Better address setup

[afilini]

Description

Instead of using the derivation index we should store for each address whether or not it has been "setup" with the blockchain backend. We could merge Blockchain::setup() with Blockchain::sync() and just let the function decide what to do with each address based on its status.

Probably depends on #359 since we would have to change the db interface.

Expected Outcomes

• Update blockchain module traits and structs as needed and fix corresponding tests
• Add add or combine tests if needed to cover updated functionality
• Update any modified public traits and structs documentation

Resources

• BDK Website
• BDK github repo
• BDK Discord

Skills Required

• Experience with git. Guide
• Basic familiarity with rust. First seven chapters of the book
• Familiarity with bitcoin descriptors and address generation
• Familiarity with changes made for schema migration in Handle database migrations #359

Mentor(s)
@notmandatory

Difficulty
Medium

Competency Test

• Install rust, compile and run all bdk examples and tests.
• Read through the BDK docs.
• Make a dummy wallet with BDK with key-value and sqlite backends.
• Familiarity with basic rust, should be able to write basic custom trait implementations on foreign structures.

[notmandatory]

Per our team chat yesterday I changed this into a summer-of-bitcoin project.

[LLFourn]

I think this one needs to be expanded up on since I don't know what I would do with it if I had to tackle it. In particular I'm not totally clear on the motivation and what the result should look like.

[notmandatory]

The genesis of this issue is #354.

My understanding is when addresses are requested (ie. get_address) prior to a sync and then setup isn't called. But by keeping track on individual addresses if they're setup or not at least the first part of #354 is solved since they can be setup if needed every time sync is called.

In my mind this change will create a table for all get_address scripts generated (or at least their indexes) with a boolean indicating if they are "setup" or not. Then when sync is called a query will be done to find any not-setup scripts, and if any are found call setup on them. I think the reason to not setup each script as it's requested is that it would be a performance hit. Also we allow random indexes to be requested so keeping only a range of index not setup wouldn't work.

But maybe there's a better way to address the two issues from #354 and we can explore that with whoever takes this on.

[notmandatory]

Closing this issue since sync/scan works differently now with the bdk_chain crate in v1.0.0.