Bug (?) in atom_naming_convention

[paulo-ferraz-oliveira]

Bug Description

atom_naming_convention might be wrongly implemented (the default regex part).

I recently noticed non200_ was a possible atom / function_name, but this is probably undesired.

Now, if we fix it as per our initial intention (which was...?), it is a breaking change, so we'd need to bump major.

To Reproduce

Have non200_ in your code as an atom, or function name.

Expected Behavior

The regex should be improved, but also... why would non_200 not be accepted?

[elbrujohalcon]

non_200 should be accepted. Let's improve the regex

[paulo-ferraz-oliveira]

Yeah, tentatively I'm leaving this here: ^([a-z][a-z0-9]*(?:_[a-z0-9]+)*)(?:_SUITE)?$.

We force:

• starting with [a-z]
• potentially following with 0 or more [a-z0-9]
• potentially following with 0 or more "_ forcefully followed by [a-z0-9]"
• eventually ending in _SUITE