MaxGekk commented on code in PR #55326: URL: https://github.com/apache/spark/pull/55326#discussion_r3398939883
########## sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/parquet/types/ops/TimeTypeParquetOps.scala: ########## @@ -0,0 +1,120 @@ +/* + * 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.spark.sql.execution.datasources.parquet.types.ops + +import org.apache.parquet.io.api.RecordConsumer +import org.apache.parquet.schema.{LogicalTypeAnnotation, Type, Types} +import org.apache.parquet.schema.LogicalTypeAnnotation.TimeUnit +import org.apache.parquet.schema.PrimitiveType.PrimitiveTypeName.INT64 +import org.apache.parquet.schema.Type.Repetition + +import org.apache.spark.sql.catalyst.expressions.SpecializedGetters +import org.apache.spark.sql.catalyst.util.DateTimeUtils +import org.apache.spark.sql.errors.QueryExecutionErrors +import org.apache.spark.sql.execution.datasources.parquet.{HasParentContainerUpdater, ParentContainerUpdater, ParquetPrimitiveConverter} +import org.apache.spark.sql.internal.SQLConf +import org.apache.spark.sql.types.{DataType, TimeType} + +/** + * Parquet operations for TimeType. + * + * TimeType is a primitive Long-backed type stored in Parquet as INT64 with the + * TIME(isAdjustedToUTC=false, unit=MICROS) logical type annotation. + * + * IMPORTANT - internal vs Parquet representation: + * - Spark internal: nanoseconds since midnight (Long) + * - Parquet storage: microseconds since midnight (INT64) + * - Write path: nanos -> micros (DateTimeUtils.nanosToMicros) + * - Read path: micros -> nanos (DateTimeUtils.microsToNanos) + * + * @param t the TimeType with precision information + * @since 4.3.0 + */ +case class TimeTypeParquetOps(t: TimeType) extends ParquetTypeOps { + + override def dataType: DataType = t + + // ==================== Schema Conversion ==================== + + override def convertToParquetType( + fieldName: String, repetition: Repetition, inShredded: Boolean): Type = + Types.primitive(INT64, repetition) + .as(LogicalTypeAnnotation.timeType(false, TimeUnit.MICROS)) + .named(fieldName) + + // ==================== Value Write ==================== + + override def makeWriter( + recordConsumer: () => RecordConsumer, + makeFieldWriter: DataType => (SpecializedGetters, Int) => Unit + ): (SpecializedGetters, Int) => Unit = + // Evaluate the supplier at write time (not creation time) because recordConsumer + // is null during init() and set later in prepareForWrite(). + (row: SpecializedGetters, ordinal: Int) => + recordConsumer().addLong(DateTimeUtils.nanosToMicros(row.getLong(ordinal))) + + // ==================== Row-Based Read ==================== + + override def newConverter( + parquetType: org.apache.parquet.schema.Type, Review Comment: @stevomitric Could you address this one as well? Same cleanup as in `ParquetTypeOps` after my earlier note — import `Converter` and use the already-imported `Type` unqualified. ########## sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/parquet/types/ops/TimeTypeParquetOpsSuite.scala: ########## @@ -0,0 +1,129 @@ +/* + * 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.spark.sql.execution.datasources.parquet.types.ops + +import org.apache.parquet.schema.{LogicalTypeAnnotation, Type, Types} +import org.apache.parquet.schema.LogicalTypeAnnotation.TimeUnit +import org.apache.parquet.schema.PrimitiveType.PrimitiveTypeName.{INT32, INT64} +import org.apache.parquet.schema.Type.Repetition.REQUIRED + +import org.apache.spark.{SparkFunSuite, SparkRuntimeException} +import org.apache.spark.sql.types.TimeType + +/** + * Unit tests for [[TimeTypeParquetOps.requireCompatibleParquetType]]. + * + * TimeType is stored in Parquet as INT64 TIME(MICROS, isAdjustedToUTC=false). + * Any other encoding is rejected so that reading fails loudly rather than + * silently mis-decoding (e.g. interpreting NANOS as MICROS, which would be off + * by 1000x). + * + * These tests document the exact reject set and pin the intended behavior of + * the read-path guard. They also serve as a regression hook for the Review Comment: @stevomitric Please address this one too: reword the scaladoc and the comment at line 80 to state the current invariant as the intended behavior, and point to a JIRA ticket for the pending `isAdjustedToUTC` resolution instead of the PR-review framing. -- 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]
