----- Original Message -----
From: "Mohun Biswas" <m_biswas_at_mailinator.com>
To: <curl-library_at_cool.haxx.se>
Sent: Monday, April 19, 2004 7:40 PM
Subject: Re: URL parsing

> codemastr wrote:
> > The problem with this approach is, it is not portable. MSVC++, for
example,
> > does not have regex. So it would fail. So you'd need to include a GNU
regex
> > lib with your program, which I doubt most people want to do. It also
relies
> > on header files that are not available everywhere.
> >
> > It's a nice idea, but it's not very portable.
>
> PCRE (http://www.pcre.org/) is portable to Windows/Mac/Unix and has an
> unrestricted license.

Yeah, PCRE is portable. However, you've just removed the advantage of this
method. The reason to do this, is to avoid (large) parsing code such as used
by liburi. If I need to include PCRE, which is something like ~500KB, well
then I might as well just include liburi instead. And as I said in my email,
you could include GNU regex. It is also portable, it it is not 500KB. But
personally, I don't even want to do that.

Dominick Meglio
Received on 2004-04-21