Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Drink backend: support runtime call #1891

Merged
merged 10 commits into from
Sep 7, 2023

Conversation

pmikolajczyk41
Copy link
Member

Since inkdevhub/drink#36 is done, we can finally support arbitrary runtime calls in e2e tests. However, the solution is kinda tricky.

Problem

In drink! we have strongly typed runtime calls. This is because we keep the runtime directly, in-memory, and thus we have access to the RuntimeCall enum family. And thus there's no need of dynamic dispatch.

In standard e2e backend (i.e. node+subxt) we must have only dynamic dispatch. This is because we support running a node with an arbitrary runtime, and therefore we cannot statically generate runtime types, including RuntimeCall enum family.

But in order to support using both backends with the same test code, we have to provide a common interface...

Solution

We can't make ChainBackend::runtime_call statically typed, since for subxt_client::Client there is no way of providing correct types. Hence, we have to leave the API as it is (i.e. dynamic dispatch with pallet and call as strings and call data as subxt::Value wrappers).

For the drink! implementation we do following mumbo-jumbo:

  • we get metadata from the runtime
  • using subxt we create a valid call object using arguments and metadata
  • we encode the call object
  • we decode it to the RuntimeCall object
    🤷‍♂️

Note

Since drink::MinimalRuntime and substrate-contracts-node's runtime use different addressing, the testcases that relate to runtime calling are compatible with only one backend. However, notice that this is the incompatibility of used runtimes, not backends, and thus it's okay.

Copy link
Contributor

@SkymanOne SkymanOne left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You could potentially do some hacky type inference here, but current solutions looks good enough to me.

# Conflicts:
#	crates/ink/tests/ui/contract/fail/message-returns-non-codec.stderr
#	crates/ink/tests/ui/trait_def/fail/message_output_non_codec.stderr
@@ -33,6 +33,7 @@ tracing = { workspace = true }
tracing-subscriber = { workspace = true, features = ["env-filter"] }
scale = { package = "parity-scale-codec", workspace = true }
subxt = { workspace = true }
subxt-metadata = { workspace = true }
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we make this one optional too and dependant on the drink feature?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

oh, sure, my bad; done

Copy link
Collaborator

@ascjones ascjones left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In standard e2e backend (i.e. node+subxt) we must have only dynamic dispatch. This is because we support running a node with an arbitrary runtime, and therefore we cannot statically generate runtime types, including RuntimeCall enum family.

Yes and no, I think it would be beneficial to define a minimal set of hardcodeed common calls that are useful and supported by both substrate-contracts-node and drink. Balances::transfer etc... (in fact we already do this for the pallet_contracts calls instantiate, call etc.

Of course people can still target their own custom node and then if they don't satisfy that API then calls simply return an error "Not implemented for target node".

Either way this dynamic API is required and very useful.

@SkymanOne SkymanOne merged commit d540cdd into use-ink:master Sep 7, 2023
@pmikolajczyk41
Copy link
Member Author

Either way this dynamic API is required and very useful.

Sure it is, I'm not attacking it. Just saying that it is not (and won't be) natively supported in drink

@pmikolajczyk41 pmikolajczyk41 deleted the drink-runtime-call branch September 7, 2023 17:12
@SkymanOne SkymanOne mentioned this pull request Mar 4, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants