = web_returns.wr_order_number))
otherCondition=()
-----------------------hashJoin[INNER_JOIN] hashCondition=((ws1.ws_web_site_sk
= web_site.web_site_sk)) otherCondition=() build RFs:RF5
web_site_sk->[ws_web_site_sk]
-------------------------hashJoin[LEFT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = ws_wh.ws_order_number)) otherCondition=()
build RFs:RF4 ws_order_number->[ws_order_number]
---------------------------PhysicalDistribute
-----------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_addr_sk = customer_address.ca_address_sk))
otherCondition=() build RFs:RF3 ca_address_sk->[ws_ship_addr_sk]
-------------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_date_sk = date_dim.d_date_sk)) otherCondition=()
build RFs:RF2 d_date_sk->[ws_ship_date_sk]
+--------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = ws_wh.ws_order_number)) otherCondition=()
+----------------------PhysicalDistribute
+------------------------PhysicalProject
+--------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
+----------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((ws1.ws_order_number = web_returns.wr_order_number))
otherCondition=() build RFs:RF5 ws_order_number->[wr_order_number]
+------------------------PhysicalProject
+--------------------------hashJoin[INNER_JOIN]
hashCondition=((web_returns.wr_order_number = ws_wh.ws_order_number))
otherCondition=() build RFs:RF4 wr_order_number->[ws_order_number]
+----------------------------PhysicalDistribute
+------------------------------PhysicalProject
+--------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
+----------------------------PhysicalDistribute
+------------------------------PhysicalProject
+--------------------------------PhysicalOlapScan[web_returns] apply RFs: RF5
+------------------------PhysicalDistribute
+--------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_web_site_sk = web_site.web_site_sk)) otherCondition=()
build RFs:RF3 web_site_sk->[ws_web_site_sk]
+----------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_addr_sk = customer_address.ca_address_sk))
otherCondition=() build RFs:RF2 ca_address_sk->[ws_ship_addr_sk]
+------------------------------hashJoin[INNER_JOIN]
hashCondition=((ws1.ws_ship_date_sk = date_dim.d_date_sk)) otherCondition=()
build RFs:RF1 d_date_sk->[ws_ship_date_sk]
--------------------------------PhysicalProject
-----------------------------------PhysicalOlapScan[web_sales] apply RFs: RF2
RF3 RF4 RF5
+----------------------------------PhysicalOlapScan[web_sales] apply RFs: RF1
RF2 RF3
--------------------------------PhysicalDistribute
----------------------------------PhysicalProject
------------------------------------filter((date_dim.d_date <= '1999-04-02')
and (date_dim.d_date >= '1999-02-01'))
@@ -35,19 +46,8 @@ PhysicalCteAnchor ( cteId=CTEId#0 )
--------------------------------PhysicalProject
----------------------------------filter((customer_address.ca_state = 'NC'))
------------------------------------PhysicalOlapScan[customer_address]
---------------------------PhysicalDistribute
-----------------------------PhysicalProject
-------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
-------------------------PhysicalDistribute
---------------------------PhysicalProject
-----------------------------filter((web_site.web_company_name = 'pri'))
-------------------------------PhysicalOlapScan[web_site]
-----------------------PhysicalProject
-------------------------hashJoin[INNER_JOIN]
hashCondition=((web_returns.wr_order_number = ws_wh.ws_order_number))
otherCondition=() build RFs:RF1 wr_order_number->[ws_order_number];RF6
wr_order_number->[ws_order_number,ws_order_number]
---------------------------PhysicalDistribute
-----------------------------PhysicalProject
-------------------------------PhysicalCteConsumer ( cteId=CTEId#0 )
---------------------------PhysicalDistribute
-----------------------------PhysicalProject
-------------------------------PhysicalOlapScan[web_returns]
+----------------------------PhysicalDistribute
+------------------------------PhysicalProject
+--------------------------------filter((web_site.web_company_name = 'pri'))
+----------------------------------PhysicalOlapScan[web_site]
diff --git
a/regression-test/data/nereids_tpch_shape_sf1000_p0/nostats_rf_prune/q21.out
b/regression-test/data/nereids_tpch_shape_sf1000_p0/nostats_rf_prune/q21.out
index b248bc83d64..d744f431889 100644
--- a/regression-test/data/nereids_tpch_shape_sf1000_p0/nostats_rf_prune/q21.out
+++ b/regression-test/data/nereids_tpch_shape_sf1000_p0/nostats_rf_prune/q21.out
@@ -8,27 +8,27 @@ PhysicalResultSink
----------PhysicalDistribute
------------hashAgg[LOCAL]
--------------PhysicalProject
-----------------hashJoin[INNER_JOIN] hashCondition=((supplier.s_nationkey =
nation.n_nationkey)) otherCondition=() build RFs:RF3 n_nationkey->[s_nationkey]
+----------------hashJoin[INNER_JOIN] hashCondition=((supplier.s_nationkey =
nation.n_nationkey)) otherCondition=() build RFs:RF4 n_nationkey->[s_nationkey]
------------------PhysicalProject
---------------------hashJoin[INNER_JOIN] hashCondition=((supplier.s_suppkey =
l1.l_suppkey)) otherCondition=() build RFs:RF2 s_suppkey->[l_suppkey]
+--------------------hashJoin[INNER_JOIN] hashCondition=((supplier.s_suppkey =
l1.l_suppkey)) otherCondition=() build RFs:RF3 s_suppkey->[l_suppkey]
----------------------PhysicalDistribute
-------------------------hashJoin[INNER_JOIN] hashCondition=((orders.o_orderkey
= l1.l_orderkey)) otherCondition=() build RFs:RF1 o_orderkey->[l_orderkey]
---------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((l2.l_orderkey = l1.l_orderkey)) otherCondition=(( not
(l_suppkey = l_suppkey))) build RFs:RF0 l_orderkey->[l_orderkey]
+------------------------hashJoin[INNER_JOIN] hashCondition=((orders.o_orderkey
= l1.l_orderkey)) otherCondition=() build RFs:RF2 o_orderkey->[l_orderkey]
+--------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((l2.l_orderkey = l1.l_orderkey)) otherCondition=(( not
(l_suppkey = l_suppkey))) build RFs:RF1 l_orderkey->[l_orderkey]
----------------------------PhysicalProject
-------------------------------PhysicalOlapScan[lineitem] apply RFs: RF0
-----------------------------hashJoin[LEFT_ANTI_JOIN]
hashCondition=((l3.l_orderkey = l1.l_orderkey)) otherCondition=(( not
(l_suppkey = l_suppkey)))
-------------------------------PhysicalProject
---------------------------------filter((l1.l_receiptdate > l1.l_commitdate))
-----------------------------------PhysicalOlapScan[lineitem] apply RFs: RF1 RF2
+------------------------------PhysicalOlapScan[lineitem] apply RFs: RF1
+----------------------------hashJoin[RIGHT_ANTI_JOIN]
hashCondition=((l3.l_orderkey = l1.l_orderkey)) otherCondition=(( not
(l_suppkey = l_suppkey))) build RFs:RF0 l_orderkey->[l_orderkey]
------------------------------PhysicalProject
--------------------------------filter((l3.l_receiptdate > l3.l_commitdate))
-----------------------------------PhysicalOlapScan[lineitem]
+----------------------------------PhysicalOlapScan[lineitem] apply RFs: RF0
+------------------------------PhysicalProject
+--------------------------------filter((l1.l_receiptdate > l1.l_commitdate))
+----------------------------------PhysicalOlapScan[lineitem] apply RFs: RF2 RF3
--------------------------PhysicalProject
----------------------------filter((orders.o_orderstatus = 'F'))
------------------------------PhysicalOlapScan[orders]
----------------------PhysicalDistribute
------------------------PhysicalProject
---------------------------PhysicalOlapScan[supplier] apply RFs: RF3
+--------------------------PhysicalOlapScan[supplier] apply RFs: RF4
------------------PhysicalDistribute
--------------------PhysicalProject
----------------------filter((nation.n_name = 'SAUDI ARABIA'))
diff --git
a/regression-test/data/nereids_tpch_shape_sf1000_p0/shape_no_stats/q21.out
b/regression-test/data/nereids_tpch_shape_sf1000_p0/shape_no_stats/q21.out
index b248bc83d64..d744f431889 100644
--- a/regression-test/data/nereids_tpch_shape_sf1000_p0/shape_no_stats/q21.out
+++ b/regression-test/data/nereids_tpch_shape_sf1000_p0/shape_no_stats/q21.out
@@ -8,27 +8,27 @@ PhysicalResultSink
----------PhysicalDistribute
------------hashAgg[LOCAL]
--------------PhysicalProject
-----------------hashJoin[INNER_JOIN] hashCondition=((supplier.s_nationkey =
nation.n_nationkey)) otherCondition=() build RFs:RF3 n_nationkey->[s_nationkey]
+----------------hashJoin[INNER_JOIN] hashCondition=((supplier.s_nationkey =
nation.n_nationkey)) otherCondition=() build RFs:RF4 n_nationkey->[s_nationkey]
------------------PhysicalProject
---------------------hashJoin[INNER_JOIN] hashCondition=((supplier.s_suppkey =
l1.l_suppkey)) otherCondition=() build RFs:RF2 s_suppkey->[l_suppkey]
+--------------------hashJoin[INNER_JOIN] hashCondition=((supplier.s_suppkey =
l1.l_suppkey)) otherCondition=() build RFs:RF3 s_suppkey->[l_suppkey]
----------------------PhysicalDistribute
-------------------------hashJoin[INNER_JOIN] hashCondition=((orders.o_orderkey
= l1.l_orderkey)) otherCondition=() build RFs:RF1 o_orderkey->[l_orderkey]
---------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((l2.l_orderkey = l1.l_orderkey)) otherCondition=(( not
(l_suppkey = l_suppkey))) build RFs:RF0 l_orderkey->[l_orderkey]
+------------------------hashJoin[INNER_JOIN] hashCondition=((orders.o_orderkey
= l1.l_orderkey)) otherCondition=() build RFs:RF2 o_orderkey->[l_orderkey]
+--------------------------hashJoin[RIGHT_SEMI_JOIN]
hashCondition=((l2.l_orderkey = l1.l_orderkey)) otherCondition=(( not
(l_suppkey = l_suppkey))) build RFs:RF1 l_orderkey->[l_orderkey]
----------------------------PhysicalProject
-------------------------------PhysicalOlapScan[lineitem] apply RFs: RF0
-----------------------------hashJoin[LEFT_ANTI_JOIN]
hashCondition=((l3.l_orderkey = l1.l_orderkey)) otherCondition=(( not
(l_suppkey = l_suppkey)))
-------------------------------PhysicalProject
---------------------------------filter((l1.l_receiptdate > l1.l_commitdate))
-----------------------------------PhysicalOlapScan[lineitem] apply RFs: RF1 RF2
+------------------------------PhysicalOlapScan[lineitem] apply RFs: RF1
+----------------------------hashJoin[RIGHT_ANTI_JOIN]
hashCondition=((l3.l_orderkey = l1.l_orderkey)) otherCondition=(( not
(l_suppkey = l_suppkey))) build RFs:RF0 l_orderkey->[l_orderkey]
------------------------------PhysicalProject
--------------------------------filter((l3.l_receiptdate > l3.l_commitdate))
-----------------------------------PhysicalOlapScan[lineitem]
+----------------------------------PhysicalOlapScan[lineitem] apply RFs: RF0
+------------------------------PhysicalProject
+--------------------------------filter((l1.l_receiptdate > l1.l_commitdate))
+----------------------------------PhysicalOlapScan[lineitem] apply RFs: RF2 RF3
--------------------------PhysicalProject
----------------------------filter((orders.o_orderstatus = 'F'))
------------------------------PhysicalOlapScan[orders]
----------------------PhysicalDistribute
------------------------PhysicalProject
---------------------------PhysicalOlapScan[supplier] apply RFs: RF3
+--------------------------PhysicalOlapScan[supplier] apply RFs: RF4
------------------PhysicalDistribute
--------------------PhysicalProject
----------------------filter((nation.n_name = 'SAUDI ARABIA'))
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@doris.apache.org
For additional commands, e-mail: commits-h...@doris.apache.org
