[SPARK-8199] day_of_month alias

python/pyspark/sql/functions.py

@@ -670,7 +670,7 @@ def date_format(dateCol, format):
     [Row(date=u'04/08/2015')]
     """
     sc = SparkContext._active_spark_context
-    return Column(sc._jvm.functions.dateFormat(dateCol, format))
+    return Column(sc._jvm.functions.date_format(dateCol, format))
 
 
 @since(1.5)
@@ -715,14 +715,26 @@ def month(col):
 @since(1.5)
 def day(col):
     """
-    Extract the day of a given date as integer.
+    Extract the day of the month of a given date as integer.
     >>> sqlContext.createDataFrame([('2015-04-08',)], ['a']).select(day('a').alias('day')).collect()
     [Row(day=8)]
     """
     sc = SparkContext._active_spark_context
     return Column(sc._jvm.functions.day(col))
 
 
+@since(1.5)
+def day_of_month(col):
+    """
+    Extract the day of the month of a given date as integer.
+    >>> df0 = sqlContext.createDataFrame([('2015-04-08',)], ['a'])
+    >>> df0.select(day_of_month('a').alias('day')).collect()
+    [Row(day=8)]
+    """
+    sc = SparkContext._active_spark_context
+    return Column(sc._jvm.functions.day_of_month(col))
+
+
 @since(1.5)
 def day_in_year(col):
     """
@@ -732,7 +744,7 @@ def day_in_year(col):
     [Row(day=98)]
     """
     sc = SparkContext._active_spark_context
-    return Column(sc._jvm.functions.day(col))
+    return Column(sc._jvm.functions.day_in_year(col))
 
 
 @since(1.5)
@@ -784,7 +796,7 @@ def week_of_year(col):
     [Row(week=15)]
     """
     sc = SparkContext._active_spark_context
-    return Column(sc._jvm.functions.weekOfYear(col))
+    return Column(sc._jvm.functions.week_of_year(col))
 
 
 class UserDefinedFunction(object):

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

@@ -184,6 +184,7 @@ object FunctionRegistry {
     expression[Quarter]("quarter"),
     expression[Month]("month"),
     expression[Day]("day"),
+    expression[Day]("day_of_month"),
     expression[DayInYear]("day_in_year"),
     expression[Hour]("hour"),
     expression[Minute]("minute"),

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

@@ -1779,19 +1779,33 @@ object functions {
   def month(columnName: String): Column = month(Column(columnName))
 
   /**
-   * Extracts the day as an integer from a given date/timestamp/string.
+   * Extracts the day of the month as an integer from a given date/timestamp/string.
    * @group datetime_funcs
    * @since 1.5.0
    */
   def day(e: Column): Column = Day(e.expr)
 
   /**
-   * Extracts the day as an integer from a given date/timestamp/string.
+   * Extracts the day of the month as an integer from a given date/timestamp/string.
    * @group datetime_funcs
    * @since 1.5.0
    */
   def day(columnName: String): Column = day(Column(columnName))
 
+  /**
+   * Extracts the day of the month as an integer from a given date/timestamp/string.
+   * @group datetime_funcs
+   * @since 1.5.0
+   */
+  def day_of_month(e: Column): Column = Day(e.expr)
+
+  /**
+   * Extracts the day of the month as an integer from a given date/timestamp/string.
+   * @group datetime_funcs
+   * @since 1.5.0
+   */
+  def day_of_month(columnName: String): Column = day_of_month(Column(columnName))
+
   /**
    * Extracts the day of the year as an integer from a given date/timestamp/string.
    * @group datetime_funcs

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

@@ -95,6 +95,18 @@ class DateExpressionsSuite extends QueryTest {
       Row(8, 8, 8))
   }
 
+  test("day of month") {
+    val df = Seq((d, sdfDate.format(d), ts)).toDF("a", "b", "c")
+
+    checkAnswer(
+      df.select(day_of_month("a"), day_of_month("b"), day_of_month("c")),
+      Row(8, 8, 8))
+
+    checkAnswer(
+      df.selectExpr("day_of_month(a)", "day_of_month(b)", "day_of_month(c)"),
+      Row(8, 8, 8))
+  }
+
   test("day in year") {
     val df = Seq((d, sdfDate.format(d), ts)).toDF("a", "b", "c")