feat: Update to use the newest XDR defitinions

[willemneal]

What

Use newest XDR.

Why

[TODO: Why this change is being made. Include any context required to understand the why.]

Known limitations

N/A

Issues

• soroban-cli: value overhaul update - strval json/xdr parsing/serialization #478
• soroban-cli: value overhaul update - contract deploy/install commands #479
• soroban-cli: value overhaul update - contract code fetching for invoke arg parsing, inspect, read #480
• soroban-cli: value overhaul update - inspect ScSpecTypeDef formatting #481
• soroban-cli: token wrap ScVals & footprint for the contract code #482

[paulbellamy]

Looking great so far!

[willemneal]

The last failing test is when trying to return a complex enum that includes a case with a tuple including a tuple struct.

To see the failing test pull the latest and run

cargo test --package soroban-cli --test it -- custom_types::complex_tuple_output --exact --nocapture

You should get:

-----------Result:
Vec(
    Some(
        ScVec(
            VecM(
                [
                    Symbol(
                        ScSymbol(
                            StringM(Tuple),
                        ),
                    ),
                    Vec(
                        Some(
                            ScVec(
                                VecM(
                                    [],
                                ),
                            ),
                        ),
                    ),
                ],
            ),
        ),
    ),
)

A new method was added to highlight this error:

    pub fn complex_tuple(_env: Env) -> ComplexEnum {
        ComplexEnum::Tuple(TupleStruct(
            Test {
                a: 1,
                b: true,
                c: Symbol::short("test"),
            },
            SimpleEnum::First,
        ))
    }

At first I thought this was a bug in the enum parsing on my end, but this is the value that is returned from the call.

Note: you need to rebuild the examples with make build-test-wasms.

[graydon]

@willemneal apologies, a bit of a hard landing of a lot of new code and this was obviously a broken path we just didn't have a test for. fix is in stellar/rs-soroban-env#726 which should merge shortly.

[willemneal]

@paulbellamy Currently this depends on stellar/rs-soroban-sdk#886 so it shouldn't merge until then, and I'm getting some errors in CI because it currently depends on AhaLab's branch.

[paulbellamy]

(changed base back-and-forth so github would stop showing a diff that was already in the base branch)

[paulbellamy]

Missing test cases, and sdk bug aside, LGTM.

[paulbellamy]

typo?

[willemneal]

Nope an update from the sdk added a new implicit address method. Should I add a comment?

[willemneal]

I forgot this was pointing at a different branch. I guess we have less stress merging it. Then we could have another PR to add the tests and then merge to main? Or should we just change the branch to main now?

[paulbellamy]

Yeah, let's do it. So long as we don't forget the fixes.