org.kitesdk.data

Class PartitionStrategy.Builder

java.lang.Object

org.kitesdk.data.PartitionStrategy.Builder

Enclosing class:

PartitionStrategy

public static class PartitionStrategy.Builder
extends Object

A fluent builder to aid in the construction of PartitionStrategys.

Constructor Summary

Constructors

Constructor and Description
PartitionStrategy.Builder()

Method Summary

Methods

Modifier and Type	Method and Description
PartitionStrategy	build() Build a configured PartitionStrategy instance.
PartitionStrategy.Builder	dateFormat(String sourceName, String name, String format) Configure a partitioner that applies a custom date format to a timestamp field.
PartitionStrategy.Builder	day(String sourceName) Configure a partitioner for extracting the day from a timestamp field.
PartitionStrategy.Builder	day(String sourceName, String name) Configure a partitioner for extracting the day from a timestamp field.
PartitionStrategy.Builder	hash(String sourceName, int buckets) Configure a hash partitioner with the specified number of buckets.
PartitionStrategy.Builder	hash(String sourceName, String name, int buckets) Configure a hash partitioner with the specified number of buckets.
PartitionStrategy.Builder	hour(String sourceName) Configure a partitioner for extracting the hour from a timestamp field.
PartitionStrategy.Builder	hour(String sourceName, String name) Configure a partitioner for extracting the hour from a timestamp field.
PartitionStrategy.Builder	identity(String sourceName) Configure an identity partitioner.
PartitionStrategy.Builder	identity(String sourceName, int cardinalityHint) Configure an identity partitioner with a cardinality hint of cardinalityHint.
PartitionStrategy.Builder	identity(String sourceName, String name) Configure an identity partitioner.
PartitionStrategy.Builder	identity(String sourceName, String name, int cardinalityHint) Configure an identity partitioner with a cardinality hint of cardinalityHint.
PartitionStrategy.Builder	minute(String sourceName) Configure a partitioner for extracting the minute from a timestamp field.
PartitionStrategy.Builder	minute(String sourceName, String name) Configure a partitioner for extracting the minute from a timestamp field.
PartitionStrategy.Builder	month(String sourceName) Configure a partitioner for extracting the month from a timestamp field.
PartitionStrategy.Builder	month(String sourceName, String name) Configure a partitioner for extracting the month from a timestamp field.
PartitionStrategy.Builder	provided(String name) Configure a partitioner that uses values always provided at runtime.
PartitionStrategy.Builder	provided(String name, String valuesType) Configure a partitioner that uses values always provided at runtime.
PartitionStrategy.Builder	range(String sourceName, int... upperBounds) Configure a range partitioner with a set of upperBounds.
PartitionStrategy.Builder	range(String sourceName, String... upperBounds) Configure a range partitioner for strings with a set of upperBounds.
PartitionStrategy.Builder	year(String sourceName) Configure a partitioner for extracting the year from a timestamp field.
PartitionStrategy.Builder	year(String sourceName, String name) Configure a partitioner for extracting the year from a timestamp field.

Methods inherited from class java.lang.Object

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

Constructor Detail

PartitionStrategy.Builder

public PartitionStrategy.Builder()

Method Detail

hash

public PartitionStrategy.Builder hash(String sourceName,
                             int buckets)

Configure a hash partitioner with the specified number of buckets. The partition name is the source field name with a "_hash" suffix. For example, hash("color", 34) creates "color_hash" partitions.

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

buckets - The number of buckets into which data is to be partitioned.

Returns:

An instance of the builder for method chaining.

hash

public PartitionStrategy.Builder hash(String sourceName,
                             @Nullable
                             String name,
                             int buckets)

Configure a hash partitioner with the specified number of buckets. If name is null, the partition name will be the source field name with a "_hash" suffix. For example, hash("color", null, 34) will create "color_hash" partitions.

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

name - The entity field name of the partition.

buckets - The number of buckets into which data is to be partitioned.

Returns:

An instance of the builder for method chaining.

Since:

0.3.0

identity

public PartitionStrategy.Builder identity(String sourceName)

Configure an identity partitioner. The partition name is the source field name with a "_copy" suffix. For example, identity("color", String.class, 34) creates "color_copy" partitions.

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

Returns:

An instance of the builder for method chaining.

Since:

0.14.0

See Also:

IdentityFieldPartitioner

identity

public PartitionStrategy.Builder identity(String sourceName,
                                 String name)

Configure an identity partitioner. If name is null, the partition name will be the source field name with a "_copy" suffix. For example, identity("color", null, ...) will create "color_copy" partitions.

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

name - A name for the partition field

Returns:

An instance of the builder for method chaining.

Since:

0.14.0

See Also:

IdentityFieldPartitioner

identity

public PartitionStrategy.Builder identity(String sourceName,
                                 int cardinalityHint)

Configure an identity partitioner with a cardinality hint of cardinalityHint. The partition name is the source field name with a "_copy" suffix. For example, identity("color", String.class, 34) creates "color_copy" partitions.

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

cardinalityHint - A hint as to the number of partitions that will be created (i.e. the number of discrete values for the field name in the data).

Returns:

An instance of the builder for method chaining.

Since:

0.14.0

See Also:

IdentityFieldPartitioner

identity

public PartitionStrategy.Builder identity(String sourceName,
                                 String name,
                                 int cardinalityHint)

Configure an identity partitioner with a cardinality hint of cardinalityHint. If name is null, the partition name will be the source field name with a "_copy" suffix. For example, identity("color", null, ...) will create "color_copy" partitions.

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

name - A name for the partition field

cardinalityHint - A hint as to the number of partitions that will be created (i.e. the number of discrete values for the field name in the data).

Returns:

An instance of the builder for method chaining.

Since:

0.14.0

See Also:

IdentityFieldPartitioner

range

public PartitionStrategy.Builder range(String sourceName,
                              int... upperBounds)

Configure a range partitioner with a set of upperBounds. The partition name will be the source field name with a "_bound" suffix. For example, range("number", 5, 10) creates "number_bound" partitions.

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

upperBounds - A variadic list of upper bounds of each partition.

Returns:

An instance of the builder for method chaining.

See Also:

IntRangeFieldPartitioner

range

public PartitionStrategy.Builder range(String sourceName,
                              String... upperBounds)

Configure a range partitioner for strings with a set of upperBounds. The partition name will be the source field name with a "_bound" suffix. For example, range("color", "blue", "green") creates "color_bound" partitions.

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

upperBounds - A variadic list of upper bounds of each partition.

Returns:

An instance of the builder for method chaining.

year

public PartitionStrategy.Builder year(String sourceName,
                             @Nullable
                             String name)

Configure a partitioner for extracting the year from a timestamp field. The UTC timezone is assumed. If name is null, the partition entity name will be "year".

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

name - The entity field name of the partition.

Returns:

An instance of the builder for method chaining.

Since:

0.3.0

year

public PartitionStrategy.Builder year(String sourceName)

Configure a partitioner for extracting the year from a timestamp field. The UTC timezone is assumed. The partition entity name is "year".

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

Returns:

An instance of the builder for method chaining.

Since:

0.8.0

month

public PartitionStrategy.Builder month(String sourceName,
                              @Nullable
                              String name)

Configure a partitioner for extracting the month from a timestamp field. The UTC timezone is assumed. If name is null, the partition entity name will be "month".

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

name - The entity field name of the partition.

Returns:

An instance of the builder for method chaining.

Since:

0.3.0

month

public PartitionStrategy.Builder month(String sourceName)

Configure a partitioner for extracting the month from a timestamp field. The UTC timezone is assumed. The partition entity name is "month".

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

Returns:

An instance of the builder for method chaining.

Since:

0.8.0

day

public PartitionStrategy.Builder day(String sourceName,
                            @Nullable
                            String name)

Configure a partitioner for extracting the day from a timestamp field. The UTC timezone is assumed. If name is null, the partition entity name will be "day".

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

name - The entity field name of the partition.

Returns:

An instance of the builder for method chaining.

Since:

0.3.0

day

public PartitionStrategy.Builder day(String sourceName)

Configure a partitioner for extracting the day from a timestamp field. The UTC timezone is assumed. The partition entity name is "day".

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

Returns:

An instance of the builder for method chaining.

Since:

0.8.0

hour

public PartitionStrategy.Builder hour(String sourceName,
                             @Nullable
                             String name)

Configure a partitioner for extracting the hour from a timestamp field. The UTC timezone is assumed. If name is null, the partition entity name will be "hour".

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

name - The entity field name of the partition.

Returns:

An instance of the builder for method chaining.

Since:

0.3.0

hour

public PartitionStrategy.Builder hour(String sourceName)

Configure a partitioner for extracting the hour from a timestamp field. The UTC timezone is assumed. The partition entity name is "hour".

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

Returns:

An instance of the builder for method chaining.

Since:

0.8.0

minute

public PartitionStrategy.Builder minute(String sourceName,
                               @Nullable
                               String name)

Configure a partitioner for extracting the minute from a timestamp field. The UTC timezone is assumed. If name is null, the partition entity name will be "minute".

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

name - The entity field name of the partition.

Returns:

An instance of the builder for method chaining.

Since:

0.3.0

minute

public PartitionStrategy.Builder minute(String sourceName)

Configure a partitioner for extracting the minute from a timestamp field. The UTC timezone is assumed. The partition entity name is "minute".

Parameters:

sourceName - The entity field name from which to get values to be partitioned.

Returns:

An instance of the builder for method chaining.

Since:

0.8.0

dateFormat

public PartitionStrategy.Builder dateFormat(String sourceName,
                                   String name,
                                   String format)

Configure a partitioner that applies a custom date format to a timestamp field. The UTC timezone is assumed.

Parameters:

sourceName - The entity field name of the timestamp to format

name - A name for the partitions created by the format (e.g. "day")

format - A SimpleDateFormat format-string.

Returns:

This builder for method chaining.

Since:

0.9.0

provided

public PartitionStrategy.Builder provided(String name)

Configure a partitioner that uses values always provided at runtime.

The partitioner created by this method will expect String values.

Parameters:

name - A name for the partitions

Returns:

This builder for method chaining

Since:

0.17.0

provided

public PartitionStrategy.Builder provided(String name,
                                 @Nullable
                                 String valuesType)

Configure a partitioner that uses values always provided at runtime.

The partitioner created by this method will expect values based on the given valuesType: "string", "int", or "long".

Parameters:

name - A name for the partitions

valuesType - A type string for values this partitioner will expect; one of "string", "int", or "long". If null, the default is "string".

Returns:

This builder for method chaining

Since:

0.17.0

build

public PartitionStrategy build()

Build a configured PartitionStrategy instance. This builder should be considered single use and discarded after a call to this method.

Returns:

The configured instance of PartitionStrategy.

Since:

0.9.0