Package org.apache.flink.client.python

Interface PythonFunctionFactory

All Known Implementing Classes:

PythonFunctionFactoryImpl

public interface PythonFunctionFactory

The factory which creates the PythonFunction objects from given module name and object name.

Nested Class Summary

Nested Classes

Modifier and Type	Interface	Description
static class	PythonFunctionFactory.CacheKey	The cache key.

Field Summary

Fields

Modifier and Type	Field	Description
static final ScheduledExecutorService	CACHE_CLEANUP_EXECUTOR_SERVICE
static final AtomicReference<Boolean>	CACHE_CLEANUP_EXECUTOR_SERVICE_STARTED
static final org.apache.flink.shaded.guava32.com.google.common.cache.LoadingCache<PythonFunctionFactory.CacheKey,PythonFunctionFactory>	PYTHON_FUNCTION_FACTORY_CACHE

Method Summary

Modifier and Type	Method	Description
static PythonFunctionFactory	createPythonFunctionFactory(org.apache.flink.configuration.ReadableConfig config)
static void	ensureCacheCleanupExecutorServiceStarted()
org.apache.flink.table.functions.python.PythonFunction	getPythonFunction(String moduleName, String objectName)	Returns PythonFunction according to moduleName and objectName.
static org.apache.flink.table.functions.python.PythonFunction	getPythonFunction(String fullyQualifiedName, org.apache.flink.configuration.ReadableConfig config, ClassLoader classLoader)	Returns PythonFunction according to the fully qualified name of the Python UDF i.e ${moduleName}.

Field Details

CACHE_CLEANUP_EXECUTOR_SERVICE

static final ScheduledExecutorService CACHE_CLEANUP_EXECUTOR_SERVICE

CACHE_CLEANUP_EXECUTOR_SERVICE_STARTED

static final AtomicReference<Boolean> CACHE_CLEANUP_EXECUTOR_SERVICE_STARTED

PYTHON_FUNCTION_FACTORY_CACHE

static final org.apache.flink.shaded.guava32.com.google.common.cache.LoadingCache<PythonFunctionFactory.CacheKey,PythonFunctionFactory> PYTHON_FUNCTION_FACTORY_CACHE

Method Details

getPythonFunction

org.apache.flink.table.functions.python.PythonFunction getPythonFunction(String moduleName,
 String objectName)

Returns PythonFunction according to moduleName and objectName.

Parameters:

moduleName - The module name of the Python UDF.

objectName - The function name / class name of the Python UDF.

Returns:

The PythonFunction object which represents the Python UDF.

getPythonFunction

static org.apache.flink.table.functions.python.PythonFunction getPythonFunction(String fullyQualifiedName,
 org.apache.flink.configuration.ReadableConfig config,
 ClassLoader classLoader)
                                                                         throws ExecutionException

Returns PythonFunction according to the fully qualified name of the Python UDF i.e ${moduleName}.${functionName} or ${moduleName}.${className}.

Parameters:

fullyQualifiedName - The fully qualified name of the Python UDF.

config - The configuration of python dependencies.

classLoader - The classloader which is used to identify different jobs.

Returns:

The PythonFunction object which represents the Python UDF.

Throws:

ExecutionException

ensureCacheCleanupExecutorServiceStarted

static void ensureCacheCleanupExecutorServiceStarted()

createPythonFunctionFactory

static PythonFunctionFactory createPythonFunctionFactory(org.apache.flink.configuration.ReadableConfig config)
                                                  throws ExecutionException,
InterruptedException,
IOException

Throws:

ExecutionException

InterruptedException

IOException