[move-compiler] Support casting without parentheses

[macalinao]

Motivation

Easier to read and type if casts do not require enclosing parentheses.

Have you read the Contributing Guidelines on pull requests?

Yes

Test Plan

Updated language tests, and ran:

cargo test --workspace --exclude move-prover --exclude move-to-yul --exclude evm-exec-utils -- --skip prove

[wrwg]

I'm not sure what the consequences are. There is likely a reason why there are braces (like certain syntax constructs become ambiguous otherwise), but perhaps also not. Do you remember @tnowacki ?

Unfortunately, these small changes look nice and harmless isolated but may have more indirect impact than we understand.

[macalinao]

I'm not sure what the consequences are. There is likely a reason why there are braces (like certain syntax constructs become ambiguous otherwise), but perhaps also not. Do you remember @tnowacki ?

Unfortunately, these small changes look nice and harmless isolated but may have more indirect impact than we understand.

Is there a way we can figure out what the impact could be via test cases? If so, what are some test cases we can construct to test this?

[wrwg]

I'm not sure what the consequences are. There is likely a reason why there are braces (like certain syntax constructs become ambiguous otherwise), but perhaps also not. Do you remember @tnowacki ? Unfortunately, these small changes look nice and harmless isolated but may have more indirect impact than we understand.

Is there a way we can figure out what the impact could be via test cases? If so, what are some test cases we can construct to test this?

We need @tnowacki input here because he was present when this grammar was designed.

[sblackshear]

Agreed that we need @tnowacki's input here; I don't have enough context on the grammar. I do have some ideas for precedence stress tests:

• my_num / 2 as u64
• my_num as u64 as u64
• a as u64 / b as u64

[wrwg]

This PR has been orphan for a while. Everybody seems to want it but concerns are around breaking changes. @macalinao are you intending to continue work on this? Then please feel free to re-open. @tnowacki do you want to take this issue on your todo list? Then we may want to turn this into an issue instead.

[tnowacki]

Filed an issue so at least this work won't be lost

[oxade]

This PR has been orphan for a while. Everybody seems to want it but concerns are around breaking changes. @macalinao are you intending to continue work on this? Then please feel free to re-open. @tnowacki do you want to take this issue on your todo list? Then we may want to turn this into an issue instead.

I spoke with @macalinao a few weeks back about closing another PR he had open. Turns out he's very busy right now so we can probably carry on the work.