Re: [lmi] Which is the best C++ wrapper for libxml2?

[Vadim Zeitlin]

On Mon, 27 Nov 2006 16:47:53 +0000 Greg Chicares <address@hidden> wrote:

GC> On 2006-11-8 14:42 UTC, Vadim Zeitlin wrote:
GC> > > On Wed, 08 Nov 2006 04:27:01 +0000 Greg Chicares <address@hidden> wrote:
GC> [snip example comparing two libraries]
GC> > >  The code seems pretty similar to me in both cases to be honest...
GC> 
GC> The difference is the responsibility they place on a programmer
GC> who uses one library or the other. This:
GC>   parser.get_document()->get_root_node() // libxml++
GC> dereferences a pointer without checking it, and therefore looks
GC> unsafe to me, whereas this:
GC>   parser.get_document().get_root_node()  // xmlwrapp
GC> would not raise that code-review objection.

 Yes, I certainly agree with this and with the corollary that xmlwrapp API
is better than libxml++.

GC> > >  In any case, I still think libxml++ has more chances of being developed
GC> > > [again] than xmlwrapp.
GC> 
GC> I'd agree. But I can't imagine they'd accept a wholesale rewrite
GC> of their API that would break every application that uses it.

 Not the current maintainers, no.

GC> I suppose it's just a matter of time before someone develops a
GC> better C++ xml library. It's been discussed by boost, e.g.:
GC>   http://lists.boost.org/Archives/boost/2005/10/96129.php

 Hmm, this was a year ago -- has anything happened since then? It really
seems incredible that there is no good, maintained, free XML C++ library.


GC> I can't justify the additional work it would take to put all of
GC> this in place and maintain it. We'll use xmlwrapp-0.5.0 because
GC> it provides the features and safety we need with far less work;
GC> we'll switch to a different library when something that's truly
GC> better comes along.

 Again, I certainly can't disagree with this decision. I tried to find some
arguments in favour of libxml++ just for the sake of balance but I do
prefer xmlwrapp API myself.

 Regards,
VZ