Example usage for org.apache.lucene.search IndexSearcher getSimilarity

List of usage examples for org.apache.lucene.search IndexSearcher getSimilarity

  1. HOME
  2. Java
  3. org.apache.lucene
  4. org.apache.lucene.search.*
  5. IndexSearcher
  6. getSimilarity

Introduction

In this page you can find the example usage for org.apache.lucene.search IndexSearcher getSimilarity.

Prototype

public Similarity getSimilarity()

Source Link

Document

Expert: Get the Similarity to use to compute scores.

Usage

From source file:com.gentics.cr.lucene.indexaccessor.DefaultIndexAccessor.java

License:Apache License

/**
 * Fetches a double checked Searcher that has been checked for the presence of a reopen file
 * Note that it may occure that a prioritized Searcher may be reopened twice.
 * @param indexLocation //from w w  w .  j  a v  a 2 s  .co m
 * @return
 * @throws IOException
 */
public IndexSearcher getPrioritizedSearcher() throws IOException {
    boolean reopened = this.numReopening > 0;
    IndexSearcher searcher = (IndexSearcher) getSearcher();

    if (reopened) {
        //REOPEN SEARCHER AS IT WAS PRIORITIZED
        synchronized (DefaultIndexAccessor.this) {
            IndexReader reader = searcher.getIndexReader();
            IndexSearcher oldSearcher = searcher;
            IndexReader newReader = reader.reopen();
            if (newReader != reader) {
                searcher = new IndexSearcher(newReader);
                searcher.setSimilarity(oldSearcher.getSimilarity());
                oldSearcher.getIndexReader().close();
                for (Map.Entry<Similarity, IndexSearcher> e : cachedSearchers.entrySet()) {
                    if (e.getValue() == oldSearcher) {
                        cachedSearchers.put(e.getKey(), searcher);
                    }
                }
            }
        }
    }

    return searcher;
}

From source file:com.gentics.cr.lucene.indexaccessor.DefaultIndexAccessor.java

License:Apache License

/**
 * Reopens all of the Searchers in the Searcher cache. This method is invoked
 * in a synchronized context.//from ww w.j av a 2 s  . c  om
 */
private void reopenCachedSearchers() {
    LOGGER.debug(
            "reopening cached searchers (" + cachedSearchers.size() + "):" + Thread.currentThread().getId());
    Set<Similarity> keys = cachedSearchers.keySet();
    for (Similarity key : keys) {
        IndexSearcher searcher = cachedSearchers.get(key);
        try {
            IndexReader oldReader = searcher.getIndexReader();
            IndexSearcher oldSearcher = searcher;
            IndexReader newReader = oldReader.reopen();

            if (newReader != oldReader) {

                cachedSearchers.remove(key);
                searcher = new IndexSearcher(newReader);
                searcher.setSimilarity(oldSearcher.getSimilarity());
                oldSearcher.getIndexReader().close();
                cachedSearchers.put(key, searcher);
            }

        } catch (IOException e) {
            LOGGER.error("error reopening cached Searcher", e);
        }
    }

}

From source file:com.xiaomi.linden.lucene.query.flexiblequery.FlexibleWeight.java

License:Apache License

public FlexibleWeight(FlexibleQuery query, IndexSearcher searcher) throws IOException {
    this.query = query;
    this.similarity = searcher.getSimilarity();
    final IndexReaderContext context = searcher.getTopReaderContext();

    int[] maxDocFreqs = null;
    long[] maxTotalTermFreqs = null;
    Map<Term, TermContext> builtTermMap = new HashMap<>();
    if (query.enableGlobalIDF()) {
        FlexibleQuery.FlexibleTerm[][] globalTerms = query.getGlobalTerms();
        TermContext[][] globalStates = new TermContext[globalTerms.length][];
        for (int i = 0; i < globalTerms.length; ++i) {
            globalStates[i] = new TermContext[globalTerms[i].length];
            for (int j = 0; j < globalTerms[i].length; ++j) {
                Term term = globalTerms[i][j].term;
                TermContext termContext = builtTermMap.get(term);
                if (termContext != null) {
                    globalStates[i][j] = termContext;
                } else {
                    globalStates[i][j] = TermContext.build(context, globalTerms[i][j].term);
                    builtTermMap.put(term, globalStates[i][j]);
                }//ww w .jav  a  2 s .c  om
            }
        }
        maxDocFreqs = new int[globalTerms[0].length];
        maxTotalTermFreqs = new long[globalTerms[0].length];
        int fieldLength = globalTerms.length;
        int termLength = globalTerms[0].length;
        for (int i = 0; i < termLength; ++i) {
            int maxDocFreq = 0;
            long maxTotalTermFreq = 0;
            for (int j = 0; j < fieldLength; ++j) {
                maxDocFreq = Math.max(globalStates[j][i].docFreq(), maxDocFreq);
                maxTotalTermFreq = Math.max(globalStates[j][i].totalTermFreq(), maxTotalTermFreq);
            }
            maxDocFreqs[i] = maxDocFreq;
            maxTotalTermFreqs[i] = maxTotalTermFreq;
        }
    }

    FlexibleQuery.FlexibleTerm[][] terms = query.getTerms();
    TermContext[][] states = new TermContext[terms.length][];
    for (int i = 0; i < terms.length; ++i) {
        states[i] = new TermContext[terms[i].length];
        for (int j = 0; j < terms[i].length; ++j) {
            Term term = terms[i][j].term;
            TermContext termContext = builtTermMap.get(term);
            if (termContext != null) {
                states[i][j] = termContext;
            } else {
                states[i][j] = TermContext.build(context, terms[i][j].term);
                builtTermMap.put(term, states[i][j]);
            }
        }
    }
    termStatsMatrix = new TermStats[terms.length][];
    for (int i = 0; i < terms.length; ++i) {
        termStatsMatrix[i] = new TermStats[terms[i].length];
        for (int j = 0; j < terms[i].length; ++j) {
            FlexibleQuery.FlexibleTerm term = terms[i][j];
            TermContext state = states[i][j];
            TermStatistics termStats;
            if (query.enableGlobalIDF()) {
                termStats = new TermStatistics(term.term.bytes(), maxDocFreqs[j], maxTotalTermFreqs[j]);
            } else {
                termStats = searcher.termStatistics(term.term, state);
            }
            Similarity.SimWeight stats = similarity.computeWeight(term.boost,
                    searcher.collectionStatistics(term.term.field()), termStats);
            TermStats termStatsInfo = new TermStats();
            termStatsInfo.stats = stats;
            termStatsInfo.term = term.term;
            termStatsInfo.termContext = state;
            termStatsMatrix[i][j] = termStatsInfo;
        }
    }
}

From source file:in.student.project.queryexpansion.SearchFilesLDA.java

License:Apache License

/**
 *
 * args/* w  w w  .  j a va 2s .  c om*/
 * 0 - properties_file
 * 
 * # index-dir
 * index-dir = index
 * 
 * # query-file
 * query-file = queries.txt
 * 
 * # out-count
 * docs-per-query = 1000
 * 
 * # query-term-count (0 - any count) (ex. queries with only 3 terms - 3 )
 * query-terms-count = 0
 * 
 * # out-file - name of the file where results will be written
 * out-file = search.result
 *
 * #query-expansion
 * For details see <code> QueryExpansion Constants</code> 
 * @throws Exception 
 *
 */
public static void main(String[] args) throws Exception {
    // Load Properties
    Properties properties = new Properties();
    properties.load(new FileInputStream(args[0]));
    String runTag = args[0];
    String remoteQuery = "";
    if (args.length > 1) {
        remoteQuery = args[1];
        expandedQuery = "";
    }

    properties.setProperty(Defs.RUN_TAG_FLD, runTag);
    String indexDir = properties.getProperty("index-dir");
    String queryFile = properties.getProperty("query-file");
    int termCount = Integer.valueOf(properties.getProperty("query-terms-count")).intValue();
    int outCount = Integer.valueOf(properties.getProperty("docs-per-query")).intValue();
    String queryFileName = properties.getProperty("query-file");
    String outFileName = runTag + properties.getProperty("out-file");
    String queryExpansionFlag = properties.getProperty(QueryExpansion.METHOD_FLD, "");

    IndexReader idxReader = DirectoryReader.open(FSDirectory.open(new File(indexDir)));
    IndexSearcher searcher = new IndexSearcher(idxReader);
    Analyzer analyzer = new StandardAnalyzer(Version.LUCENE_48);
    BufferedWriter writer = new BufferedWriter(new FileWriter(new File(outFileName)));
    String query_num = null;
    BufferedReader in = new BufferedReader(new FileReader(queryFileName));
    TFIDFSimilarity similarity = null;
    int hitsCount = 50;

    while (true) {
        String line;
        line = in.readLine();
        try {
            if (line.length() == -1)
                break;
        } catch (Exception e) {
            return;
        }
        StringTokenizer tknzr = new StringTokenizer(line);
        query_num = tknzr.nextToken();
        line = line.substring(query_num.length()).trim();
        if (remoteQuery.length() > 0)
            line = remoteQuery;
        String queryStr = line;
        queryStr = QueryParser.escape(queryStr);
        QueryParser parser = new QueryParser(Version.LUCENE_48, Defs.FLD_TEXT, analyzer);
        Query query = parser.parse(queryStr);
        QueryTermVector queryTermVector = new QueryTermVector(line, analyzer);
        String[] terms = queryTermVector.getTerms();
        similarity = (TFIDFSimilarity) searcher.getSimilarity();

        TopDocs hits = searcher.search(query, hitsCount);
        System.out.println("query" + " : " + query.toString());
        System.out.println(hits.totalHits + " total matching documents");

        // Query Expansion with LDA
        QueryExpansion queryExpansion;
        queryExpansion = new QueryExpansion(analyzer, searcher, similarity, properties);
        query = queryExpansion.expandQueryLDA(queryStr, hits, properties);
        expandedQuery = query.toString("contents");
        logger.info("Expanded Query: " + query);
        System.out.println("Expanded Query: " + query);
        hits = searcher.search(query, hitsCount);
        logger.info(hits.totalHits + " total matching documents");
        Vector<TermQuery> expandedQueryTerms = queryExpansion.getExpandedTerms();
        generateOutput(hits, expandedQueryTerms, query_num, writer, termCount, outCount, searcher, similarity,
                idxReader);

        writer.flush();
        if (remoteQuery.length() > 0)
            break; // onetime call.
    }
    writer.close();

}

From source file:in.student.project.queryexpansion.SearchFilesRocchio.java

License:Apache License

/**
 *
 * args/*from  w ww  .jav a2s.co  m*/
 * 0 - properties_file
 * 
 * # index-dir
 * index-dir = index
 * 
 * # query-file
 * query-file = queries.txt
 * 
 * # out-count
 * docs-per-query = 1000
 * 
 * # query-term-count (0 - any count) (ex. queries with only 3 terms - 3 )
 * query-terms-count = 0
 * 
 * # out-file - name of the file where results will be written
 * out-file = search.result
 *
 * #query-expansion
 * For details see <code> QueryExpansion Constants</code> 
 * @throws Exception 
 *
 */
public static void main(String[] args) throws Exception {
    // Load Properties
    Properties properties = new Properties();
    properties.load(new FileInputStream(args[0]));
    String runTag = args[0];
    String remoteQuery = "";
    if (args.length > 1) {
        remoteQuery = args[1];
        expandedQuery = "";
    }

    properties.setProperty(Defs.RUN_TAG_FLD, runTag);
    String indexDir = properties.getProperty("index-dir");
    String queryFile = properties.getProperty("query-file");
    int termCount = Integer.valueOf(properties.getProperty("query-terms-count")).intValue();
    int outCount = Integer.valueOf(properties.getProperty("docs-per-query")).intValue();
    String queryFileName = properties.getProperty("query-file");
    String outFileName = runTag + properties.getProperty("out-file");
    String queryExpansionFlag = properties.getProperty(QueryExpansion.METHOD_FLD, "");

    IndexReader idxReader = DirectoryReader.open(FSDirectory.open(new File(indexDir)));
    IndexSearcher searcher = new IndexSearcher(idxReader);
    Analyzer analyzer = new StandardAnalyzer(Version.LUCENE_48);
    BufferedWriter writer = new BufferedWriter(new FileWriter(new File(outFileName)));
    String query_num = null;
    BufferedReader in = new BufferedReader(new FileReader(queryFileName));
    TFIDFSimilarity similarity = null;
    int hitsCount = 50;

    while (true) {
        String line;
        line = in.readLine();
        try {
            if (line.length() == -1)
                break;
        } catch (Exception e) {
            return;
        }
        StringTokenizer tknzr = new StringTokenizer(line);
        query_num = tknzr.nextToken();
        line = line.substring(query_num.length()).trim();
        if (remoteQuery.length() > 0)
            line = remoteQuery;
        String queryStr = line;
        queryStr = QueryParser.escape(queryStr);
        QueryParser parser = new QueryParser(Version.LUCENE_48, Defs.FLD_TEXT, analyzer);
        Query query = parser.parse(queryStr);
        QueryTermVector queryTermVector = new QueryTermVector(line, analyzer);
        String[] terms = queryTermVector.getTerms();
        similarity = (TFIDFSimilarity) searcher.getSimilarity();

        TopDocs hits = searcher.search(query, hitsCount);
        System.out.println("query" + " : " + query.toString());
        System.out.println(hits.totalHits + " total matching documents");

        // Query Expansion with Rocchio algorithm
        if (queryExpansionFlag.equals(QueryExpansion.ROCCHIO_METHOD)) {
            QueryExpansion queryExpansion;
            queryExpansion = new QueryExpansion(analyzer, searcher, similarity, properties);
            query = queryExpansion.expandQuery(queryStr, hits, properties);
            expandedQuery = query.toString("contents");
            System.out.println("Expanded Query: " + query);
            hits = searcher.search(query, hitsCount);
            Vector<TermQuery> expandedQueryTerms = queryExpansion.getExpandedTerms();
            generateOutput(hits, expandedQueryTerms, query_num, writer, termCount, outCount, searcher,
                    similarity, idxReader);
        }

        writer.flush();
        if (remoteQuery.length() > 0)
            break; // onetime call.
    }
    writer.close();
}

From source file:org.apache.blur.lucene.search.AbstractWrapperQuery.java

License:Apache License

public Similarity getSimilarity(IndexSearcher searcher) {
    return searcher.getSimilarity();
}

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

License:Apache License

@Override
public DocValues getValues(Map context, AtomicReaderContext readerContext) throws IOException {
    IndexSearcher searcher = (IndexSearcher) context.get("searcher");
    Similarity sim = searcher.getSimilarity();
    // todo: we need docFreq that takes a BytesRef
    String strVal = ByteUtils.UTF8toUTF16(indexedBytes);
    int docfreq = searcher.docFreq(new Term(indexedField, strVal));
    float idf = sim.idf(docfreq, searcher.maxDoc());
    return new ConstDoubleDocValues(idf, this);
}

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

License:Apache License

@Override
public DocValues getValues(Map context, AtomicReaderContext readerContext) throws IOException {
    IndexSearcher searcher = (IndexSearcher) context.get("searcher");
    final Similarity similarity = searcher.getSimilarity();
    final byte[] norms = readerContext.reader.norms(field);
    if (norms == null) {
        return new ConstDoubleDocValues(0.0, this);
    }/*w w  w  . jav  a 2s  . co m*/

    return new FloatDocValues(this) {
        @Override
        public float floatVal(int doc) {
            return similarity.decodeNormValue(norms[doc]);
        }
    };
}

From source file:org.elasticsearch.common.lucene.search.ExtendedIndexSearcher.java

License:Apache License

public ExtendedIndexSearcher(IndexSearcher searcher) {
    super(searcher.getIndexReader());
    setSimilarity(searcher.getSimilarity());
}

From source file:org.elasticsearch.index.search.child.ChildrenConstantScoreQuery.java

License:Apache License

@Override
public Weight createWeight(IndexSearcher searcher) throws IOException {
    SearchContext searchContext = SearchContext.current();
    BytesRefHash parentIds = new BytesRefHash(512, searchContext.pageCacheRecycler());
    ParentIdCollector collector = new ParentIdCollector(parentType, parentChildIndexFieldData, parentIds);
    final Query childQuery;
    if (rewrittenChildQuery == null) {
        childQuery = rewrittenChildQuery = searcher.rewrite(originalChildQuery);
    } else {// ww w. j  av a2 s. c  o m
        assert rewriteIndexReader == searcher.getIndexReader();
        childQuery = rewrittenChildQuery;
    }
    IndexSearcher indexSearcher = new IndexSearcher(searcher.getIndexReader());
    indexSearcher.setSimilarity(searcher.getSimilarity());
    indexSearcher.search(childQuery, collector);

    long remaining = parentIds.size();
    if (remaining == 0) {
        return Queries.newMatchNoDocsQuery().createWeight(searcher);
    }

    Filter shortCircuitFilter = null;
    if (remaining == 1) {
        BytesRef id = parentIds.get(0, new BytesRef());
        shortCircuitFilter = new TermFilter(
                new Term(UidFieldMapper.NAME, Uid.createUidAsBytes(parentType, id)));
    } else if (remaining <= shortCircuitParentDocSet) {
        shortCircuitFilter = new ParentIdsFilter(parentType, nonNestedDocsFilter, parentIds);
    }

    ParentWeight parentWeight = new ParentWeight(parentFilter, shortCircuitFilter, parentIds);
    searchContext.addReleasable(parentWeight);
    return parentWeight;
}