[MINOR][SQL][DOCS] Add notes of the deterministic assumption on UDF functions

## What changes were proposed in this pull request?

Spark assumes that UDF functions are deterministic. This PR adds explicit notes about that.

## How was this patch tested?

It's only about docs.

Author: Dongjoon Hyun <[email protected]>

Closes #13087 from dongjoon-hyun/SPARK-15282.

(cherry picked from commit 37c617e)
Signed-off-by: Michael Armbrust <[email protected]>

Author: dongjoon-hyun

python/pyspark/sql/functions.py

@@ -1756,6 +1756,9 @@ def __call__(self, *cols):
 @since(1.3)
 def udf(f, returnType=StringType()):
     """Creates a :class:`Column` expression representing a user defined function (UDF).
+    Note that the user-defined functions must be deterministic. Due to optimization,
+    duplicate invocations may be eliminated or the function may even be invoked more times than
+    it is present in the query.
 
     >>> from pyspark.sql.types import IntegerType
     >>> slen = udf(lambda s: len(s), IntegerType())

sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/ScalaUDF.scala

@@ -23,6 +23,7 @@ import org.apache.spark.sql.types.DataType
 
 /**
  * User-defined function.
+ * Note that the user-defined functions must be deterministic.
  * @param function  The user defined scala function to run.
  *                  Note that if you use primitive parameters, you are not able to check if it is
  *                  null or not, and the UDF will return null for you if the primitive input is

sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala

@@ -199,6 +199,9 @@ class SQLContext private[sql](
 
   /**
    * A collection of methods for registering user-defined functions (UDF).
+   * Note that the user-defined functions must be deterministic. Due to optimization,
+   * duplicate invocations may be eliminated or the function may even be invoked more times than
+   * it is present in the query.
    *
    * The following example registers a Scala closure as UDF:
    * {{{

sql/core/src/main/scala/org/apache/spark/sql/SparkSession.scala

@@ -145,6 +145,9 @@ class SparkSession private(
 
   /**
    * A collection of methods for registering user-defined functions (UDF).
+   * Note that the user-defined functions must be deterministic. Due to optimization,
+   * duplicate invocations may be eliminated or the function may even be invoked more times than
+   * it is present in the query.
    *
    * The following example registers a Scala closure as UDF:
    * {{{

sql/core/src/main/scala/org/apache/spark/sql/UDFRegistration.scala

@@ -32,6 +32,7 @@ import org.apache.spark.sql.types.DataType
 
 /**
  * Functions for registering user-defined functions. Use [[SQLContext.udf]] to access this.
+ * Note that the user-defined functions must be deterministic.
  *
  * @since 1.3.0
  */

sql/core/src/main/scala/org/apache/spark/sql/expressions/UserDefinedFunction.scala

@@ -25,6 +25,9 @@ import org.apache.spark.sql.types.DataType
 
 /**
  * A user-defined function. To create one, use the `udf` functions in [[functions]].
+ * Note that the user-defined functions must be deterministic. Due to optimization,
+ * duplicate invocations may be eliminated or the function may even be invoked more times than
+ * it is present in the query.
  * As an example:
  * {{{
  *   // Defined a UDF that returns true or false based on some numeric score.

sql/core/src/main/scala/org/apache/spark/sql/internal/SessionState.scala

@@ -100,6 +100,7 @@ private[sql] class SessionState(sparkSession: SparkSession) {
 
   /**
    * Interface exposed to the user for registering user-defined functions.
+   * Note that the user-defined functions must be deterministic.
    */
   lazy val udf: UDFRegistration = new UDFRegistration(functionRegistry)