@retrivora-ai/rag-engine 0.1.4 → 0.1.5

package/dist/DocumentChunker-D9-fObJp.d.mts

@@ -1,155 +0,0 @@
-import { e as RagConfig, C as ChatMessage, b as ChatResponse, d as IngestDocument, g as VectorDBConfig, f as UpsertDocument, V as VectorMatch, L as LLMConfig, c as EmbeddingConfig, I as ILLMProvider } from './Pipeline-Bo6CUCox.mjs';
-
-/**
- * VectorPlugin — main orchestrator class.
- * This is the primary interface for host applications.
- */
-declare class VectorPlugin {
-    private config;
-    private pipeline;
-    /**
-     * Initializes the plugin with the host configuration.
-     * @param hostConfig - Configuration object passed from the host application.
-     */
-    constructor(hostConfig?: Partial<RagConfig>);
-    /**
-     * Get the current resolved configuration.
-     */
-    getConfig(): RagConfig;
-    /**
-     * Run a chat query.
-     */
-    chat(message: string, history?: ChatMessage[], namespace?: string): Promise<ChatResponse>;
-    /**
-     * Ingest documents into the vector database.
-     */
-    ingest(documents: IngestDocument[], namespace?: string): Promise<Array<{
-        docId: string;
-        chunksIngested: number;
-    }>>;
-    /**
-     * Check the health of the connected providers.
-     */
-    health(): Promise<Record<string, boolean>>;
-}
-
-/**
- * ConfigResolver — validates and normalizes host configuration.
- * It merges host-provided config with environment defaults.
- */
-declare class ConfigResolver {
-    /**
-     * Resolves the final configuration by merging host-provided config with defaults.
-     * @param hostConfig - Partial configuration passed from the host application.
-     */
-    static resolve(hostConfig?: Partial<RagConfig>): RagConfig;
-    /**
-     * Validates the configuration for required fields.
-     */
-    static validate(config: RagConfig): void;
-}
-
-/**
- * BaseVectorProvider — Abstract base class for all vector database providers.
- * Each provider (Pinecone, Milvus, Redis, etc.) must extend this class.
- */
-declare abstract class BaseVectorProvider {
-    protected readonly config: VectorDBConfig;
-    protected readonly indexName: string;
-    constructor(config: VectorDBConfig);
-    /**
-     * Initialise the connection (create tables, verify index, etc.)
-     */
-    abstract initialize(): Promise<void>;
-    /**
-     * Insert or update a single vector + content pair.
-     */
-    abstract upsert(doc: UpsertDocument, namespace?: string): Promise<void>;
-    /**
-     * Batch upsert for efficient ingestion of many documents.
-     */
-    abstract batchUpsert(docs: UpsertDocument[], namespace?: string): Promise<void>;
-    /**
-     * Find the top-K most similar vectors to the query vector.
-     */
-    abstract query(vector: number[], topK: number, namespace?: string, filter?: Record<string, unknown>): Promise<VectorMatch[]>;
-    /**
-     * Delete a stored vector by ID.
-     */
-    abstract delete(id: string, namespace?: string): Promise<void>;
-    /**
-     * Delete all vectors in a namespace (full data reset for a project).
-     */
-    abstract deleteNamespace(namespace: string): Promise<void>;
-    /**
-     * Check if the underlying DB is reachable.
-     */
-    abstract ping(): Promise<boolean>;
-    /**
-     * Gracefully close the connection.
-     */
-    abstract disconnect(): Promise<void>;
-}
-
-/**
- * ProviderRegistry — dynamic provider loader for Vector DBs and LLMs.
- */
-declare class ProviderRegistry {
-    private static vectorProviders;
-    /**
-     * Register a custom vector provider.
-     */
-    static registerVectorProvider(name: string, providerClass: new (config: VectorDBConfig) => BaseVectorProvider): void;
-    /**
-     * Creates a vector database provider based on the configuration.
-     */
-    static createVectorProvider(config: VectorDBConfig): Promise<BaseVectorProvider>;
-    /**
-     * Creates an LLM provider based on the configuration.
-     */
-    static createLLMProvider(llmConfig: LLMConfig, embeddingConfig?: EmbeddingConfig): ILLMProvider;
-}
-
-/**
- * DocumentChunker — splits long text into overlapping chunks
- * suitable for vector embedding and retrieval.
- *
- * Strategy: sentence-aware sliding window
- *   1. Split on sentence boundaries
- *   2. Accumulate sentences into chunks up to `chunkSize` characters
- *   3. Carry over `chunkOverlap` characters from the previous chunk
- */
-interface Chunk {
-    id: string;
-    content: string;
-    metadata?: Record<string, unknown>;
-}
-interface ChunkOptions {
-    /** Target chunk size in characters (default 1000) */
-    chunkSize?: number;
-    /** Overlap between consecutive chunks in characters (default 200) */
-    chunkOverlap?: number;
-    /** Source document identifier used as ID prefix */
-    docId?: string;
-    /** Extra metadata to attach to every chunk */
-    metadata?: Record<string, unknown>;
-}
-declare class DocumentChunker {
-    private readonly chunkSize;
-    private readonly chunkOverlap;
-    constructor(chunkSize?: number, chunkOverlap?: number);
-    /**
-     * Split a single text string into overlapping chunks.
-     */
-    chunk(text: string, options?: ChunkOptions): Chunk[];
-    /**
-     * Chunk multiple documents at once.
-     */
-    chunkMany(documents: Array<{
-        content: string;
-        docId?: string;
-        metadata?: Record<string, unknown>;
-    }>): Chunk[];
-}
-
-export { BaseVectorProvider as B, type Chunk as C, DocumentChunker as D, ProviderRegistry as P, VectorPlugin as V, type ChunkOptions as a, ConfigResolver as b };

package/dist/chunk-FGGSVVSY.mjs

@@ -1,162 +0,0 @@
-import {
-  VectorPlugin
-} from "./chunk-AIAB2IEE.mjs";
-import {
-  __spreadProps,
-  __spreadValues
-} from "./chunk-I4E63NIC.mjs";
-
-// src/handlers/index.ts
-import { NextResponse } from "next/server";
-
-// src/utils/DocumentParser.ts
-var DocumentParser = class {
-  /**
-   * Extract text from a File or Buffer based on its type.
-   */
-  static async parse(file, fileName, mimeType) {
-    var _a;
-    const extension = (_a = fileName.split(".").pop()) == null ? void 0 : _a.toLowerCase();
-    if (extension === "txt" || extension === "md" || mimeType === "text/plain" || mimeType === "text/markdown") {
-      return this.readAsText(file);
-    }
-    if (extension === "json" || mimeType === "application/json") {
-      const text = await this.readAsText(file);
-      try {
-        const obj = JSON.parse(text);
-        return JSON.stringify(obj, null, 2);
-      } catch (e) {
-        return text;
-      }
-    }
-    if (extension === "csv" || mimeType === "text/csv") {
-      return this.readAsText(file);
-    }
-    if (extension === "pdf" || mimeType === "application/pdf") {
-      try {
-        const pdf = await import("pdf-parse/lib/pdf-parse.js");
-        const buffer = Buffer.isBuffer(file) ? file : Buffer.from(await file.arrayBuffer());
-        const data = await pdf.default(buffer);
-        return data.text;
-      } catch (err) {
-        console.warn('[DocumentParser] PDF parsing failed. Make sure "pdf-parse" is installed.', err);
-        return `[PDF Parsing Error for ${fileName}]`;
-      }
-    }
-    if (extension === "docx" || mimeType === "application/vnd.openxmlformats-officedocument.wordprocessingml.document") {
-      try {
-        const mammoth = await import("mammoth");
-        const buffer = Buffer.isBuffer(file) ? file : Buffer.from(await file.arrayBuffer());
-        const result = await mammoth.extractRawText({ buffer });
-        return result.value;
-      } catch (err) {
-        console.warn('[DocumentParser] DOCX parsing failed. Make sure "mammoth" is installed.', err);
-        return `[DOCX Parsing Error for ${fileName}]`;
-      }
-    }
-    try {
-      return await this.readAsText(file);
-    } catch (e) {
-      throw new Error(`[DocumentParser] Unsupported file format: ${fileName} (${mimeType})`);
-    }
-  }
-  static async readAsText(file) {
-    if (Buffer.isBuffer(file)) {
-      return file.toString("utf-8");
-    }
-    return await file.text();
-  }
-};
-
-// src/handlers/index.ts
-function createChatHandler(config) {
-  const plugin = new VectorPlugin(config);
-  return async function POST(req) {
-    try {
-      const body = await req.json();
-      const { message, history = [], namespace } = body;
-      if (!(message == null ? void 0 : message.trim())) {
-        return NextResponse.json({ error: "message is required" }, { status: 400 });
-      }
-      const result = await plugin.chat(message, history, namespace);
-      return NextResponse.json(result);
-    } catch (err) {
-      const message = err instanceof Error ? err.message : "Internal server error";
-      return NextResponse.json({ error: message }, { status: 500 });
-    }
-  };
-}
-function createIngestHandler(config) {
-  const plugin = new VectorPlugin(config);
-  return async function POST(req) {
-    try {
-      const body = await req.json();
-      const { documents, namespace } = body;
-      if (!Array.isArray(documents) || documents.length === 0) {
-        return NextResponse.json({ error: "documents array is required" }, { status: 400 });
-      }
-      const results = await plugin.ingest(documents, namespace);
-      return NextResponse.json({ results });
-    } catch (err) {
-      const message = err instanceof Error ? err.message : "Internal server error";
-      return NextResponse.json({ error: message }, { status: 500 });
-    }
-  };
-}
-function createHealthHandler(config) {
-  const plugin = new VectorPlugin(config);
-  return async function GET() {
-    try {
-      const health = await plugin.health();
-      const status = health.vectorDB && health.llm ? "ok" : "degraded";
-      return NextResponse.json(__spreadProps(__spreadValues({
-        status
-      }, health), {
-        timestamp: (/* @__PURE__ */ new Date()).toISOString()
-      }));
-    } catch (err) {
-      const message = err instanceof Error ? err.message : "Unknown error";
-      return NextResponse.json({ status: "error", error: message }, { status: 500 });
-    }
-  };
-}
-function createUploadHandler(config) {
-  const plugin = new VectorPlugin(config);
-  return async function POST(req) {
-    try {
-      const formData = await req.formData();
-      const files = formData.getAll("files");
-      const namespace = formData.get("namespace") || void 0;
-      if (!files || files.length === 0) {
-        return NextResponse.json({ error: "No files provided" }, { status: 400 });
-      }
-      const documents = await Promise.all(
-        files.map(async (file) => {
-          const content = await DocumentParser.parse(file, file.name, file.type);
-          return {
-            docId: file.name,
-            content,
-            metadata: {
-              fileName: file.name,
-              fileSize: file.size,
-              fileType: file.type,
-              uploadedAt: (/* @__PURE__ */ new Date()).toISOString()
-            }
-          };
-        })
-      );
-      const results = await plugin.ingest(documents, namespace);
-      return NextResponse.json({ message: "Upload successful", results });
-    } catch (err) {
-      const message = err instanceof Error ? err.message : "Upload failed";
-      return NextResponse.json({ error: message }, { status: 500 });
-    }
-  };
-}
-
-export {
-  createChatHandler,
-  createIngestHandler,
-  createHealthHandler,
-  createUploadHandler
-};