langchain.js
    Preparing search index...

    Interface TypesenseConfig

    Typesense vector store configuration.

    interface TypesenseConfig {
        columnNames?: {
            metadataColumnNames?: string[];
            pageContent?: string;
            vector?: string;
        };
        schemaName: string;
        searchParams?: MultiSearchRequestSchema;
        typesenseClient: Client;
        import?<T extends Record<string, unknown> = Record<string, unknown>>(
            data: T[],
            collectionName: string,
        ): Promise<void>;
    }

    Hierarchy (View Summary)

    Index

    Properties

    columnNames? schemaName searchParams? typesenseClient

    Methods

    import?

    Properties

    columnNames?: {
        metadataColumnNames?: string[];
        pageContent?: string;
        vector?: string;
    }

    Column names.

    Type Declaration

    • OptionalmetadataColumnNames?: string[]

      Metadata column names.

      []
    • OptionalpageContent?: string

      Page content column name.

      'text'
    • Optionalvector?: string

      Vector column name.

      'vec'
    schemaName: string

    Typesense schema name in which documents will be stored and searched.

    searchParams?: MultiSearchRequestSchema

    Typesense search parameters.

    { q: '*', per_page: 5, query_by: '' }
    typesenseClient: Client

    Typesense client.

    Methods

    • Replace default import function. Default import function will update documents if there is a document with the same id.

      Type Parameters

      • T extends Record<string, unknown> = Record<string, unknown>

      Parameters

      • data: T[]
      • collectionName: string

      Returns Promise<void>

    Settings

    Member Visibility

    On This Page

    Properties
    Methods