Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
tree: 8f30621f8e

Fetching latest commit…

Cannot retrieve the latest commit at this time

..
Failed to load latest commit information.
.externalToolBuilders
org/wikimedia/lsearch
.classpath
.cvsignore
.project
Makefile
README.txt
build.xml
buildindex
mwsearch.conf.example

README.txt

  Lucene Search extension for MediaWiki
  =====================================

Note: This is at best beta-quality code.  Be aware that it may contain bugs
and not work as intended. If you want to use it, please report any issues 
to the authors (see end of file).

To use:

  1. Compile the sources using Eclipse or Ant, or download the binary
     distribution [not yet available].
  2. Build the initial search index:
  
       java -Djdbc.drivers=com.mysql.jdbc.Driver org.wikimedia.lsearch.MWSearch 
               -rebuild jdbc:mysql://192.168.0.160/searchdb
  
  3. Copy mwsearch.conf.example to mwsearch.conf and modify as appropriate.
  
  4. Start the MWDaemon:
  
       java org.wikimedia.lsearch.MWDaemon
  
  5. Install the LuceneSearch extension from the `extensions' module in your
     MediaWiki.  Make sure $wgDisableInternalSearch is enabled.

Requirements:
* Jakarta Lucene: http://lucene.apache.org/
* BerkeleyDB Java Edition: http://www.sleepycat.com/
  (only required if you use TitlePrefixMatcher).
* MySQL Connector/J: http://www.mysql.com/products/connector/j/
* MediaWiki 1.4 or 1.5

Should now work with Java 1.4 as well as 1.5.
Testing w/GCJ has been done, and except for the prefix matcher, it appears to work.
(BerkeleyDB does not work under GCJ at the moment, and the matcher requires it).

GCJ 4.0 is required for the native build.

Todo:
* Speed up index building
* Incremental index updates
* Fix title prefix matcher for GCJ

Send feedback to wikitech-l@mail.wikimedia.org.
Something went wrong with that request. Please try again.