refactor `get_next_internal_address` to give only 1 address

[KnowWhoami]

Edit: see https://github.com/citadel-tech/coinswap/pull/245#discussion_r1732905255 Fixes #242

[KnowWhoami]

Add an assert to check if the count is zero or not.

We don't require that since derive_address rpc call itself take care of it. I call get_next_internal_address with count =0 and got this error:

[KnowWhoami]

Should I write doc comments to mention this?

[Shourya742]

Add an assert to check if the count is zero or not.

We don't require that since derive_address rpc call itself take care of it. I call get_next_internal_address with count =0 and got this error:

Not exactly. We need to capture deterministic errors in the error propagation chain as early as possible. In this case, capturing the error is especially important because we're making a network call, and we want to avoid intentionally sending a buggy payload. What you're suggesting isn't a good practice.

Now that I think about it, this must be why Belcher didn’t use this approach in the first place—to avoid errors in this situation. Given how vast the address space is, we’re unlikely to exhaust it. It might be better to leave it as is and simply document it.

[KnowWhoami]

• After making required changes as told in https://github.com/citadel-tech/coinswap/pull/245#discussion_r1732905255
• I am getting this error because in this function implementation -> get_next_hd_index is called to get the next unused index for the given keychain which derives it from the corresponding descriptor.
• But after deriving the address ->. the index upto which all the address are used is not updated.

https://github.com/citadel-tech/coinswap/blob/73e17a80ac10d42b97966430d745d1bf90155563/src/wallet/api.rs#L923

• Infact we don't have any api, variable etc to update this index for every keychain as we do have for External keychain.

https://github.com/citadel-tech/coinswap/blob/73e17a80ac10d42b97966430d745d1bf90155563/src/wallet/api.rs#L908

[mojoX911]

Parked for BDK.