RE: [boost] Proposal: Regression, XML, and Unicode Game Plan

From: Jeremy Siek (jsiek_at_[hidden])
Date: 2001-09-26 22:24:58

I'd have to agree with Jeff on this. In addition, I expect it would not be
difficult to implement a much higher performance parser using modern C++
style instead of Java-style C++.

On Wed, 26 Sep 2001, Jeff Garland wrote:
jeff>
jeff> I have a lot of respect for the Apache tools, but I would
jeff> still like to see boosters look at building XML parsing tools.
jeff> XML is being used extensively on many projects.  The main
jeff> problem with xerces-c implementation is that it is a JAVA
jeff> implementation remapped to C++.  To maximize portability they
jeff> avoid 'difficult' C++ features like templates.  Needless to
jeff> say things get ugly fast with all sorts of nasty casting and
jeff> character conversions.  In fact, I would be all for XML
jeff> parsing tools that don't follow SAX or DOM interfaces at all,
jeff> but just provide a library that uses the C++ standard library
jeff> or extensions thereof.
jeff>
jeff> Jeff

----------------------------------------------------------------------
 Jeremy Siek                          http://php.indiana.edu/~jsiek/
 Ph.D. Student, Indiana Univ. B'ton   email: jsiek_at_[hidden]
 C++ Booster (http://www.boost.org)   office phone: [phone]
----------------------------------------------------------------------