2.0.1 ----- o Improve speed of verifying unique constraints on constraint creation. o Reduce size on disk for numeric indexes. 1.9.4 (2013-09-19) ------------------ o Fixed issue where querying an index for "*:*" while having removed entities from it in the same transaction might result in exceptions. 1.9.3 ----- o Stop keeping norms in Lucene indexes, saving space and memory 1.9.RC1 (2013-04-15) -------------------- o Lucene upgraded to 3.6.2 1.9.M05 (2013-03-05) -------------------- o LuceneIndexProvider is now replaced by LuceneKernelExtension 1.8.M05 (2012-06-21) -------------------- o Removed lucene_writers_cache_size and have lucene_searcher_cache_size decide for both, since it's doesn't make sense to have a writer w/o a searcher and isn't possible to have a searcher w/o its writer. o Loosened contention regarding getting index searchers for querying. 1.7.M01 (2012-03-12) -------------------- o Added LowerCaseKeywordAnalyzer and a clear test for it which shows how to configure a case-insensitive exact index. 1.6 (2012-01-19) ---------------- o Fixed a ConcurrentModificationException problem when rotating logs and listing store files. 1.6.M03 (2012-01-12) -------------------- o Upgraded to Lucene version 3.5. o Implemented the new Index#putIfAbsent. 1.6.M02 (2011-12-16) -------------------- o Better AddRelationshipCommand#toString and added script for dumping index logs o Added tests to ensure that the host header is used when creating link URIs in responses o Fix auto indexing bug on removal of non-loaded heavy properties 1.6.M01 (2011-11-24) -------------------- o Added LRU cache for open indexes so that a limit can be set. Thanks vivekprahlad (github id) for the patch. o When using ImpermanentGraphDatabase RAMDirectory is used instead of FSDirectory. Useful for speed during testing. o Added a LowerCaseKeywordAnalyzer to use for having an exact case-insensitive index. 1.5.M02 (2011-10-10) -------------------- o Fixed an issue where null would be returned for consecutive calls after it got an initial IllegalStateException (if snapshoting an empty index) 1.5.M01 (2011-08-31) -------------------- o Auto indexing hooks in on each operation instead of transaction events. o Fixed an issue where some not-yet-instantiated indexes wouldn't be copied during a full backup. o Fix for numeric values not being indexed properly in batch inserter. 1.4 (2011-07-08) ---------------- o Fix for numeric values not being indexed properly in fulltext indexes. 1.4.M05 (2011-06-27) -------------------- o Added the tests and documentation for auto indexing 1.4.M04 (2011-06-09) -------------------- o The LuceneBatchInserterIndex now accepts a batch size (defaults to 50K) after which it automatically commits, as a way of dealing with out-of-memory issues. 1.4.M03 (2011-05-26) -------------------- o Moved LuceneIndexProvider to public package to work better with OSGI. o Tweaked lucene batch inserter settings for better performance. 1.4.M02 (2011-05-12) -------------------- o Upgraded lucene version to 3.1 1.3.M05 (2011-03-24) -------------------- o Updated documentation o Moved QueryContext and ValueContext to be part of API 1.3.M04 (2011-03-10) -------------------- o Fixed an issue where batch inserter indexes would ignore numeric values. o Fixed an issue which could result in batch inserted indexes trying to grab too much memory. 1.3.M03 (2011-02-24) ------------------- o Fixed a bug where the use of ValueContext was ignored during batch insertion. o Added some more convenient Index#remove methods. 1.3.M02 (2011-02-10) ------------------------ o Broke neo4j-shell dependency. o Fixes a bug with the execution of RemoveCommand that could cause NPE during recovery. o Added a timeline index implementation based on lucene. o Uncommitted changes to index can now be searched within the same transaction. 0.5-1.3.M01 (2011-01-27) ------------------------ o Can participate in online backups, where the full index files are requested. o Fix for a NPE that could occur in IndexHits#size() if that was the first method called on the instance. o Fix for a recovery issue which occured when there had only been modifications to relationship indexes. o IndexHits#close() is run in the finalizer as a safety net, although best practice is to call it your self. 0.2-1.2 (2010-12-29) -------------------- o No changes. 0.2-1.2.M06 (2010-12-21) ------------------------ o More flexible caching (can instantiate an arbitrary Cache object). o Merged the fast-writes branch. Basically it makes committing changes to lucene indexes faster and more throughput-friendly for multiple threads. The performance improvement for committing lucene transactions shows best for small transactions, but will improve all modifying operations to lucene indexes. o Fixed bug which could close an IndexSearcher before it was supposed to when there was exactly one item in the result set, two or more threads did this at the same time as there was another thread doing an update... AND the reading threads all called IndexHits#getSingle() on the result. 0.2-1.2.M05 (2010-12-02) ------------------------ o Added IndexHits#currentScore which exposes the most recently fetched IndexHits item's score from the search, if any. 0.2-1.2.M04 (2010-11-18) ------------------------ o Index configuration handling moved to IndexManager (graphDb.index()) and added setConfiguration/removeConfiguration which can modify index configuration, but could render existing index data invalid if wrong settings are modified, so consider those advanced. o Made config recognition more flexible for forNodes(indexName,config) and forRelationships(indexName,config) so that the config can contain differences, just as long as they yield the same storage format. 0.2-1.2.M03 (2010-11-04) ------------------------ o Added support for modifying index configurations as well as listing and deleting indexes and index configurations. -API o Added Index#getName() and Index#getConfiguration() o Added IndexManager#nodeIndexNames() and #relationshipIndexNames() 0.2-1.2.M02 (2010-10-21) ------------------------ o Moved to be an official and supported neo4j component. o The Index API moved to kernel, this component merely provides an implementation.