Class WebSearchContentRetriever

java.lang.Object

dev.langchain4j.rag.content.retriever.WebSearchContentRetriever

All Implemented Interfaces:

ContentRetriever

public class WebSearchContentRetriever
extends Object
implements ContentRetriever

A ContentRetriever that retrieves relevant Content from the web using a WebSearchEngine.
It returns one Content for each result that a WebSearchEngine has returned for a given Query.
Depending on the WebSearchEngine implementation, the Content.textSegment() can contain either a snippet of a web page or a complete content of a web page.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	WebSearchContentRetriever.WebSearchContentRetrieverBuilder

Constructor Summary

Constructors

Constructor	Description
WebSearchContentRetriever(WebSearchEngine webSearchEngine, Integer maxResults)

Method Summary

Modifier and Type	Method	Description
static WebSearchContentRetriever.WebSearchContentRetrieverBuilder	builder()
List<Content>	retrieve(Query query)	Retrieves relevant Contents using a given Query.
CompletableFuture<List<Content>>	retrieveAsync(Query query)	Non-blocking counterpart of ContentRetriever.retrieve(Query), invoked by the asynchronous (CompletableFuture/CompletionStage) and reactive (Flow.Publisher) AI Service modes when RAG is configured.

Methods inherited from class Object

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

Methods inherited from interface ContentRetriever

addListener, addListeners

Constructor Details

WebSearchContentRetriever

public WebSearchContentRetriever(WebSearchEngine webSearchEngine,
 Integer maxResults)

Method Details

builder

public static WebSearchContentRetriever.WebSearchContentRetrieverBuilder builder()

retrieve

public List<Content> retrieve(Query query)

Description copied from interface: ContentRetriever

Retrieves relevant Contents using a given Query. The Contents are sorted by relevance, with the most relevant Contents appearing at the beginning of the returned List<Content>.

Specified by:

retrieve in interface ContentRetriever

Parameters:

query - The Query to use for retrieval.

Returns:

A list of retrieved Contents.

retrieveAsync

public CompletableFuture<List<Content>> retrieveAsync(Query query)

Description copied from interface: ContentRetriever

Non-blocking counterpart of ContentRetriever.retrieve(Query), invoked by the asynchronous (CompletableFuture/CompletionStage) and reactive (Flow.Publisher) AI Service modes when RAG is configured.

The default implementation returns a failed future carrying AsyncNotSupportedException: a retriever backed by blocking I/O (embedding-model call, vector-store query, web search, etc.) must opt in by overriding this method so it can return a genuinely non-blocking future instead of secretly tying up a worker thread. A retriever that cannot be made non-blocking is still usable from these modes via DefaultRetrievalAugmentor, which offloads the blocking ContentRetriever.retrieve(Query) to its executor.

An implementation that honors cancellation should abort its in-flight I/O when the returned future is cancelled (best-effort); Java cannot safely interrupt arbitrary code, so this is not guaranteed.

Specified by:

retrieveAsync in interface ContentRetriever

Parameters:

query - The Query to use for retrieval.

Returns:

A CompletableFuture of the retrieved Contents, sorted by relevance.