szehon-ho commented on code in PR #12494: URL: https://github.com/apache/iceberg/pull/12494#discussion_r2056830271
########## spark/v4.0/spark/src/main/java/org/apache/iceberg/spark/procedures/CreateChangelogViewProcedure.java: ########## @@ -0,0 +1,321 @@ +/* + * 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.iceberg.spark.procedures; + +import java.util.Arrays; +import java.util.List; +import java.util.Map; +import java.util.Set; +import java.util.function.Predicate; +import java.util.stream.Collectors; +import org.apache.iceberg.MetadataColumns; +import org.apache.iceberg.Table; +import org.apache.iceberg.relocated.com.google.common.base.Preconditions; +import org.apache.iceberg.relocated.com.google.common.collect.ImmutableMap; +import org.apache.iceberg.relocated.com.google.common.collect.Lists; +import org.apache.iceberg.relocated.com.google.common.collect.Sets; +import org.apache.iceberg.spark.ChangelogIterator; +import org.apache.iceberg.spark.source.SparkChangelogTable; +import org.apache.spark.api.java.function.MapPartitionsFunction; +import org.apache.spark.sql.Column; +import org.apache.spark.sql.Dataset; +import org.apache.spark.sql.Encoders; +import org.apache.spark.sql.Row; +import org.apache.spark.sql.catalyst.InternalRow; +import org.apache.spark.sql.catalyst.expressions.OrderUtils; +import org.apache.spark.sql.connector.catalog.Identifier; +import org.apache.spark.sql.connector.catalog.TableCatalog; +import org.apache.spark.sql.connector.iceberg.catalog.ProcedureParameter; +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; + +/** + * A procedure that creates a view for changed rows. + * + * <p>The procedure always removes the carry-over rows. Please query {@link SparkChangelogTable} + * instead when carry-over rows are required. + * + * <p>The procedure doesn't compute the pre/post update images by default. If you want to compute + * them, you can set "compute_updates" to be true in the options. + * + * <p>Carry-over rows are the result of a removal and insertion of the same row within an operation + * because of the copy-on-write mechanism. For example, given a file which contains row1 (id=1, + * data='a') and row2 (id=2, data='b'). A copy-on-write delete of row2 would require erasing this + * file and preserving row1 in a new file. The changelog table would report this as (id=1, data='a', + * op='DELETE') and (id=1, data='a', op='INSERT'), despite it not being an actual change to the + * table. The procedure finds the carry-over rows and removes them from the result. + * + * <p>Pre/post update images are converted from a pair of a delete row and an insert row. Identifier + * columns are used for determining whether an insert and a delete record refer to the same row. If + * the two records share the same values for the identity columns they are considered to be before + * and after states of the same row. You can either set identifier fields in the table schema or + * input them as the procedure parameters. Here is an example of pre/post update images with an + * identifier column(id). A pair of a delete row and an insert row with the same id: + * + * <ul> + * <li>(id=1, data='a', op='DELETE') + * <li>(id=1, data='b', op='INSERT') + * </ul> + * + * <p>will be marked as pre/post update images: + * + * <ul> + * <li>(id=1, data='a', op='UPDATE_BEFORE') + * <li>(id=1, data='b', op='UPDATE_AFTER') + * </ul> + */ +public class CreateChangelogViewProcedure extends BaseProcedure { + + private static final ProcedureParameter TABLE_PARAM = + ProcedureParameter.required("table", DataTypes.StringType); + private static final ProcedureParameter CHANGELOG_VIEW_PARAM = + ProcedureParameter.optional("changelog_view", DataTypes.StringType); + private static final ProcedureParameter OPTIONS_PARAM = + ProcedureParameter.optional("options", STRING_MAP); + private static final ProcedureParameter COMPUTE_UPDATES_PARAM = + ProcedureParameter.optional("compute_updates", DataTypes.BooleanType); + private static final ProcedureParameter IDENTIFIER_COLUMNS_PARAM = + ProcedureParameter.optional("identifier_columns", STRING_ARRAY); + private static final ProcedureParameter NET_CHANGES = + ProcedureParameter.optional("net_changes", DataTypes.BooleanType); + + private static final ProcedureParameter[] PARAMETERS = + new ProcedureParameter[] { + TABLE_PARAM, + CHANGELOG_VIEW_PARAM, + OPTIONS_PARAM, + COMPUTE_UPDATES_PARAM, + IDENTIFIER_COLUMNS_PARAM, + NET_CHANGES, + }; + + private static final StructType OUTPUT_TYPE = + new StructType( + new StructField[] { + new StructField("changelog_view", DataTypes.StringType, false, Metadata.empty()) + }); + + public static SparkProcedures.ProcedureBuilder builder() { + return new BaseProcedure.Builder<CreateChangelogViewProcedure>() { + @Override + protected CreateChangelogViewProcedure doBuild() { + return new CreateChangelogViewProcedure(tableCatalog()); + } + }; + } + + private CreateChangelogViewProcedure(TableCatalog tableCatalog) { + super(tableCatalog); + } + + @Override + public ProcedureParameter[] parameters() { + return PARAMETERS; + } + + @Override + public StructType outputType() { + return OUTPUT_TYPE; + } + + @Override + public InternalRow[] call(InternalRow args) { + ProcedureInput input = new ProcedureInput(spark(), tableCatalog(), PARAMETERS, args); + + Identifier tableIdent = input.ident(TABLE_PARAM); + + // load insert and deletes from the changelog table + Identifier changelogTableIdent = changelogTableIdent(tableIdent); + Dataset<Row> df = loadRows(changelogTableIdent, options(input)); + + boolean netChanges = input.asBoolean(NET_CHANGES, false); + String[] identifierColumns = identifierColumns(input, tableIdent); + Set<String> unorderableColumnNames = + Arrays.stream(df.schema().fields()) + .filter(field -> !OrderUtils.isOrderable(field.dataType())) + .map(StructField::name) + .collect(Collectors.toSet()); + + Preconditions.checkArgument( + identifierColumns.length > 0 || unorderableColumnNames.isEmpty(), + "Identifier field is required as table contains unorderable columns: %s", + unorderableColumnNames); + + if (shouldComputeUpdateImages(input)) { + Preconditions.checkArgument(!netChanges, "Not support net changes with update images"); + df = computeUpdateImages(identifierColumns, df); + } else { + df = removeCarryoverRows(df, netChanges); + } + + String viewName = viewName(input, tableIdent.name()); + + df.createOrReplaceTempView(viewName); + + return toOutputRows(viewName); + } + + private Dataset<Row> computeUpdateImages(String[] identifierColumns, Dataset<Row> df) { + Preconditions.checkArgument( + identifierColumns.length > 0, + "Cannot compute the update images because identifier columns are not set"); + + Column[] repartitionSpec = new Column[identifierColumns.length + 1]; + for (int i = 0; i < identifierColumns.length; i++) { + repartitionSpec[i] = df.col(CreateChangelogViewProcedure.delimitedName(identifierColumns[i])); Review Comment: never mind, i see it: https://github.com/apache/iceberg/pull/12494#discussion_r2040874901 -- 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: issues-unsubscr...@iceberg.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For additional commands, e-mail: issues-h...@iceberg.apache.org
