Lucene and Solr 4.0 released
After entering beta mid-August, the final 4.0 versions of the Apache project's Lucene search engine library and Solr, the search platform built on top of Lucene, have now been released. Solr allows users to create a full-text web-accessible, dynamically clustered search engine that is capable of ingesting rich documents like Word or PDF files and indexing them for complex searching.
Powering Solr is the Lucene Core library, a Java-based indexing and search technology with support for spell checking and hit highlighting that is capable of analysing or tokenising the content it is fed with. Lucene can be plugged into projects that need those kinds of capabilities, as a standalone library.
Since August's beta, Lucene 4.0 has been enhanced with a new "Block" PostingsFormat which improves search performance and index compression such that it is likely to become the default format in the future. The developers have also moved all of Lucene's non-default codecs out into their own codecs module.
The Solr 4.0 release is focused on bringing easy scaling to Solr through the introduction of SolrCloud, which allows sharded clusters to be quickly and simply created. Solr has also had a number of features added to allow it to work as a NoSQL primary store. Lucene 4.0's focus has been on developing better indexing, faster fuzzy queries and more efficient in-memory storage.
Solr 4.0 has been optimised during the beta with performance enhancements being made to SolrClouds transaction log, sync recover, leader election and cluster state caching. There have also been improvements to Solr's Admin UI and the addition of new spatial field types with "polygon support". Both Lucene and Solr 4.0 have been the recipients of many bug fixes and optimisations in the run up to the final release.