itests: add event matrix tests for realtime eth filters and subscriptions

[iand]

Related Issues

Part of #9849

Proposed Changes

Bugs fixed:

• realtime filters/subs were using old method of checking for indexed values
• realtime filters/subs were not ignoring empty criteria for topics
• realtime filters/subs were comparing raw against cbor-encoded values. This is now internally consistent, but will change again , see Logs: Indexing/Filtering bugs ref-fvm#1345
• subscriptions were not matching alternate values of a topic due to byte slice backing array reuse causing them to use the same vale for each alternate

New/revised tests using the event matrix to cover wide variety of filter conditions:

• TestEthSubscribeLogs
• TestEthGetFilterLogs - tests EthNewFilter/EthGetFilterChanges combo
• TestEthGetFilterChanges - tests EthNewFilter/EthGetFilterLogs combo

New test:

• TestEthNewFilterMergesHistoricWithRealtime

Some noisy reorganisation of event matrix test cases to make them usable to the new tests

Additional Info

Checklist

Before you mark the PR ready for review, please make sure that:

• Commits have a clear commit message.
• PR title is in the form of of <PR type>: <area>: <change being made>
  • example: fix: mempool: Introduce a cache for valid signatures
  • PR type: fix, feat, build, chore, ci, docs, perf, refactor, revert, style, test
  • area, e.g. api, chain, state, market, mempool, multisig, networking, paych, proving, sealing, wallet, deps
• New features have usage guidelines and / or documentation updates in
  • Lotus Documentation
  • Discussion Tutorials
• Tests exist for new functionality or change in behavior
• CI is green

[magik6k]

Note - #10027 changes how the rpc works / is used, probably worth landing those tests first so that we can have some confidence in the rpc changes later

[iand]

Note - #10027 changes how the rpc works / is used, probably worth landing those tests first so that we can have some confidence in the rpc changes later

Do you mean landing the tests in this PR or waiting for the tests in 10027?

[iand]

Added cbor decoding to EthGetTransactionReceipt since we're doing it in the other places we construct an EthLog

[geoff-vball]

Just two comments for you!

[geoff-vball]

I know it wasn't written in this PR, but invokeAndWaitUntilAllOnChain sends messages into msgChan whether they're reverted or applied. It sets a reverted flag in the msgInTipset object, but this value is never checked. This won't cause any problems if your test setup only uses one miner because there will be no reverts, but this could cause problems with more complicated setups. We should either limit the use of this function to single-miner setups, or we should do something with the reverted flag, so we know a message might not actually be on chain.

[iand]

Good point. The ethereum event apis can use this flag, but these single miner tests aren't exercising it. We could add further tests that cause reversions and ensure they are emitted in the ethereum event log apis.

[iand]

Thanks for the review @geoff-vball. I've removed the debug statement. Are you able to ok this for merging?

[geoff-vball]

Yep, go ahead and merge after the last test passes.