validator/tests: add vector_similarity() function test

Add tests to validate the results of vector_similarity() function.

Refs: scylladb/scylladb#25993

Author: QuerthDP

crates/validator/src/tests/mod.rs

@@ -8,6 +8,7 @@ mod crud;
 mod full_scan;
 mod reconnect;
 mod serde;
+mod vector_similarity;
 
 use crate::ServicesSubnet;
 use crate::dns::Dns;
@@ -221,6 +222,7 @@ pub(crate) async fn register() -> Vec<(String, TestCase)> {
         ("full_scan", full_scan::new().await),
         ("reconnect", reconnect::new().await),
         ("serde", serde::new().await),
+        ("vector_similarity", vector_similarity::new().await),
     ]
     .into_iter()
     .map(|(name, test_case)| (name.to_string(), test_case))

crates/validator/src/tests/vector_similarity.rs

@@ -0,0 +1,89 @@
+/*
+ * Copyright 2025-present ScyllaDB
+ * SPDX-License-Identifier: LicenseRef-ScyllaDB-Source-Available-1.0
+ */
+
+use crate::common::*;
+use crate::tests::*;
+use std::time::Duration;
+use tracing::info;
+
+pub(crate) async fn new() -> TestCase {
+    let timeout = Duration::from_secs(30);
+    TestCase::empty()
+        .with_init(timeout, init)
+        .with_cleanup(timeout, cleanup)
+        .with_test(
+            "vector_similarity_function_returns_expected_results",
+            timeout,
+            vector_similarity_function_returns_expected_results,
+        )
+}
+
+async fn vector_similarity_function_returns_expected_results(actors: TestActors) {
+    info!("started");
+
+    let (session, client) = prepare_connection(&actors).await;
+
+    let keyspace = create_keyspace(&session).await;
+    let table = create_table(&session, "pk INT PRIMARY KEY, v VECTOR<FLOAT, 3>", None).await;
+
+    let embeddings: Vec<Vec<f32>> = vec![
+        vec![1.0, 2.0, 3.0],
+        vec![4.0, 5.0, 6.0],
+        vec![7.0, 8.0, 9.0],
+    ];
+    for (i, embedding) in embeddings.into_iter().enumerate() {
+        session
+            .query_unpaged(
+                format!("INSERT INTO {table} (pk, v) VALUES (?, ?)"),
+                (i as i32, &embedding),
+            )
+            .await
+            .expect("failed to insert data");
+    }
+
+    let index = create_index(
+        &session,
+        &client,
+        &table,
+        "v",
+        Some("{'similarity_function' : 'EUCLIDEAN'}"),
+    )
+    .await;
+
+    wait_for(
+        || async { client.count(&index.keyspace, &index.index).await == Some(3) },
+        "Waiting for 3 vectors to be indexed",
+        Duration::from_secs(5),
+    )
+    .await;
+
+    // Check if the query returns the expected results (recall at least 85%)
+    let rows = get_query_results(
+        format!(
+            "SELECT pk, vector_similarity() FROM {table} ORDER BY v ANN OF [0.0, 0.0, 0.0] LIMIT 5"
+        ),
+        &session,
+    )
+    .await;
+    assert_eq!(rows.len(), 3);
+    let results: Vec<(i32, f32)> = vec![(0, 14.0), (1, 77.0), (2, 194.0)];
+    for (i, row) in rows.iter().enumerate() {
+        let pk: i32 = row.columns[0].as_ref().unwrap().as_int().unwrap();
+        let similarity: f32 = row.columns[1].as_ref().unwrap().as_float().unwrap();
+        assert_eq!(
+            (pk, similarity),
+            results[i],
+            "Row {i} does not match expected result"
+        );
+    }
+
+    // Drop keyspace
+    session
+        .query_unpaged(format!("DROP KEYSPACE {keyspace}"), ())
+        .await
+        .expect("failed to drop a keyspace");
+
+    info!("finished");
+}