[SPARK-11736][SQL] Add monotonically_increasing_id to function registry.

https://issues.apache.org/jira/browse/SPARK-11736

Author: Yin Huai <yhuai@databricks.com>

Closes #9703 from yhuai/MonotonicallyIncreasingID.

(cherry picked from commit d83c2f9)
Signed-off-by: Reynold Xin <rxin@databricks.com>

Author: yhuai

sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/FunctionRegistry.scala

@@ -281,7 +281,8 @@ object FunctionRegistry {
     expression[Sha1]("sha1"),
     expression[Sha2]("sha2"),
     expression[SparkPartitionID]("spark_partition_id"),
-    expression[InputFileName]("input_file_name")
+    expression[InputFileName]("input_file_name"),
+    expression[MonotonicallyIncreasingID]("monotonically_increasing_id")
   )
 
   val builtin: SimpleFunctionRegistry = {

sql/core/src/test/scala/org/apache/spark/sql/ColumnExpressionSuite.scala

@@ -563,6 +563,10 @@ class ColumnExpressionSuite extends QueryTest with SharedSQLContext {
       df.select(monotonicallyIncreasingId()),
       Row(0L) :: Row(1L) :: Row((1L << 33) + 0L) :: Row((1L << 33) + 1L) :: Nil
     )
+    checkAnswer(
+      df.select(expr("monotonically_increasing_id()")),
+      Row(0L) :: Row(1L) :: Row((1L << 33) + 0L) :: Row((1L << 33) + 1L) :: Nil
+    )
   }
 
   test("sparkPartitionId") {