Skip to content

GH-3160: Handle element types correctly in CassandraFilterExpressionC… #3197

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 1 commit into from
Closed

GH-3160: Handle element types correctly in CassandraFilterExpressionC… #3197

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 14 additions & 1 deletion ...java/org/springframework/ai/vectorstore/cassandra/CassandraFilterExpressionConverter.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,11 @@
import java.util.stream.Collectors;

import com.datastax.oss.driver.api.core.metadata.schema.ColumnMetadata;
import com.datastax.oss.driver.api.core.type.DataType;
import com.datastax.oss.driver.api.core.type.DataTypes;
import com.datastax.oss.driver.api.core.type.ListType;
import com.datastax.oss.driver.api.core.type.MapType;
import com.datastax.oss.driver.api.core.type.SetType;
import com.datastax.oss.driver.api.core.type.codec.registry.CodecRegistry;
import com.datastax.oss.driver.shaded.guava.common.base.Preconditions;

Expand Down Expand Up @@ -118,10 +122,19 @@ private void doListValue(ColumnMetadata column, Object v, StringBuilder context)
}

private void doValue(ColumnMetadata column, Object v, StringBuilder context) {

DataType dataType = column.getType();

// Check if we're handling an element inside a collection for an IN clause
if ((dataType instanceof ListType) && !(v instanceof Collection)) {
// Extract the element type from the collection type
dataType = ((ListType) dataType).getElementType();
}

if (DataTypes.SMALLINT.equals(column.getType())) {
v = ((Number) v).shortValue();
}
context.append(CodecRegistry.DEFAULT.codecFor(column.getType()).format(v));
context.append(CodecRegistry.DEFAULT.codecFor(dataType).format(v));
}

private Optional<ColumnMetadata> getColumn(String name) {
Expand Down
58 changes: 58 additions & 0 deletions ...re/src/test/java/org/springframework/ai/vectorstore/cassandra/CassandraVectorStoreIT.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,6 +29,8 @@

import com.datastax.oss.driver.api.core.CqlSession;
import com.datastax.oss.driver.api.core.CqlSessionBuilder;
import com.datastax.oss.driver.api.core.metadata.schema.KeyspaceMetadata;
import com.datastax.oss.driver.api.core.metadata.schema.TableMetadata;
import com.datastax.oss.driver.api.core.servererrors.InvalidQueryException;
import com.datastax.oss.driver.api.core.servererrors.SyntaxError;
import com.datastax.oss.driver.api.core.type.DataTypes;
Expand Down Expand Up @@ -57,6 +59,7 @@
import org.springframework.core.io.DefaultResourceLoader;

import static org.assertj.core.api.Assertions.assertThat;
import static org.junit.Assert.fail;

/**
* Use `mvn failsafe:integration-test -Dit.test=CassandraVectorStoreIT`
Expand Down Expand Up @@ -522,6 +525,61 @@ void getNativeClientTest() {
});
}

@Test
void searchWithCollectionFilter() {
this.contextRunner.run(context -> {
try (CassandraVectorStore store = createTestStore(context,
new SchemaColumn("currencies", DataTypes.listOf(DataTypes.TEXT), SchemaColumnTags.INDEXED))) {

// Create test documents with different currency lists
var btcDocument = new Document("BTC_doc", "Bitcoin document", Map.of("currencies", List.of("BTC")));
var ethDocument = new Document("ETH_doc", "Ethereum document", Map.of("currencies", List.of("ETH")));
var multiCurrencyDocument = new Document("MULTI_doc", "Multi-currency document",
Map.of("currencies", List.of("BTC", "ETH", "SOL")));

store.add(List.of(btcDocument, ethDocument, multiCurrencyDocument));

// Verify initial state
List<Document> results = store
.similaritySearch(SearchRequest.builder().query("document").topK(5).build());
assertThat(results).hasSize(3);

try {
// Test filtering with IN operator on a collection field
Filter.Expression filterExpression = new Filter.Expression(Filter.ExpressionType.IN,
new Filter.Key("currencies"), new Filter.Value(List.of("BTC")));

// Search using programmatic filter
store.similaritySearch(SearchRequest.builder()
.query("document")
.topK(5)
.similarityThresholdAll()
.filterExpression(filterExpression)
.build());

// If we get here without an exception, it means Cassandra
// unexpectedly accepted the query,
// which is surprising since Cassandra doesn't support the IN operator
// on collection columns.
// This would indicate a potential change in Cassandra's behavior.
Assertions.fail("Expected InvalidQueryException from Cassandra");
}
catch (InvalidQueryException e) {
// This is the expected outcome: Cassandra rejects the query with a
// specific error
// indicating that collection columns cannot be used with IN
// operators, which is
// a documented limitation of Cassandra's query language. Support for
// collection
// filtering via CONTAINS would be needed for this type of query to
// work.
assertThat(e.getMessage()).contains("Collection column 'currencies'");
assertThat(e.getMessage()).contains("cannot be restricted by a 'IN' relation");
}
}
});
}

@SpringBootConfiguration
@EnableAutoConfiguration(exclude = { DataSourceAutoConfiguration.class })
public static class TestApplication {
Expand Down