CHANGES.txt

Lucene Change Log

$Id$

1.4.2

 1. Fixed bug #31241: Sorting could lead to incorrect results (documents
    missing, others duplicated) if the sort keys were not unique and there
    were more than 100 matches. (Daniel Naber)

 2. Memory leak in Sort code (bug #31240) eliminated.
    (Rafal Krzewski via Christoph and Daniel)
    
 3. FuzzyQuery now takes an additional parameter that specifies the
    minimum similarity that is required for a term to match the query.
    The QueryParser syntax for this is term~x, where x is a floating 
    point number between 0 and 1 (a bigger number means that a higher
    similarity is required). Furthermore, a prefix can be specified
    for FuzzyQuerys so that only those terms are considered similar that 
    start with this prefix. This can speed up FuzzyQuery greatly.
    (Daniel Naber, Christoph Goller)
    
 4. PhraseQuery and PhrasePrefixQuery now allow the explicit specification
    of relative positions. (Christoph Goller)
    
 5. QueryParser changes: Fix for ArrayIndexOutOfBoundsExceptions 
    (patch #9110); some unused method parameters removed; The ability
    to specify a minimum similarity for FuzzyQuery has been added.
    (Christoph Goller)

 6. IndexSearcher optimization: a new ScoreDoc is no longer allocated
    for every non-zero-scoring hit.  This makes 'OR' queries that
    contain common terms substantially faster.  (cutting)


1.4.1

 1. Fixed a performance bug in hit sorting code, where values were not
    correctly cached.  (Aviran via cutting)

 2. Fixed errors in file format documentation. (Daniel Naber)


1.4 final

 1. Added "an" to the list of stop words in StopAnalyzer, to complement
    the existing "a" there.  Fix for bug 28960
     (http://issues.apache.org/bugzilla/show_bug.cgi?id=28960). (Otis)

 2. Added new class FieldCache to manage in-memory caches of field term
    values.  (Tim Jones)

 3. Added overloaded getFieldQuery method to QueryParser which
    accepts the slop factor specified for the phrase (or the default
    phrase slop for the QueryParser instance).  This allows overriding
    methods to replace a PhraseQuery with a SpanNearQuery instead,
    keeping the proper slop factor. (Erik Hatcher)

 4. Changed the encoding of GermanAnalyzer.java and GermanStemmer.java to
    UTF-8 and changed the build encoding to UTF-8, to make changed files
    compile. (Otis Gospodnetic)

 5. Removed synchronization from term lookup under IndexReader methods
    termFreq(), termDocs() or termPositions() to improve
    multi-threaded performance.  (cutting)

 6. Fix a bug where obsolete segment files were not deleted on Win32.


1.4 RC3

 1. Fixed several search bugs introduced by the skipTo() changes in
    release 1.4RC1.  The index file format was changed a bit, so
    collections must be re-indexed to take advantage of the skipTo()
    optimizations.  (Christoph Goller)

 2. Added new Document methods, removeField() and removeFields().
    (Christoph Goller)

 3. Fixed inconsistencies with index closing.  Indexes and directories
    are now only closed automatically by Lucene when Lucene opened
    them automatically.  (Christoph Goller)

 4. Added new class: FilteredQuery.  (Tim Jones)

 5. Added a new SortField type for custom comparators.  (Tim Jones)

 6. Lock obtain timed out message now displays the full path to the lock
    file. (Daniel Naber via Erik)

 7. Fixed a bug in SpanNearQuery when ordered. (Paul Elschot via cutting)

 8. Fixed so that FSDirectory's locks still work when the
    java.io.tmpdir system property is null.  (cutting)

 9. Changed FilteredTermEnum's constructor to take no parameters,
    as the parameters were ignored anyway (bug #28858)

1.4 RC2

 1. GermanAnalyzer now throws an exception if the stopword file
    cannot be found (bug #27987). It now uses LowerCaseFilter
    (bug #18410) (Daniel Naber via Otis, Erik)

 2. Fixed a few bugs in the file format documentation. (cutting)


1.4 RC1

 1. Changed the format of the .tis file, so that:

    - it has a format version number, which makes it easier to
      back-compatibly change file formats in the future.

    - the term count is now stored as a long.  This was the one aspect
      of the Lucene's file formats which limited index size.

    - a few internal index parameters are now stored in the index, so
      that they can (in theory) now be changed from index to index,
      although there is not yet an API to do so.

    These changes are back compatible.  The new code can read old
    indexes.  But old code will not be able read new indexes. (cutting)

 2. Added an optimized implementation of TermDocs.skipTo().  A skip
    table is now stored for each term in the .frq file.  This only
    adds a percent or two to overall index size, but can substantially
    speedup many searches.  (cutting)

 3. Restructured the Scorer API and all Scorer implementations to take
    advantage of an optimized TermDocs.skipTo() implementation.  In
    particular, PhraseQuerys and conjunctive BooleanQuerys are
    faster when one clause has substantially fewer matches than the
    others.  (A conjunctive BooleanQuery is a BooleanQuery where all
    clauses are required.)  (cutting)

 4. Added new class ParallelMultiSearcher.  Combined with
    RemoteSearchable this makes it easy to implement distributed
    search systems.  (Jean-Francois Halleux via cutting)

 5. Added support for hit sorting.  Results may now be sorted by any
    indexed field.  For details see the javadoc for
    Searcher#search(Query, Sort).  (Tim Jones via Cutting)

 6. Changed FSDirectory to auto-create a full directory tree that it
    needs by using mkdirs() instead of mkdir().  (Mladen Turk via Otis)

 7. Added a new span-based query API.  This implements, among other
    things, nested phrases.  See javadocs for details.  (Doug Cutting)

 8. Added new method Query.getSimilarity(Searcher), and changed
    scorers to use it.  This permits one to subclass a Query class so
    that it can specify it's own Similarity implementation, perhaps
    one that delegates through that of the Searcher.  (Julien Nioche
    via Cutting)

 9. Added MultiReader, an IndexReader that combines multiple other
    IndexReaders.  (Cutting)

10. Added support for term vectors.  See Field#isTermVectorStored().
    (Grant Ingersoll, Cutting & Dmitry)

11. Fixed the old bug with escaping of special characters in query
    strings: http://issues.apache.org/bugzilla/show_bug.cgi?id=24665
    (Jean-Francois Halleux via Otis)

12. Added support for overriding default values for the following,
    using system properties:
      - default commit lock timeout
      - default maxFieldLength
      - default maxMergeDocs
      - default mergeFactor
      - default minMergeDocs
      - default write lock timeout
    (Otis)

13. Changed QueryParser.jj to allow '-' and '+' within tokens:
    http://issues.apache.org/bugzilla/show_bug.cgi?id=27491
    (Morus Walter via Otis)

14. Changed so that the compound index format is used by default.
    This makes indexing a bit slower, but vastly reduces the chances
    of file handle problems.  (Cutting)


1.3 final

 1. Added catch of BooleanQuery$TooManyClauses in QueryParser to
    throw ParseException instead. (Erik Hatcher)

 2. Fixed a NullPointerException in Query.explain(). (Doug Cutting)

 3. Added a new method IndexReader.setNorm(), that permits one to
    alter the boosting of fields after an index is created.

 4. Distinguish between the final position and length when indexing a
    field.  The length is now defined as the total number of tokens,
    instead of the final position, as it was previously.  Length is
    used for score normalization (Similarity.lengthNorm()) and for
    controlling memory usage (IndexWriter.maxFieldLength).  In both of
    these cases, the total number of tokens is a better value to use
    than the final token position.  Position is used in phrase
    searching (see PhraseQuery and Token.setPositionIncrement()).

 5. Fix StandardTokenizer's handling of CJK characters (Chinese,
    Japanese and Korean ideograms).  Previously contiguous sequences
    were combined in a single token, which is not very useful.  Now
    each ideogram generates a separate token, which is more useful.


1.3 RC3

 1. Added minMergeDocs in IndexWriter.  This can be raised to speed
    indexing without altering the number of files, but only using more
    memory.  (Julien Nioche via Otis)

 2. Fix bug #24786, in query rewriting. (bschneeman via Cutting)

 3. Fix bug #16952, in demo HTML parser, skip comments in
    javascript. (Christoph Goller)

 4. Fix bug #19253, in demo HTML parser, add whitespace as needed to
    output (Daniel Naber via Christoph Goller)

 5. Fix bug #24301, in demo HTML parser, long titles no longer
    hang things. (Christoph Goller)

 6. Fix bug #23534, Replace use of file timestamp of segments file
    with an index version number stored in the segments file.  This
    resolves problems when running on file systems with low-resolution
    timestamps, e.g., HFS under MacOS X.  (Christoph Goller)

 7. Fix QueryParser so that TokenMgrError is not thrown, only
    ParseException.  (Erik Hatcher)

 8. Fix some bugs introduced by change 11 of RC2.  (Christoph Goller)

 9. Fixed a problem compiling TestRussianStem.  (Christoph Goller)

10. Cleaned up some build stuff.  (Erik Hatcher)


1.3 RC2

 1. Added getFieldNames(boolean) to IndexReader, SegmentReader, and
    SegmentsReader. (Julien Nioche via otis)

 2. Changed file locking to place lock files in
    System.getProperty("java.io.tmpdir"), where all users are
    permitted to write files.  This way folks can open and correctly
    lock indexes which are read-only to them.

 3. IndexWriter: added a new method, addDocument(Document, Analyzer),
    permitting one to easily use different analyzers for different
    documents in the same index.

 4. Minor enhancements to FuzzyTermEnum.
    (Christoph Goller via Otis)

 5. PriorityQueue: added insert(Object) method and adjusted IndexSearcher
    and MultiIndexSearcher to use it.
    (Christoph Goller via Otis)

 6. Fixed a bug in IndexWriter that returned incorrect docCount().
    (Christoph Goller via Otis)

 7. Fixed SegmentsReader to eliminate the confusing and slightly different
    behaviour of TermEnum when dealing with an enumeration of all terms,
    versus an enumeration starting from a specific term.
    This patch also fixes incorrect term document frequences when the same term
    is present in multiple segments.
    (Christoph Goller via Otis)

 8. Added CachingWrapperFilter and PerFieldAnalyzerWrapper. (Erik Hatcher)

 9. Added support for the new "compound file" index format (Dmitry
    Serebrennikov)

10. Added Locale setting to QueryParser, for use by date range parsing.

11. Changed IndexReader so that it can be subclassed by classes
    outside of its package.  Previously it had package-private
    abstract methods.  Also modified the index merging code so that it
    can work on an arbitrary IndexReader implementation, and added a
    new method, IndexWriter.addIndexes(IndexReader[]), to take
    advantage of this. (cutting)

12. Added a limit to the number of clauses which may be added to a
    BooleanQuery.  The default limit is 1024 clauses.  This should
    stop most OutOfMemoryExceptions by prefix, wildcard and fuzzy
    queries which run amok. (cutting)

13. Add new method: IndexReader.undeleteAll().  This undeletes all
    deleted documents which still remain in the index. (cutting)


1.3 RC1

 1. Fixed PriorityQueue's clear() method.
    Fix for bug 9454, http://nagoya.apache.org/bugzilla/show_bug.cgi?id=9454
    (Matthijs Bomhoff via otis)

 2. Changed StandardTokenizer.jj grammar for EMAIL tokens.
    Fix for bug 9015, http://nagoya.apache.org/bugzilla/show_bug.cgi?id=9015
    (Dale Anson via otis)

 3. Added the ability to disable lock creation by using disableLuceneLocks
    system property.  This is useful for read-only media, such as CD-ROMs.
    (otis)

 4. Added id method to Hits to be able to access the index global id.
    Required for sorting options.
    (carlson)

 5. Added support for new range query syntax to QueryParser.jj.
    (briangoetz)

 6. Added the ability to retrieve HTML documents' META tag values to
    HTMLParser.jj.
    (Mark Harwood via otis)

 7. Modified QueryParser to make it possible to programmatically specify the
    default Boolean operator (OR or AND).
    (Péter Halácsy via otis)

 8. Made many search methods and classes non-final, per requests.
    This includes IndexWriter and IndexSearcher, among others.
    (cutting)

 9. Added class RemoteSearchable, providing support for remote
    searching via RMI.  The test class RemoteSearchableTest.java
    provides an example of how this can be used.  (cutting)

 10. Added PhrasePrefixQuery (and supporting MultipleTermPositions).  The
     test class TestPhrasePrefixQuery provides the usage example.
     (Anders Nielsen via otis)

 11. Changed the German stemming algorithm to ignore case while
     stripping. The new algorithm is faster and produces more equal
     stems from nouns and verbs derived from the same word.
     (gschwarz)

 12. Added support for boosting the score of documents and fields via
     the new methods Document.setBoost(float) and Field.setBoost(float).

     Note: This changes the encoding of an indexed value.  Indexes
     should be re-created from scratch in order for search scores to
     be correct.  With the new code and an old index, searches will
     yield very large scores for shorter fields, and very small scores
     for longer fields.  Once the index is re-created, scores will be
     as before. (cutting)

 13. Added new method Token.setPositionIncrement().

     This permits, for the purpose of phrase searching, placing
     multiple terms in a single position.  This is useful with
     stemmers that produce multiple possible stems for a word.

     This also permits the introduction of gaps between terms, so that
     terms which are adjacent in a token stream will not be matched by
     and exact phrase query.  This makes it possible, e.g., to build
     an analyzer where phrases are not matched over stop words which
     have been removed.

     Finally, repeating a token with an increment of zero can also be
     used to boost scores of matches on that token.  (cutting)

 14. Added new Filter class, QueryFilter.  This constrains search
     results to only match those which also match a provided query.
     Results are cached, so that searches after the first on the same
     index using this filter are very fast.

     This could be used, for example, with a RangeQuery on a formatted
     date field to implement date filtering.  One could re-use a
     single QueryFilter that matches, e.g., only documents modified
     within the last week.  The QueryFilter and RangeQuery would only
     need to be reconstructed once per day. (cutting)

 15. Added a new IndexWriter method, getAnalyzer().  This returns the
     analyzer used when adding documents to this index. (cutting)

 16. Fixed a bug with IndexReader.lastModified().  Before, document
     deletion did not update this.  Now it does.  (cutting)

 17. Added Russian Analyzer.
     (Boris Okner via otis)

 18. Added a public, extensible scoring API.  For details, see the
     javadoc for org.apache.lucene.search.Similarity.

 19. Fixed return of Hits.id() from float to int. (Terry Steichen via Peter).

 20. Added getFieldNames() to IndexReader and Segment(s)Reader classes.
     (Peter Mularien via otis)

 21. Added getFields(String) and getValues(String) methods.
     Contributed by Rasik Pandey on 2002-10-09
     (Rasik Pandey via otis)

 22. Revised internal search APIs.  Changes include:

       a. Queries are no longer modified during a search.  This makes
       it possible, e.g., to reuse the same query instance with
       multiple indexes from multiple threads.

       b. Term-expanding queries (e.g. PrefixQuery, WildcardQuery,
       etc.)  now work correctly with MultiSearcher, fixing bugs 12619
       and 12667.

       c. Boosting BooleanQuery's now works, and is supported by the
       query parser (problem reported by Lee Mallabone).  Thus a query
       like "(+foo +bar)^2 +baz" is now supported and equivalent to
       "(+foo^2 +bar^2) +baz".

       d. New method: Query.rewrite(IndexReader).  This permits a
       query to re-write itself as an alternate, more primitive query.
       Most of the term-expanding query classes (PrefixQuery,
       WildcardQuery, etc.) are now implemented using this method.

       e. New method: Searchable.explain(Query q, int doc).  This
       returns an Explanation instance that describes how a particular
       document is scored against a query.  An explanation can be
       displayed as either plain text, with the toString() method, or
       as HTML, with the toHtml() method.  Note that computing an
       explanation is as expensive as executing the query over the
       entire index.  This is intended to be used in developing
       Similarity implementations, and, for good performance, should
       not be displayed with every hit.

       f. Scorer and Weight are public, not package protected.  It now
       possible for someone to write a Scorer implementation that is
       not in the org.apache.lucene.search package.  This is still
       fairly advanced programming, and I don't expect anyone to do
       this anytime soon, but at least now it is possible.

       g. Added public accessors to the primitive query classes
       (TermQuery, PhraseQuery and BooleanQuery), permitting access to
       their terms and clauses.

     Caution: These are extensive changes and they have not yet been
     tested extensively.  Bug reports are appreciated.
     (cutting)

 23. Added convenience RAMDirectory constructors taking File and String
     arguments, for easy FSDirectory to RAMDirectory conversion.
     (otis)

 24. Added code for manual renaming of files in FSDirectory, since it
     has been reported that java.io.File's renameTo(File) method sometimes
     fails on Windows JVMs.
     (Matt Tucker via otis)

 25. Refactored QueryParser to make it easier for people to extend it.
     Added the ability to automatically lower-case Wildcard terms in
     the QueryParser.
     (Tatu Saloranta via otis)


1.2 RC6

 1. Changed QueryParser.jj to have "?" be a special character which
    allowed it to be used as a wildcard term. Updated TestWildcard
    unit test also. (Ralf Hettesheimer via carlson)

1.2 RC5

 1. Renamed build.properties to default.properties and updated
    the BUILD.txt document to describe how to override the
    default.property settings without having to edit the file. This
    brings the build process closer to Scarab's build process.
    (jon)

 2. Added MultiFieldQueryParser class. (Kelvin Tan, via otis)

 3. Updated "powered by" links. (otis)

 4. Fixed instruction for setting up JavaCC - Bug #7017 (otis)

 5. Added throwing exception if FSDirectory could not create diectory
    - Bug #6914 (Eugene Gluzberg via otis)

 6. Update MultiSearcher, MultiFieldParse, Constants, DateFilter,
    LowerCaseTokenizer javadoc (otis)

 7. Added fix to avoid NullPointerException in results.jsp
    (Mark Hayes via otis)

 8. Changed Wildcard search to find 0 or more char instead of 1 or more
    (Lee Mallobone, via otis)

 9. Fixed error in offset issue in GermanStemFilter - Bug #7412
    (Rodrigo Reyes, via otis)

 10. Added unit tests for wildcard search and DateFilter (otis)

 11. Allow co-existence of indexed and non-indexed fields with the same name
     (cutting/casper, via otis)

 12. Add escape character to query parser.
     (briangoetz)

 13. Applied a patch that ensures that searches that use DateFilter
     don't throw an exception when no matches are found. (David Smiley, via
     otis)

 14. Fixed bugs in DateFilter and wildcardquery unit tests. (cutting, otis, carlson)


1.2 RC4

 1. Updated contributions section of website.
    Add XML Document #3 implementation to Document Section.
    Also added Term Highlighting to Misc Section. (carlson)

 2. Fixed NullPointerException for phrase searches containing
    unindexed terms, introduced in 1.2RC3.  (cutting)

 3. Changed document deletion code to obtain the index write lock,
    enforcing the fact that document addition and deletion cannot be
    performed concurrently.  (cutting)

 4. Various documentation cleanups.  (otis, acoliver)

 5. Updated "powered by" links.  (cutting, jon)

 6. Fixed a bug in the GermanStemmer.  (Bernhard Messer, via otis)

 7. Changed Term and Query to implement Serializable.  (scottganyo)

 8. Fixed to never delete indexes added with IndexWriter.addIndexes().
    (cutting)

 9. Upgraded to JUnit 3.7. (otis)

1.2 RC3

 1. IndexWriter: fixed a bug where adding an optimized index to an
    empty index failed.  This was encountered using addIndexes to copy
    a RAMDirectory index to an FSDirectory.

 2. RAMDirectory: fixed a bug where RAMInputStream could not read
    across more than across a single buffer boundary.

 3. Fix query parser so it accepts queries with unicode characters.
    (briangoetz)

 4. Fix query parser so that PrefixQuery is used in preference to
    WildcardQuery when there's only an asterisk at the end of the
    term.  Previously PrefixQuery would never be used.

 5. Fix tests so they compile; fix ant file so it compiles tests
    properly.  Added test cases for Analyzers and PriorityQueue.

 6. Updated demos, added Getting Started documentation. (acoliver)

 7. Added 'contributions' section to website & docs. (carlson)

 8. Removed JavaCC from source distribution for copyright reasons.
    Folks must now download this separately from metamata in order to
    compile Lucene.  (cutting)

 9. Substantially improved the performance of DateFilter by adding the
    ability to reuse TermDocs objects.  (cutting)

10. Added IndexReader methods:
      public static boolean indexExists(String directory);
      public static boolean indexExists(File directory);
      public static boolean indexExists(Directory directory);
      public static boolean isLocked(Directory directory);
      public static void unlock(Directory directory);
    (cutting, otis)

11. Fixed bugs in GermanAnalyzer (gschwarz)


1.2 RC2, 19 October 2001:
 - added sources to distribution
 - removed broken build scripts and libraries from distribution
 - SegmentsReader: fixed potential race condition
 - FSDirectory: fixed so that getDirectory(xxx,true) correctly
   erases the directory contents, even when the directory
   has already been accessed in this JVM.
 - RangeQuery: Fix issue where an inclusive range query would
   include the nearest term in the index above a non-existant
   specified upper term.
 - SegmentTermEnum: Fix NullPointerException in clone() method
   when the Term is null.
 - JDK 1.1 compatibility fix: disabled lock files for JDK 1.1,
   since they rely on a feature added in JDK 1.2.

1.2 RC1 (first Apache release), 2 October 2001:
  - packages renamed from com.lucene to org.apache.lucene
  - license switched from LGPL to Apache
  - ant-only build -- no more makefiles
  - addition of lock files--now fully thread & process safe
  - addition of German stemmer
  - MultiSearcher now supports low-level search API
  - added RangeQuery, for term-range searching
  - Analyzers can choose tokenizer based on field name
  - misc bug fixes.

1.01b (last Sourceforge release), 2 July 2001
 . a few bug fixes
 . new Query Parser
 . new prefix query (search for "foo*" matches "food")

1.0, 2000-10-04

This release fixes a few serious bugs and also includes some
performance optimizations, a stemmer, and a few other minor
enhancements.

0.04 2000-04-19

Lucene now includes a grammar-based tokenizer, StandardTokenizer.

The only tokenizer included in the previous release (LetterTokenizer)
identified terms consisting entirely of alphabetic characters.  The
new tokenizer uses a regular-expression grammar to identify more
complex classes of terms, including numbers, acronyms, email
addresses, etc.

StandardTokenizer serves two purposes:

 1. It is a much better, general purpose tokenizer for use by
    applications as is.

    The easiest way for applications to start using
    StandardTokenizer is to use StandardAnalyzer.

 2. It provides a good example of grammar-based tokenization.

    If an application has special tokenization requirements, it can
    implement a custom tokenizer by copying the directory containing
    the new tokenizer into the application and modifying it
    accordingly.

0.01, 2000-03-30

First open source release.

The code has been re-organized into a new package and directory
structure for this release.  It builds OK, but has not been tested
beyond that since the re-organization.