cli: display multiwallet balances in -getinfo

[jonatack]

This PR is a client-side version of #18453, per review feedback there and review club discussions. It updates bitcoin-cli -getinfo on the client side to display wallet name and balance for the loaded wallets when more than one is loaded (e.g. you are in "multiwallet mode") and -rpcwallet= is not passed; otherwise, behavior is unchanged.

before

$ bitcoin-cli -getinfo -regtest
{
  "version": 199900,
  "blocks": 15599,
  "headers": 15599,
  "verificationprogress": 1,
  "timeoffset": 0,
  "connections": 0,
  "proxy": "",
  "difficulty": 4.656542373906925e-10,
  "chain": "regtest",
  "balance": 0.00001000,
  "relayfee": 0.00001000
}

after

$ bitcoin-cli -getinfo -regtest
{
  "version": 199900,
  "blocks": 15599,
  "headers": 15599,
  "verificationprogress": 1,
  "timeoffset": 0,
  "connections": 0,
  "proxy": "",
  "difficulty": 4.656542373906925e-10,
  "chain": "regtest",
  "balances": {
    "": 0.00001000,
    "Encrypted": 0.00003500,
    "day-to-day": 0.00000120,
    "side project": 0.00000094
  }
}

---

Review club discussion about this PR is here: https://bitcoincore.reviews/18453

This PR can be manually tested by building, creating/loading/unloading several wallets with bitcoin-cli createwallet/loadwallet/unloadwallet and running bitcoin-cli -getinfo and bitcoin-cli -rpcwallet=<wallet-name> -getinfo.

wallet_multiwallet.py --usecli provides regression test coverage on this change, along with interface_bitcoin_cli.py where this PR adds test coverage.

Credit to Wladimir J. van der Laan for the idea in #17314 and #18453 (comment).

[promag]

Concept ACK

interface_bitcoin_cli.py fails in no wallet job.

[fanquake]

It seems this is in part reverting changes that were just merged in #18574 (split out of #18453), and refactoring to handle multiwallet.

I understand splitting up changes, and maybe it doesn't matter so much in this instance, but it's a bit of code/review churn if we're PR'ing and merging refactors only to essentially undo the changes and do something else in a different PR a few hours later.

[jonatack]

I empathise. See #18453 (comment). #18574 was the only change that seemed to have consensus, so it seemed best to split it out for merge and propose client-side code in this PR to test and compare with the server-side code in #18453.

EDIT: This PR no longer touches the changes merged in #18574.

[DrahtBot]

The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.

Conflicts

Reviewers, this pull request conflicts with the following ones:

• cli/gui: support "@height" in place of blockhash for getblock on client side #16439 (RPC: support "@height" in place of blockhash for getblock etc by ajtowns)

If you consider this pull request important, please also help to review the conflicting pull requests. Ideally, start with the one that should be merged first.

[jonatack]

Updated with the following:

• ensure it works when built without the wallet
• extract the connection try/wait/failure logic to be callable for each connection, so we can use this for calling listwallets and getbalances to be more robust
• drop an unused arg in JSONRPCProcessBatchReply as requested in cli: call getbalances.ismine.trusted instead of getwalletinfo.balance #18574 (comment)

[jnewbery]

Concept and approach ACK.

[michaelfolkson]

ACK ea59b18387c4b26240e1f087c568408922b3c940

Built, ran tests, tested manually by creating new wallet, transferring funds to it and unloading. Light code review.

[jonatack]

Agree that a release note followup which can include a few other things would be good

Thanks -- done in #19089 and #19354.