sunjincheng121 commented on a change in pull request #7920:
[FLINK-11844][table-api] Simplify over window API classes and improve
documentation
URL: https://github.com/apache/flink/pull/7920#discussion_r264090981
##########
File path:
flink-table/flink-table-planner/src/main/scala/org/apache/flink/table/api/windows.scala
##########
@@ -20,90 +20,180 @@ package org.apache.flink.table.api
import org.apache.flink.table.expressions._
import org.apache.flink.table.plan.logical._
-import org.apache.flink.table.typeutils.{RowIntervalTypeInfo,
TimeIntervalTypeInfo}
-import org.apache.flink.table.api.scala.{CURRENT_RANGE, CURRENT_ROW}
/**
- * Over window is similar to the traditional OVER SQL.
+ * A over-window specification.
+ *
+ * Similar to SQL, over window aggregates compute an aggregate for each input
row over a range
+ * of its neighboring rows.
+ */
+class OverWindow(
+ alias: Expression,
+ partitionBy: Seq[Expression],
+ orderBy: Expression,
+ preceding: Expression,
+ following: Option[Expression]) {
+
+ def getAlias: Expression = alias
+
+ def getPartitioning: Seq[Expression] = partitionBy
+
+ def getOrder: Expression = orderBy
+
+ def getPreceding: Expression = preceding
+
+ def getFollowing: Option[Expression] = following
+}
+
+//
------------------------------------------------------------------------------------------------
+// Over windows
+//
------------------------------------------------------------------------------------------------
+
+/**
+ * Partially defined over window with partitioning.
+ */
+class OverWindowPartitioned(partitionBy: Seq[Expression]) {
+
+ /**
+ * Specifies the time attribute on which rows are ordered.
+ *
+ * For streaming tables, reference a rowtime or proctime time attribute here
+ * to specify the time mode.
+ *
+ * For batch tables, refer to a timestamp or long attribute.
+ *
+ * @param orderByExpr field reference
+ * @return an over window with defined order
+ */
+ def orderBy(orderByExpr: String): OverWindowPartitionedOrdered = {
+ orderBy(ExpressionParser.parseExpression(orderByExpr))
+ }
+
+ /**
+ * Specifies the time attribute on which rows are ordered.
+ *
+ * For streaming tables, reference a rowtime or proctime time attribute here
+ * to specify the time mode.
+ *
+ * For batch tables, refer to a timestamp or long attribute.
+ *
+ * @param orderBy field reference
+ * @return an over window with defined order
+ */
+ def orderBy(orderBy: Expression): OverWindowPartitionedOrdered = {
+ new OverWindowPartitionedOrdered(partitionBy, orderBy)
+ }
+}
+
+/**
+ * Partially defined over window with (optional) partitioning and order.
*/
-case class OverWindow(
- private[flink] val alias: Expression,
- private[flink] val partitionBy: Seq[Expression],
- private[flink] val orderBy: Expression,
- private[flink] val preceding: Expression,
- private[flink] val following: Expression)
+class OverWindowPartitionedOrdered(partitionBy: Seq[Expression], orderBy:
Expression) {
+
+ /**
+ * Set the preceding offset (based on time or row-count intervals) for over
window.
+ *
+ * @param precedingExpr preceding offset relative to the current row.
+ * @return an over window with defined preceding
+ */
+ def preceding(precedingExpr: String): OverWindowPartitionedOrderedPreceding
= {
Review comment:
precedingExpr -> preceding
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
