From: Sean Chittenden <sean@...>
Date: 2002-10-23T13:01:59+09:00
Subject: Re: XMLParser, NQXML, REXML, ...

> Hi XML-freaks,
> 
> 	Who has used XMLParser, NQXML, REXML, ...?
> 
> I am right now looking at the faqtotoum
> http://www.rubygarden.org/iowa/faqtotum
> and try to work on pending questions.
> 
> Feedback/Experience would be welcome.
> Why did you individually choose parser X and not
> parser Y?

http://www.rubynet.org/modules/xml/libxml/

I'm writing libxml for Ruby.  It's actually pretty stable at this
point and very fast compared to the other XML parsers out there.
Right now there is full XPath support, it does do validation, can
write XML documents using the DOM interface, and can read/write
gzipped XML documents.  A SAX interface will likely come sometime in
mid-November, however the xslt libraries are getting the majority of
my time at the moment.  XSLT variables are coming shortly, maybe even
tonight sometime.

Markus and I are working on rubydoc which is now able to automatically
document the API and interface for libxml according to the rubydoc
DTD.  As soon as libxml2 has support for schemas, I will move all of
our DTD's to schemas.  In the mean time, the spec for rubydoc can be
found here:

http://cvs.rubynet.org/index.cgi/projects/rubydoc/src/doc/

With libxslt, I'm working on the rubydoc CLI that will merge in hand
written documentation (from an inline source or from an external
rubydoc XML file) with the auto-generated document to form a complete
map of the source, regardless of whether or not it is C module or
written in Ruby.

I like libxml because it's very fast, you can do xpath queries on any
node in the document, and the XML parser is based on libxml2 which is
spec complete in most cases and released under the MIT license.  If
you have any problems with code, please let me know.

-sc

-- 
Sean Chittenden