chore: upgrade jsonrpsee to 0.11

[tarrencev]

i've started the process of upgrading to 0.11 to support configuring the hyper instance to handle health checks. it's mostly just adjusting imports/namespaces. im pretty new to rust, so would appreciate any feedback / guidance. in particular, i'm struggling to case the errors to the new type

[koivunej]

in particular, i'm struggling to case the errors to the new type

If you mean this error:

error[E0277]: the trait bound `jsonrpsee::jsonrpsee_types::error::ErrorCode: From<rpc::types::reply::ErrorCode>` is not satisfied
   --> crates/pathfinder/src/rpc/types.rs:404:29
    |
404 |                 code: ecode.into(),
    |                             ^^^^ the trait `From<rpc::types::reply::ErrorCode>` is not implemented for `jsonrpsee::jsonrpsee_types::error::ErrorCode`
    |
    = help: the following implementations were found:
              <jsonrpsee::jsonrpsee_types::error::ErrorCode as From<i32>>
    = note: required because of the requirements on the impl of `Into<jsonrpsee::jsonrpsee_types::error::ErrorCode>` for `rpc::types::reply::ErrorCode`

Looking at it, it's quite difficult to understand from the error message. Since it used to be that CallError::Custom { code: i32, ... }. To get past this on to the next error, you can just:

$ git diff
diff --git a/crates/pathfinder/src/rpc/types.rs b/crates/pathfinder/src/rpc/types.rs
index 88ed24b..1bd52f2 100644
--- a/crates/pathfinder/src/rpc/types.rs
+++ b/crates/pathfinder/src/rpc/types.rs
@@ -400,8 +400,9 @@ pub mod reply {

     impl From<ErrorCode> for Error {
         fn from(ecode: ErrorCode) -> Self {
+            let error: i32 = ecode as i32;
             Error::Call(CallError::Custom(ErrorObject {
-                code: ecode.into(),
+                code: error.into(),
                 message: Cow::Owned(ecode.to_string()),
                 data: None,
             }))

So our ErrorCode can be converted to i32 with as i32 conversion because it is defined as enum Foo { Variant = 1, ... }, and the jsonrpcsee ErrorCode has an From<i32> conversion, so it sounds like this would be at least a temporary measure to get forward. There will be a next error with the introduced ErrorObject which has ErrorObject::owned for creating it.

Hopefully this helps you forwards! Please keep asking questions but as a reminder we haven't done the upgrade yet so there might be more troublesome parts ahead!

[koivunej]

Using ErrorObject::owned is a bit difficult as well since we need to pass the None. This seems to work ErrorObject::owned(error, Cow::Owned(ecode.to_string()), Option::<()>::None) in the setting after introducing the error: i32 from previous message, and let you on to perhaps more clear errors.

[tarrencev]

@koivunej thanks so much for your feedback! I made some progress and the binary is compiling now!

[koivunej]

You can just add commits for now and rebase when done, it would be easier to follow to help if need be.

[koivunej]

@tarrencev I found the cause for rpc::tests::events::positional_args::get_events_with_invalid_page_size failure and fixed it in 1e86e79. The previous version looked very much the same but since the ErrorObject::owned takes a serializable object now in 0.11 (instead of RawValue in 0.6), the serde_json::value::Value needs to be given instead of it's stringified version. Pushed and cleaned up most of the tarrancev_updaterpsee branch.

[tarrencev]

@tarrencev I found the cause for rpc::tests::events::positional_args::get_events_with_invalid_page_size failure and fixed it in 1e86e79. The previous version looked very much the same but since the ErrorObject::owned takes a serializable object now in 0.11 (instead of RawValue in 0.6), the serde_json::value::Value needs to be given instead of it's stringified version. Pushed and cleaned up most of the tarrancev_updaterpsee branch.

awesome. thanks for taking it on. should we close this in favor of your pr?

[koivunej]

Let's just keep it these both open, I can handle the merging of the work when we are done reviewing. Thanks a lot for getting the ball rolling!

[kkovaacs]

LGTM

[CHr15F0x]

Looking good 👍

[tarrencev]

@koivunej anything i can help with to push this across the line? :)

[tarrencev]

@koivunej @kkovaacs @CHr15F0x thanks for your help on this. i've rebased + migrated some new code to be compatible with 0.11. would be great to get this merged so we can support #194

[Mirko-von-Leipzig]

Some clippy, test and doc fixes required.

[koivunej]

@tarrencev We will be postponing this until we get the next release out, so nothing at the moment. Of course, feel free to handle any of the clippy/doc failures, which should be quite straight forward.

[tarrencev]

@tarrencev We will be postponing this until we get the next release out, so nothing at the moment. Of course, feel free to handle any of the clippy/doc failures, which should be quite straight forward.

ok thanks for the update! i'll keep this branch in shape and work on another on top to expose the health check

[koivunej]

The failed update::tests::fetch_latest_github_release must be a transient one.

[koivunej]

The remaining test failures should work okay, the issue is with secrets.

[koivunej]

Rebased to the latest, secrets are still the only issue.

[CHr15F0x]

LGTM!

[koivunej]

Allright, thanks a lot @tarrencev! Apologies for the delay, we had some interesting times with releases and bugs in the meantime :)