taprpc: move ImportProof from tapdevrpc to taprpc

[bhandras]

This commit turns theImportProof API to be publicly available as it is required for some use cases.

[coveralls]

Pull Request Test Coverage Report for Build 11240704007

Details

• 0 of 24 (0.0%) changed or added relevant lines in 3 files are covered.
• 35 unchanged lines in 5 files lost coverage.
• Overall coverage decreased (-0.006%) to 40.379%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
rpcserver.go	0	1	0.0%
itest/utils.go	0	2	0.0%
cmd/tapcli/proofs.go	0	21	0.0%

Files with Coverage Reduction	New Missed Lines	%
tapdb/addrs.go	2	79.04%
asset/asset.go	2	81.8%
commitment/tap.go	4	84.17%
tapgarden/caretaker.go	8	68.5%
universe/interface.go	19	47.09%

Totals
Change from base Build 11218781069:	-0.006%
Covered Lines:	24259
Relevant Lines:	60078

---

💛 - Coveralls

[jharveyb]

Relevant existing discourse:

#804 (comment)

Is the use case manually importing a proof for an asset that the importing node actually owns?

In that case, maybe we can improve the existing InsertProof call to 'take ownership' of the inserted assets, if we know that we own the relevant internal and script keys.

Not sure if that covers your use or you're looking at something more involved? Either way, sidestepping the local universe will probably cause issues later.

[dstadulis]

Core user need: when importing a proof and recognize and if the node owns the asset that's being imported
Custom channels side steps the introspection need

[bhandras]

Relevant existing discourse:

#804 (comment)

Is the use case manually importing a proof for an asset that the importing node actually owns?

In that case, maybe we can improve the existing InsertProof call to 'take ownership' of the inserted assets, if we know that we own the relevant internal and script keys.

Not sure if that covers your use or you're looking at something more involved? Either way, sidestepping the local universe will probably cause issues later.

IIUC InsertProof wouldn't work in our case as what we need is that a receiver who doesn't own the internal and script keys to be able to spend the corresponding anchor and sweep the asset (an HTLC sweep).

[bhandras]

Closing this as we managed to emulate the same functionality by iterating on the whole proof with InsertProof plus setting the anchor derivation upon sweep. Thanks again to @guggero for all the insights and debugging help!