Handle multiple identical discs in one release

[Freso]

Came upon this issue in https://github.com/JonnyJD/musicbrainz-isrcsubmit/issues/118

I haven't actually tested this in whipper, but since the most specific we can be on the command-line in case of multiple releases having the same Disc ID attached, I don't see how whipper would be able to deal with a single release having multiple discs that resolve to the same Disc ID.

See https://github.com/JoeLametta/whipper/issues/167 for more thoughts/suggestions.

[luziferius]

There is at least one disambiguation available, if 2 discs have the same DiscID: The ISRC. If you get a release with duplicate DiscIDs when querying the MB database, you could read the ISRCs stored on the disc in the tray and look them up in the MB database. If the disc has ISRCs and those ISRCs are actually submitted, it should be possible to disambiguate the disc.

[JoeLametta]

Blocked on #167.

[Hasshu]

It seems that I've encountered this very issue, so... ahem!

There are two similar releases of Legacy of the Dark Lands by Blind Guardian Twilight Orchestra:

1. https://musicbrainz.org/release/876e371d-6198-4d38-8e77-bebb56c9ebf4/discids (let's call it "A");
2. https://musicbrainz.org/release/f02c2ba4-9e65-4231-aed1-4257fab8d1c0/discids (let's call it "B").

As you can see, both Disc IDs of release A are identical to those of release B. In other words:

A1 = B1 A2 = B2

But that's not all: since they contain different versions of the same pieces (of the same duration), B2 and B3 have identical Disc IDs as well! Which means:

A2 = B2 = B3

And that's where it collides with the impeller... When I try to rip B2 (whipper cd rip -pUk), whipper offers me only two choices: A2 or B3; whereas B2 (which I'm actually trying to rip) is not listed as an option.

Now, I don't know whether it's a whipper issue or a MusicBrainz issue, but I believe that whipper shouldn't rely on a third-party service as much as it does — however convenient MusicBrainz may be as a metadata source. If I may be so bold, the issues like #342 and #425 should be given higher priority.

P.S.: Oh, and when I'm trying to rip B3, whipper claims that it has picked the right disc... but then produces what appears to be a mixture of data from both B3 and B2, for whatever reason.