Add API for internal addresses

[LLFourn]

There are good reasons for applications to need to get internal
addresses too. For example creating a transactions that splits an output
into several smaller ones.

Checklists

All Submissions:

• I've signed all my commits
• I followed the contribution guidelines
• I ran cargo fmt and cargo clippy before committing

New Features:

• I've added tests for the new feature
• I've added docs for the new feature
• I've updated CHANGELOG.md

[rajarshimaitra]

Concept + tACK 63f1ec0

I think it makes sense and the code looks good to me.

[w0xlt]

tACK 63f1ec0

I tested this commit with two wallets:

1 - with internal and external descriptors: worked as expected.

2- with external descriptor only: get_internal_address returns the same addresses as get_address.

Maybe if there is not an internal descriptor, a possible better approach would be to throw an error when calling get_internal_address .

But I think this behavior is not related to this PR. It happens because get_descriptor_for_keychain(keychain) apparently returns the same descriptor for internal and external keychain when the wallet is created with external descriptor only,

[rajarshimaitra]

Maybe if there is not an internal descriptor, a possible better approach would be to throw an error when calling get_internal_address .
But I think this behavior is not related to this PR. It happens because get_descriptor_for_keychain(keychain) apparently returns the same descriptor for internal and external keychain when the wallet is created with external descriptor only,

I think the get_descriptor_for_keychain() is written that way to make wallet api agnostic of change descriptor. Weather its there or not, all the logic will simply follow "use external if there's no internal". This simplifies internal logic handling at many places and doesn't force us to handle "no change desc case" separately. and the lib user just keep this in mind while using the wallet. Its also documented in many places.

So it makes sense to me to keep the behavior intact in peek_address too.

[rajarshimaitra]

ACK 89f88ce

Sorry I missed few others comment nits in last review, of the previous methods.

[LLFourn]

comments addressed ready for further review.

[notmandatory]

I think docs are breaking because of a regression in rust +nightly, there seems to be a bug fix PR in the works so I think we should wait a couple days and see if that fixes it. See #538

[notmandatory]

#538 is merged and fixes the docs issue, once you rebase this PR should be good to go.

[notmandatory]

Looks good, one small thing to fix in the new test.

[notmandatory]

Shouldn't you be verifying that the internal address is the same as the external here?

Suggested change

	wallet.get_address(AddressIndex::New).unwrap().address,
	wallet.get_internal_address(AddressIndex::New).unwrap().address,

[LLFourn]

thanks fixed @notmandatory

[notmandatory]

#538 is merged and fixes the docs issue, once you rebase this PR should be good to go.

I think you still need to rebase since this went in, that will fix the rust docs CI job.

[LLFourn]

@notmandatory new CI problem. Looks like tokio increased MSRV without major version bump. I think we should set MSRV for bdk to v1.56 (2021 edition) and leave it there.

[notmandatory]

I guess we need to revisit and not close #331 .. at this point I agree the best option is set MSRV to rust 1.56, it's not the current stable for Debian.. but is being tested and should be shipped with next stable debian "bookworm" release. https://tracker.debian.org/pkg/rustc

[LLFourn]

Rebased. Should be ready for merge after CI (hopefully) passes.

[notmandatory]

ACK 022256c

Requested changes look to be resolved.