did it again .. somebody smack me upside the head!
---------- Forwarded message ---------- From: Drew Crampsie drew.crampsie@gmail.com Date: 2009/9/1 Subject: Re: [cltl3-devel] RFC: CLtL3 Charter To: Gustavo gugamilare@gmail.com
2009/9/1 Gustavo gugamilare@gmail.com:
2009/9/1 drew.crampsie@gmail.com
I'd personally much prefer a 'lispy' (read : verbose and understandable) implementation of regexps then the one from perl, and still wouldn't want it included as part of CLtL3..
cl-ppcre allow the use of sexps as regexps. I think that they are "verbose" and "understandable".
So it might ... but as per section 4 "Preference will be given to topics that cannot be implemented portably and have multiple existing implementations.".
There will be a library folks, and that is what you as a coder will likely use. CLtL3 is not for coders, but for implementors and library authors (who, yes, are also coders but you get the idea). It's for things we can't currently do in portable common lisp, not for things that have been done and already see wide use. There is nothing holding back the acceptance of cl-ppcre as the 'standard' portable implementation of perl compatible regular expressions, so why waste time documenting and discussing it?
It will take an incredible amount of effort to describe in any detail the interface of cl-ppcre, and for little to no gain. When a new perl comes out, do we go ahead and change the standard? do we freeze cl-ppcre at the time we publish CLtL3?
I understand that what a lot of people want is a 'standard library', and not an updated description. I'd like that too. But we have to get there somehow, and as it stands we cannot build our 'standard library' on Standard Common Lisp. That is what CLtL3 is trying to fix.
Cheers,
drewc