mansuf / mangadex-downloader

A command-line tool to download manga from MangaDex, written in Python.
https://mangadex-dl.mansuf.link/
MIT License
477 stars 39 forks source link

Latest v2.5.1 breaks some previous features #37

Closed rokibhasansagar closed 1 year ago

rokibhasansagar commented 1 year ago

What happened ?

After the latest tagged version of v2.5.1, while trying to fix #36, I found that it broke some useful features.

The options which I found not working are:

  1. --use-chapter-title
  2. --no-group-name

Using the same Reproducible command, explained at #36,,

mangadex-downloader "https://mangadex.org/title/296cbc31-af1a-4b5b-a34b-fee2b4cad542/-oshi-no-ko" --language "en" --cover "original" --use-compressed-image --start-chapter 1 --end-chapter 6 --save-as raw --use-chapter-title --no-group-name

If run, results somewhat like this: Screenshot_20221121_183049

Where the above mentioned command options are ignored by the app: Screenshot_20221121_183128

If I try to download "Chapter 11" from the manga, which has 2 scanlator uploads, both are being downloaded with scanlator's name prefixed in the folder name, but chapter title is still ignored. Screenshot_20221121_185140 Screenshot_20221121_185224

So it's basically forcing --group all to be an absolute option as well. But selecting a single group with --group XXX is still operational.

So, I get Downloaded Folders (for raw format) as: <Scanlator Group/Uploader Name> <Volume> <Chapter>, which is unexpected.

What did you expect to happen ?

I should've got folders/files prefixed as <Volume> <Chapter> - <Title>, as it did in the previous tagged versions.

OS version

Ubuntu 22.04

App version

mangadex-downloader v2.5.1 (https://github.com/mansuf/mangadex-downloader)
Python: 3.10.6
arch: x64
bundled executable: no

Installation origin

PyPI (Python Package Index)

Installation origin (other sources)

No response

Reproducible command

mangadex-downloader "https://mangadex.org/title/296cbc31-af1a-4b5b-a34b-fee2b4cad542/-oshi-no-ko" --language "en" --cover "original" --use-compressed-image --start-chapter 1 --end-chapter 6 --save-as raw --use-chapter-title --no-group-name

And for testing the chapter download with dual scanlator (chapter 11)

mangadex-downloader "https://mangadex.org/title/296cbc31-af1a-4b5b-a34b-fee2b4cad542/-oshi-no-ko" --language "en" --cover "original" --use-compressed-image --start-chapter 10 --end-chapter 12 --save-as raw --use-chapter-title --no-group-name

Additional context

No response

rokibhasansagar commented 1 year ago

Side note: While figuring out this issue, if you could set the app to skip/ignore chapter links which redirects to non-mangadex links would be helpful (those are basically Licensed Uploader's website links). I don't know how the mangadex API works, but if there is a way to filter the links before downloading them, I'd appreciate if you could implement it too in the near future.

mansuf commented 1 year ago

Alright, this should be fixed in https://github.com/mansuf/mangadex-downloader/commit/9cebc81b84df9b588d8f308a6b9fafd876ee3b44

I tested in my pc and it's fixed. Now if you could test it would be helpful. I will release this patch if you're satisfied with the results (due to obvious reason: Fixed one bug and another bug appeared)

Installation command:


pip install git+https://github.com/mansuf/mangadex-downloader.git@9cebc81b84df9b588d8f308a6b9fafd876ee3b44
rokibhasansagar commented 1 year ago

I'm testing it right now. I'll inform about the new changes in a minute.

rokibhasansagar commented 1 year ago

Thanks for fixing it quickly 😃

Both --use-chapter-title and --no-group-name inclusion or exclusion, with or without using --group option works perfectly.

Now, #36 is actually fixed without breaking anything. So #37 can be resolved with the latest change.

You can push the new change to upstream for release.

mansuf commented 1 year ago

Side note: While figuring out this issue, if you could set the app to skip/ignore chapter links which redirects to non-mangadex links would be helpful (those are basically Licensed Uploader's website links). I don't know how the mangadex API works, but if there is a way to filter the links before downloading them, I'd appreciate if you could implement it too in the near future.

This has been implemented in v2.5.1. The app will look for images not referral links. If the chapter doesn't have images it will simply skipped and will be not downloaded

mansuf commented 1 year ago

Released in v2.5.2