org.apache.drill.exec.store.easy.text.reader

Class RepeatedVarCharOutput

java.lang.Object

org.apache.drill.exec.store.easy.text.reader.BaseFieldOutput

org.apache.drill.exec.store.easy.text.reader.RepeatedVarCharOutput

public class RepeatedVarCharOutput
extends BaseFieldOutput

Class is responsible for generating record batches for text file inputs. We generate a record batch with a single vector of type repeated varchar vector. Each record is a single value within the vector containing all the fields in the record as individual array elements.

Field Summary

Fields inherited from class org.apache.drill.exec.store.easy.text.reader.BaseFieldOutput

currentDataPointer, currentFieldIndex, fieldBytes, fieldProjected, fieldWriteCount, maxField, writer

Constructor Summary

Constructors

Constructor and Description
RepeatedVarCharOutput(RowSetLoader loader, boolean[] projectionMask) Provide the row set loader (which must have just one repeated Varchar column) and an optional array projection mask.

Method Summary

Modifier and Type	Method and Description
protected ScalarWriter	columnWriter()
boolean	endField() Write the value into an array position.

Methods inherited from class org.apache.drill.exec.store.easy.text.reader.BaseFieldOutput

append, endEmptyField, finishRecord, getRecordCount, isFull, startField, startRecord, writeToVector

Methods inherited from class java.lang.Object

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

Constructor Detail

RepeatedVarCharOutput

public RepeatedVarCharOutput(RowSetLoader loader,
                             boolean[] projectionMask)

Provide the row set loader (which must have just one repeated Varchar column) and an optional array projection mask.

Parameters:

loader - row set loader

projectionMask - array projection mask

Method Detail

endField

public boolean endField()

Write the value into an array position. Rules:

• If there is no projection mask, collect all columns.
• If a selection mask is present, we previously found the index of the last projection column (maxField). If the current column is beyond that number, ignore the data and stop accepting columns.
• If the column is projected, add the data to the array.
• If the column is not projected, add a blank value to the array.

The above ensures that we leave no holes in the portion of the array that is projected (by adding blank columns where needed), and we just ignore columns past the end of the projected part of the array. (No need to fill holes at the end.)

Overrides:

endField in class BaseFieldOutput

Returns:

true if engine should continue processing record. false if rest of record can be skipped.

columnWriter

protected ScalarWriter columnWriter()

Specified by:

columnWriter in class BaseFieldOutput