Missing test(s) for message recovery

[sitetester]

What was the problem?

This PR resolves #8193

How was it solved?

Update code since LIP is updated

How was it tested?

Corresponding tests added

[codecov]

Codecov Report

Merging #9137 (bc3dc83) into release/6.1.0 (7d8301a) will not change coverage.
The diff coverage is n/a.

Additional details and impacted files

@@              Coverage Diff               @@
##           release/6.1.0    #9137   +/-   ##
==============================================
  Coverage          84.35%   84.35%           
==============================================
  Files                655      655           
  Lines              24113    24113           
  Branches            3497     3497           
==============================================
  Hits               20341    20341           
  Misses              3772     3772           

Files	Coverage Δ
...roperability/mainchain/commands/recover_message.ts	99.38% <ø> (ø)

[shuse2]

im not sure why this is added 🤔

[gkoumout]

This check is not needed in the mainchain. As the comment in the LIP says

        # Check that sending chain is valid. This check is relevant only in case direct sidechain channels are enabled
        if OWN_CHAIN_ID != getMainchainID() and ccm.sendingChainID != OWN_CHAIN_ID:
            raise  Exception("Cross-chain message sending chain ID is not valid.")

Currently (no direct channels) the message recovery command exists only in mainchain and this extra check is not needed. It is anyways pointless since the OWN_CHAIN_ID != getMainchainID() would never be true in mainchain, so adding it would never change the outcome of the verification.

In the future, in case direct sidechain channels get enabled, then the message recovery command would also need to be supported in the sidechains. There, we would need this check to ensure the validity (this check essentially says that sidechains can never forward messages from other chains).

So this check would be needed only in case message recovery gets enabled in sidechains.

[AndreasKendziorra]

All changes look good, but there is one open point in issue #8193 , namely the very last one:

Tests of lines 594, 633: I am not sure that it is checked that recoveredCCMs array is correctly updated.

The tests were refactored quite a bit in the meantime, but the corresponding ones in the current version are the ones in line 572 and 603. So far, it is only tested that terminatedOutboxStore.set is called, but is not tested that it is called with the correct value. So I think the right way to do this is

1. testing that RMTCalculateRoot was called with the right values (namely recoveredCCMs as Greg commented), and
2. that terminatedOutboxStore.set is called with the return value of RMTCalculateRoot.

[AndreasKendziorra]

due to this, the expectation in line 597 will be met, even though it was not triggered from the function you are testing, right? Alternatively, you could create a copy of ctx, change the status of ccm to CCM_STATUS_CODE_RECOVERED and swap ccm.sendingChainID and ccm.receivingChainID.

[AndreasKendziorra]

same as above