Zeedif
87aae46a1f
* fix: correct chapter facets URL to include /chapters endpoint
Update addChapterSortAndFilterFacets to use the correct URL path
from `/manga/{id}` to `/manga/{id}/chapters` for proper routing.
* feat(opds): restructure feeds and add exploration capabilities
This commit completely refactors the OPDS v1.2 implementation to align it more closely with the WebUI experience, separating "Library" browsing from "Explore" functionality.
Key changes include:
- The root feed is now a navigation feed directing to distinct "Library" and "Explore" sections.
- A new "History" feed has been added to the root to show recently read chapters.
- The "Explore" section now allows browsing all available sources, not just those with manga in the library.
- Feeds for exploring a source now support faceting by "Popular" and "Latest", mirroring the WebUI.
- The "Library" section retains all previous browsing methods (by category, genre, status, etc.).
- Facet link generation has been corrected to use the proper base URL, fixing broken navigation in chapter lists.
- The `OpdsFeedBuilder.kt` file has been refactorized into smaller, more manageable helper files (`OpdsEntryBuilder.kt`, `OpdsFeedHelper.kt`) to resolve a `java.lang.OutOfMemoryError: GC overhead limit exceeded` error during compilation.
- All OPDS-related strings (`strings.xml`) have been updated to reflect the new structure and improve clarity.
This new structure provides a much more intuitive and powerful browsing experience for OPDS clients, enabling content discovery in addition to library management.
* feat(opds)!: implement advanced filtering and sorting for library feeds
This commit significantly enhances the OPDS library feeds by introducing advanced sorting and filtering capabilities, mirroring the features available in the WebUI. It also standardizes the terminology from "manga" to "series" across all user-facing OPDS feeds for better clarity and consistency.
Key Features & Changes:
- **Library Facets:** All library feeds (All Series, By Source, By Category, By Genre, etc.) now include OPDS facets for:
- **Sorting:** By title (A-Z, Z-A), last read, latest chapter, date added, and total unread chapters.
- **Filtering:** By content status including unread, downloaded, ongoing, and completed.
- **Terminology Update:** The term "manga" has been replaced with "series" in all user-facing OPDS titles, descriptions, and endpoints to align with the frontend terminology.
- **Code Refactoring:**
- `MangaRepository` has been updated with the correct Exposed SQL syntax (`Case`/`sum` for conditional counts, `having` clause for filtering on aggregates) to support the new facets.
- `OpdsEntryBuilder` now includes a new function `addLibraryMangaSortAndFilterFacets` to generate the facet links.
- `OpdsV1Controller` and `OpdsFeedBuilder` have been updated to handle the new `sort` and `filter` parameters and to call the new facet generation logic.
BREAKING CHANGE: The API endpoints for manga have been renamed to use 'series'. Any client implementation will need to update its routes.
For example, `/api/opds/v1.2/manga/{id}/chapters` is now `/api/opds/v1.2/series/{id}/chapters`.
* feat(opds): add item counts (thr:count) to navigation and facet links
This change enhances the OPDS feeds by including the number of items for various navigation links and filter facets, adhering to the OPDS 1.2 specification.
The `thr:count` attribute provides a hint to clients about the number of entries in a linked feed, significantly improving the user experience by showing counts upfront.
- Navigation Feeds (Categories, Sources, Genres, Statuses, Languages) now display the total number of manga for each entry in their respective links.
- Acquisition Feeds for the library and chapters now include counts for their filter facets (e.g., Unread, Downloaded, Completed).
This required updating DTOs to carry count data, modifying repository queries to calculate these counts efficiently, and adjusting the feed builders to include the `thr:count` attribute in the generated XML.
* refactor(opds)!: simplify root feed by removing library sub-level
The OPDS feed navigation was previously nested, requiring users to first select "Library" and then navigate to a subsection like "All Series" or "Categories". This extra step is cumbersome for OPDS clients and complicates the user experience.
This change elevates all library-related navigation entries directly to the root feed, flattening the hierarchy and making content more accessible.
As part of this refactoring:
- The `getLibraryFeed` builder and its corresponding controller/API endpoints have been removed.
- Unused string resources for the "Library" entry have been deleted.
BREAKING CHANGE: The `/api/opds/v1.2/library` endpoint has been removed. Clients should now discover library sections directly from the root feed at `/api/opds/v1.2`.
* feat(opds): enhance feeds with comprehensive manga and chapter details
This commit significantly enriches the OPDS feeds to provide a more detailed and compliant user experience.
- Refactored `OpdsMangaAcqEntry` and `OpdsChapterMetadataAcqEntry` to include additional fields such as status, source information, author, description, and web URLs.
- The OPDS entry builder (`OpdsEntryBuilder`) now populates entries with this richer metadata, including summaries, content descriptions, authors, and categories, aligning more closely with the OPDS Catalog specification.
- Added OPDS constants for 'popular' and 'new' sort relations to align with the specification.
- Included "alternate" links for both manga and chapters, allowing clients to open the item on its source website ("View on web").
- Updated internationalization strings and constants to support the new features and metadata.
* fix(opds): fetch chapters for non-library manga in feed
Previously, when accessing the OPDS chapter feed for a manga discovered via the "Explore" feature (and thus not yet in the library), the feed would be empty. This was because the feed generation logic only queried the local database, which had no chapter entries for these manga.
This commit resolves the issue by modifying `getSeriesChaptersFeed` to be a suspend function. It now implements a fallback mechanism:
- It first attempts to load chapters from the local database.
- If no chapters are found, it triggers an online fetch from the source to populate the database.
- It then re-queries the local data to build the complete chapter feed.
This ensures that chapter lists are correctly displayed for all manga, whether they are in the library or being explored for the first time.
Additionally, this commit includes a minor correction to the URN identifier for the root feed to better align with its path.
* feat(opds): provide direct stream and acquisition links when page count is known
Previously, the OPDS chapter feed always provided a single link to a separate metadata feed for each chapter. This was done to defer the costly operation of fetching the page count for undownloaded chapters, ensuring the main chapter list loaded quickly.
This commit introduces a more efficient, conditional approach. If a chapter's page count is already known (e.g., because it's downloaded or has been previously fetched), the chapter feed entry now includes direct links for:
- OPDS-PSE page streaming (`pse:stream`).
- CBZ file acquisition (`acquisition/open-access`).
- Chapter cover image (`image`).
If the page count is not known, the entry falls back to the previous behavior, linking to the metadata feed to perform the page count lookup on-demand.
This significantly improves the user experience for OPDS clients by reducing the number of requests needed to start reading or downloading chapters that are already available on the server, making navigation faster and more fluid.
* fix(opds): resolve suspend calls and add missing lastReadAt for OPDS feeds
The OPDS feed generation was failing to compile due to two main issues:
1. The `OpdsChapterListAcqEntry` DTO was missing the `lastReadAt` property, which is required for the OPDS-PSE `lastReadDate` attribute.
2. Several functions in `OpdsFeedBuilder` were attempting to call the `suspend` function `createChapterListEntry` from a non-coroutine context.
This commit resolves these issues by:
- Adding the `lastReadAt` field to `OpdsChapterListAcqEntry` and populating it correctly from the database in the `ChapterRepository`.
- Refactoring `getHistoryFeed`, `getLibraryUpdatesFeed`, and `getSeriesChaptersFeed` in `OpdsFeedBuilder` to be `suspend` functions.
- Wrapping the entry creation logic in `withContext(Dispatchers.IO)` to provide the necessary coroutine scope for the suspend call and to perform the mapping on a background thread.
* refactor(opds): standardize library feed generation and enhance facets
This commit refactors the OPDS v1.2 feed generation logic to improve code structure, correctness, and feature capability.
The primary changes include:
- A new private `getLibraryFeed` helper function in `OpdsV1Controller` has been introduced to centralize and DRY up the logic for creating library-specific acquisition feeds.
- A new `OpdsMangaFilter` DTO now encapsulates all filtering, sorting, and pagination parameters, simplifying the controller handlers and making them more maintainable.
- URL generation for category, genre, status, and language feeds has been corrected. Links now correctly point to root-level paths (e.g., `/opds/v1.2/genre/{name}`) instead of being incorrectly nested under `/library/`.
- The OPDS facet system is enhanced with more specific facet groups and "All" links for a better user experience when clearing filters.
Associated changes:
- i18n strings in `strings.xml` have been reorganized with comments and new strings have been added to support the enhanced facet groups.
- The route for the publication status feed has been renamed from `/status/{id}` to `/statuses` for consistency.
- KDoc comments have been added and improved throughout the affected files for better code documentation.
* fix(opds): revert direct acquisition links in chapter feeds to improve performance
This reverts commit 33cdc0d534292760a3225cee18e274df542f0778.
The previous change introduced direct stream and download links in chapter list feeds when the page count was known. While convenient, this caused a significant performance degradation on feeds with many chapters, as it required checking for the existence of a CBZ file for every single entry.
This commit restores the original behavior where chapter list entries always link to a dedicated metadata feed. This approach defers expensive I/O operations until a user explicitly requests a single chapter's details, ensuring that chapter list feeds load quickly and efficiently. Direct acquisition and streaming links are now exclusively generated within the metadata feed.
| Build | Stable | Preview | Support Server |
|---|---|---|---|
 |
 |
 |
 |
Table of Content
- What is Suwayomi?
- Suwayomi client projects
- Downloading and Running the app
What is Suwayomi?
A free and open source manga reader server that runs extensions built for Mihon (Tachiyomi).
Suwayomi is an independent Mihon (Tachiyomi) compatible software and is not a Fork of Mihon (Tachiyomi).
Suwayomi-Server is as multi-platform as you can get. Any platform that runs java and/or has a modern browser can run it. This includes Windows, Linux, macOS, chrome OS, etc. Follow Downloading and Running the app for installation instructions.
You can use Mihon (Tachiyomi) to access your Suwayomi-Server. For more info look here.
Features
Note
These are capabilities of Suwayomi-Server, the actual working support is provided by each front-end app, checkout their respective readme for more info.
- Installing and executing Mihon (Tachiyomi)'s Extensions, So you'll get the same sources
- Searching and browsing installed sources
- A library to save your mangas and categories to put them into
- Automated library updates to check for new chapters
- Automated download of new chapters
- Viewing latest updated chapters
- Ability to download Manga for offline read
- Backup and restore support powered by Mihon (Tachiyomi)-compatible Backups
- Automated backup creations
- Tracking via MyAnimeList, AniList, MangaUpdates, etc.
- FlareSolverr support to bypass Cloudflare protection
- Automated WebUI updates (supports the default WebUI and VUI)
- OPDS and OPDS-PSE support (endpoint:
/api/opds/v1.2)
Suwayomi client projects
You need a client/user interface app as a front-end for Suwayomi-Server, if you Directly Download Suwayomi-Server you'll get a bundled version of Suwayomi-WebUI with it.
Here's a list of known clients/user interfaces for Suwayomi-Server (checkout the respective GitHub repository for their features):
Actively Developed Clients
- Suwayomi-WebUI: The web front-end that Suwayomi-Server ships with by default.
- Suwayomi-VUI: A Suwayomi-Server preview focused web frontend built with svelte
- Tachidesk-VaadinUI: A Web front-end for Suwayomi-Server built with Vaadin.
Inactive Clients (functional but outdated)
- Tachidesk-JUI: The native desktop front-end for Suwayomi-Server.
- Tachidesk-Sorayomi: A Flutter front-end for Desktop(Linux, windows, etc.), Web and Android with a User Interface inspired by Mihon (Tachiyomi).
Abandoned Clients (functionality unknown)
- Tachidesk-qtui: A C++/Qt front-end for mobile devices(Android/linux), feature support is basic.
- Tachidesk-GTK: A native Rust/GTK desktop client.
- Equinox: A web user interface made with Vue.js.
Downloading and Running the app
Using Operating System Specific Bundles
To facilitate the use of Suwayomi we provide bundle releases that include The Java Runtime Environment, ElectronJS and the Suwayomi-Launcher.
If a bundle for your operating system or cpu architecture is not provided then refer to Advanced Methods
Windows
Download the latest win64(Windows 64-bit) release from the releases section or a preview one from the preview repository.
Unzip the downloaded file and double-click on one of the launcher scripts.
macOS
Download the latest macOS-x64(older macOS systems) or macOS-arm64(Apple M1 and newer) release from the releases section or a preview one from the preview repository.
Unzip the downloaded file and double-click on one of the launcher scripts.
GNU/Linux
Download the latest linux-x64(x86_64) release from the releases section or a preview one from the preview repository.
tar xvf the downloaded file and double-click on one of the launcher scripts or run them using the terminal.
WebView support (GNU/Linux)
WebView support is implemented via KCEF. This is optional, and is only necessary to support some extensions.
To have a functional WebView, several dependencies are required; aside from X11 libraries necessary for rendering Chromium, some JNI bindings are necessary: gluegen and jogl (found in Ubuntu as libgluegen2-jni and libjogl2-jni).
Note that on some systems (e.g. Ubuntu), the JNI libraries are not automatically found, see below.
A KCEF server is launched on startup, which loads the X11 libraries.
If those are missing, you should see "Could not load 'jcef' library".
If so, use ldd ~/.local/share/Tachidesk/bin/kcef/libjcef.so | grep not to figure out which libraries are not found on your system.
The JNI bindings are only loaded when a browser is actually launched. This is done by extensions that rely on WebView, not by Suwayomi itself. If there is a problem loading the JNI libraries, you should see a message indicating the library and the search path. This search path includes the current working directory, if you do not want to modify system directories.
Refer to the Dockerfile for more details.
Other methods of getting Suwayomi
Docker
Check our Official Docker release Suwayomi Container for running Suwayomi Server in a docker container. Source code for our container is available at docker-tachidesk, an example compose file can also be found there. By default, the server will be running on http://localhost:4567 open this url in your browser.
Arch Linux
You can install Suwayomi from the AUR:
yay -S tachidesk
Debian/Ubuntu
Download the latest deb package from the release section.
Caution
These options are outdated and unmaintained (relevant issue)
MPR
git clone https://mpr.makedeb.org/tachidesk-server.git cd tachidesk-server makedeb -siUbuntu
sudo add-apt-repository ppa:suwayomi/tachidesk-server sudo apt update sudo apt install tachidesk-server
NixOS
You can deploy Suwayomi on NixOS using the module services.suwayomi-server in your configuration:
{
services.suwayomi-server = {
enable = true;
};
}
For more information, see the NixOS manual.
You can also directly use the package from nixpkgs.
Advanced Methods
Running the jar release directly
In order to run the app you need the following:
- The jar release of Suwayomi-Server
- The Java Runtime Environment(JRE) 21 or newer
- A Browser like Google Chrome, Firefox, Edge, etc.
- ElectronJS (optional)
Download the latest .jar release from the releases section or a preview jar build from the preview repository.
Make sure you have The Java Runtime Environment installed on your system, Double-click on the jar file or run java -jar Suwayomi-Server-vX.Y.Z-rxxxx.jar from a Terminal/Command Prompt window to run the app which will open a new browser window automatically.
Using Suwayomi Remotely
You can run Suwayomi on your computer or a server and connect to it remotely through one of our clients or the bundled web interface with a web browser. This method of using Suwayomi is requiring a bit of networking/firewall/port forwarding/server configuration/etc. knowledge on your side, if you can run a Minecraft server and configure it, then you are good to go.
Check out this wiki page for a guide on configuring Suwayomi-Server.
If you face issues with your setup then we are happy to provide help, just join our discord server(a discord badge is on the top of the page, you are just a click-clack away!).
Syncing With Mihon (Tachiyomi) and Neko
The Suwayomi extension and tracker
- You can install and configure the
Suwayomiextension inside Mihon (Tachiyomi) and forks. - The extension will load your Suwayomi library.
- By manipulating extension search filters you can browse your categories.
- You can enable the Suwayomi tracker to track reading progress with your Suwayomi server.
- Note: to sync from
- Mihon (Tachiyomi) to Suwayomi: Mihon (Tachiyomi) automatically updates the chapters read status when it's updating the tracker (e.g. while reading)
- Suwayomi to Mihon (Tachiyomi): To sync Mihon (Tachiyomi) with Suwayomi, you have to open the manga's track information, then, Mihon (Tachiyomi) will automatically update its chapter list with the state from Suwayomi
- Note: to sync from
The Suwayomi merge source in Neko
- You can enable the
Suwayomisource in the Merge Source settings - You can merge titles in Neko with titles from your Suwayomi library.
- You can enable 2-way automatic sync to track reading progress with your Suwayomi server.
- Note: only applies to merged titles
- Neko automatically updates the chapters read status in Suwayomi
- During updates, Neko will automatically update its chapter list with the read state from Suwayomi
- This only pulls if the status is read, to prevent marking read chapters as unread in Neko
- Note: only applies to merged titles
Other methods
Checkout this issue for tracking progress.
Troubleshooting and Support
See this troubleshooting wiki page.
Contributing and Technical info
See CONTRIBUTING.md.
Translation
Feel free to translate the project on Weblate
Translation Progress
Credit
This project is a spiritual successor of TachiWeb-Server, Many of the ideas and the groundwork adopted in this project comes from TachiWeb.
The AndroidCompat module was originally developed by @null-dev for TachiWeb-Server and is licensed under Apache License Version 2.0 and Copyright 2019 Andy Bao and contributors.
Parts of Mihon (Tachiyomi) is adopted into this codebase, also licensed under Apache License Version 2.0 and Copyright 2015 Javier Tomás.
You can obtain a copy of Apache License Version 2.0 from http://www.apache.org/licenses/LICENSE-2.0
Changes to both codebases is licensed under MPL v. 2.0 as the rest of this project.
License
Copyright (C) Contributors to the Suwayomi project
This Source Code Form is subject to the terms of the Mozilla Public
License, v. 2.0. If a copy of the MPL was not distributed with this
file, You can obtain one at http://mozilla.org/MPL/2.0/.
Disclaimer
The developer of this application does not have any affiliation with the content providers available.