ResumeMaker

Production-focused, minimal Next.js + TypeScript resume editor that keeps all conversions server-side.

Features

• Upload PDF resume from browser.
• Server converts PDF -> DOCX (LibreOffice headless), then DOCX -> HTML (mammoth).
• Edit HTML in a TipTap WYSIWYG editor.
• Download final PDF (preferred DOCX -> PDF via LibreOffice; fallback via Puppeteer HTML -> PDF).
• Download DOCX version.
• Optional OCR endpoint (/api/ocr) using tesseract.js.
• Temp file cleanup, max upload size validation, mime validation, basic in-memory rate limiting.

Tech stack

• Next.js pages router + API routes
• TypeScript across frontend/backend
• TipTap editor
• Conversion libraries: mammoth, html-docx-js, LibreOffice (soffice), Puppeteer fallback

Project structure

• pages/index.tsx UI for upload/edit/download
• components/Editor.tsx TipTap editor wrapper
• pages/api/upload.ts PDF -> DOCX -> HTML
• pages/api/save.ts edited HTML -> DOCX -> PDF/DOCX
• pages/api/download-docx.ts fetch stored DOCX by fileId
• pages/api/ocr.ts OCR draft endpoint
• lib/ conversion and temp-file helpers
• tests/ Jest tests for conversion endpoint flows
• scripts/smoke.ts full conversion chain script

Local run (without Docker)

1. Install dependencies:

   npm install

2. Install LibreOffice and Chromium dependencies:
   • Ubuntu/Debian:

     sudo apt-get update
     sudo apt-get install -y libreoffice fonts-dejavu-core

3. Optional env vars:

   export TEMP_DIR=./tmp
   export MAX_UPLOAD_BYTES=10485760
   export RATE_LIMIT_PER_MINUTE=30

4. Start app:

   npm run dev

5. Open http://localhost:3000

Docker run

docker compose up --build

API quick checks (cURL)

Upload resume:

curl -X POST http://localhost:3000/api/upload \
  -F "resume=@data/sample_resume.pdf"

Save edited HTML as PDF:

curl -X POST http://localhost:3000/api/save \
  -H "Content-Type: application/json" \
  -d '{"html":"<h1>Experience</h1><p>Example</p>","format":"pdf"}' \
  --output out/resume_final.pdf

Save edited HTML as DOCX:

curl -X POST http://localhost:3000/api/save \
  -H "Content-Type: application/json" \
  -d '{"html":"<h1>Education</h1><p>University</p>","format":"docx"}' \
  --output out/resume_final.docx

OCR draft endpoint:

curl -X POST http://localhost:3000/api/ocr -F "file=@scan.png"

Tests

npm test

Smoke end-to-end chain

npm run smoke

Produces out/resume_final.pdf.

Production improvement ideas

• Replace in-memory file map/rate-limiter with Redis for multi-instance deployments.
• Add queue worker (BullMQ) for heavy conversions.
• Add antivirus scanning for uploaded files.
• Improve OCR for PDF by rasterizing pages first, then running OCR per page.
• Add auth and object storage (S3/GCS) for persistent download links.