[storage] use commonware-codec to serialize/deserialize proofs
#1127
Conversation
|
Re:
Actually that comment around "<= 255 digests" is obsolete since we actually removed explicit encoding of the length of the digests vector. |
I am not too familiar with codec usage, but enforcing a reasonable limit on the # of digests sounds useful to me so someone doesn't try to spam us with massive proofs. OK with me if this requires an explicit vector length encoding to do through standard means. |
| } | ||
| } | ||
|
|
||
| impl<H: CHasher> EncodeSize for Proof<H> { |
There was a problem hiding this comment.
Not really feedback on this specific PR, but shouldn't we be able to write a macro that implements these "default" types of code serialize/deserialze implementations? This all seems very boilerplate.
There was a problem hiding this comment.
I've advocated for manually encoding to avoid silent changes from struct modification fwiw
Codecov ReportAttention: Patch coverage is
@@ Coverage Diff @@
## main #1127 +/- ##
==========================================
- Coverage 91.04% 91.03% -0.01%
==========================================
Files 215 215
Lines 57480 57471 -9
==========================================
- Hits 52330 52321 -9
Misses 5150 5150
Continue to review full report in Codecov by Sentry.
🚀 New features to boost your workflow:
|
Replaces custom
serialize/deserializeimplementation.Open questions:
u16::MAX? Or should we keep the existing assumption that there are <= 255 digests?max_lenfield in the read config? (Maybe moot if we limit number of digests to 255).