plugin-document-parser 1.0.1

package/dist/node_modules/form-data/lib/populate.js

@@ -0,0 +1,10 @@
+'use strict';
+
+// populates missing values
+module.exports = function (dst, src) {
+  Object.keys(src).forEach(function (prop) {
+    dst[prop] = dst[prop] || src[prop]; // eslint-disable-line no-param-reassign
+  });
+
+  return dst;
+};

package/dist/node_modules/form-data/package.json

@@ -0,0 +1 @@
+{"author":"Felix Geisendörfer <felix@debuggable.com> (http://debuggable.com/)","name":"form-data","description":"A library to create readable \"multipart/form-data\" streams. Can be used to submit forms and file uploads to other web applications.","version":"4.0.5","repository":{"type":"git","url":"git://github.com/form-data/form-data.git"},"main":"./lib/form_data","browser":"./lib/browser","typings":"./index.d.ts","scripts":{"pretest":"npm run lint","pretests-only":"rimraf coverage test/tmp","tests-only":"istanbul cover test/run.js","posttests-only":"istanbul report lcov text","test":"npm run tests-only","posttest":"npx npm@'>=10.2' audit --production","lint":"eslint --ext=js,mjs .","report":"istanbul report lcov text","ci-lint":"is-node-modern 8 && npm run lint || is-node-not-modern 8","ci-test":"npm run tests-only && npm run browser && npm run report","predebug":"rimraf coverage test/tmp","debug":"verbose=1 ./test/run.js","browser":"browserify -t browserify-istanbul test/run-browser.js | obake --coverage","check":"istanbul check-coverage coverage/coverage*.json","files":"pkgfiles --sort=name","get-version":"node -e \"console.log(require('./package.json').version)\"","update-readme":"sed -i.bak 's/\\/master\\.svg/\\/v'$(npm --silent run get-version)'.svg/g' README.md","postupdate-readme":"mv README.md.bak READ.ME.md.bak","restore-readme":"mv READ.ME.md.bak README.md","prepublish":"not-in-publish || npm run prepublishOnly","prepack":"npm run update-readme","postpack":"npm run restore-readme","version":"auto-changelog && git add CHANGELOG.md","postversion":"auto-changelog && git add CHANGELOG.md && git commit --no-edit --amend && git tag -f \"v$(node -e \"console.log(require('./package.json').version)\")\""},"engines":{"node":">= 6"},"dependencies":{"asynckit":"^0.4.0","combined-stream":"^1.0.8","es-set-tostringtag":"^2.1.0","hasown":"^2.0.2","mime-types":"^2.1.12"},"devDependencies":{"@ljharb/eslint-config":"^21.4.0","auto-changelog":"^2.5.0","browserify":"^13.3.0","browserify-istanbul":"^2.0.0","coveralls":"^3.1.1","cross-spawn":"^6.0.6","eslint":"^8.57.1","fake":"^0.2.2","far":"^0.0.7","formidable":"^1.2.6","in-publish":"^2.0.1","is-node-modern":"^1.0.0","istanbul":"^0.4.5","js-randomness-predictor":"^1.5.5","obake":"^0.1.2","pkgfiles":"^2.3.2","pre-commit":"^1.2.2","puppeteer":"^1.20.0","request":"~2.87.0","rimraf":"^2.7.1","semver":"^6.3.1","tape":"^5.9.0"},"license":"MIT","auto-changelog":{"output":"CHANGELOG.md","template":"keepachangelog","unreleased":false,"commitLimit":false,"backfillLimit":false,"hideCredit":true},"_lastModified":"2026-04-03T18:52:21.070Z"}

package/dist/server/collections/doc-parser-providers.js

@@ -0,0 +1,137 @@
+/**
+ * This file is part of the NocoBase (R) project.
+ * Copyright (c) 2020-2024 NocoBase Co., Ltd.
+ * Authors: NocoBase Team.
+ *
+ * This project is dual-licensed under AGPL-3.0 and NocoBase Commercial License.
+ * For more information, please refer to: https://www.nocobase.com/agreement.
+ */
+
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var doc_parser_providers_exports = {};
+__export(doc_parser_providers_exports, {
+  default: () => doc_parser_providers_default
+});
+module.exports = __toCommonJS(doc_parser_providers_exports);
+var import_database = require("@nocobase/database");
+var doc_parser_providers_default = (0, import_database.defineCollection)({
+  name: "docParserProviders",
+  title: "Document Parser Providers",
+  fields: [
+    {
+      name: "title",
+      type: "string"
+    },
+    {
+      name: "enabled",
+      type: "boolean",
+      defaultValue: true
+    },
+    // ── Endpoint ──────────────────────────────────────────────────────────────
+    {
+      name: "apiEndpoint",
+      type: "string",
+      comment: "Full URL, e.g. https://ocr.example.com/v1/parse"
+    },
+    // ── Authentication ────────────────────────────────────────────────────────
+    {
+      name: "authType",
+      type: "string",
+      defaultValue: "bearer",
+      comment: "'bearer' | 'api-key-header' | 'basic' | 'custom-headers' | 'none'"
+    },
+    {
+      // Encrypted at rest via NocoBase password field
+      name: "apiKey",
+      type: "password",
+      allowNull: true,
+      comment: "Used for bearer / api-key-header auth"
+    },
+    {
+      name: "authConfig",
+      type: "json",
+      defaultValue: {},
+      comment: JSON.stringify({
+        headerName: "X-Api-Key",
+        // for api-key-header
+        username: "",
+        // for basic auth
+        password: "",
+        // for basic auth
+        customHeaders: {}
+        // for custom-headers: { "X-Foo": "bar" }
+      })
+    },
+    // ── Request format ────────────────────────────────────────────────────────
+    {
+      name: "requestFormat",
+      type: "string",
+      defaultValue: "multipart",
+      comment: "'multipart' | 'json-base64' | 'url'"
+    },
+    {
+      name: "requestConfig",
+      type: "json",
+      defaultValue: {},
+      comment: JSON.stringify({
+        // multipart
+        fileFieldName: "file",
+        filenameFieldName: "",
+        // optional extra field for filename
+        mimetypeFieldName: "",
+        // optional extra field for mimetype
+        extraFields: {},
+        // extra form fields
+        // json-base64
+        base64FieldPath: "file",
+        // e.g. "document.content"
+        filenameFieldPath: "filename",
+        mimetypeFieldPath: "mimetype",
+        extraBody: {},
+        // url (send download URL instead of file bytes)
+        urlFieldPath: "url"
+      })
+    },
+    // ── Response extraction ───────────────────────────────────────────────────
+    {
+      name: "responseTextPath",
+      type: "string",
+      defaultValue: "text",
+      comment: "Dot-path into the JSON response body, e.g. 'data.text' or 'result.pages[0].content'"
+    },
+    // ── Scope ─────────────────────────────────────────────────────────────────
+    {
+      name: "supportedMimetypes",
+      type: "json",
+      defaultValue: [],
+      comment: 'Empty = handle everything routed to external. e.g. ["application/pdf"]'
+    },
+    {
+      name: "timeout",
+      type: "integer",
+      defaultValue: 6e4,
+      comment: "HTTP request timeout in milliseconds"
+    },
+    {
+      name: "options",
+      type: "json",
+      defaultValue: {}
+    }
+  ]
+});

package/dist/server/collections/doc-parser-settings.js

@@ -0,0 +1,85 @@
+/**
+ * This file is part of the NocoBase (R) project.
+ * Copyright (c) 2020-2024 NocoBase Co., Ltd.
+ * Authors: NocoBase Team.
+ *
+ * This project is dual-licensed under AGPL-3.0 and NocoBase Commercial License.
+ * For more information, please refer to: https://www.nocobase.com/agreement.
+ */
+
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var doc_parser_settings_exports = {};
+__export(doc_parser_settings_exports, {
+  default: () => doc_parser_settings_default
+});
+module.exports = __toCommonJS(doc_parser_settings_exports);
+var import_database = require("@nocobase/database");
+var doc_parser_settings_default = (0, import_database.defineCollection)({
+  name: "docParserSettings",
+  title: "Document Parser Settings",
+  fields: [
+    {
+      name: "mode",
+      type: "string",
+      defaultValue: "default",
+      comment: "'default' | 'internal' | 'external'"
+    },
+    {
+      // FK to docParserProviders — which external provider is active
+      name: "activeProviderId",
+      type: "bigInt",
+      allowNull: true
+    },
+    {
+      // When internal/external parsing fails, fall back to the default provider logic
+      name: "fallbackToDefault",
+      type: "boolean",
+      defaultValue: true
+    },
+    {
+      // Images are always passed through to the default provider (they don't need OCR)
+      name: "imagePassThrough",
+      type: "boolean",
+      defaultValue: true
+    },
+    {
+      // Optional: restrict which extnames this plugin handles (empty = all non-image)
+      name: "includedExtnames",
+      type: "json",
+      defaultValue: [],
+      comment: 'e.g. [".pdf", ".docx"] \u2014 empty means all non-image files'
+    },
+    {
+      name: "options",
+      type: "json",
+      defaultValue: {}
+    },
+    {
+      /**
+       * When true and plugin-docpixie is active:
+       * - Trigger docpixie:processDocument (async indexing)
+       * - Return a metadata reference block instead of full text
+       * - LLM is instructed to call docpixie:query tool for retrieval
+       */
+      name: "useDocpixie",
+      type: "boolean",
+      defaultValue: false
+    }
+  ]
+});

package/dist/server/index.js

@@ -0,0 +1,51 @@
+/**
+ * This file is part of the NocoBase (R) project.
+ * Copyright (c) 2020-2024 NocoBase Co., Ltd.
+ * Authors: NocoBase Team.
+ *
+ * This project is dual-licensed under AGPL-3.0 and NocoBase Commercial License.
+ * For more information, please refer to: https://www.nocobase.com/agreement.
+ */
+
+var __create = Object.create;
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __getProtoOf = Object.getPrototypeOf;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toESM = (mod, isNodeMode, target) => (target = mod != null ? __create(__getProtoOf(mod)) : {}, __copyProps(
+  // If the importer is in node compatibility mode or this is not an ESM
+  // file that has been converted to a CommonJS file using a Babel-
+  // compatible transform (i.e. "__esModule" has not been set), then set
+  // "default" to the CommonJS "module.exports" for node compatibility.
+  isNodeMode || !mod || !mod.__esModule ? __defProp(target, "default", { value: mod, enumerable: true }) : target,
+  mod
+));
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var server_exports = {};
+__export(server_exports, {
+  InternalParserRegistry: () => import_internal_parser_registry.InternalParserRegistry,
+  PluginDocumentParserServer: () => import_plugin2.PluginDocumentParserServer,
+  default: () => import_plugin.default
+});
+module.exports = __toCommonJS(server_exports);
+var import_plugin = __toESM(require("./plugin"));
+var import_plugin2 = require("./plugin");
+var import_internal_parser_registry = require("./services/internal-parser-registry");
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  InternalParserRegistry,
+  PluginDocumentParserServer
+});

package/dist/server/plugin.js

@@ -0,0 +1,181 @@
+/**
+ * This file is part of the NocoBase (R) project.
+ * Copyright (c) 2020-2024 NocoBase Co., Ltd.
+ * Authors: NocoBase Team.
+ *
+ * This project is dual-licensed under AGPL-3.0 and NocoBase Commercial License.
+ * For more information, please refer to: https://www.nocobase.com/agreement.
+ */
+
+var __create = Object.create;
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __getProtoOf = Object.getPrototypeOf;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toESM = (mod, isNodeMode, target) => (target = mod != null ? __create(__getProtoOf(mod)) : {}, __copyProps(
+  // If the importer is in node compatibility mode or this is not an ESM
+  // file that has been converted to a CommonJS file using a Babel-
+  // compatible transform (i.e. "__esModule" has not been set), then set
+  // "default" to the CommonJS "module.exports" for node compatibility.
+  isNodeMode || !mod || !mod.__esModule ? __defProp(target, "default", { value: mod, enumerable: true }) : target,
+  mod
+));
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var plugin_exports = {};
+__export(plugin_exports, {
+  PluginDocumentParserServer: () => PluginDocumentParserServer,
+  default: () => plugin_default
+});
+module.exports = __toCommonJS(plugin_exports);
+var import_server = require("@nocobase/server");
+var import_path = require("path");
+var import_axios = __toESM(require("axios"));
+var import_internal_parser_registry = require("./services/internal-parser-registry");
+var import_builtin_ai_handler = require("./services/builtin-ai-handler");
+var import_parse_router = require("./services/parse-router");
+var import_docParserProviders = require("./resource/docParserProviders");
+class PluginDocumentParserServer extends import_server.Plugin {
+  /**
+   * Public registry — other plugins register their format handlers here:
+   *
+   *   const docParser = this.pm.get(PluginDocumentParserServer);
+   *   docParser.internalParserRegistry.register({ name, supports, parse });
+   */
+  internalParserRegistry = new import_internal_parser_registry.InternalParserRegistry();
+  parseRouter;
+  // ── Lifecycle ─────────────────────────────────────────────────────────────
+  async beforeLoad() {
+    this.internalParserRegistry.register(
+      new import_builtin_ai_handler.BuiltinAIDocumentHandler(() => {
+        const aiPlugin = this.pm.get("@nocobase/plugin-ai");
+        return aiPlugin == null ? void 0 : aiPlugin.documentLoaders;
+      })
+    );
+  }
+  async load() {
+    await this.importCollections((0, import_path.resolve)(__dirname, "collections"));
+    this.parseRouter = new import_parse_router.ParseRouter(
+      () => this.db.getRepository("docParserSettings"),
+      () => this.db.getRepository("docParserProviders"),
+      this.internalParserRegistry,
+      this.fetchFileBuffer.bind(this),
+      () => {
+        const p = this.pm.get("@nocobase/plugin-docpixie");
+        return (p == null ? void 0 : p.service) ? p : null;
+      }
+    );
+    this.wrapAIManager();
+    this.app.resourceManager.define({
+      name: "docParserProviders",
+      actions: {
+        testConnection: import_docParserProviders.testConnection
+      }
+    });
+    this.app.resourceManager.define({
+      name: "docParserSettings",
+      actions: {
+        get: import_docParserProviders.getSettings,
+        save: import_docParserProviders.saveSettings
+      }
+    });
+    this.app.acl.allow("docParserProviders", ["list", "create", "update", "destroy", "get", "testConnection"], "loggedIn");
+    this.app.acl.allow("docParserSettings", ["get", "save"], "loggedIn");
+  }
+  // ── AIManager patching ────────────────────────────────────────────────────
+  /**
+   * Wrap AIManager.registerLLMProvider so that every provider class — including
+   * those registered AFTER this plugin loads (e.g. plugin-custom-llm) — gets
+   * its `parseAttachment` intercepted.
+   *
+   * Additionally, iterate providers already registered (plugin-ai built-ins:
+   * OpenAI, Anthropic, etc.) and wrap them immediately.
+   */
+  wrapAIManager() {
+    const aiPlugin = this.pm.get("@nocobase/plugin-ai");
+    if (!(aiPlugin == null ? void 0 : aiPlugin.aiManager)) {
+      this.log.warn("[DocumentParser] plugin-ai not found \u2014 parseAttachment interception skipped");
+      return;
+    }
+    const aiManager = aiPlugin.aiManager;
+    const self = this;
+    const originalRegister = aiManager.registerLLMProvider.bind(aiManager);
+    aiManager.registerLLMProvider = (name, meta) => {
+      return originalRegister(name, { ...meta, provider: self.wrapProviderClass(meta.provider) });
+    };
+    for (const [name, meta] of aiManager.llmProviders.entries()) {
+      aiManager.llmProviders.set(name, { ...meta, provider: self.wrapProviderClass(meta.provider) });
+    }
+    this.log.info(`[DocumentParser] Wrapped ${aiManager.llmProviders.size} LLM providers`);
+  }
+  /**
+   * Create a subclass of `OriginalProviderClass` that overrides `parseAttachment`
+   * to go through our router first.  Uses `super.parseAttachment` as the default
+   * parser fallback — this correctly handles providers that already override the
+   * method (e.g. CustomLLMProvider, AnthropicProvider…).
+   */
+  wrapProviderClass(OriginalClass) {
+    const self = this;
+    return class extends OriginalClass {
+      async parseAttachment(ctx, attachment) {
+        return self.parseRouter.route(
+          ctx,
+          attachment,
+          () => super.parseAttachment(ctx, attachment)
+        );
+      }
+    };
+  }
+  // ── File buffer helper ────────────────────────────────────────────────────
+  /**
+   * Fetch the raw bytes of an attachment using the file-manager plugin,
+   * returning both the buffer and the resolved URL.
+   */
+  async fetchFileBuffer(ctx, attachment) {
+    var _a, _b;
+    const fileManager = this.app.pm.get("file-manager");
+    const rawUrl = await fileManager.getFileURL(attachment);
+    const url = decodeURIComponent(rawUrl);
+    if (url.startsWith("http://") || url.startsWith("https://")) {
+      const referer = ((_a = ctx.get) == null ? void 0 : _a.call(ctx, "referer")) || "";
+      const ua = ((_b = ctx.get) == null ? void 0 : _b.call(ctx, "user-agent")) || "";
+      const response = await import_axios.default.get(url, {
+        responseType: "arraybuffer",
+        timeout: 6e4,
+        headers: { referer, "User-Agent": ua }
+      });
+      return { buffer: Buffer.from(response.data), url };
+    }
+    const { resolve: resolvePath, sep } = require("path");
+    const { readFile } = require("fs/promises");
+    let localPath = url;
+    const appPublicPath = (process.env.APP_PUBLIC_PATH || "/").replace(/\/+$/, "");
+    if (appPublicPath && localPath.startsWith(appPublicPath + "/")) {
+      localPath = localPath.slice(appPublicPath.length);
+    }
+    const storageRoot = resolvePath(process.cwd());
+    const absPath = resolvePath(storageRoot, localPath.replace(/^\//, ""));
+    if (!absPath.startsWith(storageRoot + sep) && absPath !== storageRoot) {
+      throw new Error(`[DocumentParser] Attachment path escapes storage root: ${localPath}`);
+    }
+    const buffer = await readFile(absPath);
+    return { buffer, url };
+  }
+}
+var plugin_default = PluginDocumentParserServer;
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  PluginDocumentParserServer
+});

package/dist/server/resource/docParserProviders.js

@@ -0,0 +1,91 @@
+/**
+ * This file is part of the NocoBase (R) project.
+ * Copyright (c) 2020-2024 NocoBase Co., Ltd.
+ * Authors: NocoBase Team.
+ *
+ * This project is dual-licensed under AGPL-3.0 and NocoBase Commercial License.
+ * For more information, please refer to: https://www.nocobase.com/agreement.
+ */
+
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var docParserProviders_exports = {};
+__export(docParserProviders_exports, {
+  getSettings: () => getSettings,
+  saveSettings: () => saveSettings,
+  testConnection: () => testConnection
+});
+module.exports = __toCommonJS(docParserProviders_exports);
+var import_external_ocr_client = require("../services/external-ocr-client");
+async function testConnection(ctx, next) {
+  const { filterByTk } = ctx.action.params;
+  const repo = ctx.db.getRepository("docParserProviders");
+  const record = await repo.findById(filterByTk);
+  if (!record) {
+    ctx.throw(404, "Provider not found");
+    return;
+  }
+  const result = await (0, import_external_ocr_client.testOcrProviderConnection)({
+    apiEndpoint: record.get("apiEndpoint"),
+    authType: record.get("authType"),
+    apiKey: record.get("apiKey"),
+    authConfig: record.get("authConfig") ?? {},
+    timeout: Math.min(record.get("timeout") ?? 1e4, 15e3)
+    // cap at 15s for test
+  });
+  ctx.body = result;
+  await next();
+}
+async function getSettings(ctx, next) {
+  const repo = ctx.db.getRepository("docParserSettings");
+  let record = await repo.findOne({});
+  if (!record) {
+    record = await repo.create({
+      values: {
+        mode: "default",
+        fallbackToDefault: true,
+        imagePassThrough: true,
+        includedExtnames: []
+      }
+    });
+  }
+  ctx.body = record;
+  await next();
+}
+async function saveSettings(ctx, next) {
+  var _a, _b;
+  const repo = ctx.db.getRepository("docParserSettings");
+  const body = ctx.request.body;
+  let record = await repo.findOne({});
+  if (!record) {
+    record = await repo.create({ values: body });
+  } else {
+    await repo.update({ filter: { id: record.get("id") }, values: body });
+    record = await repo.findOne({});
+  }
+  const plugin = ctx.app.pm.get("@nocobase/plugin-document-parser");
+  (_b = (_a = plugin == null ? void 0 : plugin.parseRouter) == null ? void 0 : _a.invalidateSettingsCache) == null ? void 0 : _b.call(_a);
+  ctx.body = record;
+  await next();
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  getSettings,
+  saveSettings,
+  testConnection
+});

package/dist/server/services/builtin-ai-handler.js

@@ -0,0 +1,63 @@
+/**
+ * This file is part of the NocoBase (R) project.
+ * Copyright (c) 2020-2024 NocoBase Co., Ltd.
+ * Authors: NocoBase Team.
+ *
+ * This project is dual-licensed under AGPL-3.0 and NocoBase Commercial License.
+ * For more information, please refer to: https://www.nocobase.com/agreement.
+ */
+
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var builtin_ai_handler_exports = {};
+__export(builtin_ai_handler_exports, {
+  BuiltinAIDocumentHandler: () => BuiltinAIDocumentHandler
+});
+module.exports = __toCommonJS(builtin_ai_handler_exports);
+const AI_SUPPORTED_EXTNAMES = /* @__PURE__ */ new Set([".pdf", ".ppt", ".pptx", ".doc", ".docx", ".txt"]);
+class BuiltinAIDocumentHandler {
+  constructor(getDocumentLoaders) {
+    this.getDocumentLoaders = getDocumentLoaders;
+  }
+  name = "builtin-ai-document-loader";
+  supports(attachment) {
+    const ext = this.resolveExtname(attachment);
+    return AI_SUPPORTED_EXTNAMES.has(ext);
+  }
+  async parse(attachment, _ctx) {
+    const loaders = this.getDocumentLoaders();
+    const result = await loaders.cached.load(attachment);
+    if (!result.supported) {
+      return { text: "", handled: false };
+    }
+    return {
+      text: result.text ?? "",
+      handled: true
+    };
+  }
+  resolveExtname(attachment) {
+    if (attachment.extname) return attachment.extname.toLowerCase();
+    const name = attachment.filename ?? attachment.name ?? "";
+    const idx = name.lastIndexOf(".");
+    return idx >= 0 ? name.slice(idx).toLowerCase() : "";
+  }
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  BuiltinAIDocumentHandler
+});