Add code range to RegexpParseNode

[nirvdrum]

StrParseNode includes access to both the underlying ByteList and the logical string's code range. It would be helpful if RegexpParseNode had this information as well. Since the lexer is likely already performing a byte scan to find the bounds of the underlying ByteList, it would be ideal to capture the code range at this point and avoid an additional scan later on.

[headius]

For Truffle, I assume you'll be changing this in your copy of our parser. But either way I suppose we both could see some benefit from this.

[enebo]

This is also part of a larger initiative to make CR non-lazy so we never have an UNKNOWN CR too. This is just the regexp part of it.

[nirvdrum]

The parser change is still new to me. But yeah, this seems beneficial to both.