From 086d53d72c2ba91ff90e4e22a4e6e7881ccf2f7c Mon Sep 17 00:00:00 2001 From: Rene Date: Sun, 25 Oct 2026 18:42:59 +0100 Subject: [PATCH] fix(client): fall back to extension MIME inference when SharePoint returns octet-stream --- .../Files/__tests__/sharepoint.spec.ts | 58 +++++++++++++++++++ client/src/data-provider/Files/sharepoint.ts | 26 ++++++++- 2 files changed, 81 insertions(+), 3 deletions(-) create mode 100644 client/src/data-provider/Files/__tests__/sharepoint.spec.ts diff --git a/client/src/data-provider/Files/__tests__/sharepoint.spec.ts b/client/src/data-provider/Files/__tests__/sharepoint.spec.ts new file mode 100644 index 0000000000..d71086dba2 --- /dev/null +++ b/client/src/data-provider/Files/__tests__/sharepoint.spec.ts @@ -0,0 +1,58 @@ +jest.mock('@tanstack/react-query', () => ({ + useMutation: () => ({}), +})); + +import { resolveSharePointContentType, GENERIC_MIME_TYPE } from '../sharepoint'; + +describe('resolveSharePointContentType', () => { + it('preserves specific content-type headers and strips charset/boundary parameters', () => { + expect(resolveSharePointContentType('application/pdf; charset=UTF-8', 'document.pdf')).toBe( + 'application/pdf', + ); + expect(resolveSharePointContentType('image/png', 'photo.png')).toBe('image/png'); + expect(resolveSharePointContentType('text/plain; charset=utf-8', 'notes.txt')).toBe( + 'text/plain', + ); + }); + + it('falls back to getMimeTypeFromFileName when header is application/octet-stream', () => { + expect(resolveSharePointContentType('application/octet-stream', 'data.csv')).toBe('text/csv'); + expect( + resolveSharePointContentType('application/octet-stream; charset=binary', 'financials.xlsx'), + ).toBe('application/vnd.openxmlformats-officedocument.spreadsheetml.sheet'); + expect(resolveSharePointContentType('application/octet-stream', 'slides.pptx')).toBe( + 'application/vnd.openxmlformats-officedocument.presentationml.presentation', + ); + expect(resolveSharePointContentType('application/octet-stream', 'doc.docx')).toBe( + 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', + ); + expect(resolveSharePointContentType('application/octet-stream', 'manual.pdf')).toBe( + 'application/pdf', + ); + }); + + it('falls back to inferMimeType for source and code extensions when header is application/octet-stream', () => { + expect(resolveSharePointContentType('application/octet-stream', 'script.py')).toBe( + 'text/x-python', + ); + expect(resolveSharePointContentType('application/octet-stream', 'config.json')).toBe( + 'application/json', + ); + expect(resolveSharePointContentType('application/octet-stream', 'readme.md')).toBe( + 'text/markdown', + ); + }); + + it('falls back to extension resolution when header is null or undefined', () => { + expect(resolveSharePointContentType(null, 'table.csv')).toBe('text/csv'); + expect(resolveSharePointContentType(undefined, 'app.py')).toBe('text/x-python'); + expect(resolveSharePointContentType('', 'document.pdf')).toBe('application/pdf'); + }); + + it('returns application/octet-stream when neither header nor extension can be mapped', () => { + expect(resolveSharePointContentType('application/octet-stream', 'archive.7z')).toBe( + GENERIC_MIME_TYPE, + ); + expect(resolveSharePointContentType(null, 'unrecognized_file')).toBe(GENERIC_MIME_TYPE); + }); +}); diff --git a/client/src/data-provider/Files/sharepoint.ts b/client/src/data-provider/Files/sharepoint.ts index b55fe0efb4..2e2db50190 100644 --- a/client/src/data-provider/Files/sharepoint.ts +++ b/client/src/data-provider/Files/sharepoint.ts @@ -1,6 +1,7 @@ import { useMutation } from '@tanstack/react-query'; import { getClientLimit } from 'librechat-data-provider'; import type { UseMutationResult } from '@tanstack/react-query'; +import { inferMimeType } from 'librechat-data-provider'; export interface SharePointFile { id: string; @@ -69,8 +70,10 @@ export const useSharePointBatchDownload = (): UseMutationResult< } finally { clearTimeout(abortTimer); } - const contentType = - response.headers.get('content-type') || getMimeTypeFromFileName(file.name); + const contentType = resolveSharePointContentType( + response.headers.get('content-type'), + file.name, + ); const downloadedFile = new File([blob], file.name, { type: contentType, @@ -119,7 +122,24 @@ export const useSharePointBatchDownload = (): UseMutationResult< }); }; -function getMimeTypeFromFileName(fileName: string): string { +export const GENERIC_MIME_TYPE = 'application/octet-stream'; + +export function resolveSharePointContentType( + headerContentType: string | null | undefined, + fileName: string, +): string { + const header = headerContentType?.split(';')[0].trim().toLowerCase() ?? ''; + if (header && header !== GENERIC_MIME_TYPE) { + return header; + } + const byName = getMimeTypeFromFileName(fileName); + if (byName !== GENERIC_MIME_TYPE) { + return byName; + } + return inferMimeType(fileName, '') || GENERIC_MIME_TYPE; +} + +export function getMimeTypeFromFileName(fileName: string): string { const extension = fileName.split('.').pop()?.toLowerCase(); const mimeTypes: Record = {