voonhous commented on code in PR #18347: URL: https://github.com/apache/hudi/pull/18347#discussion_r2987199605
########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/hudi/blob/BlobTestHelpers.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.hudi.blob + +import org.apache.hudi.common.schema.{HoodieSchema, HoodieSchemaType} + +import org.apache.spark.sql.{Column, SparkSession} +import org.apache.spark.sql.functions.{lit, struct} +import org.apache.spark.sql.types._ +import org.junit.jupiter.api.Assertions.{assertEquals, assertTrue} + +import java.io.File +import java.nio.file.{Files, Path} + +object BlobTestHelpers { + def blobMetadata: Metadata = { + new MetadataBuilder() + .putString(HoodieSchema.TYPE_METADATA_FIELD, HoodieSchemaType.BLOB.name()) + .build() + } + + def inlineBlobStructCol(name: String, bytesCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.INLINE).as(HoodieSchema.Blob.TYPE), + bytesCol.as(HoodieSchema.Blob.INLINE_DATA_FIELD), + lit(null).cast("struct<externalPath:string,offset:bigint,length:bigint,managed:boolean>") + .as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def wholeFileBlobStructCol(name: String, filePathCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.OUT_OF_LINE).as(HoodieSchema.Blob.TYPE), + lit(null).cast("binary").as(HoodieSchema.Blob.INLINE_DATA_FIELD), + struct( + filePathCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_PATH), + lit(null).cast("bigint").as(HoodieSchema.Blob.EXTERNAL_REFERENCE_OFFSET), + lit(null).cast("bigint").as(HoodieSchema.Blob.EXTERNAL_REFERENCE_LENGTH), + lit(false).as(HoodieSchema.Blob.EXTERNAL_REFERENCE_IS_MANAGED) + ).as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def blobStructCol(name: String, filePathCol: Column, offsetCol: Column, lengthCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.OUT_OF_LINE).as(HoodieSchema.Blob.TYPE), + lit(null).cast("binary").as(HoodieSchema.Blob.INLINE_DATA_FIELD), + struct( + filePathCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_PATH), + offsetCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_OFFSET), + lengthCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_LENGTH), + lit(false).as(HoodieSchema.Blob.EXTERNAL_REFERENCE_IS_MANAGED) + ).as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def createTestFile(tempDir: Path, name: String, size: Int): String = { + val file = new File(tempDir.toString, name) + val bytes = (0 until size).map(i => (i % 256).toByte).toArray + Files.write(file.toPath, bytes) + file.getAbsolutePath + } + + /** + * Assert that byte array contains expected pattern (i % 256) at given offset. + * + * @param data Array of bytes to verify + * @param expectedOffset Starting offset for pattern (default 0) + */ + def assertBytesContent(data: Array[Byte], expectedOffset: Int = 0): Unit = { + for (i <- 0 until data.length) { + assertEquals((expectedOffset + i) % 256, data(i) & 0xFF, + s"Mismatch at byte $i (global offset ${expectedOffset + i})") + } + } Review Comment: Is this used? ########## hudi-spark-datasource/hudi-spark3.4.x/src/main/scala/org/apache/spark/sql/parser/HoodieSpark3_4ExtendedSqlAstBuilder.scala: ########## @@ -2689,13 +2692,24 @@ class HoodieSpark3_4ExtendedSqlAstBuilder(conf: SQLConf, delegate: ParserInterfa builder.putString("comment", _) } + val dataType = typedVisit[DataType](ctx.dataType) + + addMetadataForType(ctx.dataType(), builder) + StructField( name = colName.getText, - dataType = typedVisit[DataType](ctx.dataType), + dataType = dataType, nullable = NULL == null, metadata = builder.build()) } + private def addMetadataForType(dataType: HoodieSqlBaseParser.DataTypeContext, builder: MetadataBuilder): Unit = { + val typeText = dataType.getText + if (typeText.equalsIgnoreCase(HoodieSchemaType.BLOB.name())) { + builder.putString(HoodieSchema.TYPE_METADATA_FIELD, HoodieSchemaType.BLOB.name()) + } + } + Review Comment: Can this be placed in a shared location since logic seems duplicated? We can reduce code duplication since this doesn't seem to be version specific. ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/hudi/blob/BlobTestHelpers.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.hudi.blob + +import org.apache.hudi.common.schema.{HoodieSchema, HoodieSchemaType} + +import org.apache.spark.sql.{Column, SparkSession} +import org.apache.spark.sql.functions.{lit, struct} +import org.apache.spark.sql.types._ +import org.junit.jupiter.api.Assertions.{assertEquals, assertTrue} + +import java.io.File +import java.nio.file.{Files, Path} + +object BlobTestHelpers { + def blobMetadata: Metadata = { + new MetadataBuilder() + .putString(HoodieSchema.TYPE_METADATA_FIELD, HoodieSchemaType.BLOB.name()) + .build() + } + + def inlineBlobStructCol(name: String, bytesCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.INLINE).as(HoodieSchema.Blob.TYPE), + bytesCol.as(HoodieSchema.Blob.INLINE_DATA_FIELD), + lit(null).cast("struct<externalPath:string,offset:bigint,length:bigint,managed:boolean>") + .as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def wholeFileBlobStructCol(name: String, filePathCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.OUT_OF_LINE).as(HoodieSchema.Blob.TYPE), + lit(null).cast("binary").as(HoodieSchema.Blob.INLINE_DATA_FIELD), + struct( + filePathCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_PATH), + lit(null).cast("bigint").as(HoodieSchema.Blob.EXTERNAL_REFERENCE_OFFSET), + lit(null).cast("bigint").as(HoodieSchema.Blob.EXTERNAL_REFERENCE_LENGTH), + lit(false).as(HoodieSchema.Blob.EXTERNAL_REFERENCE_IS_MANAGED) + ).as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def blobStructCol(name: String, filePathCol: Column, offsetCol: Column, lengthCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.OUT_OF_LINE).as(HoodieSchema.Blob.TYPE), + lit(null).cast("binary").as(HoodieSchema.Blob.INLINE_DATA_FIELD), + struct( + filePathCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_PATH), + offsetCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_OFFSET), + lengthCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_LENGTH), + lit(false).as(HoodieSchema.Blob.EXTERNAL_REFERENCE_IS_MANAGED) + ).as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def createTestFile(tempDir: Path, name: String, size: Int): String = { + val file = new File(tempDir.toString, name) + val bytes = (0 until size).map(i => (i % 256).toByte).toArray + Files.write(file.toPath, bytes) + file.getAbsolutePath + } + + /** + * Assert that byte array contains expected pattern (i % 256) at given offset. + * + * @param data Array of bytes to verify + * @param expectedOffset Starting offset for pattern (default 0) + */ + def assertBytesContent(data: Array[Byte], expectedOffset: Int = 0): Unit = { + for (i <- 0 until data.length) { + assertEquals((expectedOffset + i) % 256, data(i) & 0xFF, + s"Mismatch at byte $i (global offset ${expectedOffset + i})") + } + } + + /** + * Execute code block with temporary Spark configuration. + * Automatically restores previous values after execution. + * + * @param spark SparkSession instance + * @param configs Configuration key-value pairs to set + * @param fn Code block to execute with configs + */ + def withSparkConfig[T](spark: SparkSession, configs: Map[String, String])(fn: => T): T = { + val oldValues = configs.keys.map(k => (k, spark.conf.getOption(k))).toMap + try { + configs.foreach { case (k, v) => spark.conf.set(k, v) } + fn + } finally { + oldValues.foreach { case (k, oldValue) => + oldValue match { + case Some(v) => spark.conf.set(k, v) + case None => spark.conf.unset(k) + } + } + } + } Review Comment: Same question here, is this used? ########## hudi-spark-datasource/hudi-spark3.3.x/src/main/scala/org/apache/spark/sql/parser/HoodieSpark3_3ExtendedSqlAstBuilder.scala: ########## @@ -2713,11 +2727,18 @@ class HoodieSpark3_3ExtendedSqlAstBuilder(conf: SQLConf, delegate: ParserInterfa */ override def visitComplexColType(ctx: ComplexColTypeContext): StructField = withOrigin(ctx) { import ctx._ - val structField = StructField( + val builder = new MetadataBuilder + // Add comment to metadata + Option(commentSpec()).map(visitCommentSpec).foreach { + builder.putString("comment", _) + } + addMetadataForType(ctx.dataType(), builder) + + StructField( name = identifier.getText, dataType = typedVisit(dataType()), - nullable = NULL == null) - Option(commentSpec).map(visitCommentSpec).map(structField.withComment).getOrElse(structField) Review Comment: IIUC, we can make this more succinct: ```java val builder = new MetadataBuilder addMetadataForType(ctx.dataType(), builder) val structField = StructField( name = identifier.getText, dataType = typedVisit(dataType()), nullable = NULL == null, metadata = builder.build()) Option(commentSpec).map(visitCommentSpec).map(structField.withComment).getOrElse(structField) ``` ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/hudi/blob/TestBlobSupport.scala: ########## @@ -0,0 +1,143 @@ +/* + * 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.hudi.blob + +import org.apache.hudi.SparkDatasetMixin +import org.apache.hudi.blob.BlobTestHelpers._ +import org.apache.hudi.client.SparkRDDWriteClient +import org.apache.hudi.common.model.{HoodieAvroIndexedRecord, HoodieFileFormat, HoodieKey, HoodieRecord, HoodieTableType} +import org.apache.hudi.common.schema.{HoodieSchema, HoodieSchemaField, HoodieSchemaType} +import org.apache.hudi.common.table.{HoodieTableConfig, HoodieTableMetaClient} +import org.apache.hudi.config.HoodieIndexConfig +import org.apache.hudi.index.HoodieIndex +import org.apache.hudi.testutils.HoodieClientTestBase + +import org.apache.avro.generic.{GenericData, GenericRecord, IndexedRecord} +import org.junit.jupiter.api.Assertions.{assertEquals, assertTrue} +import org.junit.jupiter.params.ParameterizedTest +import org.junit.jupiter.params.provider.EnumSource + +import java.util.{Arrays, Properties} + +import scala.collection.JavaConverters._ + +/** + * End-to-end test for blob support including schema creation, multi-commit writes, + * DataFrame reads, and SQL blob retrieval. + * + * This test validates: + * <ul> + * <li>Creating tables with blob columns</li> + * <li>Writing blob records with out-of-line references</li> + * <li>Multi-commit operations (insert + upsert)</li> + * <li>Reading blob references via DataFrame</li> + * <li>SQL read_blob() function integration</li> + * </ul> + */ +class TestBlobSupport extends HoodieClientTestBase with SparkDatasetMixin { + val SCHEMA: HoodieSchema = HoodieSchema.createRecord("test_blobs", null, null, Arrays.asList( + HoodieSchemaField.of("id", HoodieSchema.create(HoodieSchemaType.STRING), null, null), + HoodieSchemaField.of("value", HoodieSchema.create(HoodieSchemaType.INT), null, null), + HoodieSchemaField.of("data", HoodieSchema.createBlob(), null, null) + )) + + @ParameterizedTest + @EnumSource(classOf[HoodieTableType]) + def testEndToEnd(tableType: HoodieTableType): Unit = { + val filePath1 = createTestFile(tempDir, "file1.bin", 1000) + val filePath2 = createTestFile(tempDir, "file2.bin", 1000) + + val properties = new Properties() + properties.put("hoodie.datasource.write.recordkey.field", "id") + properties.put("hoodie.datasource.write.partitionpath.field", "") + properties.put(HoodieTableConfig.RECORDKEY_FIELDS.key(), "id") + properties.put(HoodieTableConfig.PARTITION_FIELDS.key(), "") + properties.setProperty(HoodieTableConfig.BASE_FILE_FORMAT.key(), HoodieFileFormat.PARQUET.toString) + + // Initialize table + HoodieTableMetaClient.newTableBuilder() + .setTableName("test_blob_table") + .setTableType(tableType) + .fromProperties(properties) + .initTable(storageConf, basePath) + + var client: SparkRDDWriteClient[IndexedRecord] = null + val config = getConfigBuilder(SCHEMA.toString()) + .withIndexConfig(HoodieIndexConfig.newBuilder.withIndexType(HoodieIndex.IndexType.SIMPLE).build) + .build() + try { + client = getHoodieWriteClient(config).asInstanceOf[SparkRDDWriteClient[IndexedRecord]] + + // First commit - insert + val commit1 = client.startCommit() + val firstBatch = createTestRecords(filePath1) + val statuses1 = client.insert(jsc.parallelize(firstBatch.asJava, 1), commit1).collect() + client.commit(commit1, jsc.parallelize(statuses1, 1)) + + // Second commit - upsert + val commit2 = client.startCommit() + val secondBatch = createTestRecords(filePath2) + val statuses2 = client.upsert(jsc.parallelize(secondBatch.asJava, 1), commit2).collect() + client.commit(commit2, jsc.parallelize(statuses2, 1)) + } finally { + if (client != null) client.close() + } + + // Read and verify DataFrame results + val table = sparkSession.read.format("hudi").load(basePath) + val rows = table.collectAsList() + assertEquals(10, rows.size()) + + rows.asScala.foreach { row => + val data = row.getStruct(row.fieldIndex("data")) + val reference = data.getStruct(data.fieldIndex(HoodieSchema.Blob.EXTERNAL_REFERENCE)) + val filePath = reference.getString(reference.fieldIndex(HoodieSchema.Blob.EXTERNAL_REFERENCE_PATH)) + assertTrue(filePath.endsWith("file2.bin")) + } + } + + private def createTestRecords(filePath: String): Seq[HoodieRecord[IndexedRecord]] = { + (0 until 10).map { i => + val id = s"id_$i" + val key = new HoodieKey(id, "") + + val dataSchema = SCHEMA.getField("data").get.schema + val fileReference = new GenericData.Record(dataSchema.getField(HoodieSchema.Blob.EXTERNAL_REFERENCE) + .get.getNonNullSchema.toAvroSchema) + fileReference.put(HoodieSchema.Blob.EXTERNAL_REFERENCE_PATH, filePath) + fileReference.put(HoodieSchema.Blob.EXTERNAL_REFERENCE_OFFSET, i * 100L) + fileReference.put(HoodieSchema.Blob.EXTERNAL_REFERENCE_LENGTH, 100L) + fileReference.put(HoodieSchema.Blob.EXTERNAL_REFERENCE_IS_MANAGED, false) + Review Comment: Nit: fix double blank line ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/hudi/blob/BlobTestHelpers.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.hudi.blob + +import org.apache.hudi.common.schema.{HoodieSchema, HoodieSchemaType} + +import org.apache.spark.sql.{Column, SparkSession} +import org.apache.spark.sql.functions.{lit, struct} +import org.apache.spark.sql.types._ +import org.junit.jupiter.api.Assertions.{assertEquals, assertTrue} + +import java.io.File +import java.nio.file.{Files, Path} + +object BlobTestHelpers { + def blobMetadata: Metadata = { + new MetadataBuilder() + .putString(HoodieSchema.TYPE_METADATA_FIELD, HoodieSchemaType.BLOB.name()) + .build() + } + + def inlineBlobStructCol(name: String, bytesCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.INLINE).as(HoodieSchema.Blob.TYPE), + bytesCol.as(HoodieSchema.Blob.INLINE_DATA_FIELD), + lit(null).cast("struct<externalPath:string,offset:bigint,length:bigint,managed:boolean>") + .as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def wholeFileBlobStructCol(name: String, filePathCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.OUT_OF_LINE).as(HoodieSchema.Blob.TYPE), + lit(null).cast("binary").as(HoodieSchema.Blob.INLINE_DATA_FIELD), + struct( + filePathCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_PATH), + lit(null).cast("bigint").as(HoodieSchema.Blob.EXTERNAL_REFERENCE_OFFSET), + lit(null).cast("bigint").as(HoodieSchema.Blob.EXTERNAL_REFERENCE_LENGTH), + lit(false).as(HoodieSchema.Blob.EXTERNAL_REFERENCE_IS_MANAGED) + ).as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def blobStructCol(name: String, filePathCol: Column, offsetCol: Column, lengthCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.OUT_OF_LINE).as(HoodieSchema.Blob.TYPE), + lit(null).cast("binary").as(HoodieSchema.Blob.INLINE_DATA_FIELD), + struct( + filePathCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_PATH), + offsetCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_OFFSET), + lengthCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_LENGTH), + lit(false).as(HoodieSchema.Blob.EXTERNAL_REFERENCE_IS_MANAGED) + ).as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def createTestFile(tempDir: Path, name: String, size: Int): String = { + val file = new File(tempDir.toString, name) + val bytes = (0 until size).map(i => (i % 256).toByte).toArray + Files.write(file.toPath, bytes) + file.getAbsolutePath + } + + /** + * Assert that byte array contains expected pattern (i % 256) at given offset. + * + * @param data Array of bytes to verify + * @param expectedOffset Starting offset for pattern (default 0) + */ + def assertBytesContent(data: Array[Byte], expectedOffset: Int = 0): Unit = { + for (i <- 0 until data.length) { + assertEquals((expectedOffset + i) % 256, data(i) & 0xFF, + s"Mismatch at byte $i (global offset ${expectedOffset + i})") + } + } + + /** + * Execute code block with temporary Spark configuration. + * Automatically restores previous values after execution. + * + * @param spark SparkSession instance + * @param configs Configuration key-value pairs to set + * @param fn Code block to execute with configs + */ + def withSparkConfig[T](spark: SparkSession, configs: Map[String, String])(fn: => T): T = { + val oldValues = configs.keys.map(k => (k, spark.conf.getOption(k))).toMap + try { + configs.foreach { case (k, v) => spark.conf.set(k, v) } + fn + } finally { + oldValues.foreach { case (k, oldValue) => + oldValue match { + case Some(v) => spark.conf.set(k, v) + case None => spark.conf.unset(k) + } + } + } + } + + /** + * Assert that DataFrame contains all specified columns. + * + * @param df DataFrame to check + * @param columnNames Variable number of column names + */ + def assertColumnsExist(df: org.apache.spark.sql.DataFrame, columnNames: String*): Unit = { Review Comment: `org.apache.spark.sql.DataFrame`, can we import this? ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/hudi/blob/BlobTestHelpers.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.hudi.blob + +import org.apache.hudi.common.schema.{HoodieSchema, HoodieSchemaType} + +import org.apache.spark.sql.{Column, SparkSession} +import org.apache.spark.sql.functions.{lit, struct} +import org.apache.spark.sql.types._ +import org.junit.jupiter.api.Assertions.{assertEquals, assertTrue} + +import java.io.File +import java.nio.file.{Files, Path} + +object BlobTestHelpers { + def blobMetadata: Metadata = { + new MetadataBuilder() + .putString(HoodieSchema.TYPE_METADATA_FIELD, HoodieSchemaType.BLOB.name()) + .build() + } + + def inlineBlobStructCol(name: String, bytesCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.INLINE).as(HoodieSchema.Blob.TYPE), + bytesCol.as(HoodieSchema.Blob.INLINE_DATA_FIELD), + lit(null).cast("struct<externalPath:string,offset:bigint,length:bigint,managed:boolean>") + .as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } Review Comment: Is this used? ########## hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/hudi/blob/BlobTestHelpers.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.hudi.blob + +import org.apache.hudi.common.schema.{HoodieSchema, HoodieSchemaType} + +import org.apache.spark.sql.{Column, SparkSession} +import org.apache.spark.sql.functions.{lit, struct} +import org.apache.spark.sql.types._ +import org.junit.jupiter.api.Assertions.{assertEquals, assertTrue} + +import java.io.File +import java.nio.file.{Files, Path} + +object BlobTestHelpers { + def blobMetadata: Metadata = { + new MetadataBuilder() + .putString(HoodieSchema.TYPE_METADATA_FIELD, HoodieSchemaType.BLOB.name()) + .build() + } + + def inlineBlobStructCol(name: String, bytesCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.INLINE).as(HoodieSchema.Blob.TYPE), + bytesCol.as(HoodieSchema.Blob.INLINE_DATA_FIELD), + lit(null).cast("struct<externalPath:string,offset:bigint,length:bigint,managed:boolean>") + .as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def wholeFileBlobStructCol(name: String, filePathCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.OUT_OF_LINE).as(HoodieSchema.Blob.TYPE), + lit(null).cast("binary").as(HoodieSchema.Blob.INLINE_DATA_FIELD), + struct( + filePathCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_PATH), + lit(null).cast("bigint").as(HoodieSchema.Blob.EXTERNAL_REFERENCE_OFFSET), + lit(null).cast("bigint").as(HoodieSchema.Blob.EXTERNAL_REFERENCE_LENGTH), + lit(false).as(HoodieSchema.Blob.EXTERNAL_REFERENCE_IS_MANAGED) + ).as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def blobStructCol(name: String, filePathCol: Column, offsetCol: Column, lengthCol: Column): Column = { + struct( + lit(HoodieSchema.Blob.OUT_OF_LINE).as(HoodieSchema.Blob.TYPE), + lit(null).cast("binary").as(HoodieSchema.Blob.INLINE_DATA_FIELD), + struct( + filePathCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_PATH), + offsetCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_OFFSET), + lengthCol.as(HoodieSchema.Blob.EXTERNAL_REFERENCE_LENGTH), + lit(false).as(HoodieSchema.Blob.EXTERNAL_REFERENCE_IS_MANAGED) + ).as(HoodieSchema.Blob.EXTERNAL_REFERENCE) + ).as(name, blobMetadata) + } + + def createTestFile(tempDir: Path, name: String, size: Int): String = { + val file = new File(tempDir.toString, name) + val bytes = (0 until size).map(i => (i % 256).toByte).toArray + Files.write(file.toPath, bytes) + file.getAbsolutePath + } + + /** + * Assert that byte array contains expected pattern (i % 256) at given offset. + * + * @param data Array of bytes to verify + * @param expectedOffset Starting offset for pattern (default 0) + */ + def assertBytesContent(data: Array[Byte], expectedOffset: Int = 0): Unit = { + for (i <- 0 until data.length) { + assertEquals((expectedOffset + i) % 256, data(i) & 0xFF, + s"Mismatch at byte $i (global offset ${expectedOffset + i})") + } + } + + /** + * Execute code block with temporary Spark configuration. + * Automatically restores previous values after execution. + * + * @param spark SparkSession instance + * @param configs Configuration key-value pairs to set + * @param fn Code block to execute with configs + */ + def withSparkConfig[T](spark: SparkSession, configs: Map[String, String])(fn: => T): T = { + val oldValues = configs.keys.map(k => (k, spark.conf.getOption(k))).toMap + try { + configs.foreach { case (k, v) => spark.conf.set(k, v) } + fn + } finally { + oldValues.foreach { case (k, oldValue) => + oldValue match { + case Some(v) => spark.conf.set(k, v) + case None => spark.conf.unset(k) + } + } + } + } + + /** + * Assert that DataFrame contains all specified columns. + * + * @param df DataFrame to check + * @param columnNames Variable number of column names + */ + def assertColumnsExist(df: org.apache.spark.sql.DataFrame, columnNames: String*): Unit = { + columnNames.foreach { colName => + assertTrue(df.columns.contains(colName), + s"DataFrame missing expected column: $colName. Available: ${df.columns.mkString(", ")}") + } + } Review Comment: Is this used? -- 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]
