Refactor tx kernel events errors

[Dominik1999]

Closes #768

This PR moves (almost) all events into api.masm. All slot-check are now in the subsequent files from api.masm

Two open questions about events that are not in api.masm yet:

• The event ACCOUNT_PUSH_PROCEDURE_INDEX_EVENT seems to be needed in account.masm or we need to rewrite this test

  miden-base/miden-tx/src/tests/kernel_tests/test_account.rs

  Lines 506 to 510 in 13724e9

  	fn test_authenticate_procedure() {
  	let tx_context = TransactionContextBuilder::with_standard_account(
  	ACCOUNT_ID_REGULAR_ACCOUNT_UPDATABLE_CODE_OFF_CHAIN,
  	ONE,
  	)

  Rewriting the test is not hard, just not sure, if this is what we want.

• Some events are still emitted in tx.masm regarding note creation. I can move those to api.masm, too.

# EVENTS
# =================================================================================================

# Event emitted before a new note is created.
const.NOTE_BEFORE_CREATED_EVENT=131083
# Event emitted after a new note is created.
const.NOTE_AFTER_CREATED_EVENT=131084

# Event emitted before an ASSET is added to a note
const.NOTE_BEFORE_ADD_ASSET_EVENT=131085
# Event emitted after an ASSET is added to a note
const.NOTE_AFTER_ADD_ASSET_EVENT=131086

Let me know what you think @bobbinth, then I can finalize the PR and we can merge this week.

[bobbinth]

Looks good! Thank you! I left a few comments line. I think the biggest issue is that moving storage events into api.masm would break how faucets work. I don't have a great solution to this yet - but also, we should probably modify a test or two to make sure these errors get caught (or confirm that there are no errors with the current changes).

[bobbinth]

This should probably be un-commented?

[Dominik1999]

Indeed, I forgot to work on this. The test also passes when un-commented.

[bobbinth]

I am not sure this is going to work. If I'm understanding the flow correctly:

• account::set_item_raw will not fire STORAGE_ON_AFTER_SET_ITEM event.
• This means that Host::on_account_storage_after_set_item() won't get executed and the AccountDeltaTracker won't be updated.
• This means that changes to the account storage for the faucet reserved slot won't be included in account delta.

We should probably modify one of the tests to test for this (i.e., when we mint a new asset, account storage delta should include the update incrementing the total number of assets minted).

[bobbinth]

One potential way to handle this is to introduce new events which get fired in mint_asset and burn_asset procedures in api.masm. Then handlers of these events would modify the account delta tracker accordingly.

Though, this may be introducing too much complexity.

[Dominik1999]

I was thinking about the same. If this change ("firing all events from api.masm") is purely aesthetics, we should not introduce this extra complexity. If the change adds readability to the transaction kernel, we should make the change.

One problem is that in api.masm, we don't have the information that we want to update yet. To update the account_storage_delta in the correct way, we need the TOTAL_ISSUANCE, which we compute at a later stage. So, to fire all events in api.masm, we need to rewrite the way how the faucet works.

[bobbinth]

Similar comment to the one above and this is also relevant to the other changes in this file.

[bobbinth]

nit: I would probably move drop dropw to line 203 below and update the comments accordingly.

[bobbinth]

Similar nit as above.