org.apache.hadoop.hive.ql.io.orc

Class OrcRecordUpdater

java.lang.Object

org.apache.hadoop.hive.ql.io.orc.OrcRecordUpdater

All Implemented Interfaces:

RecordUpdater

public class OrcRecordUpdater
extends Object
implements RecordUpdater

A RecordUpdater where the files are stored as ORC. A note on various record structures: the row coming in (as in insert(long, Object) for example), is a struct like <RecordIdentifier, f1, ... fn> but what is written to the file * is <op, owid, writerId, rowid, cwid, <f1, ... fn>> (see createEventObjectInspector(ObjectInspector)) So there are OIs here to make the translation.

Field Summary

Fields

Modifier and Type	Field and Description
static Collection	ALL_ACID_ROW_NAMES
static int	ROW

Method Summary

Modifier and Type	Method and Description
void	close(boolean abort) Close this updater.
void	delete(long currentWriteId, Object row) Delete a row from the table.
void	flush() Flush the current set of rows to the underlying file system, so that they are available to readers.
long	getBufferedRowCount() Returns the number of rows in memory before flush().
org.apache.hadoop.fs.Path	getDeleteFilePath()
SerDeStats	getStats() Returns the statistics information
org.apache.hadoop.fs.Path	getUpdatedFilePath() Returns the path of the file this updater wrote to
void	insert(long currentWriteId, Object row) Insert a new record into the table.
String	toString()
void	update(long currentWriteId, Object row) Update an old record with a new set of values.

Methods inherited from class java.lang.Object

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

Field Detail

ROW

public static final int ROW

See Also:

Constant Field Values

ALL_ACID_ROW_NAMES

public static final Collection ALL_ACID_ROW_NAMES

Method Detail

toString

public String toString()

Overrides:

toString in class Object

insert

public void insert(long currentWriteId,
                   Object row)
            throws IOException

Description copied from interface: RecordUpdater

Insert a new record into the table.

Specified by:

insert in interface RecordUpdater

Parameters:

currentWriteId - the table write id of the current transaction.

row - the row of data to insert

Throws:

IOException

update

public void update(long currentWriteId,
                   Object row)
            throws IOException

Description copied from interface: RecordUpdater

Update an old record with a new set of values.

Specified by:

update in interface RecordUpdater

Parameters:

currentWriteId - the current write id

row - the new values for the row

Throws:

IOException

delete

public void delete(long currentWriteId,
                   Object row)
            throws IOException

Description copied from interface: RecordUpdater

Delete a row from the table.

Specified by:

delete in interface RecordUpdater

Parameters:

currentWriteId - the current write id

Throws:

IOException

flush

public void flush()
           throws IOException

Description copied from interface: RecordUpdater

Flush the current set of rows to the underlying file system, so that they are available to readers. Most implementations will need to write additional state information when this is called, so it should only be called during streaming when a transaction is finished, but the RecordUpdater can't be closed yet.

Specified by:

flush in interface RecordUpdater

Throws:

IOException

close

public void close(boolean abort)
           throws IOException

Description copied from interface: RecordUpdater

Close this updater. No further calls are legal after this.

Specified by:

close in interface RecordUpdater

Parameters:

abort - Can the data since the last flush be discarded?

Throws:

IOException

getStats

public SerDeStats getStats()

Description copied from interface: RecordUpdater

Returns the statistics information

Specified by:

getStats in interface RecordUpdater

Returns:

SerDeStats

getBufferedRowCount

public long getBufferedRowCount()

Description copied from interface: RecordUpdater

Returns the number of rows in memory before flush().

Specified by:

getBufferedRowCount in interface RecordUpdater

Returns:

- buffered row count

getUpdatedFilePath

public org.apache.hadoop.fs.Path getUpdatedFilePath()

Description copied from interface: RecordUpdater

Returns the path of the file this updater wrote to

Specified by:

getUpdatedFilePath in interface RecordUpdater

getDeleteFilePath

public org.apache.hadoop.fs.Path getDeleteFilePath()