feat: support formatter alternate display

[tdelabro]

Rust formatting for hex representation used the alternate formatted attribute to decide if the 0x prefix should be used or not.
You are not doing it.
This causes an issue downstream in https://github.com/starknet-io/types-rs, forcing me to do intermediate string allocation to strip the prefix and then really it if actually specified in the formater

See: https://github.com/starknet-io/types-rs/blob/main/crates/starknet-types-core/src/felt/alloc_impls.rs#L29-L35

Also, this is a breaking change, as all the previous uses without {:#} will now print without 0x, and only the ones with {:#} will print 0x...
So maybe we want to do the opposite: 0x being the default and no 0x being the alternate. It's a bit quirky, as it opposes the way Rust lang's LowerHex and UpperHex traits handle this, but it would still work.

Also, I fixed this single instance because it is the one I depend on, but it may be a more widespread issue across the rest of your codebase.

Also, you could probably implement LowerHex and UpperHex and handle sign, padding, alternate, etc correctly

[diegokingston]

There is an inconsistency for no-std. Fixing that should allow us to merge it

[tdelabro]

Hey, I added LowerHex and UpperHex requirements to the trait IsUnsignedInteger.
It really helps later on in from_reduced_big_uint, by not having to guess if the generic type passed used hex or decimal, it's Display implementation. We always use LowerHex without prefix instead.
Less code, better performance.

Display actually doesn't have an alternate form according to the Rust docs, so we always go for 0x... This is arbitrary, but coherent with the previous implementation (which makes it a non-breaking change after all).
For Lower/UpperHex, we handle the alternate flag accordingly.
This way, everything is covered and up to rust standards

edit: it is actually a breaking change because the public IsUnsignedInteger type now has increased requirement.

[tdelabro]

@pablodeymo @diegokingston wdyt of the new impl?

[pablodeymo]

🚀

[diegokingston]

Looks good! We need all the commits to be signed and we can merge it!

[tdelabro]

@diegokingston my commits are not signed by default? How do I fix this?

Also, codecov failed, probably because I'm pushing from a fork, wdyt?

[diegokingston]

I guess you need to configure your signing key https://docs.github.com/en/authentication/managing-commit-signature-verification/signing-commits or check if you're using one password https://developer.1password.com/docs/ssh/git-commit-signing/. You have a way of signing everything "git rebase --exec 'git commit --amend --no-edit -n -S' -i main" though you may need to use "git push --force-with-lease". Uploading codecov is not a problem, that's something I can solve on our side

[tdelabro]

@diegokingston done, my commits are signed and verified