[GitHub] [flink] godfreyhe commented on a change in pull request #12964: [FLINK-17426][blink planner] Dynamic Source supportsLimit pushdown

[GitBox <gi...@apache.org>]

godfreyhe commented on a change in pull request #12964:
URL: https://github.com/apache/flink/pull/12964#discussion_r462813556



##########
File path: flink-table/flink-table-planner-blink/src/main/java/org/apache/flink/table/planner/plan/rules/logical/PushLimitIntoTableSourceScanRule.java
##########
@@ -0,0 +1,118 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.flink.table.planner.plan.rules.logical;
+
+import org.apache.flink.table.connector.source.DynamicTableSource;
+import org.apache.flink.table.connector.source.abilities.SupportsLimitPushDown;
+import org.apache.flink.table.plan.stats.TableStats;
+import org.apache.flink.table.planner.plan.nodes.logical.FlinkLogicalSort;
+import org.apache.flink.table.planner.plan.nodes.logical.FlinkLogicalTableSourceScan;
+import org.apache.flink.table.planner.plan.schema.FlinkPreparingTableBase;
+import org.apache.flink.table.planner.plan.schema.TableSourceTable;
+import org.apache.flink.table.planner.plan.stats.FlinkStatistic;
+
+import org.apache.calcite.plan.RelOptRule;
+import org.apache.calcite.plan.RelOptRuleCall;
+import org.apache.calcite.rel.core.Sort;
+import org.apache.calcite.rex.RexLiteral;
+
+import java.util.Arrays;
+import java.util.Collections;
+
+/**
+ * Planner rule that tries to push limit into a {@link FlinkLogicalTableSourceScan},
+ * which table is a {@link TableSourceTable}. And the table source in the table is a {@link SupportsLimitPushDown}.
+ * The original limit will still be retained.
+ * The reasons why the limit still be retained:
+ * 1.If the source is required to return the exact number of limit number, the implementation
+ * of the source is highly required. The source is required to accurately control the record
+ * number of split, and the parallelism setting also need to be adjusted accordingly.
+ * 2.When remove the limit, maybe filter will be pushed down to the source after limit pushed
+ * down. The source need know it should do limit first and do the filter later, it is hard to
+ * implement.
+ * 3.We can support limit with offset, we can push down offset + fetch to table source.
+ */
+public class PushLimitIntoTableSourceScanRule extends RelOptRule {
+	public static final PushLimitIntoTableSourceScanRule INSTANCE = new PushLimitIntoTableSourceScanRule();
+
+	public PushLimitIntoTableSourceScanRule() {
+		super(operand(FlinkLogicalSort.class,
+			operand(FlinkLogicalTableSourceScan.class, none())),
+			"PushLimitIntoTableSourceScanRule");
+	}
+
+	@Override
+	public boolean matches(RelOptRuleCall call) {
+		Sort sort = call.rel(0);
+		TableSourceTable tableSourceTable = call.rel(1).getTable().unwrap(TableSourceTable.class);
+
+		// a limit can be pushed down only if it satisfies the two conditions: 1) do not have order by keys, 2) have limit.
+		boolean onlyLimit = sort.getCollation().getFieldCollations().isEmpty() && sort.fetch != null;
+		return onlyLimit
+			&& tableSourceTable != null
+			&& tableSourceTable.tableSource() instanceof SupportsLimitPushDown
+			&& Arrays.stream(tableSourceTable.extraDigests()).noneMatch(str -> str.startsWith("limit=["));
+	}
+
+	@Override
+	public void onMatch(RelOptRuleCall call) {
+		Sort sort = call.rel(0);
+		FlinkLogicalTableSourceScan scan = call.rel(1);
+		TableSourceTable tableSourceTable = scan.getTable().unwrap(TableSourceTable.class);
+		int offset = sort.offset == null ? 0 : RexLiteral.intValue(sort.offset);
+		int limit = offset + RexLiteral.intValue(sort.fetch);
+
+		TableSourceTable newTableSourceTable = applyLimit(limit, tableSourceTable);
+
+		FlinkLogicalTableSourceScan newScan = FlinkLogicalTableSourceScan.create(scan.getCluster(), newTableSourceTable);
+		Sort newSort = sort.copy(sort.getTraitSet(), Collections.singletonList(newScan));
+		call.transformTo(newSort);
+	}
+
+	private TableSourceTable applyLimit(
+		long limit,
+		FlinkPreparingTableBase relOptTable) {
+		TableSourceTable oldTableSourceTable = relOptTable.unwrap(TableSourceTable.class);
+		DynamicTableSource newTableSource = oldTableSourceTable.tableSource().copy();
+		((SupportsLimitPushDown) newTableSource).applyLimit(limit);
+
+		FlinkStatistic statistic = relOptTable.getStatistic();
+		long newRowCount = 0;
+		if (statistic.getRowCount() != null) {
+			newRowCount = Math.min(limit, statistic.getRowCount().longValue());
+		} else {
+			newRowCount = limit;
+		}
+		// update TableStats after limit push down
+		TableStats newTableStats = new TableStats(newRowCount);
+		FlinkStatistic newStatistic = FlinkStatistic.builder()
+			.statistic(statistic)
+			.tableStats(newTableStats)

Review comment:
       nit:  -> `.tableStats(new TableStats(newRowCount))`

##########
File path: flink-table/flink-table-planner-blink/src/main/java/org/apache/flink/table/planner/plan/rules/logical/PushLimitIntoTableSourceScanRule.java
##########
@@ -0,0 +1,118 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.flink.table.planner.plan.rules.logical;
+
+import org.apache.flink.table.connector.source.DynamicTableSource;
+import org.apache.flink.table.connector.source.abilities.SupportsLimitPushDown;
+import org.apache.flink.table.plan.stats.TableStats;
+import org.apache.flink.table.planner.plan.nodes.logical.FlinkLogicalSort;
+import org.apache.flink.table.planner.plan.nodes.logical.FlinkLogicalTableSourceScan;
+import org.apache.flink.table.planner.plan.schema.FlinkPreparingTableBase;
+import org.apache.flink.table.planner.plan.schema.TableSourceTable;
+import org.apache.flink.table.planner.plan.stats.FlinkStatistic;
+
+import org.apache.calcite.plan.RelOptRule;
+import org.apache.calcite.plan.RelOptRuleCall;
+import org.apache.calcite.rel.core.Sort;
+import org.apache.calcite.rex.RexLiteral;
+
+import java.util.Arrays;
+import java.util.Collections;
+
+/**
+ * Planner rule that tries to push limit into a {@link FlinkLogicalTableSourceScan},
+ * which table is a {@link TableSourceTable}. And the table source in the table is a {@link SupportsLimitPushDown}.
+ * The original limit will still be retained.
+ * The reasons why the limit still be retained:
+ * 1.If the source is required to return the exact number of limit number, the implementation
+ * of the source is highly required. The source is required to accurately control the record
+ * number of split, and the parallelism setting also need to be adjusted accordingly.
+ * 2.When remove the limit, maybe filter will be pushed down to the source after limit pushed
+ * down. The source need know it should do limit first and do the filter later, it is hard to
+ * implement.
+ * 3.We can support limit with offset, we can push down offset + fetch to table source.
+ */
+public class PushLimitIntoTableSourceScanRule extends RelOptRule {
+	public static final PushLimitIntoTableSourceScanRule INSTANCE = new PushLimitIntoTableSourceScanRule();
+
+	public PushLimitIntoTableSourceScanRule() {
+		super(operand(FlinkLogicalSort.class,
+			operand(FlinkLogicalTableSourceScan.class, none())),
+			"PushLimitIntoTableSourceScanRule");
+	}
+
+	@Override
+	public boolean matches(RelOptRuleCall call) {
+		Sort sort = call.rel(0);
+		TableSourceTable tableSourceTable = call.rel(1).getTable().unwrap(TableSourceTable.class);
+
+		// a limit can be pushed down only if it satisfies the two conditions: 1) do not have order by keys, 2) have limit.
+		boolean onlyLimit = sort.getCollation().getFieldCollations().isEmpty() && sort.fetch != null;
+		return onlyLimit
+			&& tableSourceTable != null
+			&& tableSourceTable.tableSource() instanceof SupportsLimitPushDown
+			&& Arrays.stream(tableSourceTable.extraDigests()).noneMatch(str -> str.startsWith("limit=["));
+	}
+
+	@Override
+	public void onMatch(RelOptRuleCall call) {
+		Sort sort = call.rel(0);
+		FlinkLogicalTableSourceScan scan = call.rel(1);
+		TableSourceTable tableSourceTable = scan.getTable().unwrap(TableSourceTable.class);
+		int offset = sort.offset == null ? 0 : RexLiteral.intValue(sort.offset);
+		int limit = offset + RexLiteral.intValue(sort.fetch);
+
+		TableSourceTable newTableSourceTable = applyLimit(limit, tableSourceTable);
+
+		FlinkLogicalTableSourceScan newScan = FlinkLogicalTableSourceScan.create(scan.getCluster(), newTableSourceTable);
+		Sort newSort = sort.copy(sort.getTraitSet(), Collections.singletonList(newScan));
+		call.transformTo(newSort);
+	}
+
+	private TableSourceTable applyLimit(
+		long limit,
+		FlinkPreparingTableBase relOptTable) {
+		TableSourceTable oldTableSourceTable = relOptTable.unwrap(TableSourceTable.class);
+		DynamicTableSource newTableSource = oldTableSourceTable.tableSource().copy();
+		((SupportsLimitPushDown) newTableSource).applyLimit(limit);
+
+		FlinkStatistic statistic = relOptTable.getStatistic();
+		long newRowCount = 0;
+		if (statistic.getRowCount() != null) {
+			newRowCount = Math.min(limit, statistic.getRowCount().longValue());
+		} else {
+			newRowCount = limit;
+		}
+		// update TableStats after limit push down
+		TableStats newTableStats = new TableStats(newRowCount);
+		FlinkStatistic newStatistic = FlinkStatistic.builder()
+			.statistic(statistic)
+			.tableStats(newTableStats)
+			.build();
+
+		// update extraDigests
+		String[] newExtraDigests = new String[]{"limit=[" + limit + "]"};
+
+		return oldTableSourceTable.copy(
+			newTableSource,
+			newStatistic,
+			newExtraDigests

Review comment:
       nit: `newExtraDigests` is only used once, just move `new String[] {"limit=[" + limit + "]"}` into here. 




----------------------------------------------------------------
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:
users@infra.apache.org