The picturesafe team proudly announces the release of picturesafe-search 3.3.0!

This release brings the ‘sort by relevance’ feature and some improvements in indexing and analysis. There are also some minor bugfixes in this release.

picturesafe-search is an open source Java framework that allows for ultra fast adoption of Elasticsearch queries capabilities to old and new java applications.

You can find out more about picturesafe-search at

We use semantic versioning for the development of picturesafe-search.

Using this release

With Apache Maven add this dependency to your pom.xml:


Or clone the GitHub repository and build the distribution yourself.

Have fun!

— The picturesafe-search team



  • Fixed MustNotExpression.optimize()

New Feature

  • Added IndexPresetConfigurationProvider to provide the possibility to load index preset configurations dynamically.
  • Added sort by relevance feature to SortOption.


  • Refactored the ID type. Now you can use any kind of object as an ID for index objects.
  • Added elasticsearch type OBJECT and attribute withoutIndexing to field configuration.
  • Improved the configuration of the default analyzer.
  • Improved the configuration of custom analyzers, tokenizers and filters.

Breaking changes

  • Constructors of ElasticsearchServiceImpl and SingleIndexElasticsearchServiceImpl have changed
  • ID type has changed to String internally (as in Elasticsearch) and method signatures now expect Object (which is converted to String by IdFormat implementations). This should not break code compatibility unless you do not use the DocumentBuilder and build the document maps yourself.
  • FieldConfiguration interface and implementations have changed: getCharMappings() has been replaced by getCustomCharFilters().