Use packable BTreeSet impls #1240
Conversation
Co-authored-by: Thoralf-M <[email protected]>
|
Conflicts |
|
Solve conflicts and I'll review/merge today |
| @@ -337,83 +363,86 @@ impl UnlockConditions { | |||
|
|
|||
| /// Creates a new [`UnlockConditions`] from a vec. | |||
| pub fn from_vec(unlock_conditions: Vec<UnlockCondition>) -> Result<Self, Error> { | |||
There was a problem hiding this comment.
So this wouldn't error anymore on duplicates ?
There was a problem hiding this comment.
Can we somehow bring it back?
| } | ||
| Self::TooManyUnlockConditions => write!(f, "too many unlock conditions"), |
There was a problem hiding this comment.
maybe we want max/current added?
| @@ -359,7 +359,7 @@ impl DelegationOutput { | |||
|
|
|||
| // Transition, just without full ValidationContext. | |||
| pub(crate) fn transition_inner(current_state: &Self, next_state: &Self) -> Result<(), StateTransitionError> { | |||
| if !(current_state.delegation_id.is_null() && !next_state.delegation_id().is_null()) { | |||
| if !current_state.delegation_id.is_null() | next_state.delegation_id().is_null() { | |||
There was a problem hiding this comment.
nice bit stuff :D lets add a |
| @@ -67,6 +71,17 @@ impl Ord for Feature { | |||
| self.kind().cmp(&other.kind()) | |||
| } | |||
| } | |||
| impl core::hash::Hash for Feature { | |||
| fn hash<H: core::hash::Hasher>(&self, state: &mut H) { | |||
| self.kind().hash(state); | |||
There was a problem hiding this comment.
Are we sure we want the feature hash to just be the kind? that would make 2 different SenderFeatures have the same hash
There was a problem hiding this comment.
just noticed that the Eq is also just kind. for our code it works, but what if someone gets outputs and compares it with their feature, youd just expect == to work.
There was a problem hiding this comment.
We started speaking about it on Slack, we should definitely continue the discussion. I really dislike this.
|
I'm going to close this, as the concerns are valid and additionally I believe there is a better way to do this. |
Description of change
This PR uses BTreeSets rather than boxed slices where possible when unique, ordered data is required.
Related Issues
Blocked by iotaledger/common-rs#70