Tankobon

Added

* Genshin Impact manga source.

Added

* New config option `download.rate_limit`: this limits the total number of requests made when downloading chapters at any one time.
This can also be set in the GUI settings.
* Nicer chapter view that can be directly clicked on to download/view chapters.
* Method `core.Downloader.downloaded(chapter)`: Check whether the pages for a `models.Chapter` has been downloaded.

Changed

* `utils.Config` no longer implicitly creates a single instance.
It is strongly recommended to access the global instance `utils.CONFIG` to avoid stale configurations being written back to disk.

Fixed

* [26](26): The CatManga source now parses manga descriptions properly.
Implicit wrapping of `str` descriptions in `models.Metadata` to dictionaries caused descriptions to be nested.

Added

* (GUI) Support for BBCode in manga descriptions.
Any BBCode will now render correctly as HTML in the manga view.
* Dependency on bbcode.
* `.sync()` method to `utils.Config` (save config changes).

Changed

* (GUI) The visual layout has been changed a bit to prepare for viewer support.
* The `desc` field of `models.Metadata` is now a dictionary map of a language code to the localised description.

Fixed

* Configuration was not synced across `utils.Config` instances (i.e language), so newer configs would get overwritten by older ones.
`utils.Config` now will have only one instance to prevent this.
* (GUI) Cover images now scale correctly and no longer look pixelated, especially those of high resolution.

Added

* Settings dialog in the GUI (File -> Settings).
The preferred manga language can now be set through there.
* The manga info panel now shows the languages in the manga.

Changed

* Mangadex urls must now start with `mangadex.org/title`.

Removed

* Dependency on dataclasses-json.

Fixed

* The GUI now shows chapters for the current language set.
* The Mangadex source now adds all languages to the manga.
* Mangadex chapter urls will now open correctly when clicked in the GUI.
* Downloading through the CLI. `Parser.by_url` was given the shorthash instead of the url.

Changed

* All source-related abstract base classes (i.e `Parser`) now reside in `tankobon/sources/base.py`.
The source modules have been updated accordingly to reflect this.
* The method `Parser.parser` has been renamed to `Parser.by_url` to make it clearer what it is actually for.