Example usage for org.apache.lucene.queries.function FunctionQuery FunctionQuery

List of usage examples for org.apache.lucene.queries.function FunctionQuery FunctionQuery

  1. HOME
  2. Java
  3. org.apache.lucene
  4. org.apache.lucene.queries.function.*
  5. FunctionQuery
  6. FunctionQuery

Introduction

In this page you can find the example usage for org.apache.lucene.queries.function FunctionQuery FunctionQuery.

Prototype

public FunctionQuery(ValueSource func)

Source Link

Usage

From source file:com.difference.historybook.index.lucene.LuceneIndex.java

License:Apache License

@Override
public SearchResultWrapper search(String collection, String query, int offset, int size, boolean includeDebug)
        throws IndexException {
    try {/*from   w  ww  .ja v a  2  s.  c  om*/
        //TODO: make age be a component in the ranking?
        BooleanQuery.Builder queryBuilder = new BooleanQuery.Builder();
        queryBuilder.add(parser.parse(query), Occur.MUST);
        queryBuilder.add(new TermQuery(new Term(IndexDocumentAdapter.FIELD_COLLECTION, collection)),
                Occur.FILTER);
        Query baseQuery = queryBuilder.build();

        FunctionQuery boostQuery = new FunctionQuery(
                new ReciprocalFloatFunction(new DurationValueSource(new Date().getTime() / 1000,
                        new LongFieldSource(IndexDocumentAdapter.FIELD_TIMESTAMP)), RECIP, 1F, 1F));

        Query q = new CustomScoreQuery(baseQuery, boostQuery);

        QueryScorer queryScorer = new QueryScorer(q, IndexDocumentAdapter.FIELD_SEARCH);
        Fragmenter fragmenter = new SimpleSpanFragmenter(queryScorer);
        Highlighter highlighter = new Highlighter(queryScorer);
        highlighter.setTextFragmenter(fragmenter);

        GroupingSearch gsearch = new GroupingSearch(IndexDocumentAdapter.FIELD_URL_GROUP).setGroupDocsLimit(1)
                .setAllGroups(true).setIncludeMaxScore(true);
        TopGroups<?> groups = gsearch.search(searcher, q, offset, size);

        ArrayList<SearchResult> results = new ArrayList<>(size);
        for (int i = offset; i < offset + size && i < groups.groups.length; i++) {
            ScoreDoc scoreDoc = groups.groups[i].scoreDocs[0];
            Document luceneDoc = searcher.doc(scoreDoc.doc);
            IndexDocumentAdapter doc = new IndexDocumentAdapter(luceneDoc);

            TokenStream tokenStream = TokenSources.getTokenStream(IndexDocumentAdapter.FIELD_SEARCH,
                    reader.getTermVectors(scoreDoc.doc), luceneDoc.get(IndexDocumentAdapter.FIELD_SEARCH),
                    analyzer, highlighter.getMaxDocCharsToAnalyze() - 1);

            String[] snippets = highlighter.getBestFragments(tokenStream,
                    luceneDoc.get(IndexDocumentAdapter.FIELD_SEARCH), 3);
            String snippet = Arrays.asList(snippets).stream().collect(Collectors.joining("\n"));
            snippet = Jsoup.clean(snippet, Whitelist.simpleText());

            String debugInfo = null;
            if (includeDebug) {
                Explanation explanation = searcher.explain(q, scoreDoc.doc);
                debugInfo = explanation.toString();
            }

            results.add(new SearchResult(doc.getKey(), doc.getCollection(), doc.getTitle(), doc.getUrl(),
                    doc.getDomain(), doc.getTimestampText(), snippet, debugInfo, scoreDoc.score));
        }

        SearchResultWrapper wrapper = new SearchResultWrapper().setQuery(query).setOffset(offset)
                .setMaxResultsRequested(size)
                .setResultCount(groups.totalGroupCount != null ? groups.totalGroupCount : 0)
                .setResults(results);

        if (includeDebug) {
            wrapper.setDebugInfo(q.toString());
        }

        return wrapper;

    } catch (IOException | ParseException | InvalidTokenOffsetsException e) {
        LOG.error(e.getLocalizedMessage());
        throw new IndexException(e);
    }
}

From source file:com.mwired.grid.commons.commons.solr.custom.NewsDateInfluenceCustomeScoreQParsePlugin.java

@Override
public QParser createParser(String query, SolrParams sp, SolrParams sp1, SolrQueryRequest sqr) {
    return new QParser(query, sp, sp1, sqr) {
        @Override/*from   w  w  w.j a  v a 2 s  .  com*/
        public Query parse() throws SyntaxError {
            QParser parser = getParser(this.qstr, "lucene", this.req);
            Query inner = parser.parse();
            SchemaField createDate = getReq().getSchema().getField(PostPropAndColMap.CREATE_DATE);
            SchemaField influence = getReq().getSchema().getField(PostPropAndColMap.INFLUENCE_SCORE);
            ValueSource influence_source = influence.getType().getValueSource(influence, parser);
            ValueSource createDate_source = createDate.getType().getValueSource(createDate, parser);

            return new NewsDateInfluenceCustomQuery(inner, new FunctionQuery(influence_source),
                    new FunctionQuery(createDate_source));
        }
    };
}

From source file:edu.rpi.tw.linkipedia.search.query.WeightedQuery.java

License:Open Source License

public Query parse(String label, String[] contexts) throws ParseException {

    label = label.toLowerCase().trim();/*from   www . ja  va2 s  .c  om*/
    for (int i = 0; i < contexts.length; i++) {
        contexts[i] = contexts[i].toLowerCase().trim();
    }

    Query typeQuery = getTypeQuery(label);
    label = label.replaceAll("\\(.*\\)", "");
    if (debug)
        System.out.println("new label: " + label);
    BooleanQuery luceneQuery = new BooleanQuery();
    ArrayList<String> dictLabels = dict_exp.getExpandQueryString(label);
    //      ArrayList<String> temp = new ArrayList<String>( Arrays.asList(contexts));
    //      if(dictLabels != null){
    //      for(String dictLabel:dictLabels){
    //         temp.add(dictLabel);         
    //      }   
    //      }
    //      contexts = temp.toArray(new String[temp.size()]);
    ArrayList<String> weightedContext = assignContextWeight(label, contexts, contexts.length);

    //System.out.println(label+" "+weightedContext.toString());

    Query labelQuery = getLabelQuery(label);
    //Query labelQuery = getLabelSpanQuery(label);
    Query contentQuery = getContentQuery(label, weightedContext);
    Query relationQuery = getRelationQuery(label, weightedContext);

    //      if(relationQuery!=null)
    //         relationQuery.setBoost(100);

    if (typeQuery != null) {
        typeQuery.setBoost(typeBoost);
        luceneQuery.add(typeQuery, BooleanClause.Occur.SHOULD);
    }

    if (labelQuery != null) {
        //         if(relationQuery == null && labelBoost == 1){
        //            labelQuery.setBoost(10);
        //         }
        labelQuery.setBoost(labelBoost);
        luceneQuery.add(labelQuery, BooleanClause.Occur.MUST);
    }

    if (contentQuery != null) {
        contentQuery.setBoost(contentBoost);
        luceneQuery.add(contentQuery, BooleanClause.Occur.SHOULD);
    }

    if (relationQuery != null) {
        //relationQuery.setBoost(2);
        relationQuery.setBoost(relationBoost);
        luceneQuery.add(relationQuery, BooleanClause.Occur.SHOULD);
    }

    //      System.out.println("parse: labelBoost: "+labelBoost);
    //      System.out.println("parse: contentBoost: "+contentBoost);
    //      System.out.println("parse: relationBoost: "+relationBoost);
    //      System.out.println("parse: typeBoost: "+typeBoost);
    //      System.out.println("parse: defaultBoost: "+defaultBoost);

    //      Query dictQuery = getDictExpandQuery(label, weightedContext);
    //      if(dictQuery != null){
    //         luceneQuery.add(dictQuery, BooleanClause.Occur.SHOULD);   
    //      }
    FunctionQuery boostQuery = new FunctionQuery(new FloatFieldSource("boost"));
    boostQuery.setBoost(defaultBoost);

    finalQuery = new CustomScoreQuery(luceneQuery, boostQuery);

    return finalQuery;
}

From source file:edu.rpi.tw.linkipedia.search.query.WeightedQuery.java

License:Open Source License

public Query getFuzzyQuery(String label, String[] contexts) throws ParseException {
    label = label.toLowerCase().trim();/* w  ww  . j a  va  2  s .c  om*/
    for (int i = 0; i < contexts.length; i++) {
        contexts[i] = contexts[i].toLowerCase().trim();
    }

    Query typeQuery = getTypeQuery(label);
    label = label.replaceAll("\\(.*\\)", "");
    //System.out.println("new label: "+label);
    BooleanQuery luceneQuery = new BooleanQuery();
    ArrayList<String> weightedContext = assignContextWeight(label, contexts, contexts.length);

    //System.out.println(label+" "+weightedContext.toString());
    Term labelTerm = new Term("label", label);
    FuzzyQuery fuzzyLabelQuery = new FuzzyQuery(labelTerm);
    //Query labelQuery = getLabelQuery(label);   
    Query contentQuery = getContentQuery(label, weightedContext);
    Query relationQuery = getRelationQuery(label, weightedContext);

    //      if(relationQuery!=null)
    //         relationQuery.setBoost(100);

    if (typeQuery != null) {
        //typeQuery.setBoost(15);
        luceneQuery.add(typeQuery, BooleanClause.Occur.MUST);
    }

    if (fuzzyLabelQuery != null) {
        //         if(relationQuery == null)
        //            labelQuery.setBoost(10);
        luceneQuery.add(fuzzyLabelQuery, BooleanClause.Occur.MUST);
    }

    if (contentQuery != null) {
        luceneQuery.add(contentQuery, BooleanClause.Occur.SHOULD);
    }

    if (relationQuery != null) {
        //relationQuery.setBoost(2);
        luceneQuery.add(relationQuery, BooleanClause.Occur.SHOULD);
    }

    FunctionQuery boostQuery = new FunctionQuery(new FloatFieldSource("boost"));

    finalQuery = new CustomScoreQuery(luceneQuery, boostQuery);

    return finalQuery;
}

From source file:edu.rpi.tw.linkipedia.search.query.WeightedQueryFilterBackup.java

License:Open Source License

public Query parse(String label, String[] contexts) throws ParseException {

    label = label.toLowerCase().trim();/*from  w w w .j  a va  2 s  . c o m*/
    for (int i = 0; i < contexts.length; i++) {
        contexts[i] = contexts[i].toLowerCase().trim();
    }

    CachingWrapperFilter typeFilter = getTypeFilter(label);
    label = label.replaceAll("\\(.*\\)", "");
    //System.out.println("new label: "+label);
    BooleanQuery luceneQuery = new BooleanQuery();
    ArrayList<String> weightedContext = assignContextWeight(label, contexts, contexts.length);

    //System.out.println(label+" "+weightedContext.toString());

    Query labelQuery = getLabelQuery(label);
    Query contentQuery = getContentQuery(label, weightedContext);
    Query relationQuery = getRelationQuery(label, weightedContext);

    //      if(relationQuery!=null)
    //         relationQuery.setBoost(100);

    if (labelQuery != null) {

        luceneQuery.add(labelQuery, BooleanClause.Occur.MUST);
    }

    FunctionQuery boostQuery = new FunctionQuery(new FloatFieldSource("boost"));

    if (relationQuery == null && labelQuery != null) {
        labelQuery.setBoost(10);
        boostQuery.setBoost(5);
    }

    if (contentQuery != null && relationQuery != null) {
        luceneQuery.add(contentQuery, BooleanClause.Occur.SHOULD);
    }

    if (relationQuery != null) {
        //relationQuery.setBoost(2);
        luceneQuery.add(relationQuery, BooleanClause.Occur.SHOULD);
    }

    Query query = new CustomScoreQuery(luceneQuery, boostQuery);

    if (typeFilter != null) {
        finalQuery = new FilteredQuery(query, typeFilter);
        return finalQuery;
    }
    return query;

}

From source file:edu.rpi.tw.linkipedia.search.query.WeightedQueryFilterBackup.java

License:Open Source License

public Query getFuzzyQuery(String label, String[] contexts) throws ParseException {
    label = label.toLowerCase().trim();/*from w  w  w .j  a  va  2s.co  m*/

    for (int i = 0; i < contexts.length; i++) {
        contexts[i] = contexts[i].toLowerCase().trim();
    }

    CachingWrapperFilter typeFilter = getTypeFilter(label);
    label = label.replaceAll("\\(.*\\)", "");

    ArrayList<String> weightedContext = assignContextWeight(label, contexts, contexts.length);

    Term labelTerm = new Term("label", label);
    FuzzyQuery fuzzyLabelQuery = new FuzzyQuery(labelTerm);
    BooleanQuery luceneQuery = new BooleanQuery();

    if (fuzzyLabelQuery != null) {

        luceneQuery.add(fuzzyLabelQuery, BooleanClause.Occur.MUST);
    }
    Query contentQuery = getContentQuery(label, weightedContext);
    Query relationQuery = getRelationQuery(label, weightedContext);

    FunctionQuery boostQuery = new FunctionQuery(new FloatFieldSource("boost"));
    if (relationQuery == null && fuzzyLabelQuery != null) {
        fuzzyLabelQuery.setBoost(5);
        boostQuery.setBoost(5);
    }

    if (contentQuery != null && relationQuery != null) {
        luceneQuery.add(contentQuery, BooleanClause.Occur.SHOULD);
    }

    if (relationQuery != null) {
        //relationQuery.setBoost(2);
        luceneQuery.add(relationQuery, BooleanClause.Occur.SHOULD);
    }

    Query query = new CustomScoreQuery(luceneQuery, boostQuery);

    if (typeFilter != null) {
        finalQuery = new FilteredQuery(query, typeFilter);
        return finalQuery;
    }
    return query;
}

From source file:edu.rpi.tw.linkipedia.search.searching.SurfaceFormSearcher.java

License:Open Source License

public ArrayList<Annotation> mysearch(String query) {
    TopScoreDocCollector collector = TopScoreDocCollector.create(10, false);
    results = new ArrayList<Annotation>();
    try {/*from ww w  . j ava  2 s.  c o m*/
        BooleanQuery luceneQuery = new BooleanQuery();
        Term term = new Term("label", query);
        //TermQuery termQuery = new TermQuery(term);
        PayloadTermQuery ptq = new PayloadTermQuery(term, payloadFunction);
        ptq.setBoost(10);
        Analyzer analyzer = DefaultAnalyzer.getAnalyzer();
        QueryParser parser = new QueryParser(Version.LUCENE_47, "analyzedLabel", analyzer);
        Query parsedQuery = parser.parse(query);

        //luceneQuery.add(termQuery, BooleanClause.Occur.SHOULD);
        luceneQuery.add(ptq, BooleanClause.Occur.SHOULD);
        luceneQuery.add(parsedQuery, BooleanClause.Occur.SHOULD);
        FunctionQuery boostQuery = new FunctionQuery(new FloatFieldSource("boost"));
        //boostQuery.setBoost(100);
        Query finalQuery = new CustomScoreQuery(luceneQuery, boostQuery);
        System.out.println(finalQuery);
        searcher.search(finalQuery, collector);
    } catch (Exception e) {
        e.printStackTrace();
    }
    ScoreDoc[] hits = collector.topDocs().scoreDocs;
    int numTotalHits = collector.getTotalHits();

    int start = 0;
    int end = Math.min(numTotalHits, 10);

    for (int i = start; i < end; i++) {
        String thisResult = "";
        Document doc = null;
        try {
            doc = searcher.doc(hits[i].doc);
        } catch (Exception e) {
            e.printStackTrace();
        }
        String url = doc.get("url");
        thisResult += url;//+" "+hits[i].score;      
        String[] triples = doc.getValues("triple");
        //         for(String label:labels){
        //            thisResult += label+"\n";   
        //         }
        Annotation annotation = new Annotation(query, url, hits[i].score);
        annotation.setTriples(new ArrayList<String>(Arrays.asList(triples)));
        results.add(annotation);
    }
    return results;
}

From source file:org.apache.solr.schema.AbstractSpatialFieldType.java

License:Apache License

private Query getQueryFromSpatialArgs(QParser parser, SchemaField field, SpatialArgs spatialArgs) {
    T strategy = getStrategy(field.getName());

    SolrParams localParams = parser.getLocalParams();
    String score = (localParams == null ? null : localParams.get(SCORE_PARAM));
    if (score == null || "none".equals(score) || "".equals(score)) {
        //FYI Solr FieldType doesn't have a getFilter(). We'll always grab
        // getQuery() but it's possible a strategy has a more efficient getFilter
        // that could be wrapped -- no way to know.
        //See SOLR-2883 needScore
        return strategy.makeQuery(spatialArgs); //ConstantScoreQuery
    }/*www .  j a v  a  2 s.co m*/

    //We get the valueSource for the score then the filter and combine them.
    ValueSource valueSource;
    if ("distance".equals(score)) {
        double multiplier = 1.0;//TODO support units=kilometers
        valueSource = strategy.makeDistanceValueSource(spatialArgs.getShape().getCenter(), multiplier);
    } else if ("recipDistance".equals(score)) {
        valueSource = strategy.makeRecipDistanceValueSource(spatialArgs.getShape());
    } else {
        throw new SolrException(SolrException.ErrorCode.BAD_REQUEST,
                "'score' local-param must be one of 'none', 'distance', or 'recipDistance'");
    }
    FunctionQuery functionQuery = new FunctionQuery(valueSource);

    if (localParams != null && !localParams.getBool(FILTER_PARAM, true))
        return functionQuery;

    Filter filter = strategy.makeFilter(spatialArgs);
    return new FilteredQuery(functionQuery, filter);
}

From source file:org.apache.solr.search.function.TestOrdValues.java

License:Apache License

private void doTestRank(String field, boolean inOrder) throws Exception {
    IndexReader r = DirectoryReader.open(dir);
    IndexSearcher s = newSearcher(r);/*from  w ww . j av a 2  s  .  c  om*/
    ValueSource vs;
    if (inOrder) {
        vs = new OrdFieldSource(field);
    } else {
        vs = new ReverseOrdFieldSource(field);
    }

    Query q = new FunctionQuery(vs);
    log("test: " + q);
    QueryUtils.check(random(), q, s);
    ScoreDoc[] h = s.search(q, 1000).scoreDocs;
    assertEquals("All docs should be matched!", N_DOCS, h.length);
    String prevID = inOrder ? "IE" // greater than all ids of docs in this test ("ID0001", etc.)
            : "IC"; // smaller than all ids of docs in this test ("ID0001", etc.)

    for (int i = 0; i < h.length; i++) {
        String resID = s.doc(h[i].doc).get(ID_FIELD);
        log(i + ".   score=" + h[i].score + "  -  " + resID);
        log(s.explain(q, h[i].doc));
        if (inOrder) {
            assertTrue("res id " + resID + " should be < prev res id " + prevID, resID.compareTo(prevID) < 0);
        } else {
            assertTrue("res id " + resID + " should be > prev res id " + prevID, resID.compareTo(prevID) > 0);
        }
        prevID = resID;
    }
    r.close();
}

From source file:org.apache.solr.search.function.TestOrdValues.java

License:Apache License

private void doTestExactScore(String field, boolean inOrder) throws Exception {
    IndexReader r = DirectoryReader.open(dir);
    IndexSearcher s = newSearcher(r);/*w ww  .  jav  a2 s .c o m*/
    ValueSource vs;
    if (inOrder) {
        vs = new OrdFieldSource(field);
    } else {
        vs = new ReverseOrdFieldSource(field);
    }
    Query q = new FunctionQuery(vs);
    TopDocs td = s.search(q, 1000);
    assertEquals("All docs should be matched!", N_DOCS, td.totalHits);
    ScoreDoc sd[] = td.scoreDocs;
    for (int i = 0; i < sd.length; i++) {
        float score = sd[i].score;
        String id = s.getIndexReader().document(sd[i].doc).get(ID_FIELD);
        log("-------- " + i + ". Explain doc " + id);
        log(s.explain(q, sd[i].doc));
        float expectedScore = N_DOCS - i - 1;
        assertEquals("score of result " + i + " shuould be " + expectedScore + " != " + score, expectedScore,
                score, TEST_SCORE_TOLERANCE_DELTA);
        String expectedId = inOrder ? id2String(N_DOCS - i) // in-order ==> larger  values first
                : id2String(i + 1); // reverse  ==> smaller values first
        assertTrue("id of result " + i + " shuould be " + expectedId + " != " + score, expectedId.equals(id));
    }
    r.close();
}