From 23799562aefe0b8a2dfe3417360571b825397bd7 Mon Sep 17 00:00:00 2001
From: Elian Doran <contact@eliandoran.me>
Date: Thu, 2 Apr 2026 22:53:57 +0300
Subject: [PATCH] refactor(ocr): reuse office processor for PDFs

---
 apps/server/src/services/ocr/ocr_service.ts   |  2 -
 .../ocr/processors/office_processor.ts        |  2 +
 .../services/ocr/processors/pdf_processor.ts  | 40 -------------------
 apps/server/src/types.d.ts                    |  5 ---
 4 files changed, 2 insertions(+), 47 deletions(-)
 delete mode 100644 apps/server/src/services/ocr/processors/pdf_processor.ts

diff --git a/apps/server/src/services/ocr/ocr_service.ts b/apps/server/src/services/ocr/ocr_service.ts
index 12ab5d034a..0c6ec29ff3 100644
--- a/apps/server/src/services/ocr/ocr_service.ts
+++ b/apps/server/src/services/ocr/ocr_service.ts
@@ -9,7 +9,6 @@ import sql from '../sql.js';
 import { FileProcessor } from './processors/file_processor.js';
 import { ImageProcessor } from './processors/image_processor.js';
 import { OfficeProcessor } from './processors/office_processor.js';
-import { PDFProcessor } from './processors/pdf_processor.js';
 import { TIFFProcessor } from './processors/tiff_processor.js';
 
 export interface OCRResult {
@@ -37,7 +36,6 @@ class OCRService {
     constructor() {
         const imageProcessor = new ImageProcessor();
         this.processors.set('image', imageProcessor);
-        this.processors.set('pdf', new PDFProcessor());
         this.processors.set('tiff', new TIFFProcessor(imageProcessor));
         this.processors.set('office', new OfficeProcessor());
     }
diff --git a/apps/server/src/services/ocr/processors/office_processor.ts b/apps/server/src/services/ocr/processors/office_processor.ts
index cabc07b3e0..50b208e692 100644
--- a/apps/server/src/services/ocr/processors/office_processor.ts
+++ b/apps/server/src/services/ocr/processors/office_processor.ts
@@ -5,6 +5,8 @@ import { OCRProcessingOptions, OCRResult } from '../ocr_service.js';
 import { FileProcessor } from './file_processor.js';
 
 const SUPPORTED_TYPES = [
+    // PDF
+    'application/pdf',
     // Office Open XML
     'application/vnd.openxmlformats-officedocument.wordprocessingml.document',   // DOCX
     'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',         // XLSX
diff --git a/apps/server/src/services/ocr/processors/pdf_processor.ts b/apps/server/src/services/ocr/processors/pdf_processor.ts
deleted file mode 100644
index c59fe84b81..0000000000
--- a/apps/server/src/services/ocr/processors/pdf_processor.ts
+++ /dev/null
@@ -1,40 +0,0 @@
-// Import the actual library directly, bypassing the wrapper index.js
-// which tries to read a test PDF file at import time.
-import pdfParse from 'pdf-parse/lib/pdf-parse.js';
-
-import log from '../../log.js';
-import { OCRProcessingOptions, OCRResult } from '../ocr_service.js';
-import { FileProcessor } from './file_processor.js';
-
-/**
- * PDF processor for extracting embedded text from PDF files using pdf-parse.
- */
-export class PDFProcessor extends FileProcessor {
-
-    canProcess(mimeType: string): boolean {
-        return mimeType.toLowerCase() === 'application/pdf';
-    }
-
-    getSupportedMimeTypes(): string[] {
-        return ['application/pdf'];
-    }
-
-    async extractText(buffer: Buffer, options: OCRProcessingOptions = {}): Promise<OCRResult> {
-        log.info('Starting PDF text extraction...');
-
-        const data = await pdfParse(buffer);
-
-        return {
-            text: data.text.trim(),
-            confidence: 0.99,
-            extractedAt: new Date().toISOString(),
-            language: options.language || "eng",
-            pageCount: data.numpages
-        };
-    }
-
-    getProcessingType(): string {
-        return 'pdf';
-    }
-
-}
diff --git a/apps/server/src/types.d.ts b/apps/server/src/types.d.ts
index 5c859c1342..8d6048a76c 100644
--- a/apps/server/src/types.d.ts
+++ b/apps/server/src/types.d.ts
@@ -39,11 +39,6 @@ declare module "@triliumnext/share-theme/styles.css" {
     export default content;
 }
 
-declare module "pdf-parse/lib/pdf-parse.js" {
-    import pdfParse from "pdf-parse";
-    export default pdfParse;
-}
-
 declare module '*.css' {}
 declare module '*?raw' {
   const src: string