org.apache.lucene.index.memory.MemoryIndex

public class MemoryIndex extends Object

High-performance single-document main memory Apache Lucene fulltext search index.

Overview

This class is a replacement/substitute for RAM-resident Directory implementations. It is designed to enable maximum efficiency for on-the-fly matchmaking combining structured and fuzzy fulltext search in realtime streaming applications such as Nux XQuery based XML message queues, publish-subscribe systems for Blogs/newsfeeds, text chat, data acquisition and distribution systems, application level routers, firewalls, classifiers, etc. Rather than targeting fulltext search of infrequent queries over huge persistent data archives (historic search), this class targets fulltext search of huge numbers of queries over comparatively small transient realtime data (prospective search). For example as in

 float score = search(String text, Query query)

Each instance can hold at most one Lucene "document", with a document containing zero or more "fields", each field having a name and a fulltext value. The fulltext value is tokenized (split and transformed) into zero or more index terms (aka words) on addField(), according to the policy implemented by an Analyzer. For example, Lucene analyzers can split on whitespace, normalize to lower case for case insensitivity, ignore common terms with little discriminatory value such as "he", "in", "and" (stop words), reduce the terms to their natural linguistic root form such as "fishing" being reduced to "fish" (stemming), resolve synonyms/inflexions/thesauri (upon indexing and/or querying), etc. For details, see Lucene Analyzer Intro.

Arbitrary Lucene queries can be run against this class - see Lucene Query Syntax as well as Query Parser Rules. Note that a Lucene query selects on the field names and associated (indexed) tokenized terms, not on the original fulltext(s) - the latter are not stored but rather thrown away immediately after tokenization.

For some interesting background information on search technology, see Bob Wyman's Prospective Search, Jim Gray's A Call to Arms - Custom subscriptions, and Tim Bray's On Search, the Series.

Example Usage

 Analyzer analyzer = new SimpleAnalyzer(version);
 MemoryIndex index = new MemoryIndex();
 index.addField("content", "Readings about Salmons and other select Alaska fishing Manuals", analyzer);
 index.addField("author", "Tales of James", analyzer);
 QueryParser parser = new QueryParser(version, "content", analyzer);
 float score = index.search(parser.parse("+author:james +salmon~ +fish* manual~"));
 if (score > 0.0f) {
     System.out.println("it's a match");
 } else {
     System.out.println("no match found");
 }
 System.out.println("indexData=" + index.toString());

Example XQuery Usage

 (: An XQuery that finds all books authored by James that have something to do with "salmon fishing manuals", sorted by relevance :)
 declare namespace lucene = "java:nux.xom.pool.FullTextUtil";
 declare variable $query := "+salmon~ +fish* manual~"; (: any arbitrary Lucene query can go here :)

 for $book in /books/book[author="James" and lucene:match(abstract, $query) > 0.0]
 let $score := lucene:match($book/abstract, $query)
 order by $score descending
 return $book

Thread safety guarantees

MemoryIndex is not normally thread-safe for adds or queries. However, queries are thread-safe after freeze() has been called.

Performance Notes

Internally there's a new data structure geared towards efficient indexing and searching, plus the necessary support code to seamlessly plug into the Lucene framework.

This class performs very well for very small texts (e.g. 10 chars) as well as for large texts (e.g. 10 MB) and everything in between. Typically, it is about 10-100 times faster than RAM-resident directory.

Note that other Directory implementations have particularly large efficiency overheads for small to medium sized texts, both in time and space. Indexing a field with N tokens takes O(N) in the best case, and O(N logN) in the worst case.

Example throughput of many simple term queries over a single MemoryIndex: ~500000 queries/sec on a MacBook Pro, jdk 1.5.0_06, server VM. As always, your mileage may vary.

If you're curious about the whereabouts of bottlenecks, run java 1.5 with the non-perturbing '-server -agentlib:hprof=cpu=samples,depth=10' flags, then study the trace log and correlate its hotspot trailer with its call stack headers (see hprof tracing ).

Constructor Summary

Constructors

Constructor

Description

MemoryIndex()

Constructs an empty instance that will not store offsets or payloads.

MemoryIndex(boolean storeOffsets)

Constructs an empty instance that can optionally store the start and end character offset of each token term in the text.

MemoryIndex(boolean storeOffsets, boolean storePayloads)

Constructs an empty instance with the option of storing offsets and payloads.
Method Summary

Modifier and Type

Method

Description

void

addField(String fieldName, String text, Analyzer analyzer)

Convenience method; Tokenizes the given field text and adds the resulting terms to the index; Equivalent to adding an indexed non-keyword Lucene Field that is tokenized, not stored, termVectorStored with positions (or termVectorStored with positions and offsets),

void

addField(String fieldName, TokenStream stream)

Iterates over the given token stream and adds the resulting terms to the index; Equivalent to adding a tokenized, indexed, termVectorStored, unstored, Lucene Field.

void

addField(String fieldName, TokenStream stream, int positionIncrementGap)

Iterates over the given token stream and adds the resulting terms to the index; Equivalent to adding a tokenized, indexed, termVectorStored, unstored, Lucene Field.

void

addField(String fieldName, TokenStream tokenStream, int positionIncrementGap, int offsetGap)

Iterates over the given token stream and adds the resulting terms to the index; Equivalent to adding a tokenized, indexed, termVectorStored, unstored, Lucene Field.

void

addField(IndexableField field, Analyzer analyzer)

Adds a lucene IndexableField to the MemoryIndex using the provided analyzer.

IndexSearcher

createSearcher()

Creates and returns a searcher that can be used to execute arbitrary Lucene queries and to collect the resulting query results as hits.

void

freeze()

Prepares the MemoryIndex for querying in a non-lazy way.

static MemoryIndex

fromDocument(Iterable<? extends IndexableField> document, Analyzer analyzer)

Builds a MemoryIndex from a lucene Document using an analyzer

static MemoryIndex

fromDocument(Iterable<? extends IndexableField> document, Analyzer analyzer, boolean storeOffsets, boolean storePayloads)

Builds a MemoryIndex from a lucene Document using an analyzer

static MemoryIndex

fromDocument(Iterable<? extends IndexableField> document, Analyzer analyzer, boolean storeOffsets, boolean storePayloads, long maxReusedBytes)

Builds a MemoryIndex from a lucene Document using an analyzer

<T> TokenStream

keywordTokenStream(Collection<T> keywords)

Convenience method; Creates and returns a token stream that generates a token for each keyword in the given collection, "as is", without any transforming text analysis.

void

reset()

Resets the MemoryIndex to its initial state and recycles all internal buffers.

float

search(Query query)

Convenience method that efficiently returns the relevance score by matching this index against the given Lucene query expression.

void

setSimilarity(Similarity similarity)

Set the Similarity to be used for calculating field norms

String

toStringDebug()

Returns a String representation of the index data for debugging purposes.

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details
- MemoryIndex
  
  public MemoryIndex()
  
  Constructs an empty instance that will not store offsets or payloads.
- MemoryIndex
  
  public MemoryIndex(boolean storeOffsets)
  
  Constructs an empty instance that can optionally store the start and end character offset of each token term in the text. This can be useful for highlighting of hit locations with the Lucene highlighter package. But it will not store payloads; use another constructor for that.
  
  Parameters:
  
  storeOffsets - whether or not to store the start and end character offset of each token term in the text
- MemoryIndex
  
  public MemoryIndex(boolean storeOffsets, boolean storePayloads)
  
  Constructs an empty instance with the option of storing offsets and payloads.
  
  Parameters:
  
  storeOffsets - store term offsets at each position
  
  storePayloads - store term payloads at each position
Method Details
- addField
  
  public void addField(String fieldName, String text, Analyzer analyzer)
  
  Convenience method; Tokenizes the given field text and adds the resulting terms to the index; Equivalent to adding an indexed non-keyword Lucene Field that is tokenized, not stored, termVectorStored with positions (or termVectorStored with positions and offsets),
  
  Parameters:
  
  fieldName - a name to be associated with the text
  
  text - the text to tokenize and index.
  
  analyzer - the analyzer to use for tokenization
- fromDocument
  
  public static MemoryIndex fromDocument(Iterable<? extends IndexableField> document, Analyzer analyzer)
  
  Builds a MemoryIndex from a lucene Document using an analyzer
  
  Parameters:
  
  document - the document to index
  
  analyzer - the analyzer to use
  
  Returns:
  
  a MemoryIndex
- fromDocument
  
  public static MemoryIndex fromDocument(Iterable<? extends IndexableField> document, Analyzer analyzer, boolean storeOffsets, boolean storePayloads)
  
  Builds a MemoryIndex from a lucene Document using an analyzer
  
  Parameters:
  
  document - the document to index
  
  analyzer - the analyzer to use
  
  storeOffsets - true if offsets should be stored
  
  storePayloads - true if payloads should be stored
  
  Returns:
  
  a MemoryIndex
- fromDocument
  
  public static MemoryIndex fromDocument(Iterable<? extends IndexableField> document, Analyzer analyzer, boolean storeOffsets, boolean storePayloads, long maxReusedBytes)
  
  Builds a MemoryIndex from a lucene Document using an analyzer
  
  Parameters:
  
  document - the document to index
  
  analyzer - the analyzer to use
  
  storeOffsets - true if offsets should be stored
  
  storePayloads - true if payloads should be stored
  
  maxReusedBytes - the number of bytes that should remain in the internal memory pools after reset() is called
  
  Returns:
  
  a MemoryIndex
- keywordTokenStream
  
  public <T> TokenStream keywordTokenStream(Collection<T> keywords)
  
  Convenience method; Creates and returns a token stream that generates a token for each keyword in the given collection, "as is", without any transforming text analysis. The resulting token stream can be fed into addField(String, TokenStream), perhaps wrapped into another TokenFilter, as desired.
  
  Parameters:
  
  keywords - the keywords to generate tokens for
  
  Returns:
  
  the corresponding token stream
- addField
  
  public void addField(IndexableField field, Analyzer analyzer)
  
  Adds a lucene IndexableField to the MemoryIndex using the provided analyzer. Also stores doc values based on IndexableFieldType.docValuesType() if set.
  
  Parameters:
  
  field - the field to add
  
  analyzer - the analyzer to use for term analysis
- addField
  
  public void addField(String fieldName, TokenStream stream)
  
  Iterates over the given token stream and adds the resulting terms to the index; Equivalent to adding a tokenized, indexed, termVectorStored, unstored, Lucene Field. Finally closes the token stream. Note that untokenized keywords can be added with this method via keywordTokenStream(Collection), the Lucene KeywordTokenizer or similar utilities.
  
  Parameters:
  
  fieldName - a name to be associated with the text
  
  stream - the token stream to retrieve tokens from.
- addField
  
  public void addField(String fieldName, TokenStream stream, int positionIncrementGap)
  
  Iterates over the given token stream and adds the resulting terms to the index; Equivalent to adding a tokenized, indexed, termVectorStored, unstored, Lucene Field. Finally closes the token stream. Note that untokenized keywords can be added with this method via keywordTokenStream(Collection), the Lucene KeywordTokenizer or similar utilities.
  
  Parameters:
  
  fieldName - a name to be associated with the text
  
  stream - the token stream to retrieve tokens from.
  
  positionIncrementGap - the position increment gap if fields with the same name are added more than once
- addField
  
  public void addField(String fieldName, TokenStream tokenStream, int positionIncrementGap, int offsetGap)
  
  Iterates over the given token stream and adds the resulting terms to the index; Equivalent to adding a tokenized, indexed, termVectorStored, unstored, Lucene Field. Finally closes the token stream. Note that untokenized keywords can be added with this method via keywordTokenStream(Collection), the Lucene KeywordTokenizer or similar utilities.
  
  Parameters:
  
  fieldName - a name to be associated with the text
  
  tokenStream - the token stream to retrieve tokens from. It's guaranteed to be closed no matter what.
  
  positionIncrementGap - the position increment gap if fields with the same name are added more than once
  
  offsetGap - the offset gap if fields with the same name are added more than once
- setSimilarity
  
  public void setSimilarity(Similarity similarity)
  
  Set the Similarity to be used for calculating field norms
  
  Parameters:
  
  similarity - instance with custom Similarity.computeNorm(org.apache.lucene.index.FieldInvertState) implementation
- createSearcher
  
  public IndexSearcher createSearcher()
  
  Creates and returns a searcher that can be used to execute arbitrary Lucene queries and to collect the resulting query results as hits.
  
  Returns:
  
  a searcher
- freeze
  
  public void freeze()
  
  Prepares the MemoryIndex for querying in a non-lazy way.
  After calling this you can query the MemoryIndex from multiple threads, but you cannot subsequently add new data.
- search
  
  public float search(Query query)
  
  Convenience method that efficiently returns the relevance score by matching this index against the given Lucene query expression.
  
  Parameters:
  
  query - an arbitrary Lucene query to run against this index
  
  Returns:
  
  the relevance score of the matchmaking; A number in the range [0.0 .. 1.0], with 0.0 indicating no match. The higher the number the better the match.
- toStringDebug
  
  public String toStringDebug()
  
  Returns a String representation of the index data for debugging purposes.
  
  Returns:
  
  the string representation
  
  WARNING: This API is experimental and might change in incompatible ways in the next release.
- reset
  
  public void reset()
  
  Resets the MemoryIndex to its initial state and recycles all internal buffers.

Class MemoryIndex

Constructor Summary

Method Summary

Methods inherited from class java.lang.Object

Constructor Details

MemoryIndex

MemoryIndex

MemoryIndex

Method Details

addField

fromDocument

fromDocument

fromDocument

keywordTokenStream

addField

addField

addField

addField

setSimilarity

createSearcher

freeze

search

toStringDebug

reset