[SPARK-21499] [SQL] Support creating persistent function for Spark UDAF(UserDefinedAggregateFunction)

sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/catalog/SessionCatalog.scala

@@ -17,13 +17,15 @@
 
 package org.apache.spark.sql.catalyst.catalog
 
+import java.lang.reflect.InvocationTargetException
 import java.net.URI
 import java.util.Locale
 import java.util.concurrent.Callable
 import javax.annotation.concurrent.GuardedBy
 
 import scala.collection.mutable
 import scala.util.{Failure, Success, Try}
+import scala.util.control.NonFatal
 
 import com.google.common.cache.{Cache, CacheBuilder}
 import org.apache.hadoop.conf.Configuration
@@ -40,6 +42,7 @@ import org.apache.spark.sql.catalyst.plans.logical.{LogicalPlan, SubqueryAlias,
 import org.apache.spark.sql.catalyst.util.StringUtils
 import org.apache.spark.sql.internal.SQLConf
 import org.apache.spark.sql.types.StructType
+import org.apache.spark.util.Utils
 
 object SessionCatalog {
   val DEFAULT_DATABASE = "default"
@@ -1096,8 +1099,42 @@ class SessionCatalog(
    * This performs reflection to decide what type of [[Expression]] to return in the builder.
    */
   protected def makeFunctionBuilder(name: String, functionClassName: String): FunctionBuilder = {
-    // TODO: at least support UDAFs here
-    throw new UnsupportedOperationException("Use sqlContext.udf.register(...) instead.")
+    makeFunctionBuilder(name, Utils.classForName(functionClassName))
+  }
+
+  /**
+   * Construct a [[FunctionBuilder]] based on the provided class that represents a function.
+   */
+  private def makeFunctionBuilder(name: String, clazz: Class[_]): FunctionBuilder = {
+    // When we instantiate ScalaUDAF class, we may throw exception if the input
+    // expressions don't satisfy the UDAF, such as type mismatch, input number
+    // mismatch, etc. Here we catch the exception and throw AnalysisException instead.
+    (children: Seq[Expression]) => {
+      try {
+        val clsForUDAF =
+          Utils.classForName("org.apache.spark.sql.expressions.UserDefinedAggregateFunction")
+        if (clsForUDAF.isAssignableFrom(clazz)) {
+          val cls = Utils.classForName("org.apache.spark.sql.execution.aggregate.ScalaUDAF")
+          cls.getConstructor(classOf[Seq[Expression]], clsForUDAF, classOf[Int], classOf[Int])
+            .newInstance(children, clazz.newInstance().asInstanceOf[Object], Int.box(1), Int.box(1))
+            .asInstanceOf[Expression]
+        } else {
+          throw new UnsupportedOperationException("Use sqlContext.udf.register(...) instead.")
+        }
+      } catch {
+        case NonFatal(exception) =>
+          val e = exception match {
+            // Since we are using shim, the exceptions thrown by the underlying method of
+            // Method.invoke() are wrapped by InvocationTargetException
+            case i: InvocationTargetException => i.getCause
+            case o => o
+          }
+          val analysisException =
+            new AnalysisException(s"No handler for UDAF '${clazz.getCanonicalName}': $e")
+          analysisException.setStackTrace(e.getStackTrace)
+          throw analysisException
+      }
+    }
   }
 
   /**
@@ -1121,7 +1158,14 @@ class SessionCatalog(
     }
     val info = new ExpressionInfo(funcDefinition.className, func.database.orNull, func.funcName)
     val builder =
-      functionBuilder.getOrElse(makeFunctionBuilder(func.unquotedString, funcDefinition.className))
+      functionBuilder.getOrElse {
+        val className = funcDefinition.className
+        if (!Utils.classIsLoadable(className)) {
+          throw new AnalysisException(s"Can not load class '$className' when registering " +
+            s"the function '$func', please make sure it is on the classpath")
+        }
+        makeFunctionBuilder(func.unquotedString, className)
+      }
     functionRegistry.registerFunction(func, info, builder)
   }
 

sql/core/src/test/resources/sql-tests/inputs/udaf.sql

@@ -0,0 +1,13 @@
+CREATE OR REPLACE TEMPORARY VIEW t1 AS SELECT * FROM VALUES
+(1), (2), (3), (4)
+as t1(int_col1);
+
+CREATE FUNCTION myDoubleAvg AS 'test.org.apache.spark.sql.MyDoubleAvg';
+
+SELECT default.myDoubleAvg(int_col1) as my_avg from t1;
+
+SELECT default.myDoubleAvg(int_col1, 3) as my_avg from t1;
+
+CREATE FUNCTION udaf1 AS 'test.non.existent.udaf';
+
+SELECT default.udaf1(int_col1) as udaf1 from t1;

sql/core/src/test/resources/sql-tests/results/udaf.sql.out

@@ -0,0 +1,54 @@
+-- Automatically generated by SQLQueryTestSuite
+-- Number of queries: 6
+
+
+-- !query 0
+CREATE OR REPLACE TEMPORARY VIEW t1 AS SELECT * FROM VALUES
+(1), (2), (3), (4)
+as t1(int_col1)
+-- !query 0 schema
+struct<>
+-- !query 0 output
+
+
+
+-- !query 1
+CREATE FUNCTION myDoubleAvg AS 'test.org.apache.spark.sql.MyDoubleAvg'
+-- !query 1 schema
+struct<>
+-- !query 1 output
+
+
+
+-- !query 2
+SELECT default.myDoubleAvg(int_col1) as my_avg from t1
+-- !query 2 schema
+struct<my_avg:double>
+-- !query 2 output
+102.5
+
+
+-- !query 3
+SELECT default.myDoubleAvg(int_col1, 3) as my_avg from t1
+-- !query 3 schema
+struct<>
+-- !query 3 output
+java.lang.AssertionError
+assertion failed: Incorrect number of children
+
+
+-- !query 4
+CREATE FUNCTION udaf1 AS 'test.non.existent.udaf'
+-- !query 4 schema
+struct<>
+-- !query 4 output
+
+
+
+-- !query 5
+SELECT default.udaf1(int_col1) as udaf1 from t1
+-- !query 5 schema
+struct<>
+-- !query 5 output
+org.apache.spark.sql.AnalysisException
+Can not load class 'test.non.existent.udaf' when registering the function 'default.udaf1', please make sure it is on the classpath; line 1 pos 7

sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveSessionCatalog.scala

@@ -34,6 +34,8 @@ import org.apache.spark.sql.catalyst.analysis.FunctionRegistry.FunctionBuilder
 import org.apache.spark.sql.catalyst.catalog.{CatalogFunction, FunctionResourceLoader, GlobalTempViewManager, SessionCatalog}
 import org.apache.spark.sql.catalyst.expressions.{Cast, Expression}
 import org.apache.spark.sql.catalyst.parser.ParserInterface
+import org.apache.spark.sql.execution.aggregate.ScalaUDAF
+import org.apache.spark.sql.expressions.UserDefinedAggregateFunction
 import org.apache.spark.sql.hive.HiveShim.HiveFunctionWrapper
 import org.apache.spark.sql.internal.SQLConf
 import org.apache.spark.sql.types.{DecimalType, DoubleType}
@@ -95,6 +97,8 @@ private[sql] class HiveSessionCatalog(
           val udtf = HiveGenericUDTF(name, new HiveFunctionWrapper(clazz.getName), children)
           udtf.elementSchema // Force it to check input data types.
           udtf
+        } else if (classOf[UserDefinedAggregateFunction].isAssignableFrom(clazz)) {
+          ScalaUDAF(children, clazz.newInstance().asInstanceOf[UserDefinedAggregateFunction])
         } else {
           throw new AnalysisException(s"No handler for Hive UDF '${clazz.getCanonicalName}'")
         }

sql/hive/src/test/scala/org/apache/spark/sql/hive/execution/HiveUDAFSuite.scala

@@ -26,6 +26,7 @@ import org.apache.hadoop.hive.ql.util.JavaDataModel
 import org.apache.hadoop.hive.serde2.objectinspector.{ObjectInspector, ObjectInspectorFactory}
 import org.apache.hadoop.hive.serde2.objectinspector.primitive.PrimitiveObjectInspectorFactory
 import org.apache.hadoop.hive.serde2.typeinfo.TypeInfo
+import test.org.apache.spark.sql.MyDoubleAvg
 
 import org.apache.spark.sql.{AnalysisException, QueryTest, Row}
 import org.apache.spark.sql.execution.aggregate.ObjectHashAggregateExec
@@ -86,6 +87,18 @@ class HiveUDAFSuite extends QueryTest with TestHiveSingleton with SQLTestUtils {
     ))
   }
 
+  test("call JAVA UDAF") {
+    withTempView("temp") {
+      withUserDefinedFunction("myDoubleAvg" -> false) {
+        spark.range(1, 10).toDF("value").createOrReplaceTempView("temp")
+        sql(s"CREATE FUNCTION myDoubleAvg AS '${classOf[MyDoubleAvg].getName}'")
+        checkAnswer(
+          spark.sql("SELECT default.myDoubleAvg(value) as my_avg from temp"),
+          Row(105.0))
+      }
+    }
+  }
+
   test("non-deterministic children expressions of UDAF") {
     withTempView("view1") {
       spark.range(1).selectExpr("id as x", "id as y").createTempView("view1")