🔍 refactor: OCR Fully Optional with Defaults for "Upload as Text" (#9856)

* refactor: move `loadOCRConfig` from `packages/data-provider` to `packages/api` and return `undefined` if not explicitly configured

* fix: loadOCRConfig import from @librechat/api

* refactor: update defaultTextMimeTypes to support virtually all file types for text parsing

* fix: improve OCR capability check and error message for unsupported file types

* ci: remove unnecessary ocr expectation from AppService test

api/server/services/AppService.js

@@ -1,16 +1,12 @@
+const { FileSources, EModelEndpoint, getConfigDefaults } = require('librechat-data-provider');
 const {
   isEnabled,
+  loadOCRConfig,
   loadMemoryConfig,
   agentsConfigSetup,
   loadWebSearchConfig,
   loadDefaultInterface,
 } = require('@librechat/api');
-const {
-  FileSources,
-  loadOCRConfig,
-  EModelEndpoint,
-  getConfigDefaults,
-} = require('librechat-data-provider');
 const {
   checkWebSearchConfig,
   checkVariables,

api/server/services/AppService.spec.js

@@ -142,7 +142,6 @@ describe('AppService', () => {
         turnstileConfig: mockedTurnstileConfig,
         modelSpecs: undefined,
         paths: expect.anything(),
-        ocr: expect.anything(),
         imageOutputType: expect.any(String),
         fileConfig: undefined,
         secureImageLinks: undefined,

api/server/services/Files/process.js

@@ -594,10 +594,9 @@ const processAgentFileUpload = async ({ req, res, metadata }) => {
 
     const fileConfig = mergeFileConfig(appConfig.fileConfig);
 
-    const shouldUseOCR = fileConfig.checkType(
+    const shouldUseOCR =
-      file.mimetype,
+      appConfig?.ocr != null &&
-      fileConfig.ocr?.supportedMimeTypes || [],
+      fileConfig.checkType(file.mimetype, fileConfig.ocr?.supportedMimeTypes || []);
-    );
 
     if (shouldUseOCR && !(await checkCapability(req, AgentCapabilities.ocr))) {
       throw new Error('OCR capability is not enabled for Agents');
@@ -626,7 +625,7 @@ const processAgentFileUpload = async ({ req, res, metadata }) => {
     );
 
     if (!shouldUseText) {
-      throw new Error(`File type ${file.mimetype} is not supported for OCR or text parsing`);
+      throw new Error(`File type ${file.mimetype} is not supported for text parsing.`);
     }
 
     const { text, bytes } = await parseText({ req, file, file_id });

packages/api/src/files/index.ts

@@ -1,4 +1,5 @@
-export * from './mistral/crud';
 export * from './audio';
-export * from './text';
+export * from './mistral/crud';
+export * from './ocr';
 export * from './parse';
+export * from './text';

packages/api/src/files/mistral/crud.ts

@@ -303,7 +303,7 @@ async function loadAuthConfig(context: OCRContext): Promise<AuthConfig> {
 /**
  * Gets the model configuration
  */
-function getModelConfig(ocrConfig: TCustomConfig['ocr']): string {
+function getModelConfig(ocrConfig?: TCustomConfig['ocr']): string {
   const modelConfig = ocrConfig?.mistralModel || '';
 
   if (!modelConfig.trim()) {

packages/api/src/files/ocr.ts

@@ -1,7 +1,8 @@
-import type { TCustomConfig } from '../src/config';
+import { OCRStrategy } from 'librechat-data-provider';
-import { OCRStrategy } from '../src/config';
+import type { TCustomConfig } from 'librechat-data-provider';
 
-export function loadOCRConfig(config: TCustomConfig['ocr']): TCustomConfig['ocr'] {
+export function loadOCRConfig(config?: TCustomConfig['ocr']): TCustomConfig['ocr'] | undefined {
+  if (!config) return;
   const baseURL = config?.baseURL ?? '';
   const apiKey = config?.apiKey ?? '';
   const mistralModel = config?.mistralModel ?? '';

packages/data-provider/src/file-config.ts

@@ -133,7 +133,7 @@ export const defaultOCRMimeTypes = [
   /^application\/epub\+zip$/,
 ];
 
-export const defaultTextMimeTypes = [textMimeTypes];
+export const defaultTextMimeTypes = [/^[\w.-]+\/[\w.-]+$/];
 
 export const defaultSTTMimeTypes = [audioMimeTypes];
 

packages/data-provider/src/index.ts

@@ -9,7 +9,6 @@ export * from './messages';
 export * from './artifacts';
 /* schema helpers  */
 export * from './parsers';
-export * from './ocr';
 /* custom/dynamic configurations  */
 export * from './generate';
 export * from './models';