fix(mistral): name extracted images as storage does, drop inline table format, align image switch with tool param

Author: waleedlatif1

apps/docs/content/docs/integrations/mistral_parse.mdx

@@ -15,7 +15,7 @@ Use [Mistral's OCR API](https://mistral.ai/) to extract text and page structure
 
 ## Use the block [#usage-instructions]
 
-Extract text from uploaded PDF documents, or from a URL.
+Extract text, tables, and images from uploaded PDF documents, or from a URL. Images come back as files an agent can pass on.
 
 ## Actions
 

apps/sim/blocks/blocks/mistral_parse.ts

@@ -309,6 +309,8 @@ export const MistralParseV3Block: BlockConfig = {
   type: 'mistral_parse_v3',
   name: 'Mistral Parser',
   description: 'Extract text, tables, and images from PDF documents',
+  longDescription:
+    'Extract text, tables, and images from uploaded PDF documents, or from a URL. Images come back as files an agent can pass on.',
   hideFromToolbar: false,
   /* v3 renamed the canonical pair's advanced member `filePath` -> `fileReference`,
      so the inherited sentence would reference a subblock this block lacks. */
@@ -374,7 +376,7 @@ export const MistralParseV3Block: BlockConfig = {
       value: () => 'inline',
     },
     {
-      id: 'extractImages',
+      id: 'includeImageBase64',
       title: 'Extract Images',
       type: 'switch',
     },
@@ -383,15 +385,15 @@ export const MistralParseV3Block: BlockConfig = {
       title: 'Max Images',
       type: 'short-input',
       placeholder: 'Up to 100 (default 100)',
-      condition: { field: 'extractImages', value: true },
+      condition: { field: 'includeImageBase64', value: true },
       mode: 'advanced',
     },
     {
       id: 'imageMinSize',
       title: 'Min Image Size (px)',
       type: 'short-input',
       placeholder: 'Skip images smaller than this on both sides',
-      condition: { field: 'extractImages', value: true },
+      condition: { field: 'includeImageBase64', value: true },
       mode: 'advanced',
     },
     {
@@ -452,11 +454,12 @@ export const MistralParseV3Block: BlockConfig = {
           parameters.pages = pagesArray
         }
 
-        if (params.tableFormat === 'markdown' || params.tableFormat === 'html') {
-          parameters.tableFormat = params.tableFormat
-        }
+        parameters.tableFormat =
+          params.tableFormat === 'markdown' || params.tableFormat === 'html'
+            ? params.tableFormat
+            : undefined
 
-        if (params.extractImages) {
+        if (params.includeImageBase64) {
           parameters.includeImageBase64 = true
           parameters.imageLimit = parseOptionalNumberInput(params.imageLimit, 'Max Images', {
             integer: true,
@@ -479,7 +482,7 @@ export const MistralParseV3Block: BlockConfig = {
     resultType: { type: 'string', description: 'Output format type' },
     pages: { type: 'string', description: 'Page selection' },
     tableFormat: { type: 'string', description: 'Table output: inline, markdown, or html' },
-    extractImages: { type: 'boolean', description: 'Extract document images as files' },
+    includeImageBase64: { type: 'boolean', description: 'Extract document images as files' },
     imageLimit: { type: 'number', description: 'Maximum number of images to extract' },
     imageMinSize: { type: 'number', description: 'Minimum image size in pixels' },
   },

apps/sim/lib/internal/mistral/images.test.ts

@@ -1,8 +1,8 @@
 import { describe, expect, it } from 'vitest'
 import { extractOcrImageFiles, MAX_OCR_EXTRACTED_IMAGES } from '@/lib/internal/mistral/images'
 
-const PNG = Buffer.from('synthetic png bytes')
-const dataUri = (bytes: Buffer) => `data:image/png;base64,${bytes.toString('base64')}`
+const JPEG = Buffer.concat([Buffer.from([0xff, 0xd8, 0xff]), Buffer.from('synthetic jpeg body')])
+const dataUri = (bytes: Buffer) => `data:image/jpeg;base64,${bytes.toString('base64')}`
 
 function image(id: string, imageBase64: string | undefined) {
   return {
@@ -16,28 +16,28 @@ function image(id: string, imageBase64: string | undefined) {
 }
 
 describe('extractOcrImageFiles', () => {
-  it('moves image bytes into files and leaves no base64 in the page output', () => {
+  it('moves image bytes into files named as storage will name them, with no base64 left in the page output', () => {
     const { output, files } = extractOcrImageFiles({
       pages: [
         {
           index: 2,
-          markdown: '![img-0.png](img-0.png)',
-          images: [image('img-0.png', dataUri(PNG))],
+          markdown: '![img-0.jpeg](img-0.jpeg)',
+          images: [image('img-0.jpeg', dataUri(JPEG))],
         },
       ],
     })
-    expect(files).toEqual([{ buffer: PNG, name: 'page-2-img-0.png', mimeType: 'image/png' }])
-    expect(JSON.stringify(output)).not.toContain(PNG.toString('base64'))
+    expect(files).toEqual([{ buffer: JPEG, name: 'page-2-img-0.jpg', mimeType: 'image/jpeg' }])
+    expect(JSON.stringify(output)).not.toContain(JPEG.toString('base64'))
     expect(output.pages).toEqual([
       expect.objectContaining({
-        images: [expect.objectContaining({ id: 'img-0.png', file_name: 'page-2-img-0.png' })],
+        images: [expect.objectContaining({ id: 'img-0.jpeg', file_name: 'page-2-img-0.jpg' })],
       }),
     ])
   })
 
   it('stores at most the image cap and keeps the bounding boxes of the rest without a file', () => {
     const images = Array.from({ length: MAX_OCR_EXTRACTED_IMAGES + 1 }, (_, i) =>
-      image(`img-${i}.jpeg`, PNG.toString('base64'))
+      image(`img-${i}.jpeg`, JPEG.toString('base64'))
     )
     const { output, files } = extractOcrImageFiles({ pages: [{ index: 0, images }] })
     expect(files).toHaveLength(MAX_OCR_EXTRACTED_IMAGES)
@@ -51,10 +51,13 @@ describe('extractOcrImageFiles', () => {
     const huge = Buffer.alloc(21 * 1024 * 1024)
     const { output, files } = extractOcrImageFiles({
       pages: [
-        { index: 0, images: [image('img-0.png', dataUri(huge)), image('img-1.png', dataUri(PNG))] },
+        {
+          index: 0,
+          images: [image('img-0.png', dataUri(huge)), image('img-1.jpg', dataUri(JPEG))],
+        },
       ],
     })
-    expect(files.map((file) => file.name)).toEqual(['page-0-img-1.png'])
+    expect(files.map((file) => file.name)).toEqual(['page-0-img-1.jpg'])
     expect(JSON.stringify(output)).not.toContain('image_base64')
   })
 })

apps/sim/lib/internal/mistral/images.ts

@@ -4,6 +4,7 @@ import { isFileParserError } from '@/lib/file-parsers/errors'
 import type { InternalToolFile } from '@/lib/internal/tool-operations/file-result'
 import { MAX_BUFFERED_TRANSFER_BYTES } from '@/lib/uploads/shared/types'
 import { resolveEffectiveMimeType } from '@/lib/uploads/utils/file'
+import { resolveStoredFileMetadata } from '@/lib/uploads/utils/stored-file-metadata'
 
 /** Most images one parse stores; Mistral's `image_limit` is clamped to it. */
 export const MAX_OCR_EXTRACTED_IMAGES = 100
@@ -51,13 +52,15 @@ export function extractOcrImageFiles(output: unknown): OcrImageFiles {
       if (!decoded || totalBytes + decoded.buffer.length > MAX_BUFFERED_TRANSFER_BYTES) return rest
       totalBytes += decoded.buffer.length
       const id = typeof image.id === 'string' && image.id ? image.id : `img-${files.length}.jpeg`
-      const name = `page-${pageIndex}-${id}`
-      files.push({
-        buffer: decoded.buffer,
-        name,
-        mimeType: resolveEffectiveMimeType(decoded.mediaType, name),
-      })
-      return { ...rest, file_name: name }
+      /** Storage renames an image from its bytes; naming it the same way keeps `file_name` matching. */
+      const draftName = `page-${pageIndex}-${id}`
+      const { fileName, mimeType } = resolveStoredFileMetadata(
+        draftName,
+        resolveEffectiveMimeType(decoded.mediaType, draftName),
+        decoded.buffer
+      )
+      files.push({ buffer: decoded.buffer, name: fileName, mimeType })
+      return { ...rest, file_name: fileName }
     })
     return { ...page, images }
   })

apps/sim/lib/internal/mistral/operations.test.ts

@@ -256,7 +256,7 @@ describe('Mistral ingestion authorization', () => {
   )
 
   it('returns extracted images as files, never as base64 in the output', async () => {
-    const png = Buffer.from('synthetic png bytes')
+    const png = Buffer.concat([Buffer.from([0xff, 0xd8, 0xff]), Buffer.from('synthetic jpeg body')])
     submit.mockResolvedValue({
       pages: [
         {

apps/sim/tools/mistral/parser.ts

@@ -600,7 +600,7 @@ export const mistralParserV3Tool: InternalToolConfig<MistralParserV2Input, Mistr
         if (params.imageMinSize !== undefined) {
           requestBody.imageMinSize = params.imageMinSize
         }
-        if (params.tableFormat) {
+        if (params.tableFormat === 'markdown' || params.tableFormat === 'html') {
           requestBody.tableFormat = params.tableFormat
         }
 

packages/deployment-config/src/integrations.json

@@ -18362,7 +18362,7 @@
       "slug": "mistral-parser",
       "name": "Mistral Parser",
       "description": "Extract text, tables, and images from PDF documents",
-      "longDescription": "Extract text from uploaded PDF documents, or from a URL.",
+      "longDescription": "Extract text, tables, and images from uploaded PDF documents, or from a URL. Images come back as files an agent can pass on.",
       "bgColor": "#000000",
       "iconName": "MistralIcon",
       "docsUrl": "https://docs.sim.ai/integrations/mistral_parse",

scripts/check-unused-exports.baseline.json

@@ -5577,7 +5577,6 @@
   "apps/sim/tools/mistral/types.ts#MISTRAL_OCR_DIMENSIONS_OUTPUT",
   "apps/sim/tools/mistral/types.ts#MISTRAL_OCR_DIMENSIONS_OUTPUT_PROPERTIES",
   "apps/sim/tools/mistral/types.ts#MISTRAL_OCR_IMAGE_OUTPUT_PROPERTIES",
-  "apps/sim/tools/mistral/types.ts#MISTRAL_OCR_PAGE_OUTPUT_PROPERTIES",
   "apps/sim/tools/mistral/types.ts#MISTRAL_OCR_USAGE_OUTPUT",
   "apps/sim/tools/mistral/types.ts#MISTRAL_OCR_USAGE_OUTPUT_PROPERTIES",
   "apps/sim/tools/mistral/types.ts#MISTRAL_PARSER_METADATA_OUTPUT_PROPERTIES",