I only recently started using it exclusively to parse grammar, when the rules got complex so did my regexes, but when I divided my regexes into groups each on their own variable and I formed whole a regex by concatenating these variables, regex was much more concise for me.