Adds option to adjust the audio volume on a per podcast basis

[maltebaer]

As discussed in #7, this PR adds the ability to adjust the volume.

This is done using pydub, which in turn requires ffmpeg. So I guess it is not a very lightweight solution.

By the way, is there a method to list the required dependencies in a python project and install them all at once? Like with npm for Node.js projects?

[alexhartm]

By the way, is there a method to list the required dependencies in a python project and install them all at once? Like with npm for Node.js projects?

Hi @maltebaer Yes, the project uses poetry for package management. If a new dependency is consumed, it should be made available via e.g. poetry add pydub. Poetry is then also used for build and publishing it to the pypi package repository.
I think this is also something that can be automated. I'll try to set this up eventually via github actions.

[alexhartm]

On a quick test on my laptop (with the missing ffmpeg dependency), it looked like it was stuck during the download part. I think this might be because __cache_episode was stuck when doing __adjust_volume__.
I did some quick workaround via

if (ep.volume_adjustment != 0):
    adjusted_content = self.__adjust_volume__(r.content, ep.volume_adjustment)
else:
    adjusted_content=r.content

and that seemed to work.
I think it would be good if the thing keeps working without having ffmpeg available locally. (As this is something that a user has to install separately outside of pip.) But maybe there is a more clever approach than mine.
Ideally, we could also highlight the adjust_volume part at bit (in the terminal), because on constrained devices, this might take some time. However, this might also be something for a follow-up improvement.
Will try it out in the next days in more detail. And am already looking forward to this feature! 😃 👍

[maltebaer]

I updated the code to avoid potential issues with missing ffmpeg. I still get an error from the linter about S603 subprocess call: check for execution of untrusted input. I think we can ignore it in this case, however I'm not able to add the exception.

[alexhartm]

Ah, nice that you considered platform-independent'ness! I was wondering if this works on Windows as well, as I noticed another user running this on a Windows host. Unfortunately, I have no option to test it on Windows...
Do you have a Windows setup where you could try it out already?

[maltebaer]

No, I can't try it either. To be honest, I only asked Chat GPT to make it Windows safe.

[alexhartm]

This is a great change. Thanks for contributing! I'm going to release this as a new 2.1 version after the merge