feat(#297): New rule: Prefer Unquoted Atoms

[bormilan]

This is in an early stage, I have som things to fix.

[elbrujohalcon]

The code is fine, I would just simplify it by removing the options.

[elbrujohalcon]

Once you're done with the code, @bormilan … you also need to add docs to this rule (in RULES.md) and you need to include it in the appropriate ruleset(s) (I guess erl_files and hrl_files)

[bormilan]

Thanks for all the help and suggestions, currently I'm on a little vacation, but after that I will fix all of them and update the PR.

[bormilan]

Today I made some changes, but I had some questions that I commented on your suggestions above.
Besides these, I have made two hardcoded variables that I want to put somewhere else(bc they are constant values), but I don't know where to. If you could help me with that I would appreciate it. These are Regex and KeyWords.

[elbrujohalcon]

A few improvements and… I wonder what happens with 'maybe' as an atom.
I checked erlang/otp's code and wherever there is a list of keywords/reserved words… they do not include maybe, but I guess… if the feature is enabled (and I think it's enabled by default on OTP27+), that one should also be considered a keyword, right?

[elbrujohalcon]

The code is perfectly fine, @bormilan … but I still believe this rule belongs to elvis_text_style and not elvis_style.

[bormilan]

So, should we add "maybe" to the exceptions? I think we should. But because now it's not a keyword by default, we may wait and add it when erl/OTP27 comes.

[elbrujohalcon]

Personally… I would add it right now… It will be a reserved word soon enough.
I wish Erlang/OTP would have a public function that returned the list of reserved words somewhere (erl_syntax may be a great place)… 🤷🏻

[elbrujohalcon]

Excellent work, @bormilan !!! Thank you very much!

[bormilan]

Thank you so much for all the help and patience. It was a great start to my journey.

[bormilan]

I found a little problem, the formatter changes the 'maybe' to maybe in the list of exceptions. Should I turn it off to that file, or we should consider changing this in the formatter itself to make an exception for it?

[elbrujohalcon]

@bormilan The list of exceptions should have strings and not atoms… that would fix it :)

[elbrujohalcon]

But… the formatter is broken, anyway… please report an issue there.

[bormilan]

Personally… I would add it right now… It will be a reserved word soon enough. I wish Erlang/OTP would have a public function that returned the list of reserved words somewhere (erl_syntax may be a great place)… 🤷🏻

I think f_reserved_word does exactly what you want in the module erl_scan

[elbrujohalcon]

Personally… I would add it right now… It will be a reserved word soon enough. I wish Erlang/OTP would have a public function that returned the list of reserved words somewhere (erl_syntax may be a great place)… 🤷🏻

I think f_reserved_word does exactly what you want in the module erl_scan

Yeah, but it's not exported, is it?

[bormilan]

Personally… I would add it right now… It will be a reserved word soon enough. I wish Erlang/OTP would have a public function that returned the list of reserved words somewhere (erl_syntax may be a great place)… 🤷🏻

I think f_reserved_word does exactly what you want in the module erl_scan

Yeah, but it's not exported, is it?

It is:

[elbrujohalcon]

Personally… I would add it right now… It will be a reserved word soon enough. I wish Erlang/OTP would have a public function that returned the list of reserved words somewhere (erl_syntax may be a great place)… 🤷🏻

I think f_reserved_word does exactly what you want in the module erl_scan

Yeah, but it's not exported, is it?

It is:

Ah, right! I was thinking of something like f_reserved_words/0 (that would return the list) :)

But… maybe we can actually use that here or in rebar3_format… 🤔
If you want to send a PR using erl_scan:f_reserved_word/1 to verify that the atom is actually a reserved word instead of the hardcoded list of strings we now have… I'd merge it, for sure!

[bormilan]

Personally… I would add it right now… It will be a reserved word soon enough. I wish Erlang/OTP would have a public function that returned the list of reserved words somewhere (erl_syntax may be a great place)… 🤷🏻

I think f_reserved_word does exactly what you want in the module erl_scan

Yeah, but it's not exported, is it?

It is:

Ah, right! I was thinking of something like f_reserved_words/0 (that would return the list) :)

But… maybe we can actually use that here or in rebar3_format… 🤔 If you want to send a PR using erl_scan:f_reserved_word/1 to verify that the atom is actually a reserved word instead of the hardcoded list of strings we now have… I'd merge it, for sure!

Ok, I will do it shortly!