No description
  • JavaScript 100%
Find a file
Repository files (latest commit first)
Filename Latest commit message Latest commit date
2026-10-02 16:14:06 +00:00
LICENSE Update LICENSE (MIT License) 2026-10-01 08:17:19 +00:00
README.md Sync README from GitLab 2026-10-02 16:04:27 +00:00
Ridibooks Comic Ripper v1.2.user.js Release v1.2 2026-09-30 23:23:21 +00:00
Ridibooks Comic Ripper v1.3.user.js Upload Ridibooks Comic Ripper v1.3.user.js 2026-09-30 22:09:22 +00:00

Ridibooks Chapter Downloader

A custom browser script built to intercept and save high-resolution episode pages from Ridibooks directly to your computer. It monitors background network traffic to grab the original picture links and organizes them into a local directory.

✨ Main Features

  • Backend Interception: Listens for the site's internal generate network requests (via Fetch and XHR) to silently capture the raw image array as soon as the page loads.
  • Auto-Recovery System: If a network timeout occurs, the tool automatically pauses for two seconds and re-attempts the download up to three times per picture.
  • Paced Processing: Enforces a strict 200-millisecond delay between each file request to ensure your browser remains stable and does not throttle the connection.
  • Interactive Interface: Injects a dynamic, floating status button in the bottom corner of your screen. It displays a waiting status initially and turns blue once the page data is ready to process.
  • Native Directory Saving: Automatically sanitizes the book's title to create a valid folder on your hard drive, saving each file sequentially without requiring ZIP extraction.
  • Dynamic Reset: Instantly clears its cache and resets the user interface when you navigate to a new episode without needing a full page refresh.

🚀 Setup and Execution

  1. Requirements: You must have the Tampermonkey extension installed to grant the necessary local file-saving permissions.
  2. Installation: Add the code through the Greasyfork repository.
  3. Usage: Open any reading session on Ridibooks. The floating button will initially show a searching status.
  4. Save: Once the button turns blue and displays the total page count, click it to queue and save all the images to your device.

⚠️ Important Notice

This project is intended strictly for personal archiving and educational study. Please support the original authors and publishers. Do not upload, share, or redistribute the saved media under any circumstances.

For bug reports, feature suggestions, or general questions, please drop a review on Greasyfork or send an email to devjk6918@gmail.com.