FrankChen021 commented on code in PR #20729: URL: https://github.com/apache/druid/pull/20729#discussion_r4400714377
########## extensions-contrib/tsfile-extensions/pom.xml: ########## @@ -0,0 +1,98 @@ +<?xml version="1.0" encoding="UTF-8"?> +<!-- + ~ 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. + --> + +<project xmlns="http://maven.apache.org/POM/4.0.0"; xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"; xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd";> + + <groupId>org.apache.druid.extensions.contrib</groupId> + <artifactId>druid-tsfile-extensions</artifactId> + <name>druid-tsfile-extensions</name> + <description>Apache TsFile input format for Druid batch ingestion</description> + + <parent> + <artifactId>druid</artifactId> + <groupId>org.apache.druid</groupId> + <version>40.0.0-SNAPSHOT</version> + <relativePath>../../pom.xml</relativePath> + </parent> + <modelVersion>4.0.0</modelVersion> + + <dependencies> + <dependency> + <groupId>org.apache.druid</groupId> + <artifactId>druid-processing</artifactId> + <version>${project.parent.version}</version> + <scope>provided</scope> + </dependency> + <dependency> + <groupId>org.apache.tsfile</groupId> + <artifactId>tsfile</artifactId> + </dependency> Review Comment: [P1] Register the new TsFile dependencies **Finding:** Adding `org.apache.tsfile:tsfile:2.5.0` here also introduces the new `org.apache.tsfile:common` runtime dependency documented by the README, but the PR adds no `libraries` entries for either artifact to `licenses.yaml`. The contrib bundle includes this extension without excluding contrib license entries, so its generated license and notice set omits the new components. **Suggestion:** Register both artifacts with their exact versions, canonical license names, module, and any required notices before merging. ########## extensions-contrib/tsfile-extensions/src/main/java/org/apache/druid/data/input/tsfile/TsFileInputEntityReader.java: ########## @@ -0,0 +1,371 @@ +/* + * 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.druid.data.input.tsfile; + +import org.apache.druid.data.input.InputEntity; +import org.apache.druid.data.input.InputRow; +import org.apache.druid.data.input.InputRowSchema; +import org.apache.druid.data.input.IntermediateRowParsingReader; +import org.apache.druid.data.input.impl.MapInputRowParser; +import org.apache.druid.error.InvalidInput; +import org.apache.druid.java.util.common.CloseableIterators; +import org.apache.druid.java.util.common.parsers.CloseableIterator; +import org.apache.druid.java.util.common.parsers.ParseException; +import org.apache.tsfile.enums.TSDataType; +import org.apache.tsfile.exception.read.ReadProcessException; +import org.apache.tsfile.exception.write.NoMeasurementException; +import org.apache.tsfile.exception.write.NoTableException; +import org.apache.tsfile.file.metadata.TableSchema; +import org.apache.tsfile.read.query.dataset.ResultSet; +import org.apache.tsfile.read.v4.ITsFileReader; +import org.apache.tsfile.read.v4.TsFileReaderBuilder; +import org.apache.tsfile.write.schema.IMeasurementSchema; +import org.joda.time.Interval; + +import javax.annotation.Nullable; + +import java.io.File; +import java.io.IOException; +import java.io.UncheckedIOException; +import java.util.ArrayList; +import java.util.Base64; +import java.util.Collections; +import java.util.HashMap; +import java.util.Iterator; +import java.util.List; +import java.util.Locale; +import java.util.Map; +import java.util.NoSuchElementException; +import java.util.Set; +import java.util.stream.Collectors; + +/** + * Reads the rows of a TsFile file with the TsFile table-model reader and converts them into + * {@link MapBasedInputRow}-compatible maps. + * <p> + * TsFile stores its own timestamp, which is exposed to Druid under the name {@value TsFileInputFormat#TIME_COLUMN_NAME}. + * Tag, field, and attribute columns are exposed as-is. + */ +public class TsFileInputEntityReader extends IntermediateRowParsingReader<Map<String, Object>> +{ + private final InputRowSchema inputRowSchema; + private final InputEntity source; + @Nullable + private final File temporaryDirectory; + @Nullable + private final List<String> tables; + @Nullable + private final List<String> columns; + @Nullable + private final Interval interval; + + TsFileInputEntityReader( + InputRowSchema inputRowSchema, + InputEntity source, + @Nullable File temporaryDirectory, + @Nullable List<String> tables, + @Nullable List<String> columns, + @Nullable Interval interval + ) + { + this.inputRowSchema = inputRowSchema; + this.source = source; + this.temporaryDirectory = temporaryDirectory; + this.tables = tables; + this.columns = columns; + this.interval = interval; + } + + @Override + protected CloseableIterator<Map<String, Object>> intermediateRowIterator() throws IOException + { + final InputEntity.CleanableFile materialized = + source.fetch(temporaryDirectory, new byte[InputEntity.DEFAULT_FETCH_BUFFER_SIZE]); + final ITsFileReader reader; + try { + reader = new TsFileReaderBuilder().file(materialized.file()).build(); + } + catch (IOException | RuntimeException e) { + materialized.close(); + throw e; + } + + final List<String> tableNames; + try { + tableNames = resolveTableNames(reader); + } + catch (IOException | RuntimeException e) { + reader.close(); + materialized.close(); + throw e; + } + + return CloseableIterators.wrap( + new RowIterator(reader, tableNames), + () -> { + reader.close(); + materialized.close(); + } + ); + } + + @Override + protected InputEntity source() + { + return source; + } + + @Override + protected List<InputRow> parseInputRows(Map<String, Object> intermediateRow) throws ParseException + { + return Collections.singletonList(MapInputRowParser.parse(inputRowSchema, intermediateRow)); + } + + @Override + protected List<Map<String, Object>> toMap(Map<String, Object> intermediateRow) + { + return Collections.singletonList(intermediateRow); + } + + private List<String> resolveTableNames(ITsFileReader reader) throws IOException + { + final List<TableSchema> allTables = reader.getAllTableSchema(); + if (allTables.isEmpty()) { + throw InvalidInput.exception( + "TsFile[%s] contains no table schema. Only table-model TsFiles are supported; tree-model files " + + "are not.", + source.getUri() + ); + } + if (tables == null || tables.isEmpty()) { + return allTables.stream().map(TableSchema::getTableName).collect(Collectors.toList()); + } + + final Set<String> available = + allTables.stream().map(schema -> schema.getTableName().toLowerCase(Locale.ROOT)).collect(Collectors.toSet()); + for (String table : tables) { + if (!available.contains(table.toLowerCase(Locale.ROOT))) { + throw InvalidInput.exception( + "Table[%s] does not exist in the file. Available tables: %s", + table, + allTables.stream().map(TableSchema::getTableName).collect(Collectors.toList()) + ); + } + } + return tables; + } + + private List<String> selectColumns(TableSchema tableSchema) + { + if (columns != null && !columns.isEmpty()) { Review Comment: [P2] Honor the EXTERN column projection **Finding:** When `columns` is omitted, this path enumerates every physical table column. The MSQ `EXTERN` path supplies `ColumnsFilter.all()` and the new SQL example omits `columns`, so fields absent from the SQL signature are still queried and converted; an unlisted `VECTOR`, `OBJECT`, `ROW`, or `UNKNOWN` field therefore throws from `readValue`, while wide unused fields are unnecessarily read. **Suggestion:** Derive the physical projection from the EXTERN/input schema while retaining the implicit timestamp and required ingestion fields, and cover an omitted unsupported column. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
