org.apache.flink.table.filesystem

Interface PartitionComputer<T>

Type Parameters:

T - The type of the consumed records.

All Superinterfaces:

Serializable

All Known Implementing Classes:

HiveRowDataPartitionComputer, HiveRowPartitionComputer, RowDataPartitionComputer, RowPartitionComputer

@Internal
public interface PartitionComputer<T>
extends Serializable

Compute partition path from record and project non-partition columns for output writer.

See RowPartitionComputer.

Method Summary

Modifier and Type	Method and Description
LinkedHashMap<String,String>	generatePartValues(T in) Compute partition values from record.
T	projectColumnsToWrite(T in) Project non-partition columns for output writer.

Method Detail

generatePartValues

LinkedHashMap<String,String> generatePartValues(T in)
                                         throws Exception

Compute partition values from record.

Parameters:

in - input record.

Returns:

partition values.

Throws:

Exception

projectColumnsToWrite

T projectColumnsToWrite(T in)
                 throws Exception

Project non-partition columns for output writer.

Parameters:

in - input record.

Returns:

projected record.

Throws:

Exception