de.julielab.xml

Class StaxXmiSplitter

java.lang.Object

de.julielab.xml.StaxXmiSplitter

All Implemented Interfaces:

XmiSplitter

public class StaxXmiSplitter
extends Object
implements XmiSplitter

Constructor Summary

Constructors

Constructor and Description
StaxXmiSplitter(List<String> annotationModulesToExtract, boolean recursively, boolean storeBaseDocument, String docTableName, Set<String> baseDocumentAnnotations) Creates an StaxXmiSplitter that selects specified annotations and returns them separately.
StaxXmiSplitter(List<String> annotationModulesToExtract, boolean recursively, boolean storeBaseDocument, String docTableName, Set<String> baseDocumentAnnotations, int attribute_size) TODO
StaxXmiSplitter(String tableName) Creates an StaxXmiSplitter that returns the whole xmi data.
StaxXmiSplitter(String tableName, int attribute_size) TODO

Method Summary

Modifier and Type	Method and Description
int	determineMaxXmiId(byte[] ba) Determines the maximum XMI ID assigned in the XMI data given by bais plus one.
List<String>	getAnnotationsToStore() Returns the specified annotations to select.
String	getFirstAnnotationtype() Return the first annotation type that is not an element of the base document.
XmiSplitterResult	process(byte[] ba, org.apache.uima.jcas.JCas aCas, int nextPossibleId, Map<String,Integer> existingSofaIdMap) Processes the xmi data according to the parameters set for the StaxXmiSplitter.
void	setAllJavaNames(ByteArrayInputStream bais) Retrieves the types namespace uri from the xmi element in order to create the fully qualified java name for all annotations that have not been given in this form.

Methods inherited from class java.lang.Object

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

Constructor Detail

StaxXmiSplitter

public StaxXmiSplitter(List<String> annotationModulesToExtract,
                       boolean recursively,
                       boolean storeBaseDocument,
                       String docTableName,
                       Set<String> baseDocumentAnnotations,
                       int attribute_size)

TODO

StaxXmiSplitter

public StaxXmiSplitter(List<String> annotationModulesToExtract,
                       boolean recursively,
                       boolean storeBaseDocument,
                       String docTableName,
                       Set<String> baseDocumentAnnotations)

Creates an StaxXmiSplitter that selects specified annotations and returns them separately. Optionally, the base document will be returned as well. This if optimal if the xmi data already contains some annotations together with the base document, the base document has to be stored once.

Parameters:

annotationModulesToExtract - A list of annotations to select and to return separately. These should be given as the fully qualified java names of the annotation types. If not, the fully qualified java name will be retrieved from the types namespace and used as table name.

recursively - If set to true, annotations that are features of the selected annotations will also be returned.

storeBaseDocument - If set to true, the base document will be returned as well. In this case the name of the document table as well as as the first annotation type has to be given.

docTableName - The table to store the base document. Can be null if storeBaseDocument is set to false.

baseDocumentAnnotations - The names of those annotation types that should be stored together with the document.

StaxXmiSplitter

public StaxXmiSplitter(String tableName,
                       int attribute_size)

TODO

StaxXmiSplitter

public StaxXmiSplitter(String tableName)

Creates an StaxXmiSplitter that returns the whole xmi data. It will also return the highest xmi id assigned so far. This is optimal if the xmi data contains only the base document which has to be stored for the first time.

Parameters:

tableName - The table to store the xmi data.

Method Detail

process

public XmiSplitterResult process(byte[] ba,
                                 org.apache.uima.jcas.JCas aCas,
                                 int nextPossibleId,
                                 Map<String,Integer> existingSofaIdMap)
                          throws XMISplitterException

Processes the xmi data according to the parameters set for the StaxXmiSplitter.

Specified by:

process in interface XmiSplitter

Parameters:

ba - The current xmi data.

aCas - The current cas.

nextPossibleId - The next xmi id that can be assigned to the new annotations. If the base document has not yet been stored this should be 0.

Returns:

A list containing the xmi data (only unless the complete XMI document should be stored), i.e. separate annotations and optionally the base document, as output streams mapped to the table names - a HashMapString, ByteArrayOutputStream - the next possible xmi id - an Integer - to store in the database and a namespace map derived from XMI opening element where each namespace prefix is mapped to the namespace URI (only unless the complete XMI document should be stored).

Throws:

IOException

XMISplitterException

determineMaxXmiId

public int determineMaxXmiId(byte[] ba)

Determines the maximum XMI ID assigned in the XMI data given by bais plus one.

Parameters:

ba - The current xmi data as input stream.

setAllJavaNames

public void setAllJavaNames(ByteArrayInputStream bais)

Retrieves the types namespace uri from the xmi element in order to create the fully qualified java name for all annotations that have not been given in this form.

Parameters:

bais - The current xmi data as input stream.

getAnnotationsToStore

public List<String> getAnnotationsToStore()

Returns the specified annotations to select.

Returns:

getFirstAnnotationtype

public String getFirstAnnotationtype()

Return the first annotation type that is not an element of the base document.

Returns: