Package ai.preferred.cerebro.index.search.structure

Class LuIndexSearcher

  • All Implemented Interfaces:
    VersatileSearcher
    public class LuIndexSearcher
extends org.apache.lucene.search.IndexSearcher
implements VersatileSearcher
    Inherited from Lucene's IndexSearcher, this class extends Lucene's traditional full-text search to vector similarity search also. As such it shares almost all of Lucene's IndexSearcher; from thread-safety to I/O speed. Use it as you would use a Lucene searcher. Plus, it now supports vector similarity search via queryVector(double[], int).

    • Nested Class Summary

      • Nested classes/interfaces inherited from class org.apache.lucene.search.IndexSearcher

        org.apache.lucene.search.IndexSearcher.LeafSlice

    • Field Summary

      Fields 
      Modifier and Type Field Description
      protected java.util.concurrent.ExecutorService executor  
      protected org.apache.lucene.search.IndexSearcher.LeafSlice[] leafSlices  
      protected org.apache.lucene.index.IndexReader reader  

      • Fields inherited from class org.apache.lucene.search.IndexSearcher

        leafContexts, readerContext

    • Constructor Summary

      Constructors 
      Constructor Description
      LuIndexSearcher​(org.apache.lucene.index.IndexReaderContext context, java.util.concurrent.ExecutorService executor, java.lang.String splitVecPath)  
      LuIndexSearcher​(org.apache.lucene.index.IndexReader r, java.lang.String splitVecPath)
      Create a searcher from the provided index and set of hashing vectors.
      LuIndexSearcher​(org.apache.lucene.index.IndexReader r, java.util.concurrent.ExecutorService executor, java.lang.String splitVecPath)
      Runs searches for each segment separately, using the provided ExecutorService.

    • Method Summary

      All Methods Instance Methods Concrete Methods 
      Modifier and Type Method Description
      org.apache.lucene.search.ScoreDoc[] queryKeyWord​(org.apache.lucene.queryparser.classic.QueryParser queryParser, java.lang.String sQuery, int resultSize)  
      org.apache.lucene.search.ScoreDoc[] queryVector​(double[] vQuery, int resultSize)  

      • Methods inherited from class org.apache.lucene.search.IndexSearcher

        collectionStatistics, count, createNormalizedWeight, createWeight, doc, doc, doc, explain, explain, getDefaultQueryCache, getDefaultQueryCachingPolicy, getDefaultSimilarity, getIndexReader, getQueryCache, getQueryCachingPolicy, getSimilarity, getSlices, getTopReaderContext, rewrite, search, search, search, search, search, search, searchAfter, searchAfter, searchAfter, setDefaultQueryCache, setDefaultQueryCachingPolicy, setQueryCache, setQueryCachingPolicy, setSimilarity, slices, termStatistics, toString

      • Methods inherited from class java.lang.Object

        clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

    • Field Detail

      • executor

        protected final java.util.concurrent.ExecutorService executor

      • leafSlices

        protected final org.apache.lucene.search.IndexSearcher.LeafSlice[] leafSlices

      • reader

        protected org.apache.lucene.index.IndexReader reader

    • Constructor Detail

      • LuIndexSearcher

        public LuIndexSearcher​(org.apache.lucene.index.IndexReader r,
                       java.lang.String splitVecPath)
                throws java.io.IOException
        Create a searcher from the provided index and set of hashing vectors.
        Throws:
        java.io.IOException

      • LuIndexSearcher

        public LuIndexSearcher​(org.apache.lucene.index.IndexReader r,
                       java.util.concurrent.ExecutorService executor,
                       java.lang.String splitVecPath)
                throws java.io.IOException
        Runs searches for each segment separately, using the provided ExecutorService. IndexSearcher will not close/awaitTermination this ExecutorService on close; you must do so, eventually, on your own. NOTE: if you are using NIOFSDirectory, do not use the shutdownNow method of ExecutorService as this uses Thread.interrupt under-the-hood which can silently close file descriptors (see LUCENE-2239).
        Throws:
        java.io.IOException

      • LuIndexSearcher

        public LuIndexSearcher​(org.apache.lucene.index.IndexReaderContext context,
                       java.util.concurrent.ExecutorService executor,
                       java.lang.String splitVecPath)
                throws java.io.IOException
        Throws:
        java.io.IOException

    • Method Detail

      • queryKeyWord

        public org.apache.lucene.search.ScoreDoc[] queryKeyWord​(org.apache.lucene.queryparser.classic.QueryParser queryParser,
                                                        java.lang.String sQuery,
                                                        int resultSize)
                                                 throws java.lang.Exception
        Specified by:
        queryKeyWord in interface VersatileSearcher
        Parameters:
        queryParser - if null the searcher will by default carry search on the field named IndexConst.CONTENTS.
        sQuery -
        resultSize -
        Returns:
        A set of ScoreDoc of Document matching with the query.
        Throws:
        java.lang.Exception

      • queryVector

        public org.apache.lucene.search.ScoreDoc[] queryVector​(double[] vQuery,
                                                       int resultSize)
                                                throws java.lang.Exception
        Specified by:
        queryVector in interface VersatileSearcher
        Parameters:
        vQuery -
        resultSize -
        Returns:
        A set of ScoreDoc of Document having latent vector producing. the highest inner product with the query vector.
        Throws:
        java.lang.Exception