Revert: Move boolean simplification & remove rewrite.

apache · May 7, 2016 · e28bbb6 · e28bbb6
1 parent b487925
commit e28bbb6
Show file tree

Hide file tree

Showing 9 changed files with 109 additions and 131 deletions.
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
@@ -26,6 +26,7 @@ import org.apache.spark.sql.catalyst.catalog.{InMemoryCatalog, SessionCatalog}
 import org.apache.spark.sql.catalyst.encoders.OuterScopes
 import org.apache.spark.sql.catalyst.expressions._
 import org.apache.spark.sql.catalyst.expressions.aggregate._
+import org.apache.spark.sql.catalyst.optimizer.BooleanSimplification
 import org.apache.spark.sql.catalyst.planning.IntegerIndex
 import org.apache.spark.sql.catalyst.plans._
 import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, _}

diff --git a/...atalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/BooleanSimplification.scala b/...atalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/BooleanSimplification.scala
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala
@@ -22,7 +22,7 @@ import scala.collection.immutable.HashSet
 import scala.collection.mutable.ArrayBuffer
 
 import org.apache.spark.sql.catalyst.{CatalystConf, SimpleCatalystConf}
-import org.apache.spark.sql.catalyst.analysis.{BooleanSimplification, CleanupAliases, DistinctAggregationRewriter, EliminateSubqueryAliases, EmptyFunctionRegistry}
+import org.apache.spark.sql.catalyst.analysis.{CleanupAliases, DistinctAggregationRewriter, EliminateSubqueryAliases, EmptyFunctionRegistry}
 import org.apache.spark.sql.catalyst.catalog.{InMemoryCatalog, SessionCatalog}
 import org.apache.spark.sql.catalyst.expressions._
 import org.apache.spark.sql.catalyst.expressions.aggregate._
@@ -688,6 +688,108 @@ case class OptimizeIn(conf: CatalystConf) extends Rule[LogicalPlan] {
   }
 }
 
+/**
+ * Simplifies boolean expressions:
+ * 1. Simplifies expressions whose answer can be determined without evaluating both sides.
+ * 2. Eliminates / extracts common factors.
+ * 3. Merge same expressions
+ * 4. Removes `Not` operator.
+ */
+object BooleanSimplification extends Rule[LogicalPlan] with PredicateHelper {
+  def apply(plan: LogicalPlan): LogicalPlan = plan transform {
+    case q: LogicalPlan => q transformExpressionsUp {
+      case TrueLiteral And e => e
+      case e And TrueLiteral => e
+      case FalseLiteral Or e => e
+      case e Or FalseLiteral => e
+
+      case FalseLiteral And _ => FalseLiteral
+      case _ And FalseLiteral => FalseLiteral
+      case TrueLiteral Or _ => TrueLiteral
+      case _ Or TrueLiteral => TrueLiteral
+
+      case a And b if a.semanticEquals(b) => a
+      case a Or b if a.semanticEquals(b) => a
+
+      case a And (b Or c) if Not(a).semanticEquals(b) => And(a, c)
+      case a And (b Or c) if Not(a).semanticEquals(c) => And(a, b)
+      case (a Or b) And c if a.semanticEquals(Not(c)) => And(b, c)
+      case (a Or b) And c if b.semanticEquals(Not(c)) => And(a, c)
+
+      case a Or (b And c) if Not(a).semanticEquals(b) => Or(a, c)
+      case a Or (b And c) if Not(a).semanticEquals(c) => Or(a, b)
+      case (a And b) Or c if a.semanticEquals(Not(c)) => Or(b, c)
+      case (a And b) Or c if b.semanticEquals(Not(c)) => Or(a, c)
+
+      // Common factor elimination for conjunction
+      case and @ (left And right) =>
+        // 1. Split left and right to get the disjunctive predicates,
+        //   i.e. lhs = (a, b), rhs = (a, c)
+        // 2. Find the common predict between lhsSet and rhsSet, i.e. common = (a)
+        // 3. Remove common predict from lhsSet and rhsSet, i.e. ldiff = (b), rdiff = (c)
+        // 4. Apply the formula, get the optimized predicate: common || (ldiff && rdiff)
+        val lhs = splitDisjunctivePredicates(left)
+        val rhs = splitDisjunctivePredicates(right)
+        val common = lhs.filter(e => rhs.exists(e.semanticEquals))
+        if (common.isEmpty) {
+          // No common factors, return the original predicate
+          and
+        } else {
+          val ldiff = lhs.filterNot(e => common.exists(e.semanticEquals))
+          val rdiff = rhs.filterNot(e => common.exists(e.semanticEquals))
+          if (ldiff.isEmpty || rdiff.isEmpty) {
+            // (a || b || c || ...) && (a || b) => (a || b)
+            common.reduce(Or)
+          } else {
+            // (a || b || c || ...) && (a || b || d || ...) =>
+            // ((c || ...) && (d || ...)) || a || b
+            (common :+ And(ldiff.reduce(Or), rdiff.reduce(Or))).reduce(Or)
+          }
+        }
+
+      // Common factor elimination for disjunction
+      case or @ (left Or right) =>
+        // 1. Split left and right to get the conjunctive predicates,
+        //   i.e.  lhs = (a, b), rhs = (a, c)
+        // 2. Find the common predict between lhsSet and rhsSet, i.e. common = (a)
+        // 3. Remove common predict from lhsSet and rhsSet, i.e. ldiff = (b), rdiff = (c)
+        // 4. Apply the formula, get the optimized predicate: common && (ldiff || rdiff)
+        val lhs = splitConjunctivePredicates(left)
+        val rhs = splitConjunctivePredicates(right)
+        val common = lhs.filter(e => rhs.exists(e.semanticEquals))
+        if (common.isEmpty) {
+          // No common factors, return the original predicate
+          or
+        } else {
+          val ldiff = lhs.filterNot(e => common.exists(e.semanticEquals))
+          val rdiff = rhs.filterNot(e => common.exists(e.semanticEquals))
+          if (ldiff.isEmpty || rdiff.isEmpty) {
+            // (a && b) || (a && b && c && ...) => a && b
+            common.reduce(And)
+          } else {
+            // (a && b && c && ...) || (a && b && d && ...) =>
+            // ((c && ...) || (d && ...)) && a && b
+            (common :+ Or(ldiff.reduce(And), rdiff.reduce(And))).reduce(And)
+          }
+        }
+
+      case Not(TrueLiteral) => FalseLiteral
+      case Not(FalseLiteral) => TrueLiteral
+
+      case Not(a GreaterThan b) => LessThanOrEqual(a, b)
+      case Not(a GreaterThanOrEqual b) => LessThan(a, b)
+
+      case Not(a LessThan b) => GreaterThanOrEqual(a, b)
+      case Not(a LessThanOrEqual b) => GreaterThan(a, b)
+
+      case Not(a Or b) => And(Not(a), Not(b))
+      case Not(a And b) => Or(Not(a), Not(b))
+
+      case Not(Not(e)) => e
+    }
+  }
+}
+
 /**
  * Simplifies binary comparisons with semantically-equal expressions:
  * 1) Replace '<=>' with 'true' literal.

diff --git a/...atalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/CombiningLimitsSuite.scala b/...atalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/CombiningLimitsSuite.scala
@@ -17,7 +17,6 @@
 
 package org.apache.spark.sql.catalyst.optimizer
 
-import org.apache.spark.sql.catalyst.analysis.BooleanSimplification
 import org.apache.spark.sql.catalyst.dsl.expressions._
 import org.apache.spark.sql.catalyst.dsl.plans._
 import org.apache.spark.sql.catalyst.plans.logical._

diff --git a/...atalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/ConstantFoldingSuite.scala b/...atalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/ConstantFoldingSuite.scala
@@ -18,7 +18,7 @@
 package org.apache.spark.sql.catalyst.optimizer
 
 import org.apache.spark.sql.catalyst.SimpleCatalystConf
-import org.apache.spark.sql.catalyst.analysis.{BooleanSimplification, EliminateSubqueryAliases, UnresolvedExtractValue}
+import org.apache.spark.sql.catalyst.analysis.{EliminateSubqueryAliases, UnresolvedExtractValue}
 import org.apache.spark.sql.catalyst.dsl.expressions._
 import org.apache.spark.sql.catalyst.dsl.plans._
 import org.apache.spark.sql.catalyst.expressions._

diff --git a/...catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/FilterPushdownSuite.scala b/...catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/FilterPushdownSuite.scala
@@ -18,7 +18,7 @@
 package org.apache.spark.sql.catalyst.optimizer
 
 import org.apache.spark.sql.catalyst.analysis
-import org.apache.spark.sql.catalyst.analysis.{BooleanSimplification, EliminateSubqueryAliases}
+import org.apache.spark.sql.catalyst.analysis.EliminateSubqueryAliases
 import org.apache.spark.sql.catalyst.dsl.expressions._
 import org.apache.spark.sql.catalyst.dsl.plans._
 import org.apache.spark.sql.catalyst.expressions._

diff --git a/...talyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/JoinOptimizationSuite.scala b/...talyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/JoinOptimizationSuite.scala
@@ -18,7 +18,7 @@
 package org.apache.spark.sql.catalyst.optimizer
 
 import org.apache.spark.sql.catalyst.analysis
-import org.apache.spark.sql.catalyst.analysis.{BooleanSimplification, EliminateSubqueryAliases}
+import org.apache.spark.sql.catalyst.analysis.EliminateSubqueryAliases
 import org.apache.spark.sql.catalyst.dsl.expressions._
 import org.apache.spark.sql.catalyst.dsl.plans._
 import org.apache.spark.sql.catalyst.expressions.Expression

diff --git a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/LimitPushdownSuite.scala b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/LimitPushdownSuite.scala
@@ -17,7 +17,7 @@
 
 package org.apache.spark.sql.catalyst.optimizer
 
-import org.apache.spark.sql.catalyst.analysis.{BooleanSimplification, EliminateSubqueryAliases}
+import org.apache.spark.sql.catalyst.analysis.EliminateSubqueryAliases
 import org.apache.spark.sql.catalyst.dsl.expressions._
 import org.apache.spark.sql.catalyst.dsl.plans._
 import org.apache.spark.sql.catalyst.expressions.Add

diff --git a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeInSuite.scala b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeInSuite.scala
@@ -18,7 +18,7 @@
 package org.apache.spark.sql.catalyst.optimizer
 
 import org.apache.spark.sql.catalyst.SimpleCatalystConf
-import org.apache.spark.sql.catalyst.analysis.{BooleanSimplification, EliminateSubqueryAliases, UnresolvedAttribute}
+import org.apache.spark.sql.catalyst.analysis.{EliminateSubqueryAliases, UnresolvedAttribute}
 import org.apache.spark.sql.catalyst.dsl.expressions._
 import org.apache.spark.sql.catalyst.dsl.plans._
 import org.apache.spark.sql.catalyst.expressions._