JingsongLi commented on code in PR #10420: URL: https://github.com/apache/paimon/pull/10420#discussion_r4244899617
########## paimon-spark/paimon-spark-common/src/main/java/org/apache/paimon/spark/procedure/CreateTagFromWatermarkProcedure.java: ########## @@ -0,0 +1,139 @@ +/* + * 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.paimon.spark.procedure; + +import org.apache.paimon.Snapshot; +import org.apache.paimon.table.FileStoreTable; +import org.apache.paimon.utils.SnapshotManager; +import org.apache.paimon.utils.SnapshotNotExistException; +import org.apache.paimon.utils.TimeUtils; + +import org.apache.spark.sql.catalyst.InternalRow; +import org.apache.spark.sql.connector.catalog.Identifier; +import org.apache.spark.sql.connector.catalog.TableCatalog; +import org.apache.spark.sql.types.DataTypes; +import org.apache.spark.sql.types.Metadata; +import org.apache.spark.sql.types.StructField; +import org.apache.spark.sql.types.StructType; +import org.apache.spark.unsafe.types.UTF8String; + +import java.time.Duration; +import java.util.Set; + +import static org.apache.spark.sql.types.DataTypes.LongType; +import static org.apache.spark.sql.types.DataTypes.StringType; + +/** The procedure supports creating tags from snapshot watermarks. */ +public class CreateTagFromWatermarkProcedure extends BaseProcedure { + + private static final ProcedureParameter[] PARAMETERS = + new ProcedureParameter[] { + ProcedureParameter.required("table", StringType), + ProcedureParameter.required("tag", StringType), + ProcedureParameter.required("watermark", LongType), + ProcedureParameter.optional("time_retained", StringType) + }; + + private static final StructType OUTPUT_TYPE = + new StructType( + new StructField[] { + new StructField("tagName", DataTypes.StringType, true, Metadata.empty()), + new StructField("snapshot", DataTypes.LongType, true, Metadata.empty()), + new StructField("commit_time", DataTypes.LongType, true, Metadata.empty()), + new StructField("watermark", DataTypes.StringType, true, Metadata.empty()) + }); + + private CreateTagFromWatermarkProcedure(TableCatalog tableCatalog) { + super(tableCatalog); + } + + @Override + public ProcedureParameter[] parameters() { + return PARAMETERS; + } + + @Override + public StructType outputType() { + return OUTPUT_TYPE; + } + + @Override + public InternalRow[] call(InternalRow args) { + Identifier tableIdent = toIdentifier(args.getString(0), PARAMETERS[0].name()); + String tag = args.getString(1); + Long watermark = args.getLong(2); + Duration timeRetained = + args.isNullAt(3) ? null : TimeUtils.parseDuration(args.getString(3)); + + return modifyPaimonTable( + tableIdent, + table -> { + FileStoreTable fileStoreTable = (FileStoreTable) table; + SnapshotManager snapshotManager = fileStoreTable.snapshotManager(); + Snapshot snapshot = snapshotManager.laterOrEqualWatermark(watermark); Review Comment: [P2] Select the first live snapshot in an equal-watermark interval The delegated search stops at the binary-search midpoint on exact equality, so the new procedure can tag a later live snapshot even without any tags. I reproduced this with five real stream commits carrying watermarks [500,1000,1000,1000,2000] and different rows. CALL sys.create_tag_from_watermark(table => 'test.T', tag => 'plateau', watermark => 1000) returns snapshot 3; the first qualifying retained snapshot promised by the API is snapshot 2. Repeated watermarks are normal because commit watermark handling inherits/takes the maximum. With no existing tags, the subsequent tie-breaking loop cannot fix the candidate. Please use lower-bound search semantics that continue left on equality, or otherwise normalize the live candidate to the earliest matching snapshot. Add the no-tag plateau case. The existing four tests pass; adding this real SQL assertion yields four passes and one failure (expected snapshot 2, actual 3). The shared helper predates this PR, but this affects the stated contract of the new Spark procedure. -- 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. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
