Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Branch: master
Commits on Jun 5, 2013
  1. Fixed bug where quotes were not being escaped when generating URL par…

    authored
    …ams.
    
    Also changed number of pagelinks from 20 to 10 at a time.
Commits on Apr 18, 2013
  1. Fix handling of url params with empty value so that they are treated

    authored
    as null in all cases.  Previously, if multiple url param values were
    given and one was empty, then the result of getParameterValues() would
    return [value,""], which would screw-up things like date filtering.
Commits on Jul 25, 2012
  1. Simplified IndexMerger by using IndexWriter.addIndexes(IndexReader...…

    authored
    …) instead of addIndex(Directory...). Even though the Lucene docs say the latter performs better, for us it's actually faster to use the former, which will merge and optimize the input indexes much faster. Also removed command-line options that no longer apply.
Commits on Jul 21, 2012
Commits on Apr 5, 2012
Commits on Mar 27, 2012
Commits on Jan 17, 2012
  1. Changed default termIndexInterval to be the default value provided by…

    authored
    … Lucene library. Also apply limit on in-memory buffer size to avoid Lucene library aborting if the value is >2GB.
Commits on Jan 5, 2012
  1. Removed. The IndexMerger tool can optimize stand-alone indexes, so no…

    authored
    … need for two very similar command-line tools.
  2. Add use of IndexWriterConfig and TieredMergePolicy so that an optimiz…

    authored
    …e can do all the segments at once rather than sub-opts/merges. Also add command-line params for termIndexInterval and verbose output.
Commits on Dec 21, 2011
  1. Use most of available memory for buffering documents during merge. Ad…

    authored
    …d -b option to control that amount.
Commits on Dec 6, 2011
  1. Added config parameter 'positionMax' which limits the results startin…

    authored
    …g position. This prevents a user from requesting the results starting at the 12,000,000th hit (or whatever limit you set). Default is to allow anything.
Commits on Sep 22, 2011
  1. Changed hitsPerPageMax default to be Integer.MAX_VALUE. Added hitsPer…

    authored
    …PageMax to MetaOpenSearchServlet.
Commits on Sep 21, 2011
Commits on Sep 16, 2011
Commits on Sep 15, 2011
  1. Fixed handling of two hits for the same site with equal scores, so

    authored
    that the hit with the lower document id is chosen.
Commits on Sep 14, 2011
  1. Added use of indexDivisor.

    authored
Commits on Sep 13, 2011
Commits on Sep 12, 2011
Commits on Aug 28, 2011
  1. Changes to score comparitor and elsewhere so that hits with the same …

    authored
    …score are sub-sorted by document id. This mirrors the way Lucene works.
Commits on Aug 9, 2011
Commits on Jul 7, 2011
Commits on Jun 13, 2011
Commits on Mar 15, 2011
  1. Added copyright header.

    authored
Commits on Feb 22, 2011
  1. Add use of resolveIndexName() to get name of index for each result. A…

    authored
    …dd the index name to the result.
  2. Changed default Wayback to point to waybackmachine.org. Default colle…

    authored
    …ction is "", but include commented-out section to select either collection or index field in results.
  3. Add license header.

    authored
Commits on Feb 16, 2011
Something went wrong with that request. Please try again.