OTOH, all things being equal, I'd favor adding escape sequences
instead of adding new interpretations for existing characters and then
requiring new flags on the REs that use them.
Fair enough. I wonder then if Ilya's suggestion of extending the
syntax inside the braces, rather than using \m and \M is worth
investigating?
Perhaps something that matches the lookahead/lookbehind assertions:
\m{?= bold}
\m{?! bold?
\m{?< bold}
\m{?<! bold}
Or even add the 'm' after the '?' in a regular assertion:
(?m= bold)
(?m! bold)
(?m< bold)
(?m<! bold)
A
