test: StateAccount.Extra via trie.StateTrie.{Update,Get}Account()
#45
Conversation
This is a purely defensive approach in case future tests forget to clean up.
| testonly.OrPanic(func() {}) | ||
| } | ||
|
|
||
| eNil := e == nil || e.t == nil |
There was a problem hiding this comment.
This was moved from the test file, unchanged.
| ) | ||
|
|
||
| // OrPanic runs `fn` i.f.f. called from within a testing environment. | ||
| func OrPanic(fn func()) { |
There was a problem hiding this comment.
This was abstracted from params/config.libevm.go, unchanged except for the call to fn() and the size of the pc slice.
ARR4N
requested review from
a team,
darioush,
ceyonur and
michaelkaplan13
and removed request for
a team
| require.NoErrorf(t, state.UpdateAccount(addr, acct), "%T.UpdateAccount(...)", state) | ||
| assert.Equalf(t, tt.wantTrieHash, state.Hash(), "%T.Hash() after UpdateAccount()", state) |
There was a problem hiding this comment.
Can can we also add testing for Copy & Snapshot/Revert being handled properly?
For example statedb.Copy makes a shallow copy of the accounts modified, so we should ensure that setting the extras on the origin does not impact the copied statedb & the other way around.
There was a problem hiding this comment.
Snapshot/Revert covered in #48
I'm about to test Copy too. If it fails then I'll see how much needs to change and might start a new PR if the fix spreads too far.
Why this should be merged
Integration test of the new
StateAccount.Extrapayloads introduced in #44, originally with unit tests only.How this works
trie.StateTrieround trip viaUpdateAccount()and thenGetAccount(), confirming that not only are the payloads correctly echoed but that they also modify theStateTrie's root hash as expected.How this was tested
n/a