il2cpp-dump-analyzer-mcp/dist/embeddings/supabase-vector-store-fixed.d.ts

Agentic RAG system for analyzing IL2CPP dump.cs files from Unity games

import { SupabaseClient } from '@supabase/supabase-js';
import { Document } from '@langchain/core/documents';
import { Embeddings } from '@langchain/core/embeddings';
import { CodeChunk } from './chunker';
/**
 * Vector store for IL2CPP code chunks using Supabase
 * Fixed version with improved error handling and consistency
 */
export declare class SupabaseIL2CPPVectorStore {
    private embeddings;
    supabaseClient: SupabaseClient;
    private tableName;
    private dimensions;
    private isInitialized;
    private initializationPromise;
    /**
     * Initialize the Supabase vector store
     * @param embeddings Embeddings instance to use
     * @param supabaseUrl Supabase URL
     * @param supabaseKey Supabase API key
     * @param tableName Table name for vector storage
     */
    constructor(embeddings: Embeddings, supabaseUrl: string, supabaseKey: string, tableName?: string);
    /**
     * Get dimensions from embeddings instance with proper fallback
     */
    private getDimensionsFromEmbeddings;
    /**
     * Ensure the vector store is properly initialized
     */
    private ensureInitialized;
    /**
     * Initialize the Supabase table with the correct schema
     */
    private initializeTable;
    /**
     * Create a new instance of the vector store from texts
     */
    static fromTexts(texts: string[], metadatas: Record<string, any>[], embeddings: Embeddings, supabaseUrl: string, supabaseKey: string, tableName?: string): Promise<SupabaseIL2CPPVectorStore>;
    /**
     * Add documents to the vector store with improved error handling
     */
    addDocuments(documents: Document[]): Promise<void>;
    /**
     * Filter out documents that already exist in the database
     */
    private filterExistingDocuments;
    /**
     * Generate embeddings for documents with proper validation
     */
    private generateEmbeddings;
    /**
     * Validate and normalize a single embedding
     */
    private validateEmbedding;
    /**
     * Insert documents in batches with proper error handling
     */
    private insertDocumentsBatch;
    /**
     * Insert a single batch of documents
     */
    private insertSingleBatch;
    /**
     * Add code chunks to the vector store
     */
    addCodeChunks(chunks: CodeChunk[]): Promise<void>;
    /**
     * Search for similar documents based on a query string
     */
    similaritySearch(query: string, k?: number): Promise<Document[]>;
    /**
     * Search for similar documents with scores
     */
    similaritySearchWithScore(query: string, k?: number): Promise<[Document, number][]>;
    /**
     * Get the total number of documents in the vector store
     */
    getDocumentCount(): Promise<number>;
    /**
     * Delete all documents from the vector store
     */
    deleteAll(): Promise<void>;
    /**
     * Get the dimensionality of the embeddings
     */
    getDimension(): number;
    /**
     * Generate a unique hash for a document based on its content and metadata
     */
    private generateDocumentHash;
    /**
     * Check if the vector store is properly configured and accessible
     */
    healthCheck(): Promise<{
        healthy: boolean;
        message: string;
    }>;
}