Add test to assert macro hygiene for Encode and Decode derives

[Robbepop]

Uses the #![no_implicit_prelude] attribute to add UI tests to guard Encode and Decode derives against macro hygiene mistakes.

This way the PR found a minor macro hygiene bug in Decode and fixes it.
Also fixed a minor bug with MaxEncodedLen test.

[gui1117]

miri is failing because of a false positive #288 (comment)
I think you can remove the miri test

[Robbepop]

@bkchr applied your suggestion. I just copied the code and comment from the other UI test.

@thiolliere what do you mean by remove miri test? It is failing in another crate so I think it is okay to merge the PR, right?

Could we do a minor release after this? It would be nice to have the latest changes in dependencies such as ink! that also has tests for no-implicit-prelude but fails due to parity-scale-codec.

[gui1117]

@thiolliere what do you mean by remove miri test? It is failing in another crate so I think it is okay to merge the PR, right?

the CI is failing, I think if we keep a ci test failing for a false positive we will never look at it again and we will miss in case another error is raised.

By removing the miri test I mean this:

diff --git a/src/bit_vec.rs b/src/bit_vec.rs
index 1beca32..35d3123 100644
--- a/src/bit_vec.rs
+++ b/src/bit_vec.rs
@@ -191,6 +191,10 @@ mod tests {
        }
 
        #[test]
+       // Flaky test due to:
+       // * https://github.com/bitvecto-rs/bitvec/issues/135
+       // * https://github.com/rust-lang/miri/issues/1866
+       #[cfg(not(miri))]
        fn bitvec_u32() {
                for v in &test_data!(u32) {
                        let encoded = v.encode();

[Robbepop]

The miri test is removed now. Ready to merge. :)

[gui1117]

I don't mind another release, we can do another PR which update the changelog and version of parity_scale_codec.
Actually we can do a patch IMO 2.3.1

[Robbepop]

Actually we can do a patch IMO 2.3.1

👍

I don't mind another release, we can do another PR which update the changelog and version of parity_scale_codec.

Are you doing that or shall I?

[gui1117]

I'm not doing currently, you can do