com.streambase.sb.operator

Class Operator

java.lang.Object

com.streambase.sb.operator.Operator

All Implemented Interfaces:

com.streambase.sb.util.LocatedItem, java.io.Serializable

Direct Known Subclasses:

InputAdapter, OutputAdapter

public abstract class Operator
extends java.lang.Object
implements java.io.Serializable, com.streambase.sb.util.LocatedItem

Abstract base class for User code that is used as a Java Operator or an embedded Adapter in a StreamBase application. One instance will be created for each Java Operator in a StreamBase application. StreamBase Studio may operate on several StreamBase applications at a time, so Operator subclass instances may be in different applications.

Operator subclasses must have a public default constructor.

An Operator is notified of state changes through callbacks. The StreamBase runtime 'calls back' an Operator when it changes the runtime state of the Operator. These callbacks include Operator.resume(), Operator.resumed(), Operator.suspend(), Operator.suspended() and . Operator.shutdown().

Operator provides "managed threads", which are threads that run concurrently with the application, but which can synchronize with its overall state changes. These threads are started, suspended, resumed, and shut down with the application. Managed threads are registered with an Operator using the method registerRunnable. This is particularly useful for input adapters, which typically have to respond to external events asynchronously with the application. Managed threads can call sendOutput at any time.

If an Operator registers one or more managed threads, and all of its managed threads exit their run() methods, then the Operator itself will shut down.

The StreamBase runtime blocks while it waits for an Operator's managed threads to respond to a state change. This can be problematic if a managed thread is blocked on some event. However, the StreamBase runtime can be configured to interrupt a thread when it needs to change its state. This is accomplished by setting the flag shouldInterrupt to true when registering the thread with registerRunnable.

It may be that an Operator's managed thread does not respond to a state change even after it has been interrupted. If the Operator's thread does not respond to the state within a given time interval then it is considered to be in failure and it is shut down. This time interval is specified by the server configuration parameter operator-state-change-timeout-ms.

For sharing state amongst instances, Operators may use the services provided by the Operator.SharedObjectManager accessible via Operator.getRuntimeEnvironment()

Since 7.2.12, if your operator simply passes through Tuples without creating new Tuple instances, ensure you override Operator.isPassthruOperator() for maximum performance.

Special Studio Considerations:

• Call Operator.setPortHints(int, int) in your constructor to set the initial number of ports when presented in an EventFlow editor
• Call Operator.setDisplayName(String) in your constructor to set a user-friendly name for your Operator. This name is used in the Operator Name field in the Properties view, and as the details text in the Palette view's Details mode.
• Call Operator.setShortDisplayName(String) in your constructor to set a short user-friendly name for your Operator. This name is used as the name of the operator in the Palette view and above the operator in the EventFlow canvas.
• Override Operator.getPortCounts() if you change port counts based on user properties changing
• Override Operator.getIconResource(com.streambase.sb.operator.Operator.IconKind) if you wish to be displayed using your own icons
• Since 6.5, during Operator.typecheck(), subclasses should consider throwing Operator.PropertyTypecheckException instead of TypecheckException in order to allow Studio to decorate UI widgets that caused the typecheck error. See Operator.PropertyTypecheckException and Operator.getLocation(String) for details.
• See also UIHints and Parameterizable for additional control over property display in Studio
• Since 7.3.10, override Operator.getSearchKeywords() to provide user assistance when users in Studio are filtering the list of available operators and adapters.

Note: Serializations of instances of this class that are created (e.g., by using ObjectOutputStream) in one version of StreamBase in general will not be deserializable in any other version of StreamBase.

See Also:

Parameterizable, Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static interface	Operator.ConfigurationAccessor Provides access to configuration information for use by operators and adapters, available from the server's configuration file.
static class	Operator.IconKind An enumeration for the different kinds of icons that StreamBase Studio may request when displaying Operators and Adapters.
static class	Operator.LogLevel Included values are, in decreasing order of chattiness: Operator.LogLevel.ALL, Operator.LogLevel.TRACE, Operator.LogLevel.DEBUG, Operator.LogLevel.INFO, Operator.LogLevel.WARN, Operator.LogLevel.ERROR, Operator.LogLevel.OFF.
static class	Operator.OperatorStates The set of runtime states that an Operator can be in.
class	Operator.OperatorThread The thread that wraps an operator runnable
class	Operator.PropertyTypecheckException A typecheck exception associated with an Operator property (or parameter) by name.
static interface	Operator.RuntimeEnvironment This interface is used to gain access to StreamBase Server information and facilities.
static interface	Operator.SharedObject A SharedObject is an object that can be shared between Operators within a Container.
static interface	Operator.SharedObjectManager The manager for SharedObjects within a container.
static class	Operator.SuspendBehaviorStates Suspend behavior defines how an Operator handles tuples when it is suspended; meaning when it is in the SUSPENDED state.

Field Summary

Fields

Modifier and Type	Field and Description
static int	DEFAULT_STATE_CHANGE_TIMEOUT Default value for the timeout for Operator state changes, in milliseconds: 10000
static Operator.ConfigurationAccessor	OP_CONFIG_MGR_NO_CONF A configuration manager that provides no information

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	Operator() Constructs an operator.

Method Summary

Modifier and Type	Method and Description
boolean	allowsConcurrency() Override to indicate whether concurrency may be configured in Studio against this operator.
java.lang.String	evaluate(java.lang.String st, Tuple input) evaluate the given string expression in the context of the running module If an error occurs during the evaluation, the error message return value will be a string containing the error message
int[]	getAsyncInputPorts() Override to indicate to Studio which ports (0-based) might process data asynchronously, this provides different rendering on the EventFlow canvas to assist users in understanding their data flow
CaptureTransformStrategy	getCaptureStrategy() Returns the capture transform strategy that this Operator will use.
java.lang.String	getContainerName() Return the name of this operator's container.
java.lang.String	getDisplayDescription() Provides a description for this type of Operator.
java.lang.String	getDisplayName() Return the display name of this Operator.
Tuple	getDynamicVariables() Retrieves a read-only tuple describing the current value of all dynamic variables in the module this operator is contained in.
Schema	getDynamicVariablesSchema() Retrieves a Schema containing information about every dynamic variable available to the module this operator is contained in.
java.lang.String	getFullyQualifiedName() Return the fully qualified name of this operator.
java.net.URL	getIconResource(Operator.IconKind iconType) Clients should override to support custom icons.
int	getInputPortCount() Return the number of input ports.
Schema	getInputSchema(int port) Returns the schema of an input port.
com.streambase.sb.util.Location	getLocation() Return the Operator-wide location, useful for error reporting not associated to a particular property.
com.streambase.sb.util.Location	getLocation(java.lang.String property) Return a new location within this Operator, associated with the given property name.
protected org.slf4j.Logger	getLogger() Retrieves a Logger suitable for logging messages and exceptions for this Operator.
java.lang.String	getName() Return the name of this Operator.
Schema	getNamedSchema(java.lang.String name) Retrieves a named schema
Operator.ConfigurationAccessor	getOperatorConfigurationAccessor() Provides access to operator and adapter configuration information from the server's configuration file.
int	getOutputPortCount() Returns the number of output ports.
Schema	getOutputSchema(int outputPort) Return the output schema for the given output port (zero-based).
Parameterizable	getParameters() Returns the Java Bean that holds the user-configurable parameters visible in the StreamBase Studio Properties View of this operator.
PortCounts	getPortCounts() An optional method that subclasses can override to dynamically tell Studio the number of input and output ports.
Schema[]	getProposedInputSchemas(java.lang.String mainName) Override to provide schemas that Studio will offer to users to import into their projects.
Schema[]	getProposedOutputSchemas(java.lang.String mainName) Override to provide schemas that Studio will offer to users to import into their projects.
java.io.InputStream	getResourceContents(java.lang.String name) Returns an open input stream on the contents of the named resource file.
java.io.File	getResourceFile(java.lang.String name) Returns a File pointing to the regular file or directory if it exists.
boolean	getReuseTuple() Get the state of the input Tuple reuse flag.
Operator.RuntimeEnvironment	getRuntimeEnvironment() Return the Operator.RuntimeEnvironment for this StreamBase Server.
Schema	getRuntimeInputSchema(int port) Returns the schema of an input port at runtime.
Schema	getRuntimeOutputSchema(int outputPort) Return the output schema that should actually use to Operator.sendOutput(int, Tuple) or Operator.sendOutputAsync(int, Tuple).
Schema	getSchemaForCapture(java.lang.String captureName, int depth) Finds the schema for the given capture name in the context that this operator is running under.
java.lang.String[]	getSearchKeywords() Override to declare strings that assist the user in filtering for this operator when in Studio.
java.lang.String	getShortDisplayName() Return the short display name of this Operator.
int	getStateChangeTimeout()
TableAccessor	getTableAccessor(java.lang.String name) Get a TableAccessor for a table in the local module by name of that table.
TupleCaptureTransformer	getTupleCaptureTransformer(Schema s) Get a TupleCaptureTransformer capable of translating tuples with the given schema to the equivalent schema with all the capture fields expanded out, and translating expanded tuples back into tuples with the given schema This method may only be called at runtime; the exact schemas of any capture fields are not fully determined at typecheck time.
Schema	getTypecheckInputSchema(int port) Returns the schema of an input port that was set at application typecheck time.
Schema	getTypecheckOutputSchema(int outputPort) Return the output schema as set by typecheck.
boolean	hasNotYetStarted() Returns true if this Operator has not yet started running.
void	init() After the Operator has typechecked successfully, but before the application or any managed threads start, the StreamBase server will call the init method.
boolean	isDesignTime() Deprecated. please use isRuntime()
boolean	isDroppingTuples() Returns true if the Operator will drop any tuples it receives when it is suspended.
boolean	isPassthruOperator() Allows an operator to declare itself to be a "pass-thru" operator -- one that calls its sendOutput method with the unmodified tuples received by its processTuple method.
boolean	isProcessingTuples() Returns true if the Operator will process any tuples it receives when it is suspended.
boolean	isRunning() Returns true if this Operator is currently running, false otherwise.
boolean	isRuntime()
boolean	isShutdown() Returns true if this Operator is currently shut down, false otherwise.
boolean	isSuspended() Returns true if this Operator is currently suspended, false otherwise.
void	postShutdown() postShutdown is called by the StreamBase runtime just after shutting down this Operator.
abstract void	processTuple(int inputPort, Tuple tuple) This method will be called by the StreamBase server for each Tuple given to the Operator to process.
void	registerRunnable(java.lang.Runnable operatorRunnable) Deprecated. As of StreamBase version 3.7, replaced by Operator.registerRunnable(Runnable, boolean)
void	registerRunnable(java.lang.Runnable operatorRunnable, boolean shouldInterrupt) Register a Runnable object to be managed by this Operator.
void	registerRunnable(java.lang.String runnableName, java.lang.Runnable operatorRunnable, boolean shouldInterrupt) Register a Runnable object to be managed by this Operator.
void	registerRunnable(java.lang.String runnableName, java.lang.Runnable operatorRunnable, boolean shouldInterrupt, boolean synchronizedShutdown) Register a Runnable object to be managed by this Operator.
protected void	requireInputPortCount(int numPorts) Throws a PortMismatchException if the number of ports is not numPorts.
void	resume() resume() is called when an operator starts or resumes execution, before any registered runnables are started or resumed.
void	resumed() resumed() is called after all registered runnables of the operator have started or resumed.
Cancellable	scheduleRunnable(java.util.Date when, java.lang.Runnable runner) Schedules a Runnable object to be run at some future point as part of the same parallel region as this Operator.
void	sendErrorOutput(java.lang.String message) Send an error tuple from this operator's error port.
void	sendErrorOutput(java.lang.Throwable t) Send an exception via the error output port
void	sendErrorOutput(java.lang.Throwable t, int port, Tuple errorTuple) Send an exception via the error output port
void	sendOutput(int port, java.util.List<Tuple> tuples) Enqueue a List of Tuples to be sent synchronously to downstream operators.
void	sendOutput(int port, Tuple tuple) Enqueue a Tuple to be sent synchronously to downstream operators.
void	sendOutputAsync(int port, java.util.List<Tuple> tuples) Enqueue a List of Tuples to be sent asynchronously to downstream operators.
void	sendOutputAsync(int port, Tuple tuple) Enqueue a Tuple to be sent asynchronously to downstream operators.
void	setCaptureStrategy(CaptureTransformStrategy cts) Set the capture transform strategy to use for the inputs and outputs of this Operator.
protected void	setDisplayDescription(java.lang.String description) Set a description for this type of Operator.
void	setDisplayName(java.lang.String dn) The display name is the String that's shown in the Operator Name field of the Properties view, and in the detailed text for this operator in the Palette view's Details mode.
void	setDynamicVariable(java.lang.String name, java.lang.Object value) Sets the value of this module's dynamic variable.
static void	setLogLevel(org.slf4j.Logger logger, Operator.LogLevel level) Sets the log level of the given logger.
static void	setLogLevel(org.slf4j.Logger logger, java.lang.String level) Sets the log level of the given logger.
Schema	setOutputSchema(int port, Schema outputSchema) Sets the output schema for the given output port (port #'s are zero based).
void	setParameters(Parameterizable params) Sets the Java Bean that holds the user-configurable parameters visible in the StreamBase Studio Properties View of this operator.
void	setPortHints(int numInputPorts, int numOutputPorts) Used to tell StreamBase Studio how many ports to draw on the Java Operator when the operator is drawn in the IDE's canvas.
protected void	setReuseTuple(boolean reuse) Allow/disallow the runtime to reuse tuples on operator input.
void	setShortDisplayName(java.lang.String sn) The short display name is the String displayed by Studio as the name of this operator in the Palette view, and above the operator in the EventFlow canvas.
void	setSuspendBehavior(int suspendBehavior) Set the suspend behavior of this Operator.
boolean	shouldRun() Return whether or not calling operator thread is enabled and should continue running.
void	shutdown() shutdown is called by the StreamBase runtime just prior to shutting down this Operator.
int	size() Return the "size" of this operator.
void	suspend() suspend() will be called when an operator suspends, before any registered runnables are suspended.
void	suspended() suspended() will be called after all registered runnables of the operator have suspended.
abstract void	typecheck() The typecheck method is called by Studio and the StreamBase server to ensure that all the parameters for this operator are correct.

Methods inherited from class java.lang.Object

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

Field Detail

OP_CONFIG_MGR_NO_CONF

public static final Operator.ConfigurationAccessor OP_CONFIG_MGR_NO_CONF

A configuration manager that provides no information

DEFAULT_STATE_CHANGE_TIMEOUT

public static final int DEFAULT_STATE_CHANGE_TIMEOUT

Default value for the timeout for Operator state changes, in milliseconds: 10000

An Operator can receive state change events from the StreamBase runtime. In particular, when the runtime starts, suspends, resumes, or shuts down, it, in turn, applies the same state change to any Operators it holds.

When the server applies a state change to an Operator, it blocks until the Operator acknowledges that it has transitioned to the new state.

If an Operator fails to make such an acknowledgement, the StreamBase runtime might wait indefinitely. StateChangeTimeout limits the amount of time the server will wait for an Operator to respond to a requested state change.

If the Operator fails to respond within the timeout, the server will shut down the Operator and continue with its state change.

See Also:

Constant Field Values

Constructor Detail

Operator

protected Operator()

Constructs an operator.

All Operator subclasses must have a public default constructor, otherwise StreamBase Studio and the StreamBase sbd process will not be able to instantiate the Operator subclass. Nothing substantial should be done in the constructor, except for setting port hints and setting the parameters object.

See Also:

Operator.setPortHints(int, int), Operator.setParameters(com.streambase.sb.operator.Parameterizable)

Method Detail

getIconResource

public java.net.URL getIconResource(Operator.IconKind iconType)

Clients should override to support custom icons. The default implementation returns null (meaning no icon) for any kind requested.

Parameters:

iconType - the kind of icon being requested, as an Operator.IconKind enumeration value

Returns:

a URL pointing to an image resource corresponding to the type requested, or null

See Also:

Operator.IconKind

getDisplayName

public final java.lang.String getDisplayName()

Return the display name of this Operator. The display name is the String that's shown in the list of available Operators.

Returns:

The display name of the Operator.

getDisplayDescription

public final java.lang.String getDisplayDescription()

Provides a description for this type of Operator. This is used by the user interface to assist the user in selecting an Operator during authoring. The expected usage is to return a three or four line description, without embedded new lines, describing the intent, or technical details that can assist a user in ensuring they are selecting the Operator or Adapter they need.

Returns:

The description for this type of Operator, never null

Since:

7.6

setDisplayDescription

protected final void setDisplayDescription(java.lang.String description)

Set a description for this type of Operator. This method is to be called only during the zero-argument constructor for this Operator.

Parameters:

description - description for this type of Operator, never null

Since:

7.6

See Also:

Operator.getDisplayDescription()

getShortDisplayName

public final java.lang.String getShortDisplayName()

Return the short display name of this Operator. This short display name is the String displayed by Studio in the palette, and used to name new instances of this Operator to an EventFlow application.

By default, this returns the simple class name of this Operator.

Returns:

The short display name of the Operator.

Since:

6.1.2, 6.5 uses this for new instances added to an EventFlow application

getName

public java.lang.String getName()

Return the name of this Operator. Operators are named and can be managed by name. This may return null during typecheck and init phases.

Returns:

The name of the operator during runtime, or empty string during typecheck and init

getContainerName

public java.lang.String getContainerName()

Return the name of this operator's container. This may return null during typecheck and init phases.

Returns:

The name of this operator's container. This may return an empty string during typecheck and init phases.

getFullyQualifiedName

public java.lang.String getFullyQualifiedName()

Return the fully qualified name of this operator. This contains the container name, any module names and the name of the operator. This may return null during typecheck and init phases.

Returns:

The fully qualified name of this operator. This may return an empty string during typecheck and init phases.

setDisplayName

public final void setDisplayName(java.lang.String dn)

The display name is the String that's shown in the Operator Name field of the Properties view, and in the detailed text for this operator in the Palette view's Details mode. If the subclass doesn't override it, use the class name.

setShortDisplayName

public final void setShortDisplayName(java.lang.String sn)

The short display name is the String displayed by Studio as the name of this operator in the Palette view, and above the operator in the EventFlow canvas. If the subclass doesn't override it, use the class simple name.

Since:

6.1.2

setPortHints

public void setPortHints(int numInputPorts,
                         int numOutputPorts)

Used to tell StreamBase Studio how many ports to draw on the Java Operator when the operator is drawn in the IDE's canvas. The default is 1 input port, 0 output ports.

This method is only effective when called from the Java Operator's constructor. The typecheck method should call requireInputPortCount instead.

To specify the ports at runtime requireInputPortCount and setOutputSchema should be called from the typecheck method.

To specify the ports dynamically within studio, see getPortCounts.

Parameters:

numInputPorts - number of input ports. Must be non-negative.

numOutputPorts - number of output ports. Must be non-negative.

Throws:

java.lang.IllegalArgumentException - if an argument is less than zero

See Also:

Operator.requireInputPortCount(int), Operator.setOutputSchema(int, Schema)

typecheck

public abstract void typecheck()
                        throws Operator.PropertyTypecheckException,
                               TypecheckException

The typecheck method is called by Studio and the StreamBase server to ensure that all the parameters for this operator are correct. This method is also responsible for verifying input Schemas and setting output Schemas. This method must call Operator.setOutputSchema(int, Schema) method if there are any output ports.

If the parameters are not correct, or the input port Schemas are not correct, a Operator.PropertyTypecheckException or TypecheckException should be thrown. The former should always be used if the exception is related to a particular parameter.

The method requireInputPortCount() should be used to verify that the required input ports are set.

If the Operator changes the number of input ports, this method must call requireInputPortCount.

Throws:

Operator.PropertyTypecheckException - when a parameter value is unexpected.

TypecheckException - when parameters or input Schemas are not satisfied.

See Also:

Operator.requireInputPortCount(int), Operator.setOutputSchema(int, Schema), Operator.PropertyTypecheckException

init

public void init()
          throws StreamBaseException

After the Operator has typechecked successfully, but before the application or any managed threads start, the StreamBase server will call the init method. Operators should override this method to perform custom initialization.

If this operator wishes to register threads, it should call Operator.registerRunnable(Runnable) from this method.

Throws:

StreamBaseException - Prevents the application from starting.

hasNotYetStarted

public boolean hasNotYetStarted()

Returns true if this Operator has not yet started running. hasNotYetStarted() returns true if the Operator is in the runtime state of NOT_YET_STARTED, which is the initial runtime state of an Operator.

See Also:

Operator.OperatorStates.NOT_YET_STARTED

isDroppingTuples

public boolean isDroppingTuples()

Returns true if the Operator will drop any tuples it receives when it is suspended.

See Also:

Operator.SuspendBehaviorStates.DROPPING_TUPLES

isProcessingTuples

public boolean isProcessingTuples()

Returns true if the Operator will process any tuples it receives when it is suspended.

See Also:

Operator.SuspendBehaviorStates.PROCESSING_TUPLES

isRunning

public boolean isRunning()

Returns true if this Operator is currently running, false otherwise. isRunning() returns true if the Operator is in a runtime state of STARTED.

See Also:

Operator.OperatorStates.STARTED

isShutdown

public boolean isShutdown()

Returns true if this Operator is currently shut down, false otherwise. An Operator is shut down when it is in the runtime state of SHUTDOWN, which is the terminal state for Operators.

See Also:

Operator.OperatorStates.SHUTDOWN

isSuspended

public boolean isSuspended()

Returns true if this Operator is currently suspended, false otherwise. An Operator is suspended when it is in the runtime state of SUSPENDED.

See Also:

Operator.OperatorStates.SUSPENDED

processTuple

public abstract void processTuple(int inputPort,
                                  Tuple tuple)
                           throws StreamBaseException

This method will be called by the StreamBase server for each Tuple given to the Operator to process.

By default the Tuple passed in is safe for storing and mutating after the return of processTuple. The details of this Tuple's life cycle can be controlled via Operator.setReuseTuple(boolean).

Parameters:

inputPort - the input port that the tuple is from (ports are zero based)

tuple - the tuple from the given input port

Throws:

StreamBaseException - Terminates the application.

resume

public void resume()
            throws java.lang.Exception

resume() is called when an operator starts or resumes execution, before any registered runnables are started or resumed. Note that if the application is shutdown directly from a suspended state, this will not be called; instead Operator.shutdown() will be called. resume() is a callback that is called by the StreamBase runtime. Exceptions thrown during calls to Operator.resume() will be logged via Operator.logError(Exception), processed as error tuples, and trigger a shutdown of the operator.

Throws:

java.lang.Exception

resumed

public void resumed()
             throws java.lang.Exception

resumed() is called after all registered runnables of the operator have started or resumed. That is, once Operator.shouldRun() has unblocked and returned true in all registered runnables. resumed() is a callback that is called by the StreamBase runtime. Exceptions thrown during calls to Operator.resumed() will be logged via Operator.logError(Exception), processed as error tuples, and trigger a shutdown of the operator.

Throws:

java.lang.Exception

size

public int size()

Return the "size" of this operator. This size value will be displayed in sbmonitor's "size" column.

Override this method to display the size of an important data structure in this Operator. For example the size of a queue. Returns 0 by default.

This method must return promptly and never block, as this is called from a system statistics monitoring thread.

Returns:

size of this operator.

shouldRun

public final boolean shouldRun()

Return whether or not calling operator thread is enabled and should continue running.

This method returns false if the operator has not yet started or has shut down. It blocks if the operator is suspended. It returns true if the operator is running.

This method links managed operator threads with the StreamBase runtime. Every registered runnable object must repeatedly call this during its entire lifetime in order to synchronize with the main StreamBase application.

An operator (and, thus, its containing application) is not considered started until all registered runnables have called this method. Likewise, an operator is not considered suspended until all registered runnables are blocked in this method. Finally, an operator is not considered shutdown until all registered runnables have exited their run() method (either on their own or in response to this method returning false).

Returns:

true is the operator is running, false otherwise.

Throws:

java.lang.UnsupportedOperationException - If this was not called from an operator thread.

shutdown

public void shutdown()

shutdown is called by the StreamBase runtime just prior to shutting down this Operator. An implementation of shutdown should include any behavior needed to shut down the Operator - freeing resources, etc. shutdown() is a callback that is called by the StreamBase runtime.

postShutdown

public void postShutdown()

postShutdown is called by the StreamBase runtime just after shutting down this Operator. postShutdown() is a callback that is called by the StreamBase runtime.

suspend

public void suspend()
             throws java.lang.Exception

suspend() will be called when an operator suspends, before any registered runnables are suspended. suspend() is a callback that is called by the StreamBase runtime. Exceptions thrown during calls to Operator.suspend() will be logged via Operator.logError(Exception) and processed as error tuples.

Throws:

java.lang.Exception

setSuspendBehavior

public void setSuspendBehavior(int suspendBehavior)

Set the suspend behavior of this Operator. The Operator can either process or drop tuples when suspended. setSuspendBehavior() might be called in the Operator's constructor or in init().

Parameters:

suspendBehavior - The suspend behavior to set, either PROCESSING_TUPLES or DROPPING_TUPLES.

See Also:

Operator.SuspendBehaviorStates.PROCESSING_TUPLES, Operator.SuspendBehaviorStates.DROPPING_TUPLES

suspended

public void suspended()
               throws java.lang.Exception

suspended() will be called after all registered runnables of the operator have suspended. That is, once all registered runnables have called and are blocked in Operator.shouldRun(). suspended() is a callback that is called by the StreamBase runtime. Exceptions thrown during calls to Operator.suspended() will be logged via Operator.logError(Exception) and processed as error tuples.

Throws:

java.lang.Exception

getParameters

public Parameterizable getParameters()

Returns the Java Bean that holds the user-configurable parameters visible in the StreamBase Studio Properties View of this operator.

Returns:

The Java Bean that holds the user-configurable parameters visible in the StreamBase Studio Properties View of this operator.

setParameters

public void setParameters(Parameterizable params)

Sets the Java Bean that holds the user-configurable parameters visible in the StreamBase Studio Properties View of this operator.

Parameters:

params - the Java Bean that holds the user-configurable parameters visible in the StreamBase Studio Properties View of this operator.

getResourceContents

public java.io.InputStream getResourceContents(java.lang.String name)
                                        throws ResourceNotFoundException,
                                               StreamBaseException

Returns an open input stream on the contents of the named resource file. The client is responsible for closing the stream when finished.

This, or Operator.getResourceFile(String), should be called during typecheck (as opposed to waiting for init or similar run-time methods) in order to be able to surface to the authoring environment any failures locating or reading the resource. In addition, failure to call this or the alternate method during Operator.typecheck() will prevent Studio from uploading needed resources to remote servers.

Parameters:

name - simple name of requested resource

Returns:

an input stream containing the contents of the resource

Throws:

ResourceNotFoundException - if the resource could not be found

StreamBaseException - if the resource was found but could not be opened

getResourceFile

public java.io.File getResourceFile(java.lang.String name)
                             throws ResourceNotFoundException,
                                    StreamBaseException

Returns a File pointing to the regular file or directory if it exists. If this returns a regular file, Operator.getResourceContents(String) is guaranteed not to fail.

This, or Operator.getResourceContents(String), should be called during typecheck (as opposed to waiting for init or similar run-time methods) in order to be able to surface to the authoring environment any failures locating or reading the resource. In addition, failure to call this or the alternate method during Operator.typecheck() will prevent Studio from uploading needed resources to remote servers.

Parameters:

name - simple name of requested resource

Returns:

an input stream containing the contents of the resource

Throws:

ResourceNotFoundException - if the resource could not be found

StreamBaseException

Since:

6.4

registerRunnable

public final void registerRunnable(java.lang.String runnableName,
                                   java.lang.Runnable operatorRunnable,
                                   boolean shouldInterrupt,
                                   boolean synchronizedShutdown)
                            throws StreamBaseException

Register a Runnable object to be managed by this Operator. This Runnable will be started in a Thread when the Operator starts, suspended and resumed with the Operator, and stopped with the Operator.

The body of this Runnable must hook into StreamBase's thread management by repeatedly calling the Operator.shouldRun() method during its entire lifetime. This allows managed threads to be started, stopped, suspended, and resumed along with the rest of the application.

This method should be called from the operator's Operator.init() method.

Parameters:

runnableName - The name of this runnable (may be null). Names are reported on the statistics stream, and displayed by sbmonitor and sbmanager.

operatorRunnable - The Runnable to register.

shouldInterrupt - Whether this thread should be interrupted when the operator undergoes a state change. Typically true if the thread blocks on some event, input for example.

synchronizedShutdown - Whether to wait for this thead to exit before postShutdown is called.

Throws:

StreamBaseException - If it is too late in the Operator's life cycle to register a Runnable.

Since:

6.6.14

registerRunnable

public final void registerRunnable(java.lang.String runnableName,
                                   java.lang.Runnable operatorRunnable,
                                   boolean shouldInterrupt)
                            throws StreamBaseException

Register a Runnable object to be managed by this Operator. This Runnable will be started in a Thread when the Operator starts, suspended and resumed with the Operator, and stopped with the Operator.

The body of this Runnable must hook into StreamBase's thread management by repeatedly calling the Operator.shouldRun() method during its entire lifetime. This allows managed threads to be started, stopped, suspended, and resumed along with the rest of the application.

This method should be called from the operator's Operator.init() method.

Parameters:

runnableName - The name of this runnable (may be null). Names are reported on the statistics stream, and displayed by sbmonitor and sbmanager.

operatorRunnable - The Runnable to register.

shouldInterrupt - Whether this thread should be interrupted when the operator undergoes a state change. Typically true if the thread blocks on some event, input for example.

Throws:

StreamBaseException - If it is too late in the Operator's life cycle to register a Runnable.

Since:

6.5

registerRunnable

public final void registerRunnable(java.lang.Runnable operatorRunnable,
                                   boolean shouldInterrupt)
                            throws StreamBaseException

Register a Runnable object to be managed by this Operator. This Runnable will be started in a Thread when the Operator starts, suspended and resumed with the Operator, and stopped with the Operator.

The body of this Runnable must hook into StreamBase's thread management by repeatedly calling the Operator.shouldRun() method during its entire lifetime. This allows managed threads to be started, stopped, suspended, and resumed along with the rest of the application.

This method should be called from the operator's Operator.init() method.

Parameters:

operatorRunnable - The Runnable to register.

shouldInterrupt - Whether this thread should be interrupted when the operator undergoes a state change. Typically true if the thread blocks on some event, input for example. -

Throws:

StreamBaseException - If it is too late in the Operator's life cycle to register a Runnable.

registerRunnable

public final void registerRunnable(java.lang.Runnable operatorRunnable)
                            throws StreamBaseException

Deprecated. As of StreamBase version 3.7, replaced by Operator.registerRunnable(Runnable, boolean)

Register a Runnable object to be managed by this Operator. This Runnable will be started in a Thread when the Operator starts, suspended and resumed with the Operator, and stopped with the Operator.

The body of this Runnable must hook into StreamBase's thread management by repeatedly calling the Operator.shouldRun() method during its entire lifetime. This allows managed threads to be started, shut down, suspended, and resumed along with the rest of the application.

This method should be called from the adapter's Operator.init() method.

Parameters:

operatorRunnable - The Runnable to register.

Throws:

StreamBaseException - If it is too late in the Operator's life cycle to register a Runnable.

scheduleRunnable

public Cancellable scheduleRunnable(java.util.Date when,
                                    java.lang.Runnable runner)

Schedules a Runnable object to be run at some future point as part of the same parallel region as this Operator. Prefer using this to implement scheduling events by the operator rather than allocating a new thread using registerRunnable, if the parallelism is not necessary.

Parameters:

when - Time to run the runner. If it is in the past, it is run immediately

runner - The Runnable to schedule

Returns:

Handle to task for cancellation, if the task hasn't already started running. This handle is thread-safe, but attempting to cancel an already started task has no effect.

Since:

7.4

requireInputPortCount

protected void requireInputPortCount(int numPorts)

Throws a PortMismatchException if the number of ports is not numPorts. The StreamBase Studio IDE will recognize this and draw the operator box with the appropriate number of input ports.

This method should be called by the typecheck method.

Parameters:

numPorts - the number of ports

Throws:

PortMismatchException - When the number of input ports is incorrect.

getPortCounts

public PortCounts getPortCounts()
                         throws TypecheckException

An optional method that subclasses can override to dynamically tell Studio the number of input and output ports. Clients should expect this method to be called after all setters, but prior to typecheck. This method will not be called if Studio encountered any exception while calling each setter with the property values just entered by the user.

Note: This is for studio purposes only and should not be used to set any fields that you may want to use in other parts of the Operator.

Returns:

PortCounts record type

Throws:

TypecheckException

getAsyncInputPorts

public int[] getAsyncInputPorts()

Override to indicate to Studio which ports (0-based) might process data asynchronously, this provides different rendering on the EventFlow canvas to assist users in understanding their data flow

Returns:

the zero based indices of the asynchronous ports; null by default

Since:

7.6

getInputPortCount

public final int getInputPortCount()

Return the number of input ports.

Returns:

number of Input ports, if not in an application 0.

getOutputPortCount

public final int getOutputPortCount()

Returns the number of output ports. The number of output ports are set by the Operator.setOutputSchema(int, Schema) method.

See Also:

Operator.setOutputSchema(int, Schema)

sendOutput

public void sendOutput(int port,
                       Tuple tuple)
                throws StreamBaseException

Enqueue a Tuple to be sent synchronously to downstream operators.

All calls to Operator.sendOutput(int, Tuple) are well ordered with respect to other calls to Operator.sendOutput(int, Tuple). Meaning that downstream logic will receive the tuples in the order they are sent.

Calls to Operator.sendOutputAsync(int, Tuple) and Operator.sendOutput(int, List) are not well ordered with respect to each other. Meaning calls from the two methods can be received downstream in any order.

A note about reusing/caching the sent tuple.

If the given tuple is a tuple that this operator received from a processTuple() call, then the tuple can be reused if and only if getReuseTuple() is FALSE.

If the given tuple is created by this operator then it can be reused when the call to sendOutput() returns. StreamBase is done with the tuple once control returns to this operator.

Parameters:

port - The output port the Tuple is enqueued upon (ports are zero based)

tuple - The Tuple to enqueue

Throws:

StreamBaseException - if the port is invalid or the tuple argument doesn't match the schema of the output port.

sendOutput

public void sendOutput(int port,
                       java.util.List<Tuple> tuples)
                throws StreamBaseException

Enqueue a List of Tuples to be sent synchronously to downstream operators.

All calls to Operator.sendOutput(int, Tuple) are well ordered with respect to other calls to Operator.sendOutput(int, Tuple). Meaning that downstream logic will receive the tuples in the order they are sent.

Calls to Operator.sendOutputAsync(int, Tuple) and Operator.sendOutput(int, List) are not well ordered with respect to each other. Meaning calls from the two methods can be received downstream in any order.

A note about reusing/caching the sent tuples.

If the tuples in the given list are tuples that this operator received from processTuple() calls, then these tuples can be reused if and only if getReuseTuple() is FALSE.

If the given list of tuples is created by this operator, then the tuples in this list can be reused when the call to sendOutput() returns. StreamBase is done with the tuples once control returns to this operator.

Parameters:

port - The output port the Tuple is enqueued upon (ports are zero based)

tuples - The List of Tuple objects to enqueue

Throws:

StreamBaseException - if the port is invalid or any tuple in the tuples argument doesn't match the schema of the output port.

sendOutputAsync

public void sendOutputAsync(int port,
                            Tuple tuple)
                     throws StreamBaseException

Enqueue a Tuple to be sent asynchronously to downstream operators.

All calls to Operator.sendOutputAsync(int, Tuple) are well ordered with respect to other calls to Operator.sendOutputAsync(int, Tuple). Meaning that downstream logic will receive the tuples in the order they are sent.

Calls to Operator.sendOutputAsync(int, Tuple) and Operator.sendOutput(int, List) are not well ordered with respect to each other. Meaning calls from the two methods can be received downstream in any order.

Note: The Tuple will be queued into the application some time in the future. The asynchronous nature of this call could cause subtle race conditions in applications. The performance differential (if any) between sendOutputAsync and sendOutput is highly dependent upon the application and hardware. Benchmarking is the only way to ensure the fastest possible implementation.

The tuple can be reused once sendOutputAsync returns.

Parameters:

port - The output port the Tuple is enqueued upon (ports are zero based)

tuple - A tuple to enqueue asynchronously

Throws:

StreamBaseException - if the port is invalid or the tuple argument doesn't match the schema of the output port.

Since:

6.4.3

sendOutputAsync

public void sendOutputAsync(int port,
                            java.util.List<Tuple> tuples)
                     throws StreamBaseException

Enqueue a List of Tuples to be sent asynchronously to downstream operators.

All calls to Operator.sendOutputAsync(int, Tuple) are well ordered with respect to other calls to Operator.sendOutputAsync(int, Tuple). Meaning that downstream logic will receive the tuples in the order they are sent.

Calls to Operator.sendOutputAsync(int, Tuple) and Operator.sendOutput(int, List) are not well ordered with respect to each other. Meaning calls from the two methods can be received downstream in any order.

Note: The Tuples will be queued into the application some time in the future. The asynchronous nature of this call could cause subtle race conditions in applications. The performance differential (if any) between sendOutputAsync and sendOutput is highly dependent upon the application and hardware. Benchmarking is the only way to ensure the fastest possible implementation.

The tuples (and the List<Tuple>) can be reused once sendOutputAsync returns.

Parameters:

port - The output port the Tuple is enqueued upon (ports are zero based)

tuples - A List<Tuple> to enqueue asynchronously

Throws:

StreamBaseException - if the port is invalid or any tuple in the tuples argument doesn't match the schema of the output port.

Since:

6.4.3

getInputSchema

public final Schema getInputSchema(int port)

Returns the schema of an input port. At typecheck time or before, behaves like Operator.getTypecheckInputSchema(int). After typecheck time, behaves like Operator.getRuntimeInputSchema(int).

Parameters:

port - the port to return the schema for (ports are zero based)

Returns:

Schema for the port, null if not running in an appication

Throws:

java.lang.IndexOutOfBoundsException - if port not in range

getRuntimeInputSchema

public final Schema getRuntimeInputSchema(int port)

Returns the schema of an input port at runtime. This schema will never contain capture fields, as they will have been transformed according to the CaptureTransformStrategy set by Operator.setCaptureStrategy(CaptureTransformStrategy). This will be the actual schema of tuples passed to Operator.processTuple(int, Tuple)

Parameters:

port - the port to return the schema for (ports are zero based)

Returns:

Schema for the port, null if not running in an appication

Throws:

java.lang.IndexOutOfBoundsException - if port not in range

Since:

7.2.6

getTypecheckInputSchema

public final Schema getTypecheckInputSchema(int port)

Returns the schema of an input port that was set at application typecheck time. This schema may contain capture fields, and may not represent the actual schema of tuples passed into the Operator.processTuple(int, Tuple) method. For the actual schema of those tuples, see Operator.getRuntimeInputSchema(int).

Parameters:

port - the port to return the schema for (ports are zero based)

Returns:

Schema for the port, null if not running in an appication

Throws:

java.lang.IndexOutOfBoundsException - if port not in range

Since:

7.2.6

getOutputSchema

public final Schema getOutputSchema(int outputPort)

Return the output schema for the given output port (zero-based). If called before the operator is initialized, this method will act as Operator.getTypecheckOutputSchema(int). Otherwise, if called from Operator.init() or later, it will act as Operator.getRuntimeOutputSchema(int)

Parameters:

outputPort - the port to return the schema for (ports are zero based)

Returns:

Schema for the port, null if not running in an appication

Throws:

java.lang.IndexOutOfBoundsException - if port not in range

evaluate

public java.lang.String evaluate(java.lang.String st,
                                 Tuple input)
                          throws StreamBaseException

evaluate the given string expression in the context of the running module If an error occurs during the evaluation, the error message return value will be a string containing the error message

Parameters:

st - the expression to be evaluated

input - the input tuple

Returns:

the evaluated expression as a string.

Throws:

StreamBaseException

Since:

7.6

setDynamicVariable

public void setDynamicVariable(java.lang.String name,
                               java.lang.Object value)
                        throws StreamBaseException

Sets the value of this module's dynamic variable. Valid only at runtime.

Parameters:

name - name of the dynamic variable to be set

value - the new value (may be null)

Throws:

StreamBaseException - if an error occurs

Since:

7.6

getNamedSchema

public Schema getNamedSchema(java.lang.String name)

Retrieves a named schema

Parameters:

name - of the named schema

Returns:

a Schema object representing this named schema or null if there is no named schema in this module with name

Since:

7.6

getDynamicVariablesSchema

public Schema getDynamicVariablesSchema()

Retrieves a Schema containing information about every dynamic variable available to the module this operator is contained in. Never null: an empty schema is returned to indicate that there are no visible dynamic variables.

Returns:

the schema of the containing module's dynamic variables. Note that certain internal dynamic variables (such as those used by sequence operators) are not included.

Since:

7.6

getDynamicVariables

public Tuple getDynamicVariables()
                          throws TupleException

Retrieves a read-only tuple describing the current value of all dynamic variables in the module this operator is contained in. Never null: an empty tuple is returned to indicate that there are no visible dynamic variables.

Returns:

a read-only tuple containing all dynamic variables in this module.

Throws:

TupleException

Since:

7.6

getRuntimeOutputSchema

public final Schema getRuntimeOutputSchema(int outputPort)

Return the output schema that should actually use to Operator.sendOutput(int, Tuple) or Operator.sendOutputAsync(int, Tuple). Note that this schema will sometimes be different than the output schema set at typecheck time by the transformation determined by Operator.setCaptureStrategy(CaptureTransformStrategy). It will never contain capture fields. This method may not be called until after typecheck is finished; to do so will cause an IllegalStateException

Parameters:

outputPort - the port to return the schema for (ports are zero based)

Returns:

Schema for the port, null if not running in an appication

Throws:

java.lang.IndexOutOfBoundsException - if port not in range

Since:

7.2.6

getTypecheckOutputSchema

public final Schema getTypecheckOutputSchema(int outputPort)

Return the output schema as set by typecheck. Note that if this schema contains any capture fields, it may not be the same as the correct schema for tuples sent to Operator.sendOutput(int, Tuple) at runtime.

Parameters:

outputPort - the port to return the schema for (ports are zero based)

Returns:

Schema for the port, null if not running in an appication

Throws:

java.lang.IndexOutOfBoundsException - if port not in range

Since:

7.2.6

getCaptureStrategy

public final CaptureTransformStrategy getCaptureStrategy()

Returns the capture transform strategy that this Operator will use.

Returns:

the CaptureTransformStrategy to use

Since:

7.2.6

setCaptureStrategy

public final void setCaptureStrategy(CaptureTransformStrategy cts)

Set the capture transform strategy to use for the inputs and outputs of this Operator. Valid values for cts include CaptureTransformStrategy.FLATTEN and CaptureTransformStrategy.NEST. The default is CaptureTransformStrategy.FLATTEN.

Parameters:

cts - the CaptureTransformStrategy to use.

Since:

7.2.6

setOutputSchema

public final Schema setOutputSchema(int port,
                                    Schema outputSchema)
                             throws TypecheckException

Sets the output schema for the given output port (port #'s are zero based). This method should be called from the typecheck method of a Java Operator or Adapter.

Note that the actual Schema object to create tuples at runtime (for use by Operator.sendOutput(int, Tuple) or Operator.sendOutputAsync(int, Tuple)) should be retrieved using Operator.getRuntimeOutputSchema(int), as that will take into account any capture field transformations determined by Operator.getCaptureStrategy().

Parameters:

port - the port to set the given schema to (ports are zero based)

outputSchema - the schema to set the given port to

Returns:

returns the typecheck version of the output schema

Throws:

TypecheckException

Since:

7.0 returns Schema, 7.2 The returned schema should not be the Schema used to create output tuples. For the Schema to use to create output tuples, call Operator.getRuntimeOutputSchema(int) in the Operator.init() method or later.

See Also:

Operator.getOutputPortCount()

sendErrorOutput

public void sendErrorOutput(java.lang.String message)

Send an error tuple from this operator's error port.

This method will also record the message in the debug log.

Parameters:

message - A description of the error, must not be null

Throws:

java.lang.IllegalArgumentException - if message is null

Since:

5.1

sendErrorOutput

public void sendErrorOutput(java.lang.Throwable t)

Send an exception via the error output port

This method will also record the message in the debug log.

Parameters:

t - the exception to report

Throws:

java.lang.IllegalArgumentException - if exception is null

Since:

7.0

sendErrorOutput

public void sendErrorOutput(java.lang.Throwable t,
                            int port,
                            Tuple errorTuple)

Send an exception via the error output port

This method will also record the message in the debug log.

Parameters:

t - the exception to report

port - the input/output port that the error is related to

errorTuple - a tuple that caused or is related to the error

Throws:

java.lang.IllegalArgumentException - if exception is null

Since:

7.0

getTableAccessor

public TableAccessor getTableAccessor(java.lang.String name)
                               throws StreamBaseException

Get a TableAccessor for a table in the local module by name of that table.

Warning: This interface is provisional, and will likely change in upcoming versions of StreamBase.

Throws:

StreamBaseException

Since:

7.2

getLogger

protected org.slf4j.Logger getLogger()

Retrieves a Logger suitable for logging messages and exceptions for this Operator. Equivalent to LoggerFactory.getLogger(Class) with arguments this.getClass().

Returns:

a Logger instance for this Operator

Since:

6.0

setReuseTuple

protected void setReuseTuple(boolean reuse)

Allow/disallow the runtime to reuse tuples on operator input. The default is to disallow tuple reuse. Reusing operator input tuples will create fewer objects which may increase performance of the operator. To enable call: setReuseTuple(true) from the derived constructor.

Note: that if you plan to store the input Tuple as state in your operator you must copy the Tuple before you store it, if you have called setReuseTuple(true).

Parameters:

reuse - allow/disallow tuple reuse

getReuseTuple

public boolean getReuseTuple()

Get the state of the input Tuple reuse flag.

Returns:

boolean if input Tuple reuse is allowed/disallowed

getStateChangeTimeout

public int getStateChangeTimeout()

Returns:

the Operator state change timeout value, in milliseconds

See Also:

Operator.DEFAULT_STATE_CHANGE_TIMEOUT

getLocation

public com.streambase.sb.util.Location getLocation()

Return the Operator-wide location, useful for error reporting not associated to a particular property. If an error is related to a property, use Operator.getLocation(String) instead

Specified by:

getLocation in interface com.streambase.sb.util.LocatedItem

Returns:

the location for error reporting from this operator.

getLocation

public com.streambase.sb.util.Location getLocation(java.lang.String property)

Return a new location within this Operator, associated with the given property name. Equivalent to Operator.getLocation() for StreamSQL-sourced Operators. Using this as an argument to TypecheckException constructors will cause Studio to display an error indicated alongside the UI widget corresponding to the passed in property.

Parameters:

property - expected to be the bean name of an Operator property

Returns:

the location for error reporting from this operator

Since:

6.5

setLogLevel

public static void setLogLevel(org.slf4j.Logger logger,
                               java.lang.String level)
                        throws StreamBaseException

Sets the log level of the given logger. Note that this will only succeed if the default logging implementation (Logback) is used; if another logging back-end is used this call will throw an exception and the log level will remain unchanged.

Parameters:

logger - the Logger object on which to set the level.

level - the level at which to set the Logger instance.

Throws:

StreamBaseException - the log level could not be set, most likely because the current logging implementation is not Logback.

Since:

6.4.4.

setLogLevel

public static void setLogLevel(org.slf4j.Logger logger,
                               Operator.LogLevel level)
                        throws StreamBaseException

Sets the log level of the given logger. Note that this will only succeed if the default logging implementation (Logback) is used; if another logging back-end is used this call will throw an exception and the log level will remain unchanged.

Parameters:

logger - the Logger object on which to set the level.

level - the Operator.LogLevel at which to set the Logger instance.

Throws:

StreamBaseException - the log level could not be set, most likely because the current logging implementation is not Logback.

Since:

6.4.6.

getRuntimeEnvironment

public Operator.RuntimeEnvironment getRuntimeEnvironment()

Return the Operator.RuntimeEnvironment for this StreamBase Server. NOTE: Not valid during typecheck

Returns:

current RuntimeEnvironment, or null during typecheck

Since:

7.0

isDesignTime

@Deprecated
public boolean isDesignTime()

Deprecated. please use isRuntime()

Returns true when called in the context of a design-time operation (e.g. when Studio typechecks an operator), and false when called from a runtime context (i.e. application is currently running).

Returns:

true if called at design time, false if at runtime.

Since:

7.5.4

isRuntime

public boolean isRuntime()

Returns:

true if called during runtime (after the server has started), false otherwise

Since:

7.6

getOperatorConfigurationAccessor

public Operator.ConfigurationAccessor getOperatorConfigurationAccessor()

Provides access to operator and adapter configuration information from the server's configuration file. This call is intended to be made during typecheck or later.

Returns:

a Operator.ConfigurationAccessor to provide access to the aforementioned

Since:

7.5

getSchemaForCapture

public Schema getSchemaForCapture(java.lang.String captureName,
                                  int depth)

Finds the schema for the given capture name in the context that this operator is running under. For a capture field that was not nested inside any other capture field, use a depth value of 0. To expand the capture fields in a schema returned by this function with a depth of d, call this function with a depth of d+1. For example, if an output stream of this module has schema (f int, c @A), then you might call getSchemaForCapture("A", 0) to get (g int, c @B) as the schema bound to A. Then you might call getSchemaForCapture("B", 1) to get (h int) as the schema bound to B.

Note that the "nesting depth" here has nothing to do with capture fields found in lists or tuples -- the only containment that affects the nesting depth is a capture within another capture.

Parameters:

captureName - the name of the bound capture type

depth - the depth of nesting at which to look up the schema of this capture binding.

Returns:

the schema associated with the bound capture type.

Since:

7.2

getTupleCaptureTransformer

public TupleCaptureTransformer getTupleCaptureTransformer(Schema s)
                                                   throws StreamBaseException

Get a TupleCaptureTransformer capable of translating tuples with the given schema to the equivalent schema with all the capture fields expanded out, and translating expanded tuples back into tuples with the given schema This method may only be called at runtime; the exact schemas of any capture fields are not fully determined at typecheck time.

Parameters:

s - the schema with capture fields

Returns:

a TupleCaptureTransformer able to do the given transformations

Throws:

StreamBaseException - if this method is called at the wrong time, or if the transformations requested are not possible. This may occur, for example, if there is a name conflict between a captured field and a field already in the schema.

Since:

7.2

isPassthruOperator

public boolean isPassthruOperator()

Allows an operator to declare itself to be a "pass-thru" operator -- one that calls its sendOutput method with the unmodified tuples received by its processTuple method. Certain runtime optimizations are possible for pass-thru operators. In addition, the "older API" messages is suppressed for such operators. The StreamBase log adapter is an example of a pass-thru operator, as it logs, but leaves unmodified, the tuples that pass through it.

Returns:

true if the operator is pass-thru and false otherwise

Since:

7.2.12

getSearchKeywords

public java.lang.String[] getSearchKeywords()

Override to declare strings that assist the user in filtering for this operator when in Studio. Without search keywords, operators are found only by their short display name.

You may return composite words (such as "Thomson Reuters") as a single string; it has no effect on the user's filtering abilities.

Returns:

the default implementation returns no search keywords

Since:

7.3.10

allowsConcurrency

public boolean allowsConcurrency()

Override to indicate whether concurrency may be configured in Studio against this operator. Note that this applies to all instances, and is not intended to be interpreted on a per-instance basis. The default value when not overriden is true.

This has no effect on the actual runtime environment for operator instances.

Returns:

whether this operator allows Studio to set any concurrency options

Since:

7.4.0

getProposedInputSchemas

public Schema[] getProposedInputSchemas(java.lang.String mainName)

Override to provide schemas that Studio will offer to users to import into their projects. Neither property setting nor typecheck has occurred when Studio invokes this method.

Parameters:

mainName - the name the user has requested be the prefix for the top level schema, and this name should be used throughout any sub-field schemas

Returns:

an array of Schemas to offer for input ports

Since:

7.4.0

getProposedOutputSchemas

public Schema[] getProposedOutputSchemas(java.lang.String mainName)

Override to provide schemas that Studio will offer to users to import into their projects. Neither property setting nor typecheck has occurred when Studio invokes this method.

Parameters:

mainName - the name the user has requested be the prefix for the top level schema, and this name should be used throughout any sub-field schemas

Returns:

an array of Schemas to offer for output ports

Since:

7.4.0