feat: add inital flac decoding support via the symphonia crate

[KernelFreeze]

I have added the initial code path to support decoding FLAC bitstreams using the existing tooling around the symphonia crate. This needs further testing, as I don't yet have access to FLAC capabilities from the official client. Also, I have some questions regarding the required implementation: should I add a passthrough_decoder.rs code path too?
Let me know if you need any changes to be made.

[KernelFreeze]

Also, this should be needed to fully implement #1583.

[roderickvd]

Looks good to me.

[urknall]

I have added the initial code path to support decoding FLAC bitstreams using the existing tooling around the symphonia crate. This needs further testing, as I don't yet have access to FLAC capabilities from the official client. Also, I have some questions regarding the required implementation: should I add a passthrough_decoder.rs code path too? Let me know if you need any changes to be made.

passthrough decoder support would be nice, some projects based on librespot use it for other paths also.
i would prefer zero decode/re-encode and forward the flac directly to the player

[KernelFreeze]

I think I can add passthrough_decoder.rs support in another PR, so as not to increase the scope of this PR too much.

[Laxii-null]

I think I can add passthrough_decoder.rs support in another PR, so as not to increase the scope of this PR too much.

I think just doing a commit in the same PR would be good, you know, just to keep things intact

[roderickvd]

Both are fine for me, whatever is easiest.

[KernelFreeze]

I will start working on the passthrough decoder. In the meantime, it would be a great help if someone with access to the FLAC capabilities of the official client could test whether we need anything else or if it already works as is. My main concern is whether there is a need for a new decryption module.

[Laxii-null]

I will start working on the passthrough decoder. In the meantime, it would be a great help if someone with access to the FLAC capabilities of the official client could test whether we need anything else or if it already works as is. My main concern is whether there is a need for a new decryption module.

Some people in the original librespot issue gave some details about the API implemention recently that may be helpful to you