Example usage for org.apache.lucene.search.highlight QueryTermExtractor getTerms

Introduction

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

Prototype

public static final WeightedTerm[] getTerms(Query query) 

Source Link

Document

Extracts all terms texts of a given Query into an array of WeightedTerms

Usage

From source file:org.conifer.MyBean.java

License:Apache License

public static String sortOutTerms(String queryString, String resultString, String tag, String strLen)
        throws org.apache.lucene.queryParser.ParseException {
    String startTag = "<" + tag + ">";
    String endTag = "</" + tag + ">";
    StringBuffer termBuf = new StringBuffer();
    int addToLen = startTag.length() + endTag.length();

    int resultLen = Integer.parseInt(strLen);
    String displayString = resultString;

    QueryParser luceneParser = new QueryParser("text", new StandardAnalyzer());
    QueryTermExtractor luceneTerms = new QueryTermExtractor();
    Query luceneQuery = null;//from  www .j  av a 2s  .  c o m
    Query query = null;
    WeightedTerm[] weightTerms = null;
    luceneQuery = luceneParser.parse(queryString);
    weightTerms = luceneTerms.getTerms(luceneQuery);

    for (int i = 0; i < weightTerms.length; i++) {
        String term = weightTerms[i].getTerm();
        Pattern replace = Pattern.compile(term, Pattern.CASE_INSENSITIVE);
        Matcher matcher = replace.matcher(displayString);
        while (matcher.find()) {
            //displayString = matcher.replaceAll(startTag + term + endTag);
            if (termBuf.toString().indexOf(matcher.group(0)) == -1) {
                displayString = displayString.replaceAll(matcher.group(0),
                        startTag + matcher.group(0) + endTag);
                resultLen += addToLen;
            } //if
            termBuf.append(matcher.group(0));
            termBuf.append(",");
        } //while
    } //for

    int hlStart = displayString.indexOf(startTag);
    //System.out.println("hlStart: " + hlStart);

    if (hlStart > resultLen) {
        String tmpString = displayString.substring(0, hlStart);
        hlStart = tmpString.lastIndexOf(" ");
        displayString = "..." + displayString.substring(++hlStart);
        resultLen += 3;
    } //if

    if (displayString.length() > resultLen) {
        displayString = displayString.substring(0, resultLen);
        hlStart = displayString.lastIndexOf(startTag);
        int hlEnd = displayString.lastIndexOf(endTag);
        if (hlEnd < hlStart)
            displayString = displayString.substring(0, --hlStart);
        displayString = displayString + "...";
    } //if

    return displayString;
}

---

From source file:org.sd.text.lucene.SdQuery.java

License:Open Source License

public SdQuery(String queryString, QueryParser queryParser) {
    this.queryString = queryString;
    try {/*  ww  w.ja v  a 2  s.c  om*/
        this.query = queryParser.parse(queryString);

        //todo: populate fieldNames. Currently, these are left null such that
        //      if we need them we'll get a NullPointerException which will
        //      alert us that it is time to implement this method!

        // It turns out that lucene handles the functionality we need out of the box!
        // The search.highlight package handles term extraction, including exclustion of prohibited terms
        //final WeightedTerm[] queryTerms = QueryTermExtractor.getTerms(this.query.rewrite(new IndexReader()));
        // todo: implement the rewrite() call to incorporate an index reader
        final WeightedTerm[] queryTerms = QueryTermExtractor.getTerms(this.query);
        this.queryTerms = new String[queryTerms.length];
        for (int i = 0; i < queryTerms.length; i++) {
            this.queryTerms[i] = queryTerms[i].getTerm();
        }
    } catch (ParseException e) {
        throw new IllegalArgumentException(e);
    }
}

---

From source file:retriever.TermStats.java

TopDocs rerankTopDocsByWordVecSim(Query query, TopDocs topDocs) throws Exception {
    // Compute doc-query vector based similarities
    WeightedTerm[] qterms = QueryTermExtractor.getTerms(query);
    DocVecSimilarity dvecSim = new DocVecSimilarity(prop, frIndexReader, topDocs, qterms, textSimWt);
    ScoreDoc[] wvecScoreDocs = dvecSim.computeSims();

    // Combine the similarity scores of the wvecs and the text
    ScoreDoc[] combinedScoreDocs = combineSimilarities(topDocs.scoreDocs, wvecScoreDocs);

    TopDocs rerankedTopDocs = new TopDocs(topDocs.scoreDocs.length, combinedScoreDocs,
            combinedScoreDocs[0].score);

    return rerankedTopDocs;
}

---