Example usage for org.apache.lucene.search.highlight TextFragment TextFragment

Introduction

In this page you can find the example usage for org.apache.lucene.search.highlight TextFragment TextFragment.

Prototype

public TextFragment(CharSequence markedUpText, int textStartPos, int fragNum) 

Source Link

Usage

From source file:lux.search.highlight.XmlHighlighter.java

License:Mozilla Public License

private void init(TokenStream tokenStream) {
    try {//  ww w .  j a v  a2s . co m
        tokenStream.reset();
        scorer.setMaxDocCharsToAnalyze(maxDocCharsToAnalyze);
        scorerTokens = scorer.init(tokenStream);
        if (scorerTokens == null) {
            // The scorer didn't consume any tokens (it does that for PhraseQuery),
            // in which case we must give it the live token stream
            scorer.init(xmlStreamTokens);
        }
        // we score the entire document as a single fragment
        scorer.startFragment(new TextFragment("", 0, 0));
    } catch (IOException e) {
        throw new LuxException(e);
    }
}

---

From source file:org.apache.uima.lucas.ProspectiveSearchAE.java

License:Apache License

@Override
public void process(CAS aCAS) throws AnalysisEngineProcessException {

    // First create the index of the document text
    MemoryIndex index = new MemoryIndex();

    List fields = createDocument(aCAS).getFields();

    for (Iterator it = fields.iterator(); it.hasNext();) {
        Field field = (Field) it.next();

        if (field.isIndexed() && field.tokenStreamValue() != null) {
            index.addField(field.name(), field.tokenStreamValue());
        }/*from  w w  w .ja  v  a  2 s .c  o  m*/
    }

    // Search all queries against the one document index
    for (SearchQuery query : searchQueryProvider.getSearchQueries(aCAS)) {

        float score = index.search(query.query());

        if (score > matchingThreshold) {

            // Add a FS to the CAS with the search result
            FeatureStructure searchResult = aCAS.createFS(searchResultType);
            searchResult.setLongValue(searchResultIdFeature, query.id());
            aCAS.addFsToIndexes(searchResult);

            // Find matching tokens and link their annotations
            // in case the user wants search term highlighting
            if (searchResultMatchingTextFeature != null) {

                fields = createDocument(aCAS).getFields();

                for (Iterator it = fields.iterator(); it.hasNext();) {

                    Field field = (Field) it.next();

                    if (field.isIndexed() && field.tokenStreamValue() != null) {

                        TokenStream tokenStream = field.tokenStreamValue();

                        Collection<AnnotationFS> matchingTextAnnotations = new LinkedList<AnnotationFS>();

                        QueryScorer scorer = new QueryScorer(query.query(), field.name());
                        scorer.startFragment(new TextFragment(new StringBuffer(aCAS.getDocumentText()), 0, 0));

                        try {
                            scorer.init(tokenStream);

                            OffsetAttribute offsetAttr = null;
                            while (tokenStream.incrementToken()) {
                                offsetAttr = (OffsetAttribute) tokenStream.getAttribute(OffsetAttribute.class);
                                float tokenScore = scorer.getTokenScore();
                                if (tokenScore > 0) {
                                    AnnotationFS annotation = aCAS.createAnnotation(matchingTextType,
                                            offsetAttr.startOffset(), offsetAttr.endOffset());

                                    matchingTextAnnotations.add(annotation);
                                }
                            }
                        } catch (IOException e) {
                            throw new AnalysisEngineProcessException(e);
                        }

                        ArrayFS matchtingTextArray = aCAS.createArrayFS(matchingTextAnnotations.size());

                        int matchtingTextArrayIndex = 0;
                        for (AnnotationFS matchingTextAnnotation : matchingTextAnnotations) {
                            matchtingTextArray.set(matchtingTextArrayIndex++, matchingTextAnnotation);
                        }

                        searchResult.setFeatureValue(searchResultMatchingTextFeature, matchtingTextArray);
                    }
                }
            }
        }
    }
}

---