Package org.apache.flink.formats.parquet.utils

Class NestedPositionUtil

java.lang.Object

org.apache.flink.formats.parquet.utils.NestedPositionUtil

public class NestedPositionUtil
extends Object

Utils to calculate nested type position.

Constructor Summary

Constructors

Constructor	Description
NestedPositionUtil()

Method Summary

Modifier and Type	Method	Description
static CollectionPosition	calculateCollectionOffsets(ParquetField field, int[] definitionLevels, int[] repetitionLevels)	Calculate the collection's offsets according to column's max repetition level, definition level, value's repetition level and definition level.
static long[]	calculateLengthByOffsets(boolean[] collectionIsEmpty, long[] arrayOffsets)
static RowPosition	calculateRowOffsets(ParquetField field, int[] fieldDefinitionLevels, int[] fieldRepetitionLevels)	Calculate row offsets according to column's max repetition level, definition level, value's repetition level and definition level.
static boolean	isOptionalFieldValueNull(int definitionLevel, int maxDefinitionLevel)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

NestedPositionUtil

public NestedPositionUtil()

Method Details

calculateRowOffsets

public static RowPosition calculateRowOffsets(ParquetField field,
 int[] fieldDefinitionLevels,
 int[] fieldRepetitionLevels)

Calculate row offsets according to column's max repetition level, definition level, value's repetition level and definition level. Each row has three situation:

Row is not defined,because it's optional parent fields is null, this is decided by its parent's repetition level

Row is null

Row is defined and not empty.

Parameters:

field - field that contains the row column message include max repetition level and definition level.

fieldRepetitionLevels - int array with each value's repetition level.

fieldDefinitionLevels - int array with each value's definition level.

Returns:

RowPosition contains collections row count and isNull array.

calculateCollectionOffsets

public static CollectionPosition calculateCollectionOffsets(ParquetField field,
 int[] definitionLevels,
 int[] repetitionLevels)

Calculate the collection's offsets according to column's max repetition level, definition level, value's repetition level and definition level. Each collection (Array or Map) has four situation:

Collection is not defined, because optional parent fields is null, this is decided by its parent's repetition level

Collection is null

Collection is defined but empty

Collection is defined and not empty. In this case offset value is increased by the number of elements in that collection

Parameters:

field - field that contains array/map column message include max repetition level and definition level.

definitionLevels - int array with each value's repetition level.

repetitionLevels - int array with each value's definition level.

Returns:

CollectionPosition contains collections offset array, length array and isNull array.

isOptionalFieldValueNull

public static boolean isOptionalFieldValueNull(int definitionLevel,
 int maxDefinitionLevel)

calculateLengthByOffsets

public static long[] calculateLengthByOffsets(boolean[] collectionIsEmpty,
 long[] arrayOffsets)