fix(parsers): fix md, pptx, html kb uploads (#1209)

* fix md, pptx, html

* consolidate consts

apps/sim/app/workspace/[workspaceId]/knowledge/[id]/components/upload-modal/upload-modal.tsx

@@ -7,26 +7,12 @@ import { Dialog, DialogContent, DialogHeader, DialogTitle } from '@/components/u
 import { Label } from '@/components/ui/label'
 import { Progress } from '@/components/ui/progress'
 import { createLogger } from '@/lib/logs/console/logger'
+import { ACCEPT_ATTRIBUTE, ACCEPTED_FILE_TYPES, MAX_FILE_SIZE } from '@/lib/uploads/validation'
 import { getDocumentIcon } from '@/app/workspace/[workspaceId]/knowledge/components'
 import { useKnowledgeUpload } from '@/app/workspace/[workspaceId]/knowledge/hooks/use-knowledge-upload'
 
 const logger = createLogger('UploadModal')
 
-const MAX_FILE_SIZE = 100 * 1024 * 1024 // 100MB
-const ACCEPTED_FILE_TYPES = [
-  'application/pdf',
-  'application/msword',
-  'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
-  'text/plain',
-  'text/csv',
-  'application/vnd.ms-excel',
-  'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
-  'text/markdown',
-  'application/vnd.ms-powerpoint',
-  'application/vnd.openxmlformats-officedocument.presentationml.presentation',
-  'text/html',
-]
-
 interface FileWithPreview extends File {
   preview: string
 }
@@ -172,7 +158,7 @@ export function UploadModal({
 
   return (
     <Dialog open={open} onOpenChange={handleClose}>
-      <DialogContent className='flex max-h-[95vh] max-w-2xl flex-col overflow-hidden'>
+      <DialogContent className='flex max-h-[95vh] flex-col overflow-hidden sm:max-w-[600px]'>
         <DialogHeader>
           <DialogTitle>Upload Documents</DialogTitle>
         </DialogHeader>
@@ -197,7 +183,7 @@ export function UploadModal({
                 <input
                   ref={fileInputRef}
                   type='file'
-                  accept={ACCEPTED_FILE_TYPES.join(',')}
+                  accept={ACCEPT_ATTRIBUTE}
                   onChange={handleFileChange}
                   className='hidden'
                   multiple
@@ -228,7 +214,7 @@ export function UploadModal({
                   <input
                     ref={fileInputRef}
                     type='file'
-                    accept={ACCEPTED_FILE_TYPES.join(',')}
+                    accept={ACCEPT_ATTRIBUTE}
                     onChange={handleFileChange}
                     className='hidden'
                     multiple
@@ -238,7 +224,7 @@ export function UploadModal({
                   </p>
                 </div>
 
-                <div className='max-h-60 space-y-2 overflow-auto'>
+                <div className='max-h-80 space-y-2 overflow-auto'>
                   {files.map((file, index) => {
                     const fileStatus = uploadProgress.fileStatuses?.[index]
                     const isCurrentlyUploading = fileStatus?.status === 'uploading'

apps/sim/app/workspace/[workspaceId]/knowledge/components/create-modal/create-modal.tsx

@@ -14,27 +14,13 @@ import { Label } from '@/components/ui/label'
 import { Progress } from '@/components/ui/progress'
 import { Textarea } from '@/components/ui/textarea'
 import { createLogger } from '@/lib/logs/console/logger'
+import { ACCEPT_ATTRIBUTE, ACCEPTED_FILE_TYPES, MAX_FILE_SIZE } from '@/lib/uploads/validation'
 import { getDocumentIcon } from '@/app/workspace/[workspaceId]/knowledge/components'
 import { useKnowledgeUpload } from '@/app/workspace/[workspaceId]/knowledge/hooks/use-knowledge-upload'
 import type { KnowledgeBaseData } from '@/stores/knowledge/store'
 
 const logger = createLogger('CreateModal')
 
-const MAX_FILE_SIZE = 100 * 1024 * 1024 // 100MB
-const ACCEPTED_FILE_TYPES = [
-  'application/pdf',
-  'application/msword',
-  'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
-  'text/plain',
-  'text/csv',
-  'application/vnd.ms-excel',
-  'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
-  'text/markdown',
-  'application/vnd.ms-powerpoint',
-  'application/vnd.openxmlformats-officedocument.presentationml.presentation',
-  'text/html',
-]
-
 interface FileWithPreview extends File {
   preview: string
 }
@@ -498,7 +484,7 @@ export function CreateModal({ open, onOpenChange, onKnowledgeBaseCreated }: Crea
                         <input
                           ref={fileInputRef}
                           type='file'
-                          accept={ACCEPTED_FILE_TYPES.join(',')}
+                          accept={ACCEPT_ATTRIBUTE}
                           onChange={handleFileChange}
                           className='hidden'
                           multiple
@@ -540,7 +526,7 @@ export function CreateModal({ open, onOpenChange, onKnowledgeBaseCreated }: Crea
                           <input
                             ref={fileInputRef}
                             type='file'
-                            accept={ACCEPTED_FILE_TYPES.join(',')}
+                            accept={ACCEPT_ATTRIBUTE}
                             onChange={handleFileChange}
                             className='hidden'
                             multiple

apps/sim/lib/uploads/validation.ts

@@ -1,5 +1,7 @@
 import path from 'path'
 
+export const MAX_FILE_SIZE = 100 * 1024 * 1024 // 100MB
+
 export const SUPPORTED_DOCUMENT_EXTENSIONS = [
   'pdf',
   'csv',
@@ -9,21 +11,49 @@ export const SUPPORTED_DOCUMENT_EXTENSIONS = [
   'md',
   'xlsx',
   'xls',
+  'ppt',
+  'pptx',
+  'html',
+  'htm',
 ] as const
 
 export type SupportedDocumentExtension = (typeof SUPPORTED_DOCUMENT_EXTENSIONS)[number]
 
 export const SUPPORTED_MIME_TYPES: Record<SupportedDocumentExtension, string[]> = {
-  pdf: ['application/pdf'],
-  csv: ['text/csv', 'application/csv'],
-  doc: ['application/msword'],
-  docx: ['application/vnd.openxmlformats-officedocument.wordprocessingml.document'],
-  txt: ['text/plain'],
-  md: ['text/markdown', 'text/x-markdown'],
-  xlsx: ['application/vnd.openxmlformats-officedocument.spreadsheetml.sheet'],
-  xls: ['application/vnd.ms-excel'],
+  pdf: ['application/pdf', 'application/x-pdf'],
+  csv: ['text/csv', 'application/csv', 'text/comma-separated-values'],
+  doc: ['application/msword', 'application/doc', 'application/vnd.ms-word'],
+  docx: [
+    'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
+    'application/octet-stream',
+  ],
+  txt: ['text/plain', 'text/x-plain', 'application/txt'],
+  md: ['text/markdown', 'text/x-markdown', 'text/plain', 'application/markdown'],
+  xlsx: [
+    'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
+    'application/octet-stream',
+  ],
+  xls: [
+    'application/vnd.ms-excel',
+    'application/excel',
+    'application/x-excel',
+    'application/x-msexcel',
+  ],
+  ppt: ['application/vnd.ms-powerpoint', 'application/powerpoint', 'application/x-mspowerpoint'],
+  pptx: [
+    'application/vnd.openxmlformats-officedocument.presentationml.presentation',
+    'application/octet-stream',
+  ],
+  html: ['text/html', 'application/xhtml+xml'],
+  htm: ['text/html', 'application/xhtml+xml'],
 }
 
+export const ACCEPTED_FILE_TYPES = Object.values(SUPPORTED_MIME_TYPES).flat()
+
+export const ACCEPTED_FILE_EXTENSIONS = SUPPORTED_DOCUMENT_EXTENSIONS.map((ext) => `.${ext}`)
+
+export const ACCEPT_ATTRIBUTE = [...ACCEPTED_FILE_TYPES, ...ACCEPTED_FILE_EXTENSIONS].join(',')
+
 export interface FileValidationError {
   code: 'UNSUPPORTED_FILE_TYPE' | 'MIME_TYPE_MISMATCH'
   message: string