Examples with AnnotationSet gate.AnnotationSet used on opensource projects

Example 16 with AnnotationSet

use of gate.AnnotationSet in project gate-core by GateNLP.

the class CorpusAnnotationDiff method init.

/**
 * This method does the diff, Precision,Recall,FalsePositive
 * calculation and so on.
 */
@Override
public Resource init() throws ResourceInstantiationException {
    colors[DEFAULT_TYPE] = WHITE;
    colors[CORRECT_TYPE] = GREEN;
    colors[SPURIOUS_TYPE] = RED;
    colors[PARTIALLY_CORRECT_TYPE] = BLUE;
    colors[MISSING_TYPE] = YELLOW;
    // Initialize the partially sets...
    keyPartiallySet = new HashSet<Annotation>();
    responsePartiallySet = new HashSet<Annotation>();
    // Do the diff, P&R calculation and so on
    AnnotationSet keyAnnotSet = null;
    AnnotationSet responseAnnotSet = null;
    if (annotationSchema == null)
        throw new ResourceInstantiationException("No annotation schema defined !");
    if (keyCorpus == null || 0 == keyCorpus.size())
        throw new ResourceInstantiationException("No key corpus or empty defined !");
    if (responseCorpus == null || 0 == responseCorpus.size())
        throw new ResourceInstantiationException("No response corpus or empty defined !");
    // init counters and do difference for documents by pairs
    for (int type = 0; type < MAX_TYPES; type++) typeCounter[type] = 0;
    diffSet = new HashSet<DiffSetElement>();
    for (int i = 0; i < keyCorpus.size(); ++i) {
        keyDocument = keyCorpus.get(i);
        // find corresponding responce document if any
        Document doc;
        responseDocument = null;
        for (int j = 0; j < responseCorpus.size(); ++j) {
            doc = responseCorpus.get(j);
            if (0 == doc.getName().compareTo(keyDocument.getName()) || 0 == doc.getSourceUrl().getFile().compareTo(keyDocument.getSourceUrl().getFile())) {
                responseDocument = doc;
                // response corpus loop
                break;
            }
        // if
        }
        if (null == responseDocument) {
            Out.prln("There is no mach in responce corpus for document '" + keyDocument.getName() + "' from key corpus");
            // key corpus loop
            continue;
        }
        if (keyAnnotationSetName == null) {
            // Get the default key AnnotationSet from the keyDocument
            keyAnnotSet = keyDocument.getAnnotations().get(annotationSchema.getAnnotationName());
        } else {
            keyAnnotSet = keyDocument.getAnnotations(keyAnnotationSetName).get(annotationSchema.getAnnotationName());
        }
        if (keyAnnotSet == null)
            // The diff will run with an empty set.All annotations from response
            // would be spurious.
            keyAnnotList = new LinkedList<Annotation>();
        else
            // The alghoritm will modify this annotation set. It is better to make a
            // separate copy of them.
            keyAnnotList = new LinkedList<Annotation>(keyAnnotSet);
        if (responseAnnotationSetName == null)
            // Get the response AnnotationSet from the default set
            responseAnnotSet = responseDocument.getAnnotations().get(annotationSchema.getAnnotationName());
        else
            responseAnnotSet = responseDocument.getAnnotations(responseAnnotationSetName).get(annotationSchema.getAnnotationName());
        if (responseAnnotSet == null)
            // The diff will run with an empty set.All annotations from key
            // would be missing.
            responseAnnotList = new LinkedList<Annotation>();
        else
            // The alghoritm will modify this annotation set. It is better to make a
            // separate copy of them.
            responseAnnotList = new LinkedList<Annotation>(responseAnnotSet);
        // Sort them ascending on Start offset (the comparator does that)
        AnnotationSetComparator asComparator = new AnnotationSetComparator();
        Collections.sort(keyAnnotList, asComparator);
        Collections.sort(responseAnnotList, asComparator);
        // Calculate the diff Set. This set will be used later with graphic
        // visualisation.
        doDiff(keyAnnotList, responseAnnotList);
    }
    // If it runs under text mode just stop here.
    if (textMode)
        return this;
    // Show it
    // Configuring the formatter object. It will be used later to format
    // precision and recall
    formatter.setMaximumIntegerDigits(1);
    formatter.setMinimumFractionDigits(4);
    formatter.setMinimumFractionDigits(4);
    // Create an Annotation diff table model
    AnnotationDiffTableModel diffModel = new AnnotationDiffTableModel(diffSet);
    // Create a XJTable based on this model
    diffTable = new XJTable(diffModel);
    diffTable.setAlignmentX(Component.LEFT_ALIGNMENT);
    // Set the cell renderer for this table.
    AnnotationDiffCellRenderer cellRenderer = new AnnotationDiffCellRenderer();
    diffTable.setDefaultRenderer(java.lang.String.class, cellRenderer);
    diffTable.setDefaultRenderer(java.lang.Long.class, cellRenderer);
    // Put the table into a JScroll
    // Arange all components on a this JPanel
    SwingUtilities.invokeLater(new Runnable() {

        @Override
        public void run() {
            arangeAllComponents();
        }
    });
    if (DEBUG)
        printStructure(diffSet);
    return this;
}

Example 17 with AnnotationSet

use of gate.AnnotationSet in project gate-core by GateNLP.

the class DocumentImpl method getAnnotations.

// getAnnotations()
/**
 * Get a named set of annotations. Creates a new set if one with this name
 * doesn't exist yet. If the provided name is null or the empty string then
 * it returns the default annotation set.
 */
@Override
public AnnotationSet getAnnotations(String name) {
    if (name == null || "".equals(name))
        return getAnnotations();
    if (namedAnnotSets == null) {
        namedAnnotSets = new HashMap<String, AnnotationSet>();
    }
    AnnotationSet namedSet = namedAnnotSets.get(name);
    if (namedSet == null) {
        namedSet = new AnnotationSetImpl(this, name);
        namedAnnotSets.put(name, namedSet);
        DocumentEvent evt = new DocumentEvent(this, DocumentEvent.ANNOTATION_SET_ADDED, name);
        fireAnnotationSetAdded(evt);
    }
    return namedSet;
}

Example 18 with AnnotationSet

use of gate.AnnotationSet in project gate-core by GateNLP.

the class DocumentImpl method writeStartTag.

// writeStartTag
/**
 * Returns a string representing a start tag based on the input annot
 */
private String writeStartTag(Annotation annot, boolean includeFeatures, boolean includeNamespace) {
    // Get the annot feature used to store the namespace prefix, if it
    // has been defined
    String nsPrefix = null;
    if (serializeNamespaceInfo)
        nsPrefix = (String) annot.getFeatures().get(namespacePrefixFeature);
    AnnotationSet originalMarkupsAnnotSet = this.getAnnotations(GateConstants.ORIGINAL_MARKUPS_ANNOT_SET_NAME);
    StringBuffer strBuff = new StringBuffer("");
    if (annot == null)
        return strBuff.toString();
    // if (!addGatePreserveFormatTag && isRootTag){
    if (theRootAnnotation != null && annot.getId().equals(theRootAnnotation.getId())) {
        // spoil all links in an HTML file!
        if (includeFeatures) {
            strBuff.append("<");
            if (nsPrefix != null && !nsPrefix.isEmpty())
                strBuff.append(nsPrefix + ":");
            strBuff.append(annot.getType());
            strBuff.append(" ");
            if (includeNamespace) {
                // but don't add the gate ns declaration if it's already there!
                if (annot.getFeatures().get("xmlns:gate") == null)
                    strBuff.append("xmlns:gate=\"http://www.gate.ac.uk\"");
                strBuff.append(" gate:");
            }
            strBuff.append("gateId=\"");
            strBuff.append(annot.getId());
            strBuff.append("\"");
            strBuff.append(" ");
            if (includeNamespace) {
                strBuff.append("gate:");
            }
            strBuff.append("annotMaxId=\"");
            strBuff.append(nextAnnotationId);
            strBuff.append("\"");
            strBuff.append(writeFeatures(annot.getFeatures(), includeNamespace));
            strBuff.append(">");
        } else if (originalMarkupsAnnotSet.contains(annot)) {
            strBuff.append("<");
            if (nsPrefix != null && !nsPrefix.isEmpty())
                strBuff.append(nsPrefix + ":");
            strBuff.append(annot.getType());
            strBuff.append(writeFeatures(annot.getFeatures(), includeNamespace));
            strBuff.append(">");
        } else {
            strBuff.append("<");
            if (nsPrefix != null && !nsPrefix.isEmpty())
                strBuff.append(nsPrefix + ":");
            strBuff.append(annot.getType());
            strBuff.append(">");
        }
    } else {
        // spoil all links in an HTML file!
        if (includeFeatures) {
            strBuff.append("<");
            if (nsPrefix != null && !nsPrefix.isEmpty())
                strBuff.append(nsPrefix + ":");
            strBuff.append(annot.getType());
            strBuff.append(" ");
            if (includeNamespace) {
                strBuff.append("gate:");
            }
            // if includeNamespaces
            strBuff.append("gateId=\"");
            strBuff.append(annot.getId());
            strBuff.append("\"");
            strBuff.append(writeFeatures(annot.getFeatures(), includeNamespace));
            strBuff.append(">");
        } else if (originalMarkupsAnnotSet.contains(annot)) {
            strBuff.append("<");
            if (nsPrefix != null && !nsPrefix.isEmpty())
                strBuff.append(nsPrefix + ":");
            strBuff.append(annot.getType());
            strBuff.append(writeFeatures(annot.getFeatures(), includeNamespace));
            strBuff.append(">");
        } else {
            strBuff.append("<");
            if (nsPrefix != null && !nsPrefix.isEmpty())
                strBuff.append(nsPrefix + ":");
            strBuff.append(annot.getType());
            strBuff.append(">");
        }
    }
    // End if
    return strBuff.toString();
}

Example 19 with AnnotationSet

use of gate.AnnotationSet in project gate-core by GateNLP.

the class DocumentImpl method insertsSafety.

// End toXml()
/**
 * This method verifies if aSourceAnnotation can ve inserted safety into the
 * aTargetAnnotSet. Safety means that it doesn't violate the crossed over
 * contition with any annotation from the aTargetAnnotSet.
 *
 * @param aTargetAnnotSet
 *          the annotation set to include the aSourceAnnotation
 * @param aSourceAnnotation
 *          the annotation to be inserted into the aTargetAnnotSet
 * @return true if the annotation inserts safety, or false otherwise.
 */
private boolean insertsSafety(AnnotationSet aTargetAnnotSet, Annotation aSourceAnnotation) {
    if (aTargetAnnotSet == null || aSourceAnnotation == null) {
        this.crossedOverAnnotation = null;
        return false;
    }
    if (aSourceAnnotation.getStartNode() == null || aSourceAnnotation.getStartNode().getOffset() == null) {
        this.crossedOverAnnotation = null;
        return false;
    }
    if (aSourceAnnotation.getEndNode() == null || aSourceAnnotation.getEndNode().getOffset() == null) {
        this.crossedOverAnnotation = null;
        return false;
    }
    // Get the start and end offsets
    Long start = aSourceAnnotation.getStartNode().getOffset();
    Long end = aSourceAnnotation.getEndNode().getOffset();
    // Read aSourceAnnotation offsets long
    long s2 = start.longValue();
    long e2 = end.longValue();
    // Obtain a set with all annotations annotations that overlap
    // totaly or partially with the interval defined by the two provided offsets
    AnnotationSet as = aTargetAnnotSet.get(start, end);
    // Investigate all the annotations from as to see if there is one that
    // comes in conflict with aSourceAnnotation
    Iterator<Annotation> it = as.iterator();
    while (it.hasNext()) {
        Annotation ann = it.next();
        // Read ann offsets
        long s1 = ann.getStartNode().getOffset().longValue();
        long e1 = ann.getEndNode().getOffset().longValue();
        if (s1 < s2 && s2 < e1 && e1 < e2) {
            this.crossedOverAnnotation = ann;
            return false;
        }
        if (s2 < s1 && s1 < e2 && e2 < e1) {
            this.crossedOverAnnotation = ann;
            return false;
        }
    }
    // End while
    return true;
}

Example 20 with AnnotationSet

use of gate.AnnotationSet in project gate-core by GateNLP.

the class DocumentXmlUtils method toXml.

/**
 * Returns a GateXml document that is a custom XML format for wich there is a
 * reader inside GATE called gate.xml.GateFormatXmlHandler. What it does is to
 * serialize a GATE document in an XML format.
 *
 * @param doc the document to serialize.
 * @return a string representing a Gate Xml document.
 */
public static String toXml(TextualDocument doc) {
    // Initialize the xmlContent several time the size of the current document.
    // This is because of the tags size. This measure is made to increase the
    // performance of StringBuffer.
    StringBuffer xmlContent = new StringBuffer(DOC_SIZE_MULTIPLICATION_FACTOR * (doc.getContent().size().intValue()));
    // Add xml header
    xmlContent.append("<?xml version=\"1.0\" encoding=\"");
    xmlContent.append(doc.getEncoding());
    xmlContent.append("\" ?>");
    xmlContent.append(Strings.getNl());
    // Add the root element
    xmlContent.append("<GateDocument>\n");
    xmlContent.append("<!-- The document's features-->\n\n");
    xmlContent.append("<GateDocumentFeatures>\n");
    xmlContent.append(featuresToXml(doc.getFeatures(), null));
    xmlContent.append("</GateDocumentFeatures>\n");
    xmlContent.append("<!-- The document content area with serialized" + " nodes -->\n\n");
    // Add plain text element
    xmlContent.append("<TextWithNodes>");
    xmlContent.append(textWithNodes(doc, doc.getContent().toString()));
    xmlContent.append("</TextWithNodes>\n");
    // Serialize as XML all document's annotation sets
    // Serialize the default AnnotationSet
    StatusListener sListener = (StatusListener) gate.Gate.getListeners().get("gate.event.StatusListener");
    if (sListener != null)
        sListener.statusChanged("Saving the default annotation set ");
    xmlContent.append("<!-- The default annotation set -->\n\n");
    annotationSetToXml(doc.getAnnotations(), xmlContent);
    // Serialize all others AnnotationSets
    // namedAnnotSets is a Map containing all other named Annotation Sets.
    Map<String, AnnotationSet> namedAnnotSets = doc.getNamedAnnotationSets();
    if (namedAnnotSets != null) {
        Iterator<AnnotationSet> iter = namedAnnotSets.values().iterator();
        while (iter.hasNext()) {
            AnnotationSet annotSet = iter.next();
            xmlContent.append("<!-- Named annotation set -->\n\n");
            // Serialize it as XML
            if (sListener != null)
                sListener.statusChanged("Saving " + annotSet.getName() + " annotation set ");
            annotationSetToXml(annotSet, xmlContent);
        }
    // End while
    }
    // End if
    // Add the end of GateDocument
    xmlContent.append("</GateDocument>");
    if (sListener != null)
        sListener.statusChanged("Done !");
    // return the XmlGateDocument
    return xmlContent.toString();
}