Office ⇄ PDF Converter

Node.js + React app to batch-convert DOCX/PPTX to PDF and PDF back to DOCX/PPTX using LibreOffice headless mode.

Prerequisites

• Node.js 18+
• LibreOffice installed.
  • Preferred: add soffice to PATH.
  • Windows: set SOFFICE_PATH in server/.env to the full exe path, e.g. C:\\Program Files\\LibreOffice\\program\\soffice.exe.
  • If you see spawn soffice ENOENT, the backend cannot find the binary—set SOFFICE_PATH or add LibreOffice to PATH and restart npm run dev.

Quick start

1. Backend

cd server
npm install
cp .env.example .env  # on Windows: copy .env.example .env
npm run dev

2. Frontend (new terminal)

cd web
npm install
npm run dev

3. Open the UI at the URL printed by Vite (defaults to http://localhost:5173). The UI proxies API calls to http://localhost:4000.

API

• POST /api/convert
  • Form-data fields: files (one or many), targetFormat (pdf | docx | pptx)
  • Response: { jobId, targetFormat, results: [{ input, output, url, status, error? }] }
• GET /api/health

Supported directions:

• targetFormat=pdf → accepts .doc, .docx, .ppt, .pptx
• targetFormat=docx → accepts .pdf (LibreOffice draw import → DOCX)
• targetFormat=pptx → accepts .pdf (LibreOffice draw import → PPTX)

Outputs are served for download from /downloads/{jobId}/{file} and cleaned after OUTPUT_TTL_MINUTES (default 60).

Configuration

Set in server/.env:

• PORT (default 4000)
• CONCURRENCY (queue parallelism, default 1 for sequential safety)
• SOFFICE_PATH (path to LibreOffice executable)
• CONVERSION_TIMEOUT_MS (per-file cap)
• CONVERSION_RETRIES (per-file attempts, default 5)
• OUTPUT_TTL_MINUTES (how long to keep artifacts)

Notes

• PDF→DOCX/PPTX quality depends on LibreOffice import fidelity; complex layouts may shift.
• Temp uploads are removed after each job; output folders are pruned on an interval.

Key files

• Backend: server/src/server.js, server/src/converter.js
• Frontend: web/src/App.jsx, web/src/api.js

OIIAI

signed by Otaku