As patterns in offchain code #94
Conversation
choener
left a comment
choener
left a comment
There was a problem hiding this comment.
LGTM.
Locally I am still waiting on cache.iog.io, so can't yet try the changes, but they look good.
In case I see something later locally, I'll raise an issue.
| import PlutusLedgerApi.V3 (Credential) | ||
|
|
||
| data AppError era = | ||
| data ProgrammableTokensError = |
There was a problem hiding this comment.
Good, it looks like the main error type holds each suberror only once!
There was a problem hiding this comment.
Yep that is the only way to avoid issues with catch, sadly I don't think there is a way to really enforce this
| | GlobalParamsNodeNotFound -- ^ The node with the global parameters was not found | ||
| | BalancingError (CoinSelection.BalanceTxError era) | ||
| | BlockfrostErr BlockfrostError | ||
| -- TODO: The following errors are specific to the regulated stablecoin |
There was a problem hiding this comment.
Here one would wrap some RegulatedStablecoinError and these should in turn have AsRegulated... or maybe down in AppError.
There was a problem hiding this comment.
Yes I'll split up the type in another PR.
|
|
||
| makeClassyPrisms ''ProgrammableTokensError | ||
|
|
||
| data AppError era = |
There was a problem hiding this comment.
Each error type shows up once, good.
| deployAll = do | ||
| target <- ask | ||
| failOnError $ Env.withEnv $ asAdmin @C.ConwayEra $ do | ||
| failOnError @_ @(AppError C.ConwayEra) $ Env.withEnv $ asAdmin @C.ConwayEra $ do |
There was a problem hiding this comment.
I would consiver expanding Convex.MockChain a bit, and provide a MonadError instance.
Then it would not be necessary to call failOnError. A simple newtype for the unit tests in here would suffice.
Though it would not save on lines / characters since instead of MonadFail, we'd carry around a bunch of AsError. However, it would be clear what type of errors are being tested for.
|
FYI: I get a segmentation fault trying to build WSC via nix. |
|
The build passed so I'm going to merge this now, let me know if the segfault is persistent |
This seems to be weirdly specific. My AMD/Ryzen3 machine produces a segfault, but actually earlier than the wsc code, after checking. |
|
Hm, on my AMD machine I get an error when trying to |
No description provided.