Use proper SCREAMING_SNAKE_CASE conversion for EnumCount

[meithecatte]

If an enum is called FooBarBaz, EnumCount will name the constant FOOBARBAZ_COUNT. I believe FOO_BAR_BAZ_COUNT would be more expected. This can be achieved by heck, which is already a dependency of strum.

[Peternator7]

Thanks for bringing this up! I have a few thoughts on the issue.

1. This would be a breaking change, which in my mind means there's a higher threshold to reach than simply adding a new feature.
2. I definitely agree that FOO_BAR_BAZ_COUNT is cleaner in your example. I'm not sure that is true in all cases though. If the enum was called My_eNum1a2B, it wouldn't be obvious to me what the snake_case version is, but I can easily figure out the uppercase version.

I'm going to tag this as "feedback wanted" because I think more opinions on the matter would be useful.

You can also create a new constant and export that name while still getting most of the benefits of using strum

pub const FOO_BAR_BAZ_COUNT: usize = FOOBARBAZ_COUNT;

[meithecatte]

1. This would be a breaking change, which in my mind means there's a higher threshold to reach than simply adding a new feature.

I am aware that this would be a breaking change. However, strum could generate both styles of constants, and mark the "simple uppercase" ones as deprecated and doc(hidden), to facilitate a transition period.

2. I definitely agree that FOO_BAR_BAZ_COUNT is cleaner in your example. I'm not sure that is true in all cases though. If the enum was called My_eNum1a2B, it wouldn't be obvious to me what the snake_case version is, but I can easily figure out the uppercase version.

If that's what you name your enums, you did this to yourself. If it's just the fact that you're using a non-standard naming convention, heck handles this well - in fact, you don't even need to know the original "case".

[ArekPiekarz]

I would also like to see this change implemented. Currently enums with a few words in their names get hard to read generated constants for their count. Users are already expecting breaking changes because of semver rules about libraries below version 1.0.