I've been watching the Xerces-* projects with interest for a while, but
have come to realise that it operates at a too low level for what I need.
So I wonder if anyone are building utilities based on the Xerces
infrastructure? In particular, I'm looking for something analogous to James
Clark's SP toolset (specifically nsgmls). Something that works either as a
command line utility or as a Perl library, that will take XML as input and
output both some kind of document structure (preferably as ESIS, but
DOM/SAX/whatever might work too) _and_ (and this is he crux) some easily
parseable and readable list of encountered errors. ATM just Well-Formedness
errors and Validitity errors, but in the future also the results of Schema
Validation.
At the moment I'm calling SP as `nsgmls -wxml -c catalog file.xml`, and
reporting any encountered errors more or less directly to the user as:
$line, $column: $message
(if you'll forgive the pseudo-code)
But SP has some serious limitations in it's XML support and it isn't
evolving to enable such things as Namespaces and Schemas. I'm hoping I can
use Xerces as a replacement, but as yet it looks too low level for me.
Can I do that with Xerces directly? Or has anyone built something on top of
Xerces that I might be able to use? I'd prefer Xreces-P, where I have a
chance to understand the underlying code, but I can live with a
command-line util built on -C or -J if I need to.
FWIW, I'm looking to use it as a backend for
<URL:http://validator.w3.org/>.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
