Fix region nonfungible implementation

[Szegoo]

The problem with the current implementation is that minting will cause the region coremask to be set to Coremask::complete regardless of the actual coremask.

This PR fixes that.

More details about the nonfungible implementation can be found here: #3455

[seadanda]

Oof, good catch. Would be nice to add a test to interlace or partition->burn->mint and then check the mask is identical after minting. This PR should also be backported to v1.14

[Szegoo]

Would be nice to add a test to interlace or partition->burn->mint and then check the mask is identical after minting.

I added two new tests. They don't explicitly check for the coremask, but if the coremask changed when minting, the region returned would be the old one which still has the owner set to None, so the test would fail without this fix.

[seadanda]

Looks good, just needs a prdoc. I just realised it's breaking as issue is pub, so it might be an issue to backport.

[Szegoo]

I just realised it's breaking as issue is pub, so it might be an issue to backport.

But it is not exposed as an extrinsic; it can only be accessed from the runtime code, right?

[paritytech-cicd-pr]

The CI pipeline was cancelled due to failure one of the required jobs.
Job name: cargo-clippy
Logs: https://gitlab.parity.io/parity/mirrors/polkadot-sdk/-/jobs/6779083

[ggwpez]

The semver CI check should be fixed no master, please try again after merging it in.

[bkontur]

lgtm, just nits with import (can be also ignored)

[Szegoo]

The semver check was failing. Maybe merging with the master branch again will fix it?