PROJECT UNDER MIGRATION: Client-Side Version --> Client-Side & Server Side Version

This repository contains the original, fully client-side version of the Privacy-Focused PDF Editor. All PDF processing (merging, reordering, etc.) is performed directly in your browser using WebAssembly, ensuring no files ever leave your machine.

This project is no longer under active development. Development has transitioned to a new version that includes a self-hosted backend for more intensive processing.

For the latest version, please visit the new repository: Portfolio Site

Privacy-Focused PDF Editor for Legal Professionals

This is a browser-based PDF editor designed with the privacy needs of Brazilian lawyers in mind. The project delivers a simple yet comprehensive interface tailored for common legal document operations such as merging, reordering, deleting, and annotating PDFs.

The initial architecture was designed for a truly serverless environment, utilizing a simple Single-Page Application (SPA) schema within SvelteKit. The core principle was to offer a clear and secure method for handling documents and user data without storing them, preserving full user privacy.

Technical Aspects

Aside from authentication (handled via Firebase), the entire application runs client-side, with routing being handled via SvelteKit. It uses the browser's IndexedDB for local file storage, ensuring that no documents are uploaded or stored on external servers — preserving full user privacy.

Initially deployed on Cloudflare, the project was later migrated to Vercel to support potential Node.js enhancements. PDF rendering is powered by Mozilla's PDF.js, while core document operations (merge, delete, reorder) are handled by MuPDF compiled to WebAssembly — a deliberate choice for performance and device compatibility. Using a C-based engine via WASM ensures fast processing even on lower-end machines, unlike JavaScript-only libraries.

To maximize performance and maintain minimal overhead, the frontend was built with Svelte — a compiler rather than a runtime framework. This allowed for efficient state management and fast load times. The first version of this project was developed with vanilla JavaScript and PDF-Lib.js as a learning experience, before evolving into a more robust solution.