fix: use sparse index namespaces for Test 5 keyword search

fix: use sparse index namespaces for Test 5 keyword search

- Add PineconeClient.listNamespacesFromKeywordIndex() to list namespaces from the keyword (sparse) index
- Test 5 now uses a namespace from the sparse index for keywordSearch; skip with clear message if sparse index has no namespaces
- Avoids using dense-derived testNamespace for sparse-only query (CodeRabbit feedback)

Author: clean6378-max-it

scripts/test-search.ts

@@ -152,35 +152,48 @@ async function test() {
       );
     }
 
-    // Test 5: Keyword (sparse-only) search on pinecone-rag-sparse
+    // Test 5: Keyword (sparse-only) search — use namespace from sparse index, not dense
     let duration5: number | undefined;
     let test5Skipped = false;
-    console.log(`\n🔤 Test 5: Keyword search (sparse-only index)`);
-    console.log(`   Namespace: "${testNamespace}"`);
-    console.log(`   Query: "test query"`);
-    console.log(`   Top K: 3`);
-    try {
-      const startTime5 = Date.now();
-      const results5 = await client.keywordSearch({
-        query: 'test query',
-        namespace: testNamespace,
-        topK: 3,
-      });
-      duration5 = Date.now() - startTime5;
-      console.log(`✅ Keyword search returned ${results5.length} result(s) in ${duration5}ms`);
-      if (results5.length > 0) {
-        console.log(
-          `   First result score: ${results5[0].score.toFixed(4)}, reranked: ${results5[0].reranked}`
-        );
-      }
-    } catch (kwError) {
+    const sparseNamespaces = await client.listNamespacesFromKeywordIndex();
+    if (sparseNamespaces.length === 0) {
       test5Skipped = true;
+      console.log(`\n🔤 Test 5: Keyword search (sparse-only index)`);
       console.log(
-        `⚠️  Keyword search skipped: ${kwError instanceof Error ? kwError.message : String(kwError)}`
+        `⚠️  Keyword search skipped: sparse index has no namespaces (or index unavailable).`
       );
       console.log(
-        `   Ensure PINECONE_SPARSE_INDEX_NAME (e.g. pinecone-rag-sparse) exists and has data.`
+        `   Ensure PINECONE_SPARSE_INDEX_NAME (e.g. pinecone-rag-sparse or rag-hybrid-sparse) exists and has data.`
       );
+    } else {
+      const sparseTestNamespace = sparseNamespaces[0].namespace;
+      console.log(`\n🔤 Test 5: Keyword search (sparse-only index)`);
+      console.log(`   Namespace: "${sparseTestNamespace}" (from sparse index)`);
+      console.log(`   Query: "test query"`);
+      console.log(`   Top K: 3`);
+      try {
+        const startTime5 = Date.now();
+        const results5 = await client.keywordSearch({
+          query: 'test query',
+          namespace: sparseTestNamespace,
+          topK: 3,
+        });
+        duration5 = Date.now() - startTime5;
+        console.log(`✅ Keyword search returned ${results5.length} result(s) in ${duration5}ms`);
+        if (results5.length > 0) {
+          console.log(
+            `   First result score: ${results5[0].score.toFixed(4)}, reranked: ${results5[0].reranked}`
+          );
+        }
+      } catch (kwError) {
+        test5Skipped = true;
+        console.log(
+          `⚠️  Keyword search skipped: ${kwError instanceof Error ? kwError.message : String(kwError)}`
+        );
+        console.log(
+          `   Ensure PINECONE_SPARSE_INDEX_NAME exists and namespace "${sparseTestNamespace}" has data.`
+        );
+      }
     }
 
     console.log('\n✨ All tests completed successfully!');

src/pinecone-client.ts

@@ -157,6 +157,29 @@ export class PineconeClient {
     return index;
   }
 
+  /**
+   * List namespaces present on the keyword (sparse) index used for keyword_search.
+   * Use this to choose a namespace for sparse-only queries instead of the dense index list.
+   */
+  async listNamespacesFromKeywordIndex(): Promise<
+    Array<{ namespace: string; recordCount: number }>
+  > {
+    try {
+      const keywordIndex = await this.ensureKeywordIndex();
+      const stats = keywordIndex.describeIndexStats
+        ? await keywordIndex.describeIndexStats()
+        : undefined;
+      const namespaces = stats?.namespaces ?? {};
+      return Object.entries(namespaces).map(([namespace, info]) => ({
+        namespace,
+        recordCount: info?.recordCount ?? 0,
+      }));
+    } catch (error) {
+      logError('Error listing namespaces from keyword index', error);
+      return [];
+    }
+  }
+
   /**
    * List all available namespaces with their metadata information
    *