Define charset once and reference from both rules and semantic actions?

[pcafstockf]

I would like to define a charset in one location and use it both in the definition of grammar rules and within semantic actions.

This is a bit of a contrived example, so please don't take it too literally.

I have this rule in my grammar:
LineTerminator = [\n\r\u2028\u2029]

There are many other rules in my grammar which collect text (for example multi line comment blocks).

Is it possible in the semantic actions of those other rules, to test for these line terminators characters within text(), without having to retype my "regex"?

Perhaps for example checking for the difference between:
/* a one line block comment */
and
/*
a multi-line
block comment
*/

I do realize I could look at the line numbers in location() for this particular example. But like I said I'm looking for a way to re-use charset definitions, not solve comment specific problem.