in shims/spark32/src/main/scala/org/apache/spark/sql/execution/datasources/parquet/ParquetFileFormat.scala [230:425]
override def buildReaderWithPartitionValues(
sparkSession: SparkSession,
dataSchema: StructType,
partitionSchema: StructType,
requiredSchema: StructType,
filters: Seq[Filter],
options: Map[String, String],
hadoopConf: Configuration): PartitionedFile => Iterator[InternalRow] = {
hadoopConf.set(ParquetInputFormat.READ_SUPPORT_CLASS, classOf[ParquetReadSupport].getName)
hadoopConf.set(ParquetReadSupport.SPARK_ROW_REQUESTED_SCHEMA, requiredSchema.json)
hadoopConf.set(ParquetWriteSupport.SPARK_ROW_SCHEMA, requiredSchema.json)
hadoopConf.set(
SQLConf.SESSION_LOCAL_TIMEZONE.key,
sparkSession.sessionState.conf.sessionLocalTimeZone)
hadoopConf.setBoolean(
SQLConf.NESTED_SCHEMA_PRUNING_ENABLED.key,
sparkSession.sessionState.conf.nestedSchemaPruningEnabled)
hadoopConf.setBoolean(
SQLConf.CASE_SENSITIVE.key,
sparkSession.sessionState.conf.caseSensitiveAnalysis)
ParquetWriteSupport.setSchema(requiredSchema, hadoopConf)
// Sets flags for `ParquetToSparkSchemaConverter`
hadoopConf.setBoolean(
SQLConf.PARQUET_BINARY_AS_STRING.key,
sparkSession.sessionState.conf.isParquetBinaryAsString)
hadoopConf.setBoolean(
SQLConf.PARQUET_INT96_AS_TIMESTAMP.key,
sparkSession.sessionState.conf.isParquetINT96AsTimestamp)
val broadcastedHadoopConf =
sparkSession.sparkContext.broadcast(new SerializableConfiguration(hadoopConf))
// TODO: if you move this into the closure it reverts to the default values.
// If true, enable using the custom RecordReader for parquet. This only works for
// a subset of the types (no complex types).
val resultSchema = StructType(partitionSchema.fields ++ requiredSchema.fields)
val sqlConf = sparkSession.sessionState.conf
val enableOffHeapColumnVector = sqlConf.offHeapColumnVectorEnabled
val enableVectorizedReader: Boolean =
sqlConf.parquetVectorizedReaderEnabled &&
resultSchema.forall(_.dataType.isInstanceOf[AtomicType])
val enableRecordFilter: Boolean = sqlConf.parquetRecordFilterEnabled
val timestampConversion: Boolean = sqlConf.isParquetINT96TimestampConversion
val capacity = sqlConf.parquetVectorizedReaderBatchSize
val enableParquetFilterPushDown: Boolean = sqlConf.parquetFilterPushDown
// Whole stage codegen (PhysicalRDD) is able to deal with batches directly
val returningBatch = supportBatch(sparkSession, resultSchema)
val pushDownDate = sqlConf.parquetFilterPushDownDate
val pushDownTimestamp = sqlConf.parquetFilterPushDownTimestamp
val pushDownDecimal = sqlConf.parquetFilterPushDownDecimal
val pushDownStringStartWith = sqlConf.parquetFilterPushDownStringStartWith
val pushDownInFilterThreshold = sqlConf.parquetFilterPushDownInFilterThreshold
val isCaseSensitive = sqlConf.caseSensitiveAnalysis
val parquetOptions = new ParquetOptions(options, sparkSession.sessionState.conf)
val datetimeRebaseModeInRead = parquetOptions.datetimeRebaseModeInRead
val int96RebaseModeInRead = parquetOptions.int96RebaseModeInRead
(file: PartitionedFile) => {
assert(file.partitionValues.numFields == partitionSchema.size)
val filePath = new Path(new URI(file.filePath))
val split = new FileSplit(filePath, file.start, file.length, Array.empty[String])
val sharedConf = broadcastedHadoopConf.value.value
lazy val footerFileMetaData =
ParquetFooterReader.readFooter(sharedConf, filePath, SKIP_ROW_GROUPS).getFileMetaData
val datetimeRebaseSpec = DataSourceUtils.datetimeRebaseSpec(
footerFileMetaData.getKeyValueMetaData.get,
datetimeRebaseModeInRead)
// Try to push down filters when filter push-down is enabled.
val pushed = if (enableParquetFilterPushDown) {
val parquetSchema = footerFileMetaData.getSchema
val parquetFilters = new ParquetFilters(
parquetSchema,
pushDownDate,
pushDownTimestamp,
pushDownDecimal,
pushDownStringStartWith,
pushDownInFilterThreshold,
isCaseSensitive,
datetimeRebaseSpec)
filters
// Collects all converted Parquet filter predicates. Notice that not all predicates can be
// converted (`ParquetFilters.createFilter` returns an `Option`). That's why a `flatMap`
// is used here.
.flatMap(parquetFilters.createFilter(_))
.reduceOption(FilterApi.and)
} else {
None
}
// PARQUET_INT96_TIMESTAMP_CONVERSION says to apply timezone conversions to int96 timestamps'
// *only* if the file was created by something other than "parquet-mr", so check the actual
// writer here for this file. We have to do this per-file, as each file in the table may
// have different writers.
// Define isCreatedByParquetMr as function to avoid unnecessary parquet footer reads.
def isCreatedByParquetMr: Boolean =
footerFileMetaData.getCreatedBy.startsWith("parquet-mr")
val convertTz =
if (timestampConversion && !isCreatedByParquetMr) {
Some(DateTimeUtils.getZoneId(sharedConf.get(SQLConf.SESSION_LOCAL_TIMEZONE.key)))
} else {
None
}
val int96RebaseSpec = DataSourceUtils.int96RebaseSpec(
footerFileMetaData.getKeyValueMetaData.get,
int96RebaseModeInRead)
val attemptId = new TaskAttemptID(new TaskID(new JobID(), TaskType.MAP, 0), 0)
val hadoopAttemptContext =
new TaskAttemptContextImpl(broadcastedHadoopConf.value.value, attemptId)
// Try to push down filters when filter push-down is enabled.
// Notice: This push-down is RowGroups level, not individual records.
if (pushed.isDefined) {
ParquetInputFormat.setFilterPredicate(hadoopAttemptContext.getConfiguration, pushed.get)
}
val taskContext = Option(TaskContext.get())
if (enableVectorizedReader) {
val vectorizedReader = new VectorizedParquetRecordReader(
convertTz.orNull,
datetimeRebaseSpec.mode.toString,
datetimeRebaseSpec.timeZone,
int96RebaseSpec.mode.toString,
int96RebaseSpec.timeZone,
enableOffHeapColumnVector && taskContext.isDefined,
capacity
)
// SPARK-37089: We cannot register a task completion listener to close this iterator here
// because downstream exec nodes have already registered their listeners. Since listeners
// are executed in reverse order of registration, a listener registered here would close the
// iterator while downstream exec nodes are still running. When off-heap column vectors are
// enabled, this can cause a use-after-free bug leading to a segfault.
//
// Instead, we use FileScanRDD's task completion listener to close this iterator.
val iter = new RecordReaderIterator(vectorizedReader)
try {
vectorizedReader.initialize(split, hadoopAttemptContext)
logDebug(s"Appending $partitionSchema ${file.partitionValues}")
vectorizedReader.initBatch(partitionSchema, file.partitionValues)
if (returningBatch) {
vectorizedReader.enableReturningBatches()
}
// UnsafeRowParquetRecordReader appends the columns internally to avoid another copy.
iter.asInstanceOf[Iterator[InternalRow]]
} catch {
case e: Throwable =>
// SPARK-23457: In case there is an exception in initialization, close the iterator to
// avoid leaking resources.
iter.close()
throw e
}
} else {
logDebug(s"Falling back to parquet-mr")
// ParquetRecordReader returns InternalRow
val readSupport = new ParquetReadSupport(
convertTz,
enableVectorizedReader = false,
datetimeRebaseSpec,
int96RebaseSpec)
val reader = if (pushed.isDefined && enableRecordFilter) {
val parquetFilter = FilterCompat.get(pushed.get, null)
new ParquetRecordReader[InternalRow](readSupport, parquetFilter)
} else {
new ParquetRecordReader[InternalRow](readSupport)
}
val iter = new RecordReaderIterator[InternalRow](reader)
try {
reader.initialize(split, hadoopAttemptContext)
val fullSchema = requiredSchema.toAttributes ++ partitionSchema.toAttributes
val unsafeProjection = GenerateUnsafeProjection.generate(fullSchema, fullSchema)
if (partitionSchema.length == 0) {
// There is no partition columns
iter.map(unsafeProjection)
} else {
val joinedRow = new JoinedRow()
iter.map(d => unsafeProjection(joinedRow(d, file.partitionValues)))
}
} catch {
case e: Throwable =>
// SPARK-23457: In case there is an exception in initialization, close the iterator to
// avoid leaking resources.
iter.close()
throw e
}
}
}
}