FormatRBear: Allow to disable parameter-options for formatR

[Makman2]

Sometimes certain style rules shall not be checked at all, we should allow to disable them (like r_use_arrows).

CC @AsnelChristian

[gitmate-bot]

Thanks for reporting this issue!

Your aid is required, fellow coalaian. Help us triage and solving this issue!

CC @sils1297, @AbdealiJK

[AsnelChristian]

can i work on this one ?

[sils]

sure!

[Makman2]

Ah in the PR message this issue was referenced... but it's not solved yet

[gs0510]

@Makman2 @sils1297 I see the issue is still open, but using r_use_arrows = true/false did exactly what it was supposed to do. Are there other style rules that should be allowed to be disabled?

[Makman2]

I think r_use_arrows = False (for example) forced to use no arrows at all. So it's either the one or other extreme when using such a setting. Though I could be wrong with this (long time ago when this issue was opened :3)

[gs0510]

Yes what you are saying is true. So is the requirement that in some places we should use the arrow and in other places we should still be using equal to sign? For example if we have r_use_arrows=true start_line_num=10 end_line_num=20, we can have arrows in line numbers from 10 to 20. Is that what is required? Can you please explain in a little detail what the requirements are? Thanks!

[Makman2]

If the setting is not touched inside a coafile, no check shall happen at all.

For example:

# inside coafile:
bears = FormatRBear
# some settings regarding FormatRBear except 'r_use_arrows',
# meaning we don't want to care whether '=' or '<-' is used.

This can be achieved via setting the default value for r_use_arrows to None (here -> https://github.com/coala/coala-bears/blob/master/bears/r/FormatRBear.py#L36) and omitting the according command-line-parameter if r_use_arrows = None (here -> https://github.com/coala/coala-bears/blob/master/bears/r/FormatRBear.py#L83).

(Though this setting is only allowed to be either True or False, coala doesn't check the default value, that's why None does not make any problem inside the declaration as a default value.)

[gs0510]

If I do not modify the r_use_arrows, and have both = and <- in my r file, nothing changes. The <- and = are both still intact. Here's an asciinema recording that I made. Hope this closes this issue.

[Makman2]

Oh than I was wrong :3
But does this work for all other parameters too? If so we can happily close this one indeed :)

(Btw it's interesting what patch is applied in your asciinema as I can't see any change in the file :3)

[gs0510]

Are you talking about Applied 'ApplyPatchAction' on 'test.r' from 'FormatRBear' ? I just did a diff before and after the change. The bear adds a newline at the end of the file.

[Makman2]

ah alright, nothing serious then :D

[yogupta]

What else was needed in this ? I mean why is still open? Was anything missing?

[Makman2]

seems like it just wasn't continued^^

[yogupta]

If I have understood the issue then we have to remove functionalities such as r_keep_comments , r_use_arrows. Do we have to remove r_braces_on_next_line ?
Is there anything else than i would like to know . I think i can solve this.

[Makman2]

No don't remove them, allow the bear to "ignore" the settings.

I think r_use_arrows = False (for example) forced to use no arrows at all. So it's either the one or other extreme when using such a setting.

And we want by default having sth like r_use_arrows = None and we just don't pass the argument to the R formatter. This applies for other settings as well, not only for r_use_arrows.