This is a major release of indx, with lots of new features:

• A new JSON engine for structured data
  • Built in analysis of structure and data types
  • Supports nesting
  • Loads the entire JSON file for quick lookups
• Field based indexing
  • Weight profiles
  • Automatic segmentation of large bodies of text such as description fields
• New Filter method
  • ValueFilter to match a field. Supports arrays.
  • RangeFilter for numbers
• Facets
  • Histogram sorted by frequency
  • Supports arrays
  • Supports “empty search” to give facets on the entire dataset
• Boosting
  • Boost documents with any Filter. Powerful scoring method that is easy to learn and maintain.
  • Supports personalised boost lists for single user or user groups.
• Sorting
  • Second order sorting (respects search relevancy first) on every field type. Ascending or descending.
• Empty search
  • Retrieve the entire document set with facets. In the order of the json file, or sorted by a field. Supports boosting.
• Deep search
  • Set up coverage to work on the entire dataset.

Other changes

• Performance improvements on the core engine and coverage function
• .NET 9+. Planned .NET 10 LTS support when available.

Notes

• The Indx hosted RestAPI offer will be discontinued in favor of managed services.

• Indx is still an entire search engine built into a C# library. The only external dependency in use is https://www.nuget.org/packages/Newtonsoft.Json/. Indx v4 remains a lightweight search engine with minimal resource usage.

• Indx v4 can still be used with the core workflow of v3. See the migration guide.

  v3 → v4 migration guide