org.apache.uima.analysis_engine
Interface AnalysisEngineDescription

All Superinterfaces:

Cloneable, MetaDataObject, ResourceCreationSpecifier, ResourceSpecifier, Serializable, XMLizable

All Known Subinterfaces:

TaeDescription

public interface AnalysisEngineDescription
extends ResourceCreationSpecifier

An AnalysisEngineDescription contains all of the information needed to instantiate and use an AnalysisEngine.

The UIMAFramework.produceAnalysisEngine(ResourceSpecifier) factory method is used to create an AnalysisEngine instance from a AnalysisEngineDescription. This insulates applications from knowledge of the particular AnalysisEngine implementation being used.

The AnalysisEngine Description contains the following information:

For a Primitive AnalysisEngine:

• Name of main annotator class

For an Aggregate AnalysisEngine:

• A set of Resource Specifiers that specify the component AnalysisEngines that comprise the aggregate.
• Optionally, a Resource Specifier for the FlowController component that determines routing of the CAS to the component AnalysisEngines.

For All AnalysisEngines:

• A set of Resource Specifiers that specify the external resources needed by this AnalysisEngine.
• Metadata for this AnalysisEngine - this includes the capabilities of this AnalysisEngine, the TypeSystem used by this AnalysisEngine, and other parameters.

Method Summary

void	doFullValidation() Does full validation of this Analysis Engine Description.
void	doFullValidation(ResourceManager aResourceManager) Does full validation of this Analysis Engine Description.
Map<String,ResourceSpecifier>	getAllComponentSpecifiers(ResourceManager aResourceManager) For an aggregate AnalysisEngine only, gets the ResourceSpecifiers of all components in this aggregate.
AnalysisEngineMetaData	getAnalysisEngineMetaData() Retrieves the metadata that describes the AnalysisEngine.
String	getAnnotatorImplementationName() For a primitive AnalysisEngine only, retrieves the name of the annotator implementation.
ResourceSpecifier	getComponentSpecifier(String key) Gets the ResourceSpecifier of one a component of this aggregate, based on its key.
Map<String,ResourceSpecifier>	getDelegateAnalysisEngineSpecifiers() For an aggregate AnalysisEngine only, retrieves a collection of ResourceSpecifiers that indicate which delegate AnalysisEngines comprise the aggregate.
Map<String,ResourceSpecifier>	getDelegateAnalysisEngineSpecifiers(ResourceManager aResourceManager) For an aggregate AnalysisEngine only, retrieves a collection of ResourceSpecifiers that indicate which delegate AnalysisEngines comprise the aggregate.
Map<String,MetaDataObject>	getDelegateAnalysisEngineSpecifiersWithImports() Retrieves a Map whose keys are string identifiers and whose values are the either Import or ResourceSpecifier objects.
FlowControllerDeclaration	getFlowControllerDeclaration() For an aggregate AnalysisEngine only, gets the declaration of which FlowController should be used by the AnalysisEngine.
String	getFrameworkImplementation() Gets the name of the framework implementation within which the AnalysisEngine executes.
SofaMapping[]	getSofaMappings() Reserved for future use.
boolean	isPrimitive() Retrieves whether the AnalysisEngine is primitive (consisting of one annotator), as opposed to aggregate (containing multiple delegate AnalysisEngines).
void	resolveImports(Collection<String> aAlreadyImportedDelegateAeUrls, ResourceManager aResourceManager) Resolves all import declarations in this AnalysisEngineDescription.
void	resolveImports(ResourceManager aResourceManager) Resolves all import declarations in this AnalysisEngineDescription.
void	setAnnotatorImplementationName(String aImplementationName) For a primitive AnalysisEngine only, sets the name of the annotator implementation.
void	setFlowControllerDeclaration(FlowControllerDeclaration aFlowControllerDeclaration) For an aggregate AnalysisEngine only, sets the declaration of which FlowController should be used by the AnalysisEngine.
void	setFrameworkImplementation(String aFrameworkImplementation) Sets the name of the framework implementation within which the AnalysisEngine executes.
void	setPrimitive(boolean aPrimitive) Sets whether the AnalysisEngine is primitive (consisting of one annotator), as opposed to aggregate (containing multiple delegate AnalysisEngines).
void	setSofaMappings(SofaMapping[] aSofaMappings) Reserved for future use.
void	toXML(ContentHandler aContentHandler, boolean aWriteDefaultNamespaceAttribute, boolean aPreserveDelegateAnalysisEngineImports) Writes this object's XML representation by making calls on a SAX ContentHandler.
void	toXML(OutputStream aOutputStream, boolean aPreserveDelegateAnalysisEngineImports) Writes this object's XML representation as a string in UTF-8 encoding.
void	toXML(Writer aWriter, boolean aPreserveDelegateAnalysisEngineImports) Writes this object's XML representation as a string.

Methods inherited from interface org.apache.uima.resource.ResourceCreationSpecifier

getExternalResourceDependencies, getExternalResourceDependency, getImplementationName, getMetaData, getResourceManagerConfiguration, setExternalResourceDependencies, setImplementationName, setMetaData, setResourceManagerConfiguration, validate, validate

Methods inherited from interface org.apache.uima.resource.metadata.MetaDataObject

clone, equals, getAttributeValue, getSourceUrl, getSourceUrlString, isModifiable, listAttributes, setAttributeValue, setSourceUrl

Methods inherited from interface org.apache.uima.util.XMLizable

buildFromXMLElement, buildFromXMLElement, toXML, toXML, toXML, toXML

Method Detail

getFrameworkImplementation

String getFrameworkImplementation()

Gets the name of the framework implementation within which the AnalysisEngine executes. The framework name for this implementation is given by Constants.JAVA_FRAMEWORK_NAME.

Specified by:

getFrameworkImplementation in interface ResourceCreationSpecifier

Returns:

the framework implementation name

setFrameworkImplementation

void setFrameworkImplementation(String aFrameworkImplementation)

Sets the name of the framework implementation within which the AnalysisEngine executes. The framework name for this implementation is given by Constants.JAVA_FRAMEWORK_NAME.

Specified by:

setFrameworkImplementation in interface ResourceCreationSpecifier

Parameters:

aFrameworkImplementation - the framework implementation name

isPrimitive

boolean isPrimitive()

Retrieves whether the AnalysisEngine is primitive (consisting of one annotator), as opposed to aggregate (containing multiple delegate AnalysisEngines).

Some of the methods on this class apply only to one type of AnalysisEngine:
getAnnotatorImplementationName() - primitive AnalysisEngine only
getDelegateAnalysisEngineSpecifiers() - aggregate AnalysisEngine only
#getFlowController() - aggregate AnalysisEngine only

Returns:

true if and only if the AnalysisEngine is primitive

setPrimitive

void setPrimitive(boolean aPrimitive)

Sets whether the AnalysisEngine is primitive (consisting of one annotator), as opposed to aggregate (containing multiple delegate AnalysisEngines).

Parameters:

aPrimitive - true if and only if the AnalysisEngine is primitive

getAnnotatorImplementationName

String getAnnotatorImplementationName()

For a primitive AnalysisEngine only, retrieves the name of the annotator implementation. For Java annotators, this will be a fully qualified Java class name.

Returns:

the implementation name of the annotator. If the AnalysisEngine is aggregate, always returns null.

setAnnotatorImplementationName

void setAnnotatorImplementationName(String aImplementationName)

For a primitive AnalysisEngine only, sets the name of the annotator implementation. For Java annotators, this must be a fully qualified Java class name.

Parameters:

aImplementationName - the implementation name of the annotator.

getDelegateAnalysisEngineSpecifiers

Map<String,ResourceSpecifier> getDelegateAnalysisEngineSpecifiers()
                                                                  throws InvalidXMLException

For an aggregate AnalysisEngine only, retrieves a collection of ResourceSpecifiers that indicate which delegate AnalysisEngines comprise the aggregate. Each ResourceSpecifier can either:

1. completely describe how to build a AnalysisEngine instance
2. describe how to locate a distributed AnalysisEngine service, for example a specific endpoint or a JNDI name

This method returns an unmodifiable Map whose keys are string identifiers and whose values are the ResourceSpecifier objects. The string identifiers in this Map are to refer to the delegate AnalysisEngines from elsewhere in this AnalysisEngineDescription. (For example in the FlowConstraints description.)

Note that the Map returned by this method will never contain Import objects -- they will always be resolved first. If you want to get access to the original Import objects, use getDelegateAnalysisEngineSpecifiersWithImports(). Also use that method if you want to be able to make changes to the Map.

Returns:

an unmodifiable Map with String keys and ResourceSpecifier values.

Throws:

InvalidXMLException - if import resolution failed

getDelegateAnalysisEngineSpecifiers

Map<String,ResourceSpecifier> getDelegateAnalysisEngineSpecifiers(ResourceManager aResourceManager)
                                                                  throws InvalidXMLException

For an aggregate AnalysisEngine only, retrieves a collection of ResourceSpecifiers that indicate which delegate AnalysisEngines comprise the aggregate. Each ResourceSpecifier can either:

1. completely describe how to build a AnalysisEngine instance
2. describe how to locate a distributed AnalysisEngine service, for example a specific endpoint or a JNDI name

This method returns an unmodifiable Map whose keys are string identifiers and whose values are the ResourceSpecifier objects. The string identifiers in this Map are to refer to the delegate AnalysisEngines from elsewhere in this AnalysisEngineDescription. (For example in the FlowConstraints description.)

Note that the Map returned by this method will never contain Import objects -- they will always be resolved first. If you want to get access to the original Import objects, use getDelegateAnalysisEngineSpecifiersWithImports(). Also use that method if you want to be able to make changes to the Map.

Parameters:

aResourceManager - the ResourceManager to use to get the datapath needed to resolve imports

Returns:

an unmodifiable Map with String keys and ResourceSpecifier values.

Throws:

InvalidXMLException - if import resolution failed

getDelegateAnalysisEngineSpecifiersWithImports

Map<String,MetaDataObject> getDelegateAnalysisEngineSpecifiersWithImports()

Retrieves a Map whose keys are string identifiers and whose values are the either Import or ResourceSpecifier objects. These indicate the delegate AnalysisEngines that comprise the aggregate.

This is a direct representation of what is in the XML syntax for the descriptor. That is, if the XML had an <import> element, the Map will contain an Import object. If you do not want to deal with imports, use the getDelegateAnalysisEngineSpecifiers() method instead.

Returns:

a Map with String keys and ResourceSpecifier or Import objects as values. This Map may be modified to add or remove imports or specifiers.

getFlowControllerDeclaration

FlowControllerDeclaration getFlowControllerDeclaration()

For an aggregate AnalysisEngine only, gets the declaration of which FlowController should be used by the AnalysisEngine.

Returns:

an object containing either an import of a ResourceSpecifier or a ResourceSpecifier itself. This specifier will be used to create the FlowController.

setFlowControllerDeclaration

void setFlowControllerDeclaration(FlowControllerDeclaration aFlowControllerDeclaration)

For an aggregate AnalysisEngine only, sets the declaration of which FlowController should be used by the AnalysisEngine.

Parameters:

aFlowControllerDeclaration - an object containing either an import of a ResourceSpecifier or a ResourceSpecifier itself. This specifier will be used to create the FlowController.

getAllComponentSpecifiers

Map<String,ResourceSpecifier> getAllComponentSpecifiers(ResourceManager aResourceManager)
                                                        throws InvalidXMLException

For an aggregate AnalysisEngine only, gets the ResourceSpecifiers of all components in this aggregate. This includes the FlowController as well as all of the component AnalysisEngines.

This method returns an unmodifiable Map whose keys are string identifiers and whose values are the ResourceSpecifier objects. The string identifiers in this Map are to refer to the components from elsewhere in this aggregate descriptor, for example in configuration parameter overrides and resource bindings.

Note that the Map returned by this method will never contain Import objects -- they will always be resolved first. If you want to get access to the original Import objects, use getDelegateAnalysisEngineSpecifiersWithImports() and #getFlowController(). Also use those methods if you want to make changes to be able to make changes to the Map.

Parameters:

aResourceManager - the ResourceManager from which to get the datapath needed to resolve imports. Pass null to use the default ResourceManager.

Returns:

an unmodifiable Map with String keys and ResourceSpecifier values.

Throws:

InvalidXMLException - if import resolution failed

getAnalysisEngineMetaData

AnalysisEngineMetaData getAnalysisEngineMetaData()

Retrieves the metadata that describes the AnalysisEngine. This includes the AnalysisEngine's capabilties, the TypeSystem that is uses, the specified Flow information for an aggregate AnalysisEngine, and various informational attributes such as name, description, version, vendor, and copyright.

Returns:

the AnalysisEngineMetaData object containing the AnalysisEngine's metadata. This object can be modified.

getSofaMappings

SofaMapping[] getSofaMappings()

Reserved for future use.

setSofaMappings

void setSofaMappings(SofaMapping[] aSofaMappings)

Reserved for future use.

getComponentSpecifier

ResourceSpecifier getComponentSpecifier(String key)
                                        throws ResourceInitializationException

Gets the ResourceSpecifier of one a component of this aggregate, based on its key. This may be the specifier of a component (i.e. delegate) AnalysisEngine, or it may be the specifier of the FlowController.

Parameters:

key - the key of the component specifier to get

Returns:

the specifier for the component, null if there is no component with the given key

Throws:

ResourceInitializationException - if there's a problem resolving imports

doFullValidation

void doFullValidation()
                      throws ResourceInitializationException

Does full validation of this Analysis Engine Description. This essentially performs all operations necessary to instantiate an Analysis Engine from this description, except that it does not actually instantiate the Annotator classes (although it does try to load these classes). This method will also attempt to create a CAS based on the descriptor, in order to do full type system verification. If any operations fail, an exception will be thrown.

Specified by:

doFullValidation in interface ResourceCreationSpecifier

Throws:

ResourceInitializationException - if validation failed

doFullValidation

void doFullValidation(ResourceManager aResourceManager)
                      throws ResourceInitializationException

Does full validation of this Analysis Engine Description. This essentially performs all operations necessary to instantiate an Analysis Engine from this description, except that it does not actually instantiate the Annotator classes (although it does try to load these classes). This method will also attempt to create a CAS based on the descriptor, in order to do full type system verification. If any operations fail, an exception will be thrown.

Specified by:

doFullValidation in interface ResourceCreationSpecifier

Parameters:

aResourceManager - a ResourceManager instance to use to load annotator classes, external resource classes, and resolve imports by name.

Throws:

ResourceInitializationException - if validation failed

resolveImports

void resolveImports(ResourceManager aResourceManager)
                    throws InvalidXMLException

Resolves all import declarations in this AnalysisEngineDescription. For an aggregate, this is recursive, also resolving all imports in each delegate AnalysisEngine. Users do not typically need to call this method; it is called automatically when UIMAFramework.produceAnalysisEngine(ResourceSpecifier) is called.

Parameters:

aResourceManager - the Resource Manager used to locate imports by name. For example, the path in which to locate these imported descriptors can be set via the ResourceManager.setDataPath(String) method.

Throws:

InvalidXMLException - if an import target does not exist or is invalid

resolveImports

void resolveImports(Collection<String> aAlreadyImportedDelegateAeUrls,
                    ResourceManager aResourceManager)
                    throws InvalidXMLException

Resolves all import declarations in this AnalysisEngineDescription. For an aggregate, this is recursive, also resolving all imports in each delegate AnalysisEngine. Users do not typically need to call this method; it is called automatically when UIMAFramework.produceAnalysisEngine(ResourceSpecifier) is called.

This version is used internally to resolve nested imports.

Parameters:

aAlreadyImportedDelegateAeUrls - URLs of already imported AE descriptors, so we don't import them again.

aResourceManager - the Resource Manager used to locate imports by name. For example, the path in which to locate these imported descriptors can be set via the ResourceManager.setDataPath(String) method.

Throws:

InvalidXMLException - if an import target does not exist or is invalid

toXML

void toXML(Writer aWriter,
           boolean aPreserveDelegateAnalysisEngineImports)
           throws SAXException,
                  IOException

Writes this object's XML representation as a string. Note that if you want to write the XML to a file, it is highly recommended that you use XMLizable.toXML(OutputStream) instead, as it ensures that output is written in UTF-8 encoding, which is the default encoding that should be used for XML files.

Parameters:

aWriter - a Writer to which the XML string will be written

aPreserveDelegateAnalysisEngineImports - if true, XML serialization will always preserve <import> elements used to import delegate analysis engine specifiers into an aggregate. If false, the default import serialization behavior applies, which is to write <import> elements only in the case where they have not previously been resolved.

Throws:

IOException - if an I/O failure occurs

SAXException

toXML

void toXML(OutputStream aOutputStream,
           boolean aPreserveDelegateAnalysisEngineImports)
           throws SAXException,
                  IOException

Writes this object's XML representation as a string in UTF-8 encoding.

Parameters:

aOutputStream - an OutputStream to which the XML string will be written, in UTF-8 encoding.

aPreserveDelegateAnalysisEngineImports - if true, XML serialization will always preserve <import> elements used to import delegate analysis engine specifiers into an aggregate. If false, the default import serialization behavior applies, which is to write <import> elements only in the case where they have not previously been resolved.

Throws:

IOException - if an I/O failure occurs

SAXException

toXML

void toXML(ContentHandler aContentHandler,
           boolean aWriteDefaultNamespaceAttribute,
           boolean aPreserveDelegateAnalysisEngineImports)
           throws SAXException

Writes this object's XML representation by making calls on a SAX ContentHandler.

Parameters:

aContentHandler - the content handler to which this object will write events that describe its XML representation.

aWriteDefaultNamespaceAttribute - whether the namespace of this element should be written as the default namespace. This should be done only for the root element, and it defaults to false.

aPreserveDelegateAnalysisEngineImports - if true, XML serialization will always preserve <import> elements used to import delegate analysis engine specifiers into an aggregate. If false, the default import serialization behavior applies, which is to write <import> elements only in the case where they have not previously been resolved.

Throws:

IOException - if an I/O failure occurs

SAXException