1.25 - Mon Sep 5 09:24:00 UTC 2005

• Stop QueryParser turning 'notice' into 'NOT ice' (RT#14450)

1.24 - Wed 20 July 2005

• use pack and unpack to read/write frq and prx files [Minty]
• Small doc changes [Minty / TB]

1.23 - Sun 17 July 2005

• Stop rolling our own sorting in SegmentMerger/merge_term_infos [Minty]
• Replace sequence of read_byte() calls in SegmentReader with one direct call to read() [Aaron Crane]
• add a fast clone() method to Plucene::Index::TermInfo [TB]
  • and use it in preference to copy_in() [Minty]
• close the file at DESTROY in Input/Output Stream classes [Minty]
• Improve POD for Document::Field [Geoff Richards]
• minor POD fixes [Minty / TB]
• add indexing benchmark script [TB]

1.22 - Tue Jun 28 2005

• perl 5.8 required
• use pack and unpack to read/write 'segments' files in SegmentInfos [TB]
• Use string xor to find longest common string prefix in TermInfosWriter [Minty / Nicholas Clark]
• switch to direct hash access in comparison methods in Plucene::Index::Term [Minty]
• remove some indirection in SegmentTermEnum [TB]
• remove Plucene::Index::FieldInfo file [TB]
• minor speed tweaks to SegmentMergeInfo, SegmentTermPositions [TB]
• minor speed tweaks to SegmentsTermEnum (and split out into own file) [TB]
• remove Tie::Array::Sorted from SegmentMerger [TB/MP]

1.21 - Mon 24 Jan 2005

• use pack and unpack to read/write FieldInfos files (.fnm). This breaks index compatability with Lucene, as we now use BER compressed integers (but we never really had index compatability anyway). It will also break index compatability with older versions of Plucene if you have documents with more than 127 fields. Hopefully no-one actually does. [MP / TB]
• localise $\ in Plucene::Store::OutputStream so -l works [Chris Nandor]
• Removed needless new() method from FieldInfo for speed [TB]
• Speed up FieldsInfo::addinternal [TB]
• explicitly close lock files immediately after creating them [MP]
• remove all remaining catfile() calls [TB]
• tweak docs for skipping_doc() [TB]
• lots of tiny micro-optimisations [TB]

1.20 - Wed 15 Dec 2004

• Just upped the version number, to fix some issues with CGI::Kwiki

1.19 - Wed 21 July 2004

• Removed accidentally-included experimental patches

1.18 - Tue 20 July 2004

• Downgraded dependencies to work with Perl 5.6.1
• Fix to delete_term

1.17 - Tue 27 Apr 2004

• Plucene::Search::Hits::hit_doc should call getmore_hits not get_more_docs (thanks to Kjetil Fikkan for pointing it out)

1.16 - Mon 12 Apr 2004

• not released

1.15 - Mon 29 Mar 2004

• Remove some needless prereqs.
• Removal of Class::HasA completely
• Remove use of 'catfile'

1.14 - thu Mar 25 2004

• Put Encode::compat in a BEGIN block in the Store classes.
• Change the Build.PL to need Encode::compat if < 5.6

1.13 - Thu Mar 25 2004

• Remove Encode::compat.

1.12 - Tue Mar 23 2004

• Put back Encode::compat

1.11 - Tue Mar 23 2004

• More speed up patches (thanks to Marty Pauley and Tony Bowden)

1.10 - Thu Mar 18 2004

• Include the test left out for the bug in 1.08

1.09 - Wed Mar 17 2004

• Now allows documents of id 0 to be deleted correctly! (Thanks to Steve Rushe for pointing this out)

1.08 - Wed Mar 17 2004

• Some speed ups in InputStream (thanks to Marty Pauley)

1.07 - Web Feb 18 2004

• Allowed the changing of mergefactor

1.06 - Tue Feb 17 2004

• Checked $] in Build.pl to decide if to use Encode::compat
• Added create_makefile_pl => 'traditional' to Build.PL
• moved from dist_name to module_name to stop warning when running Makefile.PL

1.05 - Fri Feb 13 2004

• Remove commented out line in Plucene::Store::InputStream
• Removed hardcoded /tmp index directories, moved to File::Temp
• Removed system() calls from tests

1.04 - Fri Feb 06 2004

• Doc change to clarify what optimize() does to a writer
• Removed tell call from Plucene::Store::InputStream::read_byte

1.03 - Thu Feb 05 2004

• Added in use of Enocde::compat to ensure works with < 5.7 (thanks to Simon Cozens)
• Added Plucene::Index::IndexSearcher::close()

1.02 - Wed Feb 04 2004

• Speed up sort (thanks to Tony Bowden)
• Move from Class::Accessor to Class::Accessor::Fast (thank to Tony again)
• StandardAnalyzer should be a subclass of Analyzer!
• StandardTokenizer now tokenizes correctly
• IndexSearcher should call TopDocs with a hashref and named keys if no scorer.

1.01 - Tue Feb 03 2004

• Doc fixes to Plucene.pm, bin/dump_index and README
• Remove Lucene::QueryParser from the tests
• Added this Changes file.

1.0 - Mon Feb 02 2004

• Initial release to CPAN