io.lettuce.core.search

Class AggregationReply<K,V>

java.lang.Object

io.lettuce.core.search.AggregationReply<K,V>

Type Parameters:

K - the type of keys used in the aggregation results

V - the type of values used in the aggregation results

public class AggregationReply<K,V>
extends Object

Represents the response from a Redis Search aggregation command (FT.AGGREGATE) or an (FT.CURSOR READ) command. This class encapsulates the results of aggregation operations including grouped data, statistical computations, and cursor-based pagination for large result sets.

An aggregation reply contains:

• The number of aggregation groups returned
• A list of SearchReply objects, each representing an aggregation group or result set
• An optional cursor ID for pagination when dealing with large aggregation results

Aggregation operations can include:

• Grouping by one or more fields
• Statistical functions (COUNT, SUM, AVG, MIN, MAX, etc.)
• Sorting and limiting results
• Filtering and transformations

For cursor-based pagination, when the result set is too large to return in a single response, Redis returns a cursor ID that can be used with FT.CURSOR READ to retrieve subsequent pages.

Since:

6.8

Author:

Redis Ltd.

See Also:

SearchReply

Constructor Summary

Constructors

Constructor and Description
AggregationReply() Creates a new empty AggregationReply.

Method Summary

Modifier and Type	Method and Description
long	getAggregationGroups() Returns the number of aggregation groups in this reply.
long	getCursorId() Returns the cursor ID for pagination, if applicable.
List<SearchReply<K,V>>	getReplies() Returns the list of search replies containing the aggregation results.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

AggregationReply

public AggregationReply()

Creates a new empty AggregationReply. The reply is initialized with defaults.

Method Detail

getAggregationGroups

public long getAggregationGroups()

Returns the number of aggregation groups in this reply.

This value represents:

• For grouped aggregations: the number of distinct groups returned
• For non-grouped aggregations: typically 1, representing the entire result set
• For empty results: may be 0 or 1 depending on the aggregation type

Note: This count may be different from getReplies().size() in some cases, particularly when dealing with cursor-based pagination where not all groups are returned in a single response.

Returns:

the number of aggregation groups, typically a positive integer

getReplies

public List<SearchReply<K,V>> getReplies()

Returns the list of search replies containing the aggregation results.

Each SearchReply in the list represents:

• For grouped aggregations: one aggregation group with its computed values
• For non-grouped aggregations: typically a single reply containing all results
• For cursor-based results: the current page of results

The structure of each SearchReply depends on the aggregation operations performed:

• GROUP BY operations create separate replies for each group
• REDUCE operations add computed fields to each reply
• LOAD operations include specified fields in the results
• SORTBY operations determine the order of replies

The returned list is mutable and reflects the current state of the aggregation results. Modifying this list will affect the aggregation reply.

Returns:

a mutable list of SearchReply objects containing the aggregation results. Never null, but may be empty if no results were found.

getCursorId

public long getCursorId()

Returns the cursor ID for pagination, if applicable.

The cursor ID is used for paginating through large aggregation result sets that cannot be returned in a single response. When Redis returns a cursor ID, it indicates that there are more results available that can be retrieved using the FT.CURSOR READ command.

Cursor behavior:

• Returns -1 (NO_CURSOR) when no pagination is needed or available
• Returns a positive integer when more results are available
• Returns 0 when this is the last page of a paginated result set

To retrieve the next page of results, use the returned cursor ID with the FT.CURSOR READ command. Continue reading until the cursor ID becomes 0, indicating the end of the result set.

Note: Cursors have a timeout and will expire if not used within the configured time limit. Always check for cursor expiration when implementing pagination.

Returns:

the cursor ID for pagination. Returns -1 if no cursor is available, 0 if this is the last page, or a positive integer if more results are available.