Integrating xhtml/html validation into a web-based CMS(MS-windows)

[Guest]

Hi folks:

I'm assuming someone's been down this path before: we want to provide our
users with a tool integrated into our content-management website that will
allow them to validate their html & xhtml documents. Our servers will be MS
Win-2000.

Plug-in commercial modules considered (something in .NET? COM+? Service
exe?).

Alternatively, if we could get hold of the source for a parsing-validation
engine that can be compiled on Win-2000, we could bolt it in ourselves. My
attempts to wander around the net seem to indicate much of the source for
this sort of thing is unix-oriented. - I'm pretty unix-oriented myself, but
the workplace & client-base isn't

Yes, I know, our users could always use any number of web-hosted validators,
or stand-alone validating tools which would do the job perfectly well, but
that's much too difficult and complicated, and they'd rather pay us to fit
the validation evaluation process into a nice simple all-in-one CMS with one
URL which they won't forget.

My email is below (pardon the obfuscation), or reply to group.

Thanks

Larry

 

[Patrick TJ McPhee]

% Alternatively, if we could get hold of the source for a parsing-validation
% engine that can be compiled on Win-2000, we could bolt it in ourselves.

libxml parses and validates both XML and HTML.

 

[Nick Kew]

libxml parses and validates both XML and HTML.

The HTMLparser of libxml2 makes no attempt to validate. In certain
respects it's tag-soup: for example it applies XML comment syntax,
(which is incorrect in HTML) and doesn't grok many SGML abbreviated
forms that are technically permitted in HTML.

It's good at parsing real-world HTML (including tag-soup), but that's
totally different to validating.

 

[Nick Kew]

Plug-in commercial modules considered (something in .NET? COM+? Service
exe?).

I could supply either a COM component or a Webservice running on
Windows/Apache. In either case that'll mean porting existing Unix
software, but it won't be the first time: I've ported much of Site
Valet in the past.

Alternatively, if we could get hold of the source for a parsing-validation
engine that can be compiled on Win-2000, we could bolt it in ourselves. My

OpenSP - but it's not the easiest library to work with.

that's much too difficult and complicated, and they'd rather pay us to fit
the validation evaluation process into a nice simple all-in-one CMS with one
URL which they won't forget.

A variant I can supply is a validating spider that bugs page maintainers
(by email) when it encounters invalid pages. That was ported to Windows
over two years ago, as part of a bigger QA toolkit supplied with a CMS.