文档加载器 Document Loaders

import type { Document } from "@langchain/core/documents";
 
// 每个 Document 都是这样的形状
const doc: Document = {
  pageContent: "Langchain.js 是 JS 的 LLM 框架……",
  metadata: { source: "intro.md", page: 1 },
};

import { TextLoader } from "langchain/document_loaders/fs/text";
import { DirectoryLoader } from "langchain/document_loaders/fs/directory";
 
const loader = new TextLoader("./docs/intro.md");
const docs = await loader.load(); // Document[]
 
// 一次性加载整个目录
const dirLoader = new DirectoryLoader("./docs", {
  ".md": (path) => new TextLoader(path),
});
const allDocs = await dirLoader.load();

import { PDFLoader } from "@langchain/community/document_loaders/fs/pdf";
 
const pdfLoader = new PDFLoader("./report.pdf");
const pages = await pdfLoader.load();
console.log(pages[0].metadata.loc.pageNumber); // 第几页

import { WebBaseLoader } from "@langchain/community/document_loaders/web/cheerio";
 
const webLoader = new WebBaseLoader("https://js.langchain.com/docs/");
const webDocs = await webLoader.load();