Add "inline-comment-spacing" lint rule

[ghost]

This would check that we write f(/*argName*/ true) and not f(/* argName */ true).
I'd also like to lint for spacing on statement comments like:

/* prefer this? */
foo();

/*prefer this?*/
foo();

But we don't have a consistent style here -- thoughts?

[billti]

Personally, I find it far more readable with the spaces providing some separation. Is there a general community default/agreement on this (i.e. linter defaults)? Or just personal preference?

[ghost]

We get 5000 lines of errors if I invert the test condition and require spaces, so we've been preferring /*comment*/ within argument lists for a while. It would be possible to write a fixer though if we want a mass change. Note that this rule doesn't effect comments on statements, and is just for the f(/*argName*/ true) case.

[aozgaa]

nit: extra trailing lines?

[aozgaa]

Maybe we should have a separate rule for "inline-callExpressionArgument-comment-spacing"? Or maybe name it analogously to "boolean-trivia"? eg: "boolean-trivia-spacing"?

[ghost]

How about renaming boolean-trivia to literal-argument-comment and inline-comment-spacing to argument-comment-spacing?

[aozgaa]

Are scripts/tslint rules accessible outside our repo? Are rules in this directory a sort of test-bed for adding rules to tslint itself?

If these are only used by us, then that works for me. We probably shouldn't rename these if someone else is using the boolean-trivia rule.

[ghost]

These are our own internal rules. Some of them might be suitable to submit to tslint itself, if you want to, although there may be overlap, e.g., no-type-assertion-whitespace may already be handled by tslint's whitespace rule.

[mhegazy]

Do we need this new rule? seems like a lot of work, and limited value in my opinion.

[mhegazy]

Do not think we need this one. closing for now, please reopen if you feel otherwise.