combine projects (#1282)

src/optimizer/CMakeLists.txt

@@ -14,6 +14,7 @@ nebula_add_library(
     rule/PushFilterDownGetNbrsRule.cpp
     rule/RemoveNoopProjectRule.cpp
     rule/CombineFilterRule.cpp
+    rule/CollapseProjectRule.cpp
     rule/MergeGetVerticesAndDedupRule.cpp
     rule/MergeGetVerticesAndProjectRule.cpp
     rule/MergeGetNbrsAndDedupRule.cpp

src/optimizer/rule/CollapseProjectRule.cpp

@@ -0,0 +1,121 @@
+/* Copyright (c) 2021 vesoft inc. All rights reserved.
+ *
+ * This source code is licensed under Apache 2.0 License,
+ * attached with Common Clause Condition 1.0, found in the LICENSES directory.
+ */
+
+#include "optimizer/rule/CollapseProjectRule.h"
+
+#include "optimizer/OptContext.h"
+#include "optimizer/OptGroup.h"
+#include "planner/plan/PlanNode.h"
+#include "planner/plan/Query.h"
+
+using nebula::graph::PlanNode;
+using nebula::graph::QueryContext;
+
+namespace nebula {
+namespace opt {
+
+std::unique_ptr<OptRule> CollapseProjectRule::kInstance =
+    std::unique_ptr<CollapseProjectRule>(new CollapseProjectRule());
+
+CollapseProjectRule::CollapseProjectRule() {
+    RuleSet::QueryRules().addRule(this);
+}
+
+const Pattern& CollapseProjectRule::pattern() const {
+    static Pattern pattern = Pattern::create(graph::PlanNode::Kind::kProject,
+                                             {Pattern::create(graph::PlanNode::Kind::kProject)});
+    return pattern;
+}
+
+StatusOr<OptRule::TransformResult> CollapseProjectRule::transform(
+    OptContext* octx,
+    const MatchedResult& matched) const {
+    const auto* groupNodeAbove = matched.node;
+    const auto* projAbove = groupNodeAbove->node();
+    DCHECK_EQ(projAbove->kind(), PlanNode::Kind::kProject);
+    const auto& deps = matched.dependencies;
+    DCHECK_EQ(deps.size(), 1);
+    const auto* groupNodeBelow = deps[0].node;
+    const auto* projBelow = groupNodeBelow->node();
+    DCHECK_EQ(projBelow->kind(), PlanNode::Kind::kProject);
+    std::vector<YieldColumn*> colsBelow =
+        static_cast<const graph::Project*>(projBelow)->columns()->columns();
+    const auto* projGroup = groupNodeAbove->group();
+
+    auto* newProj = static_cast<graph::Project*>(projAbove->clone());
+    std::vector<YieldColumn*> colsAbove = newProj->columns()->columns();
+
+    // 1. collect all property reference
+    std::vector<std::string> allPropRefNames;
+    for (auto col : colsAbove) {
+        std::vector<const Expression*> propRefs = graph::ExpressionUtils::collectAll(
+            col->expr(), {Expression::Kind::kVarProperty, Expression::Kind::kInputProperty});
+        for (auto* expr : propRefs) {
+            DCHECK(expr->kind() == Expression::Kind::kVarProperty ||
+                   expr->kind() == Expression::Kind::kInputProperty);
+            allPropRefNames.emplace_back(static_cast<const PropertyExpression*>(expr)->prop());
+        }
+    }
+
+    // disable this case to avoid the expression in ProjBelow being eval multiple times
+    std::unordered_set<std::string> uniquePropRefNames;
+    for (auto p : allPropRefNames) {
+        if (!uniquePropRefNames.insert(p).second) {
+            return TransformResult::noTransform();
+        }
+    }
+
+    // 2. find link according to propRefNames and colNames in ProjBelow
+    std::unordered_map<std::string, Expression*> rewriteMap;
+    auto colNames = projBelow->colNames();
+    for (size_t i = 0; i < colNames.size(); ++i) {
+        if (uniquePropRefNames.count(colNames[i])) {
+            rewriteMap[colNames[i]] = colsBelow[i]->expr();
+        }
+    }
+
+    // 3. rewrite YieldColumns
+    auto matcher = [&rewriteMap](const Expression* e) -> bool {
+        if (e->kind() != Expression::Kind::kVarProperty &&
+            e->kind() != Expression::Kind::kInputProperty) {
+            return false;
+        }
+        auto& propName = static_cast<const PropertyExpression*>(e)->prop();
+        return rewriteMap.find(propName) != rewriteMap.end();
+    };
+    auto rewriter = [&rewriteMap](const Expression* e) -> Expression* {
+        DCHECK(e->kind() == Expression::Kind::kVarProperty ||
+               e->kind() == Expression::Kind::kInputProperty);
+        auto& propName = static_cast<const PropertyExpression*>(e)->prop();
+        return rewriteMap[propName]->clone();
+    };
+    for (auto col : colsAbove) {
+        auto* newColExpr = graph::RewriteVisitor::transform(col->expr(), matcher, rewriter);
+        col->setExpr(newColExpr);
+    }
+
+    // 4. rebuild OptGroupNode
+    newProj->setInputVar(projBelow->inputVar());
+    auto* newGroupNode = OptGroupNode::create(octx, newProj, projGroup);
+    newGroupNode->setDeps(groupNodeBelow->dependencies());
+
+    TransformResult result;
+    result.eraseAll = true;
+    result.newGroupNodes.emplace_back(newGroupNode);
+
+    return result;
+}
+
+bool CollapseProjectRule::match(OptContext* octx, const MatchedResult& matched) const {
+    return OptRule::match(octx, matched);
+}
+
+std::string CollapseProjectRule::toString() const {
+    return "CollapseProjectRule";
+}
+
+}   // namespace opt
+}   // namespace nebula

src/optimizer/rule/CollapseProjectRule.h

@@ -0,0 +1,47 @@
+/* Copyright (c) 2021 vesoft inc. All rights reserved.
+ *
+ * This source code is licensed under Apache 2.0 License,
+ * attached with Common Clause Condition 1.0, found in the LICENSES directory.
+ */
+
+#ifndef OPTIMIZER_RULE_COLLAPSEPROJECTRULE_H_
+#define OPTIMIZER_RULE_COLLAPSEPROJECTRULE_H_
+
+#include <memory>
+
+#include "optimizer/OptRule.h"
+
+namespace nebula {
+namespace opt {
+
+/**
+ * Combines two [[Project]] nodes into one
+ * Required conditions:
+ *  1. Match the pattern
+ *  2. Expressions between nodes cannot be referenced more than once
+ * Benefits:
+ *  1. reduce the copy of memory between nodes
+ *  2. reduces expression overhead in some cases(similar to column pruning)
+ */
+
+class CollapseProjectRule final : public OptRule {
+public:
+    const Pattern &pattern() const override;
+
+    StatusOr<TransformResult> transform(OptContext *ctx,
+                                        const MatchedResult &matched) const override;
+
+    bool match(OptContext *ctx, const MatchedResult &matched) const override;
+
+    std::string toString() const override;
+
+private:
+    CollapseProjectRule();
+
+    static std::unique_ptr<OptRule> kInstance;
+};
+
+}   // namespace opt
+}   // namespace nebula
+
+#endif   // OPTIMIZER_RULE_COLLAPSEPROJECTRULE_H_

src/optimizer/rule/PushFilterDownAggregateRule.cpp

@@ -79,7 +79,7 @@ StatusOr<OptRule::TransformResult> PushFilterDownAggregateRule::transform(
             return false;
         }
         auto& propName = static_cast<const VariablePropertyExpression*>(e)->prop();
-        return rewriteMap[propName];
+        return rewriteMap.find(propName) != rewriteMap.end();
     };
     auto rewriter = [&rewriteMap](const Expression* e) -> Expression* {
         DCHECK_EQ(e->kind(), Expression::Kind::kVarProperty);

tests/tck/features/optimizer/CollapseProjectRule.feature

@@ -0,0 +1,54 @@
+# Copyright (c) 2021 vesoft inc. All rights reserved.
+#
+# This source code is licensed under Apache 2.0 License,
+# attached with Common Clause Condition 1.0, found in the LICENSES directory.
+Feature: Collapse Project Rule
+
+  Background:
+    Given a graph with space named "nba"
+
+  Scenario: Collapse Project
+    When profiling query:
+      """
+      MATCH (v:player)
+      WHERE v.age > 38
+      WITH v, v.age AS age, v.age/10 AS iage, v.age%10 AS mage,v.name AS name
+      RETURN iage
+      """
+    Then the result should be, in any order:
+      | iage |
+      | 4    |
+      | 4    |
+      | 4    |
+      | 4    |
+      | 4    |
+      | 4    |
+      | 4    |
+      | 4    |
+      | 4    |
+      | 4    |
+      | 3    |
+    And the execution plan should be:
+      | id | name        | dependencies | operator info |
+      | 16 | Project     | 14           |               |
+      | 14 | Filter      | 7            |               |
+      | 7  | Project     | 6            |               |
+      | 6  | Project     | 5            |               |
+      | 5  | Filter      | 18           |               |
+      | 18 | GetVertices | 12           |               |
+      | 12 | IndexScan   | 0            |               |
+      | 0  | Start       |              |               |
+    When profiling query:
+      """
+      LOOKUP ON player
+      WHERE player.name=='Tim Duncan'
+      | YIELD $-.VertexID AS vid
+      """
+    Then the result should be, in any order:
+      | vid          |
+      | "Tim Duncan" |
+    And the execution plan should be:
+      | id | name               | dependencies | operator info |
+      | 6  | Project            | 5            |               |
+      | 5  | TagIndexPrefixScan | 0            |               |
+      | 0  | Start              |              |               |

tests/tck/features/optimizer/PushFilterDownLeftJoinRule.feature

@@ -23,25 +23,24 @@ Feature: Push Filter down LeftJoin rule
       | "Tim Duncan"   |
     And the execution plan should be:
       | id | name               | dependencies | operator info |
-      | 22 | Project            | 21           |               |
-      | 21 | Filter             | 20           |               |
-      | 20 | InnerJoin          | 19           |               |
-      | 19 | LeftJoin           | 18           |               |
-      | 18 | Project            | 17           |               |
-      | 17 | GetVertices        | 16           |               |
-      | 16 | Project            | 28           |               |
-      | 28 | GetNeighbors       | 12           |               |
-      | 12 | Project            | 11           |               |
-      | 11 | Filter             | 10           |               |
-      | 10 | InnerJoin          | 9            |               |
-      | 9  | LeftJoin           | 8            |               |
-      | 8  | Project            | 7            |               |
-      | 7  | GetVertices        | 6            |               |
-      | 6  | Project            | 27           |               |
-      | 27 | GetNeighbors       | 2            |               |
-      | 2  | Project            | 3            |               |
-      | 3  | Project            | 30           |               |
-      | 30 | TagIndexPrefixScan | 0            |               |
+      | 24 | Project            | 23           |               |
+      | 23 | Filter             | 22           |               |
+      | 22 | InnerJoin          | 21           |               |
+      | 21 | LeftJoin           | 20           |               |
+      | 20 | Project            | 19           |               |
+      | 19 | GetVertices        | 18           |               |
+      | 18 | Project            | 31           |               |
+      | 31 | GetNeighbors       | 14           |               |
+      | 14 | Project            | 13           |               |
+      | 13 | Filter             | 12           |               |
+      | 12 | InnerJoin          | 11           |               |
+      | 11 | LeftJoin           | 10           |               |
+      | 10 | Project            | 9            |               |
+      | 9  | GetVertices        | 8            |               |
+      | 8  | Project            | 30           |               |
+      | 30 | GetNeighbors       | 27           |               |
+      | 27 | Project            | 25           |               |
+      | 25 | TagIndexPrefixScan | 0            |               |
       | 0  | Start              |              |               |
     When profiling query:
       """