Search overview

Arkindex provides an optional search feature. It is disabled by default and requires some specific configuration to work. When it is set up, projects may be selectively indexed into the search engine, allowing users to search through elements, classifications, metadata, transcriptions and entities all at once.

When search is enabled, a Search button is available in the main navigation bar. Searching is also available without logging in, but only for public projects.

Search interface

Searching is distinct from filtering during navigation:

  • Search queries have some margin of error and use data from the separate search engine. The results may be incomplete, out of date, or may not be exactly what you were looking for, but they can help in exploring data.

  • Filters are less flexible, but more precise and reliable. They provide exact results from the main database and are best for exploiting data.

Enabling search on an instance

To enable the search features, system administrators must:

  1. set up an Apache Solr instance;

  2. configure Arkindex to use this Solr instance;

  3. enable the search feature flag.

Once search is available on an instance, system administrators will need to prepare projects to be indexed.

Solr may require significantly more RAM in production than other components of an Arkindex instance.