org.apache.flink.table.factories

Interface Factory

All Known Subinterfaces:

DecodingFormatFactory<I>, DeserializationFormatFactory, DynamicTableFactory, DynamicTableSinkFactory, DynamicTableSourceFactory, EncodingFormatFactory<I>, FileSystemFormatFactory, SerializationFormatFactory

All Known Implementing Classes:

AvroFileSystemFormatFactory, AvroFormatFactory, BlackHoleTableSinkFactory, CanalJsonFormatFactory, ChangelogCsvFormatFactory, CsvFileSystemFormatFactory, CsvFormatFactory, DataGenTableSourceFactory, DebeziumJsonFormatFactory, Elasticsearch6DynamicSinkFactory, Elasticsearch7DynamicSinkFactory, HBaseDynamicTableFactory, HiveDynamicTableFactory, JdbcDynamicTableFactory, JsonFileSystemFormatFactory, JsonFormatFactory, Kafka010DynamicTableFactory, Kafka011DynamicTableFactory, KafkaDynamicTableFactory, KafkaDynamicTableFactoryBase, OrcFileSystemFormatFactory, ParquetFileSystemFormatFactory, PrintTableSinkFactory, SocketDynamicTableFactory

@PublicEvolving
public interface Factory

Base interface for all kind of factories that create object instances from a list of key-value pairs in Flink's Table & SQL API.

A factory is uniquely identified by Class and factoryIdentifier().

The list of available factories is discovered using Java's Service Provider Interfaces (SPI). Classes that implement this interface can be added to META_INF/services/org.apache.flink.table.factories.Factory in JAR files.

Every factory declares a set of required and optional options. This information will not be used during discovery but is helpful when generating documentation and performing validation. A factory may discover further (nested) factories, the options of the nested factories must not be declared in the sets of this factory.

It is the responsibility of each factory to perform validation before returning an instance.

For consistency, the following style for key names of ConfigOption is recommended:

• Try to reuse key names as much as possible. Use other factory implementations as an example.
• Key names should be declared in lower case. Use "-" instead of dots or camel case to split words.
• Key names should be hierarchical where appropriate. Think about how one would define such a hierarchy in JSON or YAML file (e.g. sink.bulk-flush.max-actions).
• In case of a hierarchy, try not to use the higher level again in the key name (e.g. do sink.partitioner instead of sink.sink-partitioner) to keep the keys short.

Method Summary

Modifier and Type	Method and Description
String	factoryIdentifier() Returns a unique identifier among same factory interfaces.
Set<ConfigOption<?>>	optionalOptions() Returns a set of ConfigOption that an implementation of this factory consumes in addition to requiredOptions().
Set<ConfigOption<?>>	requiredOptions() Returns a set of ConfigOption that an implementation of this factory requires in addition to optionalOptions().

Method Detail

factoryIdentifier

String factoryIdentifier()

Returns a unique identifier among same factory interfaces.

For consistency, an identifier should be declared as one lower case word (e.g. kafka). If multiple factories exist for different versions, a version should be appended using "-" (e.g. kafka-0.10).

requiredOptions

Set<ConfigOption<?>> requiredOptions()

Returns a set of ConfigOption that an implementation of this factory requires in addition to optionalOptions().

See the documentation of Factory for more information.

optionalOptions

Set<ConfigOption<?>> optionalOptions()

Returns a set of ConfigOption that an implementation of this factory consumes in addition to requiredOptions().

See the documentation of Factory for more information.