Sorting

You can perform sorting on RAM-based retrievable fields. In the Data model property, or advanced schema index field configuration, select the RAM based option for all the fields on which you want to apply sorting.

You can specify one or several sort clauses to be treated one after the other cumulatively. For example, if our index schema has the lastname and age index fields, we could want to sort:

  • First on the lastname field by ascending order.

  • Then on the age field by descending order.

Note: Sorting can also be performed dynamically at query time using specific Search API parameters. For more information, see the Sorting and Grouping Parameters in the Search API Parameters Reference.

This task shows you how to:

See Also
About Ranking

Set Sorting in the Administration Console

  1. Go to Search > Search Logics > Sort & Relevance > Sort
  2. Click to add as many sort clauses as required.
  3. For each clause, you must specify:
    • a Name
    • a virtual Expression - click Edit and specify the expression with the Expression Builder. It can be the name of an index field only or a more complex virtual field expression.

      For example, if you want the sort to be in lowercase, use the #strlower function, to make an expression like #strlower(myAlphanumField1). If you want the sort to be normalized (that is, lowercase and unaccentuated), use the #strnormalize function, to make an expression like #strnormalize(myAlphanumField2).

    • A Sort order (ascending/descending).
    • And if the sort clause is Active, so that it is executed.

Reduce the Performance Overhead at Query-Time

Alphanumeric sorting has a high-performance overhead at query-time.

To reduce this performance issue, you can set a limit at query time using the Chars limit option, which is the maximum number of characters on which to perform the sort.

Only documents that do differ on the first limit characters are ordered.