@langchain/google-cloud-sql-pg 0.0.2 → 1.0.1

package/dist/vectorstore.cjs.map

@@ -0,0 +1 @@
+{"version":3,"file":"vectorstore.cjs","names":["VectorStore","embeddings: EmbeddingsInterface","dbConfig: VSArgs","engine: PostgresEngine","tableName: string","DEFAULT_DISTANCE_STRATEGY","columns: { [key: string]: string }","allMetadataColumns: string[]","texts: string[]","metadatas: object[] | object","dbConfig: dbConfigArgs","documents: Document[]","Document","docs: Document[]","vectors: number[][]","options?: { ids?: string[] }","ids: string[]","metadatas: Record<string, string>[]","customZip","values: { [key: string]: string | string[] }","params: { ids?: string[] }","embedding: number[]","k: number","filter?: this[\"FilterType\"]","documentsWithScores: [Document, number][]","k?: number | undefined","filter?: this[\"FilterType\"] | undefined","query: string","options: MaxMarginalRelevanceSearchOptions<this[\"FilterType\"]>","docsList: Document[]","row: { [x: string]: string }","index: BaseIndex","name?: string","concurrently: boolean","DEFAULT_INDEX_NAME_SUFFIX","indexName?: string"],"sources":["../src/vectorstore.ts"],"sourcesContent":["import { EmbeddingsInterface } from \"@langchain/core/embeddings\";\nimport {\n  MaxMarginalRelevanceSearchOptions,\n  VectorStore,\n} from \"@langchain/core/vectorstores\";\nimport { Document } from \"@langchain/core/documents\";\nimport { v4 as uuidv4 } from \"uuid\";\nimport { maximalMarginalRelevance } from \"@langchain/core/utils/math\";\nimport {\n  BaseIndex,\n  DEFAULT_DISTANCE_STRATEGY,\n  DEFAULT_INDEX_NAME_SUFFIX,\n  DistanceStrategy,\n  QueryOptions,\n} from \"./indexes.js\";\nimport PostgresEngine from \"./engine.js\";\nimport { customZip } from \"./utils/utils.js\";\n\nexport interface PostgresVectorStoreArgs {\n  schemaName?: string;\n  contentColumn?: string;\n  embeddingColumn?: string;\n  metadataColumns?: Array<string>;\n  idColumn?: string;\n  distanceStrategy?: DistanceStrategy;\n  k?: number;\n  fetchK?: number;\n  lambdaMult?: number;\n  ignoreMetadataColumns?: Array<string>;\n  metadataJsonColumn?: string;\n  indexQueryOptions?: QueryOptions;\n}\n\nexport interface dbConfigArgs {\n  engine: PostgresEngine;\n  tableName: string;\n  dbConfig?: PostgresVectorStoreArgs;\n}\n\ninterface VSArgs {\n  engine: PostgresEngine;\n  tableName: string;\n  schemaName: string;\n  contentColumn: string;\n  embeddingColumn: string;\n  metadataColumns: Array<string>;\n  idColumn: string;\n  distanceStrategy: DistanceStrategy;\n  k: number;\n  fetchK: number;\n  lambdaMult: number;\n  metadataJsonColumn: string;\n  indexQueryOptions?: QueryOptions;\n}\n\n/**\n * Google Cloud SQL for PostgreSQL vector store integration.\n *\n * Setup:\n * Install `@langchain/google-cloud-sql-pg`\n *\n * ```bash\n * npm install @langchain/google-cloud-sql-pg\n * ```\n *\n * <details open>\n * <summary><strong>Instantiate</strong></summary>\n *\n * ```typescript\n * import { Column, PostgresEngine, PostgresEngineArgs, PostgresVectorStore, VectorStoreTableArgs } from \"@langchain/google-cloud-sql-pg\";\n * // Or other embeddings\n * import { OpenAIEmbeddings } from '@langchain/openai';\n *\n *\n * const embeddings = new OpenAIEmbeddings({\n *   model: \"text-embedding-3-small\",\n * });\n *\n * const pgArgs: PostgresEngineArgs = {\n *     user: \"db-user\",\n *     password: \"password\"\n * }\n * // Create a shared connection pool\n * const engine: PostgresEngine = await PostgresEngine.fromInstance(\n *  \"project-id\",\n *  \"region\",\n *  \"instance-name\",\n *  \"database-name\",\n *  pgArgs\n * );\n * // (Optional) Specify metadata columns for filtering\n * // All other metadata will be added to JSON\n * const vectorStoreTableArgs: VectorStoreTableArgs = {\n *   metadataColumns: [new Column(\"baz\", \"TEXT\")],\n * };\n * // Create a vector store table\n * await engine.initVectorstoreTable(\"my-table\", 768, vectorStoreTableArgs);\n * // Customize the vector store\n * const pvectorArgs: PostgresVectorStoreArgs = {\n *     idColumn: \"ID_COLUMN\",\n *     contentColumn: \"CONTENT_COLUMN\",\n *     embeddingColumn: \"EMBEDDING_COLUMN\",\n *     metadataColumns: [\"baz\"]\n * }\n *\n * const vectorStore = await PostgresVectorStore.initialize(engine, embeddingService, \"my-table\", pvectorArgs);\n * ```\n * </details>\n *\n * <br />\n *\n * <details>\n * <summary><strong>Add documents</strong></summary>\n *\n * ```typescript\n * import type { Document } from '@langchain/core/documents';\n *\n * const document1 = { pageContent: \"foo\", metadata: { baz: \"bar\" } };\n * const document2 = { pageContent: \"thud\", metadata: { bar: \"baz\" } };\n * const document3 = { pageContent: \"i will be deleted :(\", metadata: {} };\n *\n * const documents: Document[] = [document1, document2, document3];\n * const ids = [\"1\", \"2\", \"3\"];\n * await vectorStore.addDocuments(documents, { ids });\n * ```\n * </details>\n *\n * <br />\n *\n * <details>\n * <summary><strong>Delete documents</strong></summary>\n *\n * ```typescript\n * await vectorStore.delete({ ids: [\"3\"] });\n * ```\n * </details>\n *\n * <br />\n *\n * <details>\n * <summary><strong>Similarity search</strong></summary>\n *\n * ```typescript\n * const results = await vectorStore.similaritySearch(\"thud\", 1);\n * for (const doc of results) {\n *   console.log(`* ${doc.pageContent} [${JSON.stringify(doc.metadata, null)}]`);\n * }\n * // Output:thud [{\"baz\":\"bar\"}]\n * ```\n * </details>\n *\n * <br />\n *\n *\n * <details>\n * <summary><strong>Similarity search with filter</strong></summary>\n *\n * ```typescript\n * const resultsWithFilter = await vectorStore.similaritySearch(\"thud\", 1, \"baz = 'bar'\");\n *\n * for (const doc of resultsWithFilter) {\n *   console.log(`* ${doc.pageContent} [${JSON.stringify(doc.metadata, null)}]`);\n * }\n * // Output:foo [{\"baz\":\"bar\"}]\n * ```\n * </details>\n *\n * <br />\n *\n *\n * <details>\n * <summary><strong>Similarity search with score</strong></summary>\n *\n * ```typescript\n * const resultsWithScore = await vectorStore.similaritySearchWithScore(\"qux\", 1);\n * for (const [doc, score] of resultsWithScore) {\n *   console.log(`* [SIM=${score.toFixed(6)}] ${doc.pageContent} [${JSON.stringify(doc.metadata, null)}]`);\n * }\n * // Output:[SIM=0.000000] qux [{\"bar\":\"baz\",\"baz\":\"bar\"}]\n * ```\n * </details>\n *\n * <br />\n *\n * <details>\n * <summary><strong>As a retriever</strong></summary>\n *\n * ```typescript\n * const retriever = vectorStore.asRetriever({\n *   searchType: \"mmr\", // Leave blank for standard similarity search\n *   k: 1,\n * });\n * const resultAsRetriever = await retriever.invoke(\"thud\");\n * console.log(resultAsRetriever);\n *\n * // Output: [Document({ metadata: { \"baz\":\"bar\" }, pageContent: \"thud\" })]\n * ```\n * </details>\n *\n * <br />\n */\nexport class PostgresVectorStore extends VectorStore {\n  declare FilterType: string;\n\n  engine: PostgresEngine;\n\n  embeddings: EmbeddingsInterface;\n\n  tableName: string;\n\n  schemaName: string;\n\n  contentColumn: string;\n\n  embeddingColumn: string;\n\n  metadataColumns: Array<string>;\n\n  idColumn: string;\n\n  metadataJsonColumn: string;\n\n  distanceStrategy: DistanceStrategy;\n\n  k: number;\n\n  fetchK: number;\n\n  lambdaMult: number;\n\n  indexQueryOptions: QueryOptions | undefined;\n\n  /**\n   * Initializes a new vector store with embeddings and database configuration.\n   *\n   * @param embeddings - Instance of `EmbeddingsInterface` used to embed queries.\n   * @param dbConfig - Configuration settings for the database or storage system.\n   */\n  constructor(embeddings: EmbeddingsInterface, dbConfig: VSArgs) {\n    super(embeddings, dbConfig);\n    this.embeddings = embeddings;\n    this.engine = dbConfig.engine;\n    this.tableName = dbConfig.tableName;\n    this.schemaName = dbConfig.schemaName;\n    this.contentColumn = dbConfig.contentColumn;\n    this.embeddingColumn = dbConfig.embeddingColumn;\n    this.metadataColumns = dbConfig.metadataColumns\n      ? dbConfig.metadataColumns\n      : [];\n    this.idColumn = dbConfig.idColumn;\n    this.metadataJsonColumn = dbConfig.metadataJsonColumn;\n    this.distanceStrategy = dbConfig.distanceStrategy;\n    this.k = dbConfig.k;\n    this.fetchK = dbConfig.fetchK;\n    this.lambdaMult = dbConfig.lambdaMult;\n    this.indexQueryOptions = dbConfig.indexQueryOptions;\n  }\n\n  /**\n   * Create a new PostgresVectorStore instance.\n   * @param {PostgresEngine} engine Required - Connection pool engine for managing connections to Cloud SQL for PostgreSQL database.\n   * @param {Embeddings} embeddings Required - Text embedding model to use.\n   * @param {string} tableName Required - Name of an existing table or table to be created.\n   * @param {string} schemaName Database schema name of the table. Defaults to \"public\".\n   * @param {string} contentColumn Column that represent a Document's page_content. Defaults to \"content\".\n   * @param {string} embeddingColumn Column for embedding vectors. The embedding is generated from the document value. Defaults to \"embedding\".\n   * @param {Array<string>} metadataColumns Column(s) that represent a document's metadata.\n   * @param {Array<string>} ignoreMetadataColumns Optional - Column(s) to ignore in pre-existing tables for a document's metadata. Can not be used with metadata_columns.\n   * @param {string} idColumn Column that represents the Document's id. Defaults to \"langchain_id\".\n   * @param {string} metadataJsonColumn Optional - Column to store metadata as JSON. Defaults to \"langchain_metadata\".\n   * @param {DistanceStrategy} distanceStrategy Distance strategy to use for vector similarity search. Defaults to COSINE_DISTANCE.\n   * @param {number} k Number of Documents to return from search. Defaults to 4.\n   * @param {number} fetchK Number of Documents to fetch to pass to MMR algorithm.\n   * @param {number} lambdaMult Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5.\n   * @param {QueryOptions} indexQueryOptions Optional - Index query option.\n   * @returns PostgresVectorStore instance.\n   */\n  static async initialize(\n    engine: PostgresEngine,\n    embeddings: EmbeddingsInterface,\n    tableName: string,\n    {\n      schemaName = \"public\",\n      contentColumn = \"content\",\n      embeddingColumn = \"embedding\",\n      metadataColumns = [],\n      ignoreMetadataColumns,\n      idColumn = \"langchain_id\",\n      metadataJsonColumn = \"langchain_metadata\",\n      distanceStrategy = DEFAULT_DISTANCE_STRATEGY,\n      k = 4,\n      fetchK = 20,\n      lambdaMult = 0.5,\n      indexQueryOptions,\n    }: PostgresVectorStoreArgs = {}\n  ): Promise<PostgresVectorStore> {\n    if (metadataColumns !== undefined && ignoreMetadataColumns !== undefined) {\n      throw Error(\n        \"Can not use both metadata_columns and ignore_metadata_columns.\"\n      );\n    }\n\n    const { rows } = await engine.pool.raw(\n      `SELECT column_name, data_type FROM information_schema.columns WHERE table_name = '${tableName}' AND table_schema = '${schemaName}'`\n    );\n    const columns: { [key: string]: string } = {};\n\n    for (const index in rows) {\n      if (rows[index]) {\n        const row = rows[index];\n        columns[row.column_name] = row.data_type;\n      }\n    }\n\n    if (!Object.prototype.hasOwnProperty.call(columns, idColumn)) {\n      throw Error(`Id column: ${idColumn}, does not exist.`);\n    }\n\n    if (!Object.prototype.hasOwnProperty.call(columns, contentColumn)) {\n      throw Error(`Content column: ${contentColumn}, does not exist.`);\n    }\n\n    const contentType = columns[contentColumn];\n\n    if (contentType !== \"text\" && !contentType.includes(\"char\")) {\n      throw Error(\n        `Content column: ${contentColumn}, is type: ${contentType}. It must be a type of character string.`\n      );\n    }\n\n    if (!Object.prototype.hasOwnProperty.call(columns, embeddingColumn)) {\n      throw Error(`Embedding column: ${embeddingColumn}, does not exist.`);\n    }\n\n    if (columns[embeddingColumn] !== \"USER-DEFINED\") {\n      throw Error(\n        `Embedding column: ${embeddingColumn} is not of type Vector.`\n      );\n    }\n\n    const jsonColumn = Object.prototype.hasOwnProperty.call(\n      columns,\n      metadataJsonColumn\n    )\n      ? metadataJsonColumn\n      : \"\";\n\n    for (const column of metadataColumns) {\n      if (!Object.prototype.hasOwnProperty.call(columns, column)) {\n        throw Error(`Metadata column: ${column}, does not exist.`);\n      }\n    }\n\n    const allColumns = columns;\n    let allMetadataColumns: string[] = [];\n    if (\n      ignoreMetadataColumns !== undefined &&\n      ignoreMetadataColumns.length > 0\n    ) {\n      for (const column of ignoreMetadataColumns) {\n        delete allColumns[column];\n      }\n\n      delete allColumns[idColumn];\n      delete allColumns[contentColumn];\n      delete allColumns[embeddingColumn];\n      allMetadataColumns = Object.keys(allColumns);\n    } else {\n      for (const column of metadataColumns) {\n        if (Object.prototype.hasOwnProperty.call(allColumns, column)) {\n          allMetadataColumns.push(column);\n        }\n      }\n    }\n    return new PostgresVectorStore(embeddings, {\n      engine,\n      tableName,\n      schemaName,\n      contentColumn,\n      embeddingColumn,\n      metadataColumns: allMetadataColumns,\n      idColumn,\n      metadataJsonColumn: jsonColumn,\n      distanceStrategy,\n      k,\n      fetchK,\n      lambdaMult,\n      indexQueryOptions,\n    });\n  }\n\n  static async fromTexts(\n    texts: string[],\n    metadatas: object[] | object,\n    embeddings: EmbeddingsInterface,\n    dbConfig: dbConfigArgs\n  ): Promise<VectorStore> {\n    const documents: Document[] = [];\n\n    for (let i = 0; i < texts.length; i += 1) {\n      const doc = new Document({\n        pageContent: texts[i],\n        metadata: Array.isArray(metadatas) ? metadatas[i] : metadatas,\n      });\n      documents.push(doc);\n    }\n\n    return PostgresVectorStore.fromDocuments(documents, embeddings, dbConfig);\n  }\n\n  static async fromDocuments(\n    docs: Document[],\n    embeddings: EmbeddingsInterface,\n    dbConfig: dbConfigArgs\n  ): Promise<VectorStore> {\n    const { engine } = dbConfig;\n    const { tableName } = dbConfig;\n    const config = dbConfig.dbConfig;\n    const vectorStore = await this.initialize(\n      engine,\n      embeddings,\n      tableName,\n      config\n    );\n\n    await vectorStore.addDocuments(docs);\n\n    return vectorStore;\n  }\n\n  async addVectors(\n    vectors: number[][],\n    documents: Document[],\n    options?: { ids?: string[] }\n  ): Promise<string[] | void> {\n    let ids: string[] = [];\n    const metadatas: Record<string, string>[] = [];\n\n    if (vectors.length !== documents.length) {\n      throw new Error(\n        \"The number of vectors must match the number of documents provided.\"\n      );\n    }\n\n    if (options?.ids) {\n      ids = options.ids;\n    } else {\n      documents.forEach((document) => {\n        if (document.id !== undefined) {\n          ids.push(document.id);\n        } else {\n          ids.push(uuidv4());\n        }\n      });\n    }\n\n    if (options && options.ids && options.ids.length !== documents.length) {\n      throw new Error(\n        \"The number of ids must match the number of documents provided.\"\n      );\n    }\n\n    documents.forEach((document) => {\n      metadatas.push(document.metadata);\n    });\n\n    const tuples = customZip(ids, documents, vectors, metadatas);\n\n    // Insert embeddings\n    for (const [id, document, embedding, metadata] of tuples) {\n      const metadataColNames =\n        this.metadataColumns.length > 0\n          ? `, \"${this.metadataColumns.join('\",\"')}\"`\n          : \"\";\n\n      let stmt = `INSERT INTO \"${this.schemaName}\".\"${this.tableName}\"(\"${this.idColumn}\", \"${this.contentColumn}\", \"${this.embeddingColumn}\" ${metadataColNames}`;\n      const values: { [key: string]: string | string[] } = {\n        id,\n        content: document.pageContent,\n        embedding: `[${embedding.toString()}]`,\n      };\n      let valuesStmt = \" VALUES (:id, :content, :embedding\";\n\n      // Add metadata\n      const extra = metadata;\n      for (const metadataColumn of this.metadataColumns) {\n        if (Object.prototype.hasOwnProperty.call(metadata, metadataColumn)) {\n          valuesStmt += `, :${metadataColumn}`;\n          values[metadataColumn] = metadata[metadataColumn];\n          delete extra[metadataColumn];\n        } else {\n          valuesStmt += \" ,null\";\n        }\n      }\n\n      // Add JSON column and/or close statement\n      stmt += this.metadataJsonColumn ? `, ${this.metadataJsonColumn})` : \")\";\n      if (this.metadataJsonColumn) {\n        valuesStmt += \", :extra)\";\n        Object.assign(values, { extra: JSON.stringify(extra) });\n      } else {\n        valuesStmt += \")\";\n      }\n\n      const query = stmt + valuesStmt;\n      await this.engine.pool.raw(query, values);\n    }\n\n    return options?.ids;\n  }\n\n  _vectorstoreType(): string {\n    return \"cloudsqlpostgresql\";\n  }\n\n  /**\n   * Adds documents to the vector store, embedding them first through the\n   * `embeddings` instance.\n   *\n   * @param documents - Array of documents to embed and add.\n   * @param options - Optional configuration for embedding and storing documents.\n   * @returns A promise resolving to an array of document IDs or void, based on implementation.\n   * @abstract\n   */\n  async addDocuments(\n    documents: Document[],\n    options?: { ids?: string[] }\n  ): Promise<string[] | void> {\n    const texts = [];\n\n    for (const doc of documents) {\n      texts.push(doc.pageContent);\n    }\n\n    const embeddings = await this.embeddings.embedDocuments(texts);\n    const results = await this.addVectors(embeddings, documents, options);\n\n    return results;\n  }\n\n  /**\n   * Deletes documents from the vector store based on the specified ids.\n   *\n   * @param params - Flexible key-value pairs defining conditions for document deletion.\n   * @param ids -  Optional: Property of {params} that contains the array of ids to be deleted\n   * @returns A promise that resolves once the deletion is complete.\n   */\n  async delete(params: { ids?: string[] }): Promise<void> {\n    if (params.ids === undefined) return;\n    const idList = params.ids.map((id) => `'${id}'`).join(\", \");\n    const query = `DELETE FROM \"${this.schemaName}\".\"${this.tableName}\" WHERE \"${this.idColumn}\" in (${idList})`;\n    await this.engine.pool.raw(query);\n  }\n\n  async similaritySearchVectorWithScore(\n    embedding: number[],\n    k: number,\n    filter?: this[\"FilterType\"]\n  ): Promise<[Document, number][]> {\n    const results = await this.queryCollection(embedding, k, filter);\n    const documentsWithScores: [Document, number][] = [];\n\n    for (const row of results) {\n      const metadata =\n        this.metadataJsonColumn && row[this.metadataJsonColumn]\n          ? row[this.metadataJsonColumn]\n          : {};\n\n      for (const col of this.metadataColumns) {\n        metadata[col] = row[col];\n      }\n\n      documentsWithScores.push([\n        new Document({ pageContent: row[this.contentColumn], metadata }),\n        row.distance,\n      ]);\n    }\n\n    return documentsWithScores;\n  }\n\n  private async queryCollection(\n    embedding: number[],\n    k?: number | undefined,\n    filter?: this[\"FilterType\"] | undefined\n  ) {\n    const fetchK = k ?? this.k;\n    const { operator } = this.distanceStrategy;\n    const { searchFunction } = this.distanceStrategy;\n    const _filter = filter !== undefined ? `WHERE ${filter}` : \"\";\n    const metadataColNames =\n      this.metadataColumns.length > 0\n        ? `, \"${this.metadataColumns.join('\",\"')}\"`\n        : \"\";\n    const metadataJsonColName = this.metadataJsonColumn\n      ? `, \"${this.metadataJsonColumn}\"`\n      : \"\";\n\n    const query = `SELECT \"${this.idColumn}\", \"${this.contentColumn}\", \"${this.embeddingColumn}\" ${metadataColNames} ${metadataJsonColName}, ${searchFunction}(\"${this.embeddingColumn}\", '[${embedding}]') as distance FROM \"${this.schemaName}\".\"${this.tableName}\" ${_filter} ORDER BY \"${this.embeddingColumn}\" ${operator} '[${embedding}]' LIMIT ${fetchK};`;\n\n    if (this.indexQueryOptions) {\n      await this.engine.pool.raw(\n        `SET LOCAL ${this.indexQueryOptions.to_string()}`\n      );\n    }\n\n    const { rows } = await this.engine.pool.raw(query);\n\n    return rows;\n  }\n\n  async maxMarginalRelevanceSearch(\n    query: string,\n    options: MaxMarginalRelevanceSearchOptions<this[\"FilterType\"]>\n  ): Promise<Document[]> {\n    const vector = await this.embeddings.embedQuery(query);\n    const results = await this.queryCollection(\n      vector,\n      options?.k,\n      options?.filter\n    );\n    const k = options?.k ? options.k : this.k;\n    const documentsWithScores: [Document, number][] = [];\n    let docsList: Document[] = [];\n\n    const embeddingList = results.map((row: { [x: string]: string }) =>\n      JSON.parse(row[this.embeddingColumn])\n    );\n    const mmrSelected = maximalMarginalRelevance(\n      vector,\n      embeddingList,\n      options?.lambda,\n      k\n    );\n\n    for (const row of results) {\n      const metadata =\n        this.metadataJsonColumn && row[this.metadataJsonColumn]\n          ? row[this.metadataJsonColumn]\n          : {};\n      for (const col of this.metadataColumns) {\n        metadata[col] = row[col];\n      }\n      documentsWithScores.push([\n        new Document({\n          pageContent: row[this.contentColumn],\n          metadata,\n        }),\n        row.distance,\n      ]);\n    }\n\n    docsList = documentsWithScores\n      .filter((_, i) => mmrSelected.includes(i))\n      .map(([doc, _]) => doc);\n\n    return docsList;\n  }\n\n  /**\n   * Create an index on the vector store table\n   * @param {BaseIndex} index\n   * @param {string} name Optional\n   * @param {boolean} concurrently Optional\n   */\n  async applyVectorIndex(\n    index: BaseIndex,\n    name?: string,\n    concurrently: boolean = false\n  ): Promise<void> {\n    if (index.constructor.name === \"ExactNearestNeighbor\") {\n      await this.dropVectorIndex();\n      return;\n    }\n\n    const filter =\n      index.partialIndexes && index.partialIndexes?.length > 0\n        ? `WHERE (${index.partialIndexes})`\n        : \"\";\n    const params = `WITH ${index.indexOptions()}`;\n    const funct = index.distanceStrategy.indexFunction;\n\n    let indexName = name;\n    if (!indexName) {\n      if (!index.name) {\n        indexName = this.tableName + DEFAULT_INDEX_NAME_SUFFIX;\n      } else {\n        indexName = index.name;\n      }\n    }\n\n    const stmt = `CREATE INDEX ${\n      concurrently ? \"CONCURRENTLY\" : \"\"\n    } ${indexName} ON \"${this.schemaName}\".\"${this.tableName}\" USING ${\n      index.indexType\n    } (${this.embeddingColumn} ${funct}) ${params} ${filter};`;\n\n    await this.engine.pool.raw(stmt);\n  }\n\n  /**\n   * Check if index exists in the table.\n   * @param {string} indexName Optional - index name\n   */\n  async isValidIndex(indexName?: string): Promise<boolean> {\n    const idxName = indexName || this.tableName + DEFAULT_INDEX_NAME_SUFFIX;\n    const stmt = `SELECT tablename, indexname\n                  FROM pg_indexes\n                  WHERE tablename = '${this.tableName}' AND schemaname = '${this.schemaName}' AND indexname = '${idxName}';`;\n    const { rows } = await this.engine.pool.raw(stmt);\n\n    return rows.length === 1;\n  }\n\n  /**\n   * Drop the vector index\n   * @param {string} indexName Optional - index name\n   */\n  async dropVectorIndex(indexName?: string): Promise<void> {\n    const idxName = indexName || this.tableName + DEFAULT_INDEX_NAME_SUFFIX;\n    const query = `DROP INDEX IF EXISTS ${idxName};`;\n    await this.engine.pool.raw(query);\n  }\n\n  /**\n   * Re-index the vector store table\n   * @param {string} indexName Optional - index name\n   */\n  async reIndex(indexName?: string) {\n    const idxName = indexName || this.tableName + DEFAULT_INDEX_NAME_SUFFIX;\n    const query = `REINDEX INDEX ${idxName};`;\n    await this.engine.pool.raw(query);\n  }\n}\n\nexport default PostgresVectorStore;\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAyMA,IAAa,sBAAb,MAAa,4BAA4BA,0CAAY;CAGnD;CAEA;CAEA;CAEA;CAEA;CAEA;CAEA;CAEA;CAEA;CAEA;CAEA;CAEA;CAEA;CAEA;;;;;;;CAQA,YAAYC,YAAiCC,UAAkB;EAC7D,MAAM,YAAY,SAAS;EAC3B,KAAK,aAAa;EAClB,KAAK,SAAS,SAAS;EACvB,KAAK,YAAY,SAAS;EAC1B,KAAK,aAAa,SAAS;EAC3B,KAAK,gBAAgB,SAAS;EAC9B,KAAK,kBAAkB,SAAS;EAChC,KAAK,kBAAkB,SAAS,kBAC5B,SAAS,kBACT,CAAE;EACN,KAAK,WAAW,SAAS;EACzB,KAAK,qBAAqB,SAAS;EACnC,KAAK,mBAAmB,SAAS;EACjC,KAAK,IAAI,SAAS;EAClB,KAAK,SAAS,SAAS;EACvB,KAAK,aAAa,SAAS;EAC3B,KAAK,oBAAoB,SAAS;CACnC;;;;;;;;;;;;;;;;;;;;CAqBD,aAAa,WACXC,QACAF,YACAG,WACA,EACE,aAAa,UACb,gBAAgB,WAChB,kBAAkB,aAClB,kBAAkB,CAAE,GACpB,uBACA,WAAW,gBACX,qBAAqB,sBACrB,mBAAmBC,2CACnB,IAAI,GACJ,SAAS,IACT,aAAa,IACb,mBACwB,GAAG,CAAE,GACD;AAC9B,MAAI,oBAAoB,UAAa,0BAA0B,OAC7D,OAAM,MACJ,iEACD;EAGH,MAAM,EAAE,MAAM,GAAG,MAAM,OAAO,KAAK,IACjC,CAAC,kFAAkF,EAAE,UAAU,sBAAsB,EAAE,WAAW,CAAC,CAAC,CACrI;EACD,MAAMC,UAAqC,CAAE;AAE7C,OAAK,MAAM,SAAS,KAClB,KAAI,KAAK,QAAQ;GACf,MAAM,MAAM,KAAK;GACjB,QAAQ,IAAI,eAAe,IAAI;EAChC;AAGH,MAAI,CAAC,OAAO,UAAU,eAAe,KAAK,SAAS,SAAS,CAC1D,OAAM,MAAM,CAAC,WAAW,EAAE,SAAS,iBAAiB,CAAC,CAAC;AAGxD,MAAI,CAAC,OAAO,UAAU,eAAe,KAAK,SAAS,cAAc,CAC/D,OAAM,MAAM,CAAC,gBAAgB,EAAE,cAAc,iBAAiB,CAAC,CAAC;EAGlE,MAAM,cAAc,QAAQ;AAE5B,MAAI,gBAAgB,UAAU,CAAC,YAAY,SAAS,OAAO,CACzD,OAAM,MACJ,CAAC,gBAAgB,EAAE,cAAc,WAAW,EAAE,YAAY,wCAAwC,CAAC,CACpG;AAGH,MAAI,CAAC,OAAO,UAAU,eAAe,KAAK,SAAS,gBAAgB,CACjE,OAAM,MAAM,CAAC,kBAAkB,EAAE,gBAAgB,iBAAiB,CAAC,CAAC;AAGtE,MAAI,QAAQ,qBAAqB,eAC/B,OAAM,MACJ,CAAC,kBAAkB,EAAE,gBAAgB,uBAAuB,CAAC,CAC9D;EAGH,MAAM,aAAa,OAAO,UAAU,eAAe,KACjD,SACA,mBACD,GACG,qBACA;AAEJ,OAAK,MAAM,UAAU,gBACnB,KAAI,CAAC,OAAO,UAAU,eAAe,KAAK,SAAS,OAAO,CACxD,OAAM,MAAM,CAAC,iBAAiB,EAAE,OAAO,iBAAiB,CAAC,CAAC;EAI9D,MAAM,aAAa;EACnB,IAAIC,qBAA+B,CAAE;AACrC,MACE,0BAA0B,UAC1B,sBAAsB,SAAS,GAC/B;AACA,QAAK,MAAM,UAAU,uBACnB,OAAO,WAAW;GAGpB,OAAO,WAAW;GAClB,OAAO,WAAW;GAClB,OAAO,WAAW;GAClB,qBAAqB,OAAO,KAAK,WAAW;EAC7C,MACC,MAAK,MAAM,UAAU,gBACnB,KAAI,OAAO,UAAU,eAAe,KAAK,YAAY,OAAO,EAC1D,mBAAmB,KAAK,OAAO;AAIrC,SAAO,IAAI,oBAAoB,YAAY;GACzC;GACA;GACA;GACA;GACA;GACA,iBAAiB;GACjB;GACA,oBAAoB;GACpB;GACA;GACA;GACA;GACA;EACD;CACF;CAED,aAAa,UACXC,OACAC,WACAR,YACAS,UACsB;EACtB,MAAMC,YAAwB,CAAE;AAEhC,OAAK,IAAI,IAAI,GAAG,IAAI,MAAM,QAAQ,KAAK,GAAG;GACxC,MAAM,MAAM,IAAIC,oCAAS;IACvB,aAAa,MAAM;IACnB,UAAU,MAAM,QAAQ,UAAU,GAAG,UAAU,KAAK;GACrD;GACD,UAAU,KAAK,IAAI;EACpB;AAED,SAAO,oBAAoB,cAAc,WAAW,YAAY,SAAS;CAC1E;CAED,aAAa,cACXC,MACAZ,YACAS,UACsB;EACtB,MAAM,EAAE,QAAQ,GAAG;EACnB,MAAM,EAAE,WAAW,GAAG;EACtB,MAAM,SAAS,SAAS;EACxB,MAAM,cAAc,MAAM,KAAK,WAC7B,QACA,YACA,WACA,OACD;EAED,MAAM,YAAY,aAAa,KAAK;AAEpC,SAAO;CACR;CAED,MAAM,WACJI,SACAH,WACAI,SAC0B;EAC1B,IAAIC,MAAgB,CAAE;EACtB,MAAMC,YAAsC,CAAE;AAE9C,MAAI,QAAQ,WAAW,UAAU,OAC/B,OAAM,IAAI,MACR;AAIJ,MAAI,SAAS,KACX,MAAM,QAAQ;OAEd,UAAU,QAAQ,CAAC,aAAa;AAC9B,OAAI,SAAS,OAAO,QAClB,IAAI,KAAK,SAAS,GAAG;QAErB,IAAI,mBAAa,CAAC;EAErB,EAAC;AAGJ,MAAI,WAAW,QAAQ,OAAO,QAAQ,IAAI,WAAW,UAAU,OAC7D,OAAM,IAAI,MACR;EAIJ,UAAU,QAAQ,CAAC,aAAa;GAC9B,UAAU,KAAK,SAAS,SAAS;EAClC,EAAC;EAEF,MAAM,SAASC,wBAAU,KAAK,WAAW,SAAS,UAAU;AAG5D,OAAK,MAAM,CAAC,IAAI,UAAU,WAAW,SAAS,IAAI,QAAQ;GACxD,MAAM,mBACJ,KAAK,gBAAgB,SAAS,IAC1B,CAAC,GAAG,EAAE,KAAK,gBAAgB,KAAK,QAAM,CAAC,CAAC,CAAC,GACzC;GAEN,IAAI,OAAO,CAAC,aAAa,EAAE,KAAK,WAAW,GAAG,EAAE,KAAK,UAAU,GAAG,EAAE,KAAK,SAAS,IAAI,EAAE,KAAK,cAAc,IAAI,EAAE,KAAK,gBAAgB,EAAE,EAAE,kBAAkB;GAC5J,MAAMC,SAA+C;IACnD;IACA,SAAS,SAAS;IAClB,WAAW,CAAC,CAAC,EAAE,UAAU,UAAU,CAAC,CAAC,CAAC;GACvC;GACD,IAAI,aAAa;GAGjB,MAAM,QAAQ;AACd,QAAK,MAAM,kBAAkB,KAAK,gBAChC,KAAI,OAAO,UAAU,eAAe,KAAK,UAAU,eAAe,EAAE;IAClE,cAAc,CAAC,GAAG,EAAE,gBAAgB;IACpC,OAAO,kBAAkB,SAAS;IAClC,OAAO,MAAM;GACd,OACC,cAAc;GAKlB,QAAQ,KAAK,qBAAqB,CAAC,EAAE,EAAE,KAAK,mBAAmB,CAAC,CAAC,GAAG;AACpE,OAAI,KAAK,oBAAoB;IAC3B,cAAc;IACd,OAAO,OAAO,QAAQ,EAAE,OAAO,KAAK,UAAU,MAAM,CAAE,EAAC;GACxD,OACC,cAAc;GAGhB,MAAM,QAAQ,OAAO;GACrB,MAAM,KAAK,OAAO,KAAK,IAAI,OAAO,OAAO;EAC1C;AAED,SAAO,SAAS;CACjB;CAED,mBAA2B;AACzB,SAAO;CACR;;;;;;;;;;CAWD,MAAM,aACJR,WACAI,SAC0B;EAC1B,MAAM,QAAQ,CAAE;AAEhB,OAAK,MAAM,OAAO,WAChB,MAAM,KAAK,IAAI,YAAY;EAG7B,MAAM,aAAa,MAAM,KAAK,WAAW,eAAe,MAAM;EAC9D,MAAM,UAAU,MAAM,KAAK,WAAW,YAAY,WAAW,QAAQ;AAErE,SAAO;CACR;;;;;;;;CASD,MAAM,OAAOK,QAA2C;AACtD,MAAI,OAAO,QAAQ,OAAW;EAC9B,MAAM,SAAS,OAAO,IAAI,IAAI,CAAC,OAAO,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,CAAC,CAAC,KAAK,KAAK;EAC3D,MAAM,QAAQ,CAAC,aAAa,EAAE,KAAK,WAAW,GAAG,EAAE,KAAK,UAAU,SAAS,EAAE,KAAK,SAAS,MAAM,EAAE,OAAO,CAAC,CAAC;EAC5G,MAAM,KAAK,OAAO,KAAK,IAAI,MAAM;CAClC;CAED,MAAM,gCACJC,WACAC,GACAC,QAC+B;EAC/B,MAAM,UAAU,MAAM,KAAK,gBAAgB,WAAW,GAAG,OAAO;EAChE,MAAMC,sBAA4C,CAAE;AAEpD,OAAK,MAAM,OAAO,SAAS;GACzB,MAAM,WACJ,KAAK,sBAAsB,IAAI,KAAK,sBAChC,IAAI,KAAK,sBACT,CAAE;AAER,QAAK,MAAM,OAAO,KAAK,iBACrB,SAAS,OAAO,IAAI;GAGtB,oBAAoB,KAAK,CACvB,IAAIZ,oCAAS;IAAE,aAAa,IAAI,KAAK;IAAgB;GAAU,IAC/D,IAAI,QACL,EAAC;EACH;AAED,SAAO;CACR;CAED,MAAc,gBACZS,WACAI,GACAC,QACA;EACA,MAAM,SAAS,KAAK,KAAK;EACzB,MAAM,EAAE,UAAU,GAAG,KAAK;EAC1B,MAAM,EAAE,gBAAgB,GAAG,KAAK;EAChC,MAAM,UAAU,WAAW,SAAY,CAAC,MAAM,EAAE,QAAQ,GAAG;EAC3D,MAAM,mBACJ,KAAK,gBAAgB,SAAS,IAC1B,CAAC,GAAG,EAAE,KAAK,gBAAgB,KAAK,QAAM,CAAC,CAAC,CAAC,GACzC;EACN,MAAM,sBAAsB,KAAK,qBAC7B,CAAC,GAAG,EAAE,KAAK,mBAAmB,CAAC,CAAC,GAChC;EAEJ,MAAM,QAAQ,CAAC,QAAQ,EAAE,KAAK,SAAS,IAAI,EAAE,KAAK,cAAc,IAAI,EAAE,KAAK,gBAAgB,EAAE,EAAE,iBAAiB,CAAC,EAAE,oBAAoB,EAAE,EAAE,eAAe,EAAE,EAAE,KAAK,gBAAgB,KAAK,EAAE,UAAU,sBAAsB,EAAE,KAAK,WAAW,GAAG,EAAE,KAAK,UAAU,EAAE,EAAE,QAAQ,WAAW,EAAE,KAAK,gBAAgB,EAAE,EAAE,SAAS,GAAG,EAAE,UAAU,SAAS,EAAE,OAAO,CAAC,CAAC;AAE9V,MAAI,KAAK,mBACP,MAAM,KAAK,OAAO,KAAK,IACrB,CAAC,UAAU,EAAE,KAAK,kBAAkB,WAAW,EAAE,CAClD;EAGH,MAAM,EAAE,MAAM,GAAG,MAAM,KAAK,OAAO,KAAK,IAAI,MAAM;AAElD,SAAO;CACR;CAED,MAAM,2BACJC,OACAC,SACqB;EACrB,MAAM,SAAS,MAAM,KAAK,WAAW,WAAW,MAAM;EACtD,MAAM,UAAU,MAAM,KAAK,gBACzB,QACA,SAAS,GACT,SAAS,OACV;EACD,MAAM,IAAI,SAAS,IAAI,QAAQ,IAAI,KAAK;EACxC,MAAMJ,sBAA4C,CAAE;EACpD,IAAIK,WAAuB,CAAE;EAE7B,MAAM,gBAAgB,QAAQ,IAAI,CAACC,QACjC,KAAK,MAAM,IAAI,KAAK,iBAAiB,CACtC;EACD,MAAM,wEACJ,QACA,eACA,SAAS,QACT,EACD;AAED,OAAK,MAAM,OAAO,SAAS;GACzB,MAAM,WACJ,KAAK,sBAAsB,IAAI,KAAK,sBAChC,IAAI,KAAK,sBACT,CAAE;AACR,QAAK,MAAM,OAAO,KAAK,iBACrB,SAAS,OAAO,IAAI;GAEtB,oBAAoB,KAAK,CACvB,IAAIlB,oCAAS;IACX,aAAa,IAAI,KAAK;IACtB;GACD,IACD,IAAI,QACL,EAAC;EACH;EAED,WAAW,oBACR,OAAO,CAAC,GAAG,MAAM,YAAY,SAAS,EAAE,CAAC,CACzC,IAAI,CAAC,CAAC,KAAK,EAAE,KAAK,IAAI;AAEzB,SAAO;CACR;;;;;;;CAQD,MAAM,iBACJmB,OACAC,MACAC,eAAwB,OACT;AACf,MAAI,MAAM,YAAY,SAAS,wBAAwB;GACrD,MAAM,KAAK,iBAAiB;AAC5B;EACD;EAED,MAAM,SACJ,MAAM,kBAAkB,MAAM,gBAAgB,SAAS,IACnD,CAAC,OAAO,EAAE,MAAM,eAAe,CAAC,CAAC,GACjC;EACN,MAAM,SAAS,CAAC,KAAK,EAAE,MAAM,cAAc,EAAE;EAC7C,MAAM,QAAQ,MAAM,iBAAiB;EAErC,IAAI,YAAY;AAChB,MAAI,CAAC,UACH,KAAI,CAAC,MAAM,MACT,YAAY,KAAK,YAAYC;OAE7B,YAAY,MAAM;EAItB,MAAM,OAAO,CAAC,aAAa,EACzB,eAAe,iBAAiB,GACjC,CAAC,EAAE,UAAU,KAAK,EAAE,KAAK,WAAW,GAAG,EAAE,KAAK,UAAU,QAAQ,EAC/D,MAAM,UACP,EAAE,EAAE,KAAK,gBAAgB,CAAC,EAAE,MAAM,EAAE,EAAE,OAAO,CAAC,EAAE,OAAO,CAAC,CAAC;EAE1D,MAAM,KAAK,OAAO,KAAK,IAAI,KAAK;CACjC;;;;;CAMD,MAAM,aAAaC,WAAsC;EACvD,MAAM,UAAU,aAAa,KAAK,YAAYD;EAC9C,MAAM,OAAO,CAAC;;qCAEmB,EAAE,KAAK,UAAU,oBAAoB,EAAE,KAAK,WAAW,mBAAmB,EAAE,QAAQ,EAAE,CAAC;EACxH,MAAM,EAAE,MAAM,GAAG,MAAM,KAAK,OAAO,KAAK,IAAI,KAAK;AAEjD,SAAO,KAAK,WAAW;CACxB;;;;;CAMD,MAAM,gBAAgBC,WAAmC;EACvD,MAAM,UAAU,aAAa,KAAK,YAAYD;EAC9C,MAAM,QAAQ,CAAC,qBAAqB,EAAE,QAAQ,CAAC,CAAC;EAChD,MAAM,KAAK,OAAO,KAAK,IAAI,MAAM;CAClC;;;;;CAMD,MAAM,QAAQC,WAAoB;EAChC,MAAM,UAAU,aAAa,KAAK,YAAYD;EAC9C,MAAM,QAAQ,CAAC,cAAc,EAAE,QAAQ,CAAC,CAAC;EACzC,MAAM,KAAK,OAAO,KAAK,IAAI,MAAM;CAClC;AACF"}

package/dist/vectorstore.d.cts

@@ -0,0 +1,300 @@
+import { PostgresEngine } from "./engine.cjs";
+import { BaseIndex, DistanceStrategy, QueryOptions } from "./indexes.cjs";
+import { EmbeddingsInterface } from "@langchain/core/embeddings";
+import { MaxMarginalRelevanceSearchOptions, VectorStore } from "@langchain/core/vectorstores";
+import { Document } from "@langchain/core/documents";
+
+//#region src/vectorstore.d.ts
+interface PostgresVectorStoreArgs {
+  schemaName?: string;
+  contentColumn?: string;
+  embeddingColumn?: string;
+  metadataColumns?: Array<string>;
+  idColumn?: string;
+  distanceStrategy?: DistanceStrategy;
+  k?: number;
+  fetchK?: number;
+  lambdaMult?: number;
+  ignoreMetadataColumns?: Array<string>;
+  metadataJsonColumn?: string;
+  indexQueryOptions?: QueryOptions;
+}
+interface dbConfigArgs {
+  engine: PostgresEngine;
+  tableName: string;
+  dbConfig?: PostgresVectorStoreArgs;
+}
+interface VSArgs {
+  engine: PostgresEngine;
+  tableName: string;
+  schemaName: string;
+  contentColumn: string;
+  embeddingColumn: string;
+  metadataColumns: Array<string>;
+  idColumn: string;
+  distanceStrategy: DistanceStrategy;
+  k: number;
+  fetchK: number;
+  lambdaMult: number;
+  metadataJsonColumn: string;
+  indexQueryOptions?: QueryOptions;
+}
+/**
+ * Google Cloud SQL for PostgreSQL vector store integration.
+ *
+ * Setup:
+ * Install `@langchain/google-cloud-sql-pg`
+ *
+ * ```bash
+ * npm install @langchain/google-cloud-sql-pg
+ * ```
+ *
+ * <details open>
+ * <summary><strong>Instantiate</strong></summary>
+ *
+ * ```typescript
+ * import { Column, PostgresEngine, PostgresEngineArgs, PostgresVectorStore, VectorStoreTableArgs } from "@langchain/google-cloud-sql-pg";
+ * // Or other embeddings
+ * import { OpenAIEmbeddings } from '@langchain/openai';
+ *
+ *
+ * const embeddings = new OpenAIEmbeddings({
+ *   model: "text-embedding-3-small",
+ * });
+ *
+ * const pgArgs: PostgresEngineArgs = {
+ *     user: "db-user",
+ *     password: "password"
+ * }
+ * // Create a shared connection pool
+ * const engine: PostgresEngine = await PostgresEngine.fromInstance(
+ *  "project-id",
+ *  "region",
+ *  "instance-name",
+ *  "database-name",
+ *  pgArgs
+ * );
+ * // (Optional) Specify metadata columns for filtering
+ * // All other metadata will be added to JSON
+ * const vectorStoreTableArgs: VectorStoreTableArgs = {
+ *   metadataColumns: [new Column("baz", "TEXT")],
+ * };
+ * // Create a vector store table
+ * await engine.initVectorstoreTable("my-table", 768, vectorStoreTableArgs);
+ * // Customize the vector store
+ * const pvectorArgs: PostgresVectorStoreArgs = {
+ *     idColumn: "ID_COLUMN",
+ *     contentColumn: "CONTENT_COLUMN",
+ *     embeddingColumn: "EMBEDDING_COLUMN",
+ *     metadataColumns: ["baz"]
+ * }
+ *
+ * const vectorStore = await PostgresVectorStore.initialize(engine, embeddingService, "my-table", pvectorArgs);
+ * ```
+ * </details>
+ *
+ * <br />
+ *
+ * <details>
+ * <summary><strong>Add documents</strong></summary>
+ *
+ * ```typescript
+ * import type { Document } from '@langchain/core/documents';
+ *
+ * const document1 = { pageContent: "foo", metadata: { baz: "bar" } };
+ * const document2 = { pageContent: "thud", metadata: { bar: "baz" } };
+ * const document3 = { pageContent: "i will be deleted :(", metadata: {} };
+ *
+ * const documents: Document[] = [document1, document2, document3];
+ * const ids = ["1", "2", "3"];
+ * await vectorStore.addDocuments(documents, { ids });
+ * ```
+ * </details>
+ *
+ * <br />
+ *
+ * <details>
+ * <summary><strong>Delete documents</strong></summary>
+ *
+ * ```typescript
+ * await vectorStore.delete({ ids: ["3"] });
+ * ```
+ * </details>
+ *
+ * <br />
+ *
+ * <details>
+ * <summary><strong>Similarity search</strong></summary>
+ *
+ * ```typescript
+ * const results = await vectorStore.similaritySearch("thud", 1);
+ * for (const doc of results) {
+ *   console.log(`* ${doc.pageContent} [${JSON.stringify(doc.metadata, null)}]`);
+ * }
+ * // Output:thud [{"baz":"bar"}]
+ * ```
+ * </details>
+ *
+ * <br />
+ *
+ *
+ * <details>
+ * <summary><strong>Similarity search with filter</strong></summary>
+ *
+ * ```typescript
+ * const resultsWithFilter = await vectorStore.similaritySearch("thud", 1, "baz = 'bar'");
+ *
+ * for (const doc of resultsWithFilter) {
+ *   console.log(`* ${doc.pageContent} [${JSON.stringify(doc.metadata, null)}]`);
+ * }
+ * // Output:foo [{"baz":"bar"}]
+ * ```
+ * </details>
+ *
+ * <br />
+ *
+ *
+ * <details>
+ * <summary><strong>Similarity search with score</strong></summary>
+ *
+ * ```typescript
+ * const resultsWithScore = await vectorStore.similaritySearchWithScore("qux", 1);
+ * for (const [doc, score] of resultsWithScore) {
+ *   console.log(`* [SIM=${score.toFixed(6)}] ${doc.pageContent} [${JSON.stringify(doc.metadata, null)}]`);
+ * }
+ * // Output:[SIM=0.000000] qux [{"bar":"baz","baz":"bar"}]
+ * ```
+ * </details>
+ *
+ * <br />
+ *
+ * <details>
+ * <summary><strong>As a retriever</strong></summary>
+ *
+ * ```typescript
+ * const retriever = vectorStore.asRetriever({
+ *   searchType: "mmr", // Leave blank for standard similarity search
+ *   k: 1,
+ * });
+ * const resultAsRetriever = await retriever.invoke("thud");
+ * console.log(resultAsRetriever);
+ *
+ * // Output: [Document({ metadata: { "baz":"bar" }, pageContent: "thud" })]
+ * ```
+ * </details>
+ *
+ * <br />
+ */
+declare class PostgresVectorStore extends VectorStore {
+  FilterType: string;
+  engine: PostgresEngine;
+  embeddings: EmbeddingsInterface;
+  tableName: string;
+  schemaName: string;
+  contentColumn: string;
+  embeddingColumn: string;
+  metadataColumns: Array<string>;
+  idColumn: string;
+  metadataJsonColumn: string;
+  distanceStrategy: DistanceStrategy;
+  k: number;
+  fetchK: number;
+  lambdaMult: number;
+  indexQueryOptions: QueryOptions | undefined;
+  /**
+   * Initializes a new vector store with embeddings and database configuration.
+   *
+   * @param embeddings - Instance of `EmbeddingsInterface` used to embed queries.
+   * @param dbConfig - Configuration settings for the database or storage system.
+   */
+  constructor(embeddings: EmbeddingsInterface, dbConfig: VSArgs);
+  /**
+   * Create a new PostgresVectorStore instance.
+   * @param {PostgresEngine} engine Required - Connection pool engine for managing connections to Cloud SQL for PostgreSQL database.
+   * @param {Embeddings} embeddings Required - Text embedding model to use.
+   * @param {string} tableName Required - Name of an existing table or table to be created.
+   * @param {string} schemaName Database schema name of the table. Defaults to "public".
+   * @param {string} contentColumn Column that represent a Document's page_content. Defaults to "content".
+   * @param {string} embeddingColumn Column for embedding vectors. The embedding is generated from the document value. Defaults to "embedding".
+   * @param {Array<string>} metadataColumns Column(s) that represent a document's metadata.
+   * @param {Array<string>} ignoreMetadataColumns Optional - Column(s) to ignore in pre-existing tables for a document's metadata. Can not be used with metadata_columns.
+   * @param {string} idColumn Column that represents the Document's id. Defaults to "langchain_id".
+   * @param {string} metadataJsonColumn Optional - Column to store metadata as JSON. Defaults to "langchain_metadata".
+   * @param {DistanceStrategy} distanceStrategy Distance strategy to use for vector similarity search. Defaults to COSINE_DISTANCE.
+   * @param {number} k Number of Documents to return from search. Defaults to 4.
+   * @param {number} fetchK Number of Documents to fetch to pass to MMR algorithm.
+   * @param {number} lambdaMult Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5.
+   * @param {QueryOptions} indexQueryOptions Optional - Index query option.
+   * @returns PostgresVectorStore instance.
+   */
+  static initialize(engine: PostgresEngine, embeddings: EmbeddingsInterface, tableName: string, {
+    schemaName,
+    contentColumn,
+    embeddingColumn,
+    metadataColumns,
+    ignoreMetadataColumns,
+    idColumn,
+    metadataJsonColumn,
+    distanceStrategy,
+    k,
+    fetchK,
+    lambdaMult,
+    indexQueryOptions
+  }?: PostgresVectorStoreArgs): Promise<PostgresVectorStore>;
+  static fromTexts(texts: string[], metadatas: object[] | object, embeddings: EmbeddingsInterface, dbConfig: dbConfigArgs): Promise<VectorStore>;
+  static fromDocuments(docs: Document[], embeddings: EmbeddingsInterface, dbConfig: dbConfigArgs): Promise<VectorStore>;
+  addVectors(vectors: number[][], documents: Document[], options?: {
+    ids?: string[];
+  }): Promise<string[] | void>;
+  _vectorstoreType(): string;
+  /**
+   * Adds documents to the vector store, embedding them first through the
+   * `embeddings` instance.
+   *
+   * @param documents - Array of documents to embed and add.
+   * @param options - Optional configuration for embedding and storing documents.
+   * @returns A promise resolving to an array of document IDs or void, based on implementation.
+   * @abstract
+   */
+  addDocuments(documents: Document[], options?: {
+    ids?: string[];
+  }): Promise<string[] | void>;
+  /**
+   * Deletes documents from the vector store based on the specified ids.
+   *
+   * @param params - Flexible key-value pairs defining conditions for document deletion.
+   * @param ids -  Optional: Property of {params} that contains the array of ids to be deleted
+   * @returns A promise that resolves once the deletion is complete.
+   */
+  delete(params: {
+    ids?: string[];
+  }): Promise<void>;
+  similaritySearchVectorWithScore(embedding: number[], k: number, filter?: this["FilterType"]): Promise<[Document, number][]>;
+  private queryCollection;
+  maxMarginalRelevanceSearch(query: string, options: MaxMarginalRelevanceSearchOptions<this["FilterType"]>): Promise<Document[]>;
+  /**
+   * Create an index on the vector store table
+   * @param {BaseIndex} index
+   * @param {string} name Optional
+   * @param {boolean} concurrently Optional
+   */
+  applyVectorIndex(index: BaseIndex, name?: string, concurrently?: boolean): Promise<void>;
+  /**
+   * Check if index exists in the table.
+   * @param {string} indexName Optional - index name
+   */
+  isValidIndex(indexName?: string): Promise<boolean>;
+  /**
+   * Drop the vector index
+   * @param {string} indexName Optional - index name
+   */
+  dropVectorIndex(indexName?: string): Promise<void>;
+  /**
+   * Re-index the vector store table
+   * @param {string} indexName Optional - index name
+   */
+  reIndex(indexName?: string): Promise<void>;
+}
+//#endregion
+export { PostgresVectorStore, PostgresVectorStoreArgs, dbConfigArgs };
+//# sourceMappingURL=vectorstore.d.cts.map