alibaba/spring-ai-alibaba · error · IllegalArgumentException

Unsupported search type:

Error message

Unsupported search type: 

What it means

doSimilaritySearch dispatches on SearchRequest.getSearchType() via a switch over SEMANTIC, FULL_TEXT, and HYBRID; any other value throws IllegalArgumentException("Unsupported search type: ..."). Since Java switch over an enum covers all constants, this default arm mainly fires for null or unexpected/custom search types.

Solutions

1. Explicitly set SearchRequest.builder().searchType(SearchType.SEMANTIC) (or FULL_TEXT/HYBRID) when constructing the request.
2. Check that the SearchRequest class version matches the vector store module (version skew).
3. Ensure searchType is not null before calling similaritySearch.
4. Only use search types documented for this Elasticsearch vector store.

Example fix

// before
SearchRequest request = SearchRequest.builder().query("q").topK(5).build();
// after
SearchRequest request = SearchRequest.builder().query("q").topK(5)
    .searchType(SearchType.SEMANTIC)
    .build();

// Java
SearchType t = searchRequest.getSearchType();
if (t != SearchType.SEMANTIC && t != SearchType.FULL_TEXT && t != SearchType.HYBRID) {
    throw new IllegalArgumentException("search type must be SEMANTIC, FULL_TEXT, or HYBRID: " + t);
}

static boolean isSupported(SearchRequest req) {
    return req != null && req.getSearchType() != null
        && (req.getSearchType() == SearchType.SEMANTIC
         || req.getSearchType() == SearchType.FULL_TEXT
         || req.getSearchType() == SearchType.HYBRID);
}

try { return vectorStore.similaritySearch(request); } catch (IllegalArgumentException e) {
    if (String.valueOf(e.getMessage()).startsWith("Unsupported search type")) { log.error("bad searchType: {}", request.getSearchType()); }
    throw e;
}

When it happens

Trigger: Calling VectorStore.similaritySearch(SearchRequest) with a SearchRequest whose searchType is not one of the supported SEMANTIC/FULL_TEXT/HYBRID values — e.g. null searchType or a type introduced in a different version of the SearchRequest model.

Common situations: Building SearchRequest without explicitly setting searchType and getting an unexpected default; version skew between the SearchRequest class and this vector store implementation; copy-paste of a search type name from another vector store.

Understand the failure class

Background: UnsupportedOperationException and "is not supported" errors: when a library deliberately refuses a call — this error's family across 30 libraries.

Related errors

• Delete operation failed
• Failed to delete documents by filter
• hybrid alpha should be between 0 ~ 1.
• Index not found
• Elastic search index name must be provided

AI-assisted analysis of alibaba/spring-ai-alibaba@f82da0b50f (2026-09-09). Data as JSON: /api/errors/eba55887a0d79031. Report an issue: GitHub.

Appendix: source

Thrown at spring-ai-alibaba-admin/spring-ai-alibaba-admin-server-core/src/main/java/org/springframework/ai/vectorstore/elasticsearch/ElasticsearchVectorStore.java:261

	private BulkResponse bulkRequest(BulkRequest bulkRequest) {
		try {
			return this.elasticsearchClient.bulk(bulkRequest);
		}
		catch (IOException e) {
			throw new RuntimeException(e);
		}
	}

	@Override
	public List<Document> doSimilaritySearch(SearchRequest searchRequest) {
		Assert.notNull(searchRequest, "The search request must not be null.");

		return switch (searchRequest.getSearchType()) {
			case SEMANTIC -> searchBySemantic(searchRequest);
			case FULL_TEXT -> searchByFullText(searchRequest);
			case HYBRID -> searchByHybrid(searchRequest);
			default -> throw new IllegalArgumentException("Unsupported search type: " + searchRequest.getSearchType());
		};
	}

	private String getElasticsearchQueryString(Filter.Expression filterExpression) {
		return Objects.isNull(filterExpression) ? "*"
				: this.filterExpressionConverter.convertExpression(filterExpression);

	}

	private Document toDocument(Hit<Document> hit, SearchType searchType) {
		Document document = hit.source();
		Document.Builder documentBuilder = document.mutate();
		if (hit.score() != null) {
			documentBuilder.metadata(DocumentMetadata.DISTANCE.value(), 1 - normalizeSimilarityScore(hit.score()));

			if (searchType == SearchType.FULL_TEXT) {
				documentBuilder.score(1 - hit.score());
			}

View on GitHub (pinned to f82da0b50f)