[Feature] Backend rewrite

[Superredstone]

This rewrite aims to completely ditch the old SpotiFLAC library.

I decided to do this rewrite in order to distance myself from the original author’s design choices in SpotiFLAC. Since it was not written as a proper library, the integration process became suboptimal due to rushed design decisions that were not oriented toward scalability and code reusability.

Features I cannot add (without some hacks) because of the original author’s implementation:

• Download in a different format
• Download at a lower quality

Questionable choices made by the original author:

• Songs that are not found in FLAC are still downloaded in M4A and then converted to FLAC, effectively nullifying FLAC’s superior quality and creating a dependency on FFmpeg
• The output file name cannot be modified

Missing features:

• Playlist download
• Album download
• Metadata fetching
• Override song flag (download even if a file already exists)
• Download from Qobuz. (PLANNED FOR FUTURE)
• Album conver embedding
• Download quality select
• A flag to write failed songs to a log file
• Move already downloaded songs skip logic before metadata fetching

[mahoganyprogrammer]

check newest spotiflac . there is logic that makes report what files failed. but it just count failed and put in separate array.

[mahoganyprogrammer]

@Superredstone i checked your code. download track broken, playlist logic exist but it seems to pass spoti id to tidal downloader

[mahoganyprogrammer]

you allowed null to pass, if no tidal link return error

[mahoganyprogrammer]

i fixed playlist downloading. it download song now. why it saves at drive root

[mahoganyprogrammer]

i fixed. outputFile was empty

[mahoganyprogrammer]

app.GetTidalIdFromSonglink(songlink) for qobuz its same just different field is fetched

[mahoganyprogrammer]

new spotiflac version released . deezer source added. i hope it wont need ffmpeg to download.

[Superredstone]

new spotiflac version released . deezer source added. i hope it wont need ffmpeg to download.

I hope too

[Superredstone]

For now i decided to not download from amazon since in the original application this feature relies on APIs made by afkarxyz, which AFIK are not open source.

If anyone finds a solution to this be sure to let me know or implement it yourself!

[Superredstone]

you allowed null to pass, if no tidal link return error

fixed

[Superredstone]

Omg, I've just encountered another great decision of the original project. Apparently downloading from deezer (which is implemented in deezer.go, at this link) does not actually download from deezer but from https://yoinkify.lol which has nothing to do with Deezer.

[kristijonas-jan]

Omg, I've just encountered another great decision of the original project. Apparently downloading from deezer (which is implemented in deezer.go, at this link) does not actually download from deezer but from https://yoinkify.lol which has nothing to do with Deezer.

I think yoinkify might download from Deezer.. still a weird as hell workaround.. https://yoinkify.lol/roadmap

[devilAPI]

is there any update on this? I would love to use the new version, as the old one is really slow for a large collection! I would love to contribute too but never written in Go before...

[Superredstone]

is there any update on this? I would love to use the new version, as the old one is really slow for a large collection! I would love to contribute too but never written in Go before...

I would like to implement at least the last checkmark A flag to write failed songs to a log file

To speed up things there should be implemented some kind of concurrency, which, TBH I'm not really familiar with, but I'll give it a try.

[devilAPI]

By speeding up I meant the yt-dlp's --download-archive equivalent.
It logs the downloaded files to a text file with ID and Name and then skips all of them at the next run.

[Superredstone]

Oh, I miss understood, it is half implemented at the moment, if not today, in a couple of days the feature should be ready.

[Superredstone]

@devilAPI the application is now capable of skipping entirely already downloaded songs before fetching metadata. This should improve download times by a lot.