Package de.julielab.xml

Class XmiBuilder

java.lang.Object

de.julielab.xml.XmiBuilder

public class XmiBuilder
extends Object

Field Summary

Fields

Modifier and Type	Field	Description
static List<String>	noViewMembers	Start tags of elements whose xmi ids should not appear in the cas:View-element.

Constructor Summary

Constructors

Constructor	Description
XmiBuilder(Map<String,String> nsAndXmiVersionMap, String[] annotationsToRetrieve)
XmiBuilder(Map<String,String> nsAndXmiVersionMap, String[] annotationsToRetrieve, long attribute_size)

Method Summary

Modifier and Type	Method	Description
ByteArrayOutputStream	buildXmi(LinkedHashMap<String,InputStream> xmiData, org.apache.uima.cas.TypeSystem ts)	Builds a single xmi representation out of the base document and the selected annotations.
void	setdocTableName(String docTableName)	The name of the table containing the base document.
void	setInputSize(int dataSize)
void	setXmiData(HashMap<String,InputStream> xmiData)	The xmi data (document and annotations) as input streams mapped to table names.
protected void	setXMIStartElementData(Map<String,String> namespacesAndXmiVersion)

Methods inherited from class java.lang.Object

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

Field Detail

noViewMembers

public static final List<String> noViewMembers

Start tags of elements whose xmi ids should not appear in the cas:View-element.

Constructor Detail

XmiBuilder

public XmiBuilder(Map<String,String> nsAndXmiVersionMap,
                  String[] annotationsToRetrieve,
                  long attribute_size)

XmiBuilder

public XmiBuilder(Map<String,String> nsAndXmiVersionMap,
                  String[] annotationsToRetrieve)

Method Detail

buildXmi

public ByteArrayOutputStream buildXmi(LinkedHashMap<String,InputStream> xmiData,
                                      org.apache.uima.cas.TypeSystem ts)

Builds a single xmi representation out of the base document and the selected annotations.

Parameters:

xmiData - The xmi data (document and annotations) as input streams mapped from storage keys. IMPORTANT: The document data must be associated with the XmiSplitter.DOCUMENT_MODULE_LABEL key. It is a LinkedHashMap because the order of the annotation data - e.g. first Sentences, then Tokens - influences the internal index order in the CAS. Thus, when serializing, the order of the LinkedHashMap will be reflected in the XMI. Since we rely on the order of appearance of annotations to determine the end of the base document, we must provide the correct order.

ts -

Returns:

The combined xmi representation as output stream.

setXMIStartElementData

protected void setXMIStartElementData(Map<String,String> namespacesAndXmiVersion)

setdocTableName

public void setdocTableName(String docTableName)

The name of the table containing the base document.

Parameters:

docTableName -

setXmiData

public void setXmiData(HashMap<String,InputStream> xmiData)

The xmi data (document and annotations) as input streams mapped to table names.

Parameters:

xmiData -

setInputSize

public void setInputSize(int dataSize)