Fix the chained UDF with Row type

src/csharp/Microsoft.Spark.E2ETest/UdfTests/UdfComplexTypesTests.cs

@@ -136,7 +136,7 @@ public void TestUdfWithReturnAsMapType()
         [Fact]
         public void TestUdfWithRowType()
         {
-            // Single Row
+            // Single Row.
             {
                 Func<Column, Column> udf = Udf<Row, string>(
                     (row) => row.GetAs<string>("city"));
@@ -149,7 +149,7 @@ public void TestUdfWithRowType()
                 Assert.Equal(expected, actual);
             }
 
-            // Multiple Rows
+            // Multiple Rows.
             {
                 Func<Column, Column, Column, Column> udf = Udf<Row, Row, string, string>(
                     (row1, row2, str) =>
@@ -173,7 +173,7 @@ public void TestUdfWithRowType()
                 Assert.Equal(expected, actual);
             }
 
-            // Nested Row
+            // Nested Rows.
             {
                 Func<Column, Column> udf = Udf<Row, string>(
                     (row) =>
@@ -197,7 +197,7 @@ public void TestUdfWithRowType()
         [Fact]
         public void TestUdfWithReturnAsRowType()
         {
-            // Single GenericRow
+            // Test UDF that returns a Row object.
             {
                 var schema = new StructType(new[]
                 {
@@ -219,7 +219,7 @@ public void TestUdfWithReturnAsRowType()
                 }
             }
 
-            // Generic row is a part of top-level column.
+            // GenericRow is a part of top-level column.
             {
                 var schema = new StructType(new[]
                 {
@@ -244,7 +244,7 @@ public void TestUdfWithReturnAsRowType()
                 }
             }
 
-            // Nested GenericRow
+            // Test UDF that returns a nested Row object.
             {
                 var subSchema1 = new StructType(new[]
                 {
@@ -295,6 +295,27 @@ public void TestUdfWithReturnAsRowType()
                         outerCol.GetAs<Row>("col3"));
                 }
             }
+
+            // Chained UDFs.
+            {
+                var schema = new StructType(new[]
+                {
+                    new StructField("col1", new IntegerType()),
+                    new StructField("col2", new StringType())
+                });
+                Func<Column, Column> udf1 = Udf<string>(
+                    str => new GenericRow(new object[] { 1, "abc" }), schema);
+
+                Func<Column, Column> udf2 = Udf<Row, string>(
+                    row => row.GetAs<string>(1));
+
+                Row[] rows = _df.Select(udf2(udf1(_df["name"]))).Collect().ToArray();
+                Assert.Equal(3, rows.Length);
+
+                var expected = new[] { "abc", "abc", "abc" };
+                string[] actual = rows.Select(x => x[0].ToString()).ToArray();
+                Assert.Equal(expected, actual);
+            }
         }
     }
 }

src/csharp/Microsoft.Spark.UnitTest/Sql/RowTests.cs

@@ -136,7 +136,7 @@ public void RowCollectorTest()
         private Pickler CreatePickler()
         {
             new StructTypePickler().Register();
-            new RowPickler().Register();
+            new TestUtils.RowPickler().Register();
             return new Pickler();
         }
 

src/csharp/Microsoft.Spark/Sql/GenericRowPickler.cs → src/csharp/Microsoft.Spark/Sql/CustomPicklers.cs

@@ -7,6 +7,17 @@
 
 namespace Microsoft.Spark.Sql
 {
+    /// <summary>
+    /// Custom pickler for Row objects.
+    /// </summary>
+    internal class RowPickler : IObjectPickler
+    {
+        public void pickle(object o, Stream outs, Pickler currentPickler)
+        {
+            currentPickler.save(((Row)o).Values);
+        }
+    }
+
     /// <summary>
     /// Custom pickler for GenericRow objects.
     /// </summary>

src/csharp/Microsoft.Spark/Sql/Functions.cs

@@ -3804,7 +3804,7 @@ public static Func<Column, Column, Column, Column, Column, Column, Column, Colum
         /// <returns>
         /// A delegate that returns a <see cref="Column"/> for the result of the UDF.
         /// </returns>
-        public static Func<Column> Udf(Func<GenericRow> udf, StructType returnType)
+        public static Func<Column> Udf(Func<Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply0;
         }
@@ -3816,7 +3816,7 @@ public static Func<Column> Udf(Func<GenericRow> udf, StructType returnType)
         /// <returns>
         /// A delegate that returns a <see cref="Column"/> for the result of the UDF.
         /// </returns>
-        public static Func<Column, Column> Udf<T>(Func<T, GenericRow> udf, StructType returnType)
+        public static Func<Column, Column> Udf<T>(Func<T, Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply1;
         }
@@ -3830,7 +3830,7 @@ public static Func<Column, Column> Udf<T>(Func<T, GenericRow> udf, StructType re
         /// A delegate that returns a <see cref="Column"/> for the result of the UDF.
         /// </returns>
         public static Func<Column, Column, Column> Udf<T1, T2>(
-            Func<T1, T2, GenericRow> udf, StructType returnType)
+            Func<T1, T2, Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply2;
         }
@@ -3845,7 +3845,7 @@ public static Func<Column, Column, Column> Udf<T1, T2>(
         /// A delegate that returns a <see cref="Column"/> for the result of the UDF.
         /// </returns>
         public static Func<Column, Column, Column, Column> Udf<T1, T2, T3>(
-            Func<T1, T2, T3, GenericRow> udf, StructType returnType)
+            Func<T1, T2, T3, Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply3;
         }
@@ -3861,7 +3861,7 @@ public static Func<Column, Column, Column, Column> Udf<T1, T2, T3>(
         /// A delegate that returns a <see cref="Column"/> for the result of the UDF.
         /// </returns>
         public static Func<Column, Column, Column, Column, Column> Udf<T1, T2, T3, T4>(
-            Func<T1, T2, T3, T4, GenericRow> udf, StructType returnType)
+            Func<T1, T2, T3, T4, Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply4;
         }
@@ -3878,7 +3878,7 @@ public static Func<Column, Column, Column, Column, Column> Udf<T1, T2, T3, T4>(
         /// A delegate that returns a <see cref="Column"/> for the result of the UDF.
         /// </returns>
         public static Func<Column, Column, Column, Column, Column, Column> Udf<T1, T2, T3, T4, T5>(
-            Func<T1, T2, T3, T4, T5, GenericRow> udf, StructType returnType)
+            Func<T1, T2, T3, T4, T5, Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply5;
         }
@@ -3896,7 +3896,7 @@ public static Func<Column, Column, Column, Column, Column, Column> Udf<T1, T2, T
         /// A delegate that returns a <see cref="Column"/> for the result of the UDF.
         /// </returns>
         public static Func<Column, Column, Column, Column, Column, Column, Column> Udf<T1, T2, T3, T4, T5, T6>(
-            Func<T1, T2, T3, T4, T5, T6, GenericRow> udf, StructType returnType)
+            Func<T1, T2, T3, T4, T5, T6, Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply6;
         }
@@ -3915,7 +3915,7 @@ public static Func<Column, Column, Column, Column, Column, Column, Column> Udf<T
         /// A delegate that returns a <see cref="Column"/> for the result of the UDF.
         /// </returns>
         public static Func<Column, Column, Column, Column, Column, Column, Column, Column> Udf<T1, T2, T3, T4, T5, T6, T7>(
-            Func<T1, T2, T3, T4, T5, T6, T7, GenericRow> udf, StructType returnType)
+            Func<T1, T2, T3, T4, T5, T6, T7, Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply7;
         }
@@ -3935,7 +3935,7 @@ public static Func<Column, Column, Column, Column, Column, Column, Column, Colum
         /// A delegate that returns a <see cref="Column"/> for the result of the UDF.
         /// </returns>
         public static Func<Column, Column, Column, Column, Column, Column, Column, Column, Column> Udf<T1, T2, T3, T4, T5, T6, T7, T8>(
-            Func<T1, T2, T3, T4, T5, T6, T7, T8, GenericRow> udf, StructType returnType)
+            Func<T1, T2, T3, T4, T5, T6, T7, T8, Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply8;
         }
@@ -3954,7 +3954,7 @@ public static Func<Column, Column, Column, Column, Column, Column, Column, Colum
         /// <param name="returnType">Schema associated with this row</param>
         /// <returns>A delegate that when invoked will return a <see cref="Column"/> for the result of the UDF.</returns>
         public static Func<Column, Column, Column, Column, Column, Column, Column, Column, Column, Column> Udf<T1, T2, T3, T4, T5, T6, T7, T8, T9>(
-            Func<T1, T2, T3, T4, T5, T6, T7, T8, T9, GenericRow> udf, StructType returnType)
+            Func<T1, T2, T3, T4, T5, T6, T7, T8, T9, Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply9;
         }
@@ -3976,7 +3976,7 @@ public static Func<Column, Column, Column, Column, Column, Column, Column, Colum
         /// A delegate that returns a <see cref="Column"/> for the result of the UDF.
         /// </returns>
         public static Func<Column, Column, Column, Column, Column, Column, Column, Column, Column, Column, Column> Udf<T1, T2, T3, T4, T5, T6, T7, T8, T9, T10>(
-            Func<T1, T2, T3, T4, T5, T6, T7, T8, T9, T10, GenericRow> udf, StructType returnType)
+            Func<T1, T2, T3, T4, T5, T6, T7, T8, T9, T10, Row> udf, StructType returnType)
         {
             return CreateUdf(udf.Method.ToString(), UdfUtils.CreateUdfWrapper(udf), returnType).Apply10;
         }

src/csharp/Microsoft.Spark/Sql/GenericRow.cs

@@ -98,6 +98,6 @@ public override bool Equals(object obj) =>
         /// Returns the hash code of the current object.
         /// </summary>
         /// <returns>The hash code of the current object</returns>
-        public override int GetHashCode() => base.GetHashCode();        
+        public override int GetHashCode() => base.GetHashCode();
     }
 }

src/csharp/Microsoft.Spark/Sql/Row.cs

@@ -3,8 +3,6 @@
 // See the LICENSE file in the project root for more information.
 
 using System;
-using System.Collections.Generic;
-using System.Linq;
 using Microsoft.Spark.Sql.Types;
 
 namespace Microsoft.Spark.Sql
@@ -36,6 +34,24 @@ internal Row(object[] values, StructType schema)
             Convert();
         }
 
+        /// <summary>
+        /// Constructor for the schema-less Row class used for chained UDFs.
+        /// </summary>
+        /// <param name="genericRow">GenericRow to convert from</param>      
+        internal Row(GenericRow genericRow)
+        {
+            _genericRow = genericRow;
+        }
+
+        /// <summary>
+        /// Returns schema-less Row which can happen within chained UDFs (same behavior as PySpark).
+        /// </summary>
+        /// <returns>schema-less Row</returns>
+        public static implicit operator Row(GenericRow genericRow)
+        {
+            return new Row(genericRow);
+        }
+
         /// <summary>
         /// Schema associated with this row.
         /// </summary>

src/csharp/Microsoft.Spark/Utils/PythonSerDe.cs

@@ -36,7 +36,8 @@ static PythonSerDe()
             Unpickler.registerConstructor(
                 "pyspark.sql.types", "_create_row_inbound_converter", s_rowConstructor);
 
-            // Register custom pickler for GenericRow objects.
+            // Register custom picklers.
+            Pickler.registerCustomPickler(typeof(Row), new RowPickler());
             Pickler.registerCustomPickler(typeof(GenericRow), new GenericRowPickler());
         }