From: Ammar Ali <ammarabuali@...>
Date: 2010-10-16T01:41:53+09:00
Subject: Re: Processing regular expressions?

On Fri, Oct 15, 2010 at 6:07 PM, Jasen J. <jasenj1@gmail.com> wrote:
> I'm interested in manipulating & munging regular expressions, not the
> matches, but the regexes themselves.
>
> Specifically, I want to parse out the groups/captures in a regex much
> like .named_captures and .names, but for ALL captures, not just the
> named ones.
>
> My end goal is to take a regex like /(one)(two)(three)four/ and split it
> up into three separate regexs
> /(one)twothreefour/
> /one(two)threefour/
> /onetwo(three)four/
>
> Any pointers would be appreciated.
>
> - Jasen.

I haven't found anything like that yet. Would love to hear about any such tools.

I recently did a couple of experiments that might be of interest of
you. It is basically a preprocessor that expands repetitions and adds
"aliases". The code and a brief description are at:
http://github.com/ammar/meta_re

Based on those experiments and some valuable feedback and ideas I
received from others, I started work on a standalone regular
expression parser. It is still in an very early stage, but might be of
interest to you as well. That project repo is at:
http://github.com/ammar/regexp_parser

I would be interested in any comments you might have about either.

Regards,
Ammar