// 0. Assuming a Vector Search index is already created in MongoDB Atlas

// 1. Configure your MongoDB connection
const collection = new MongoClient(process.env.MONGODB_URI!)
    .db("my_ecommerce_db")
    .collection("products");

// 2. Instantiate a VectorStore
const vectorStore = new MongoDBAtlasVectorSearch(
    new VoyageEmbeddings({
    modelName: "voyage-4-large",
    apiKey: process.env.VOYAGEAI_API_KEY!,
}), { collection });

// 3. Add documents to the VectorStore
await vectorStore.addDocuments([ /* Your documents here */ ]);

// 4. Perform a similarity search
const results = await vectorStore.similaritySearch("Your product text search query");
console.log(results);

// The other steps stay the same...

// 2. Instantiate a VectorStore .
// The first param (`embeddings`) will be optional now.
const vectorStore = new MongoDBAtlasVectorSearch({ collection });

declare class MongoDBAtlasVectorSearch extends VectorStore {
// -----------------------------------------------
// The following methods/constructors should be updated.
// The `embeddings` parameter will be optional and
// Auto embeddings will be used when not provided.
// -----------------------------------------------

  constructor(
    embeddings: EmbeddingsInterface,        // Optional
    args: MongoDBAtlasVectorSearchLibArgs,
  );

  static fromTexts(
    texts: string[],
    metadatas: object[] | object,
    embeddings: EmbeddingsInterface,        // Optional
    dbConfig: MongoDBAtlasVectorSearchLibArgs & {
      ids?: string[];
    },
  ): Promise<MongoDBAtlasVectorSearch>;

  static fromDocuments(
    docs: Document[],
    embeddings: EmbeddingsInterface,        // Optional
    dbConfig: MongoDBAtlasVectorSearchLibArgs & {
      ids?: string[];
    },
  ): Promise<MongoDBAtlasVectorSearch>;
  

// ------------------------------------------------------
// The following methods should fail when auto embeddings
// are used since they would be redundant/conflicting
// but are still required by the `VectorStore` interface.
// ------------------------------------------------------

  addVectors(
    vectors: number[][],
    documents: Document[],
    options?: { ids?: string[] },
  ): Promise<any[]>;

  similaritySearchVectorWithScore(
    query: number[],
    k: number,
    filter?: MongoDBAtlasFilter,
  ): Promise<[Document, number][]>;

// --------------------------------------------------------
// The following methods should keep working as expected.
// They might be using auto embeddings under the hood now.
// --------------------------------------------------------

  addDocuments(
    documents: Document[],
    options?: { ids?: string[] },
  ): Promise<any[]>;

  delete(params: { ids: any[] }): Promise<void>;

  static fixArrayPrecision(array: number[]): number[];

  similaritySearch(
    query: string,
    k?: number,
    filter?: this["FilterType"] | undefined,
    _callbacks?: Callbacks | undefined,
  ): Promise<DocumentInterface[]>;

  similaritySearchWithScore(
    query: string,
    k?: number,
    filter?: this["FilterType"] | undefined,
    _callbacks?: Callbacks | undefined,
  ): Promise<[DocumentInterface, number][]>;

  asRetriever(
    kOrFields?: number | Partial<VectorStoreRetrieverInput<this>>,
    filter?: this["FilterType"],
    callbacks?: Callbacks,
    tags?: string[],
    metadata?: Record<string, unknown>,
    verbose?: boolean,
  ): VectorStoreRetriever<this>;

  maxMarginalRelevanceSearch(
    query: string, 
    options: MaxMarginalRelevanceSearchOptions<this["FilterType"]>
  ): Promise<Document[]>; 
}