Skip to Content.
Sympa Menu

shibboleth-dev - Re: [Shib-Dev] [Patch] SP Config schema bug for <Path>

Subject: Shibboleth Developers

List archive

Re: [Shib-Dev] [Patch] SP Config schema bug for <Path>


Chronological Thread 
  • From: Daniel F Crisman <>
  • To:
  • Subject: Re: [Shib-Dev] [Patch] SP Config schema bug for <Path>
  • Date: Fri, 23 Oct 2009 12:43:24 -0400

Yes, I am validating with xmllint from libxml2.

processContents="lax" does not mean skip the contents, it means "If the
item has a uniquely determined declaration available, it must be valid
with respect to that definition, that is, validate if you can, don't
worry if you can't." [1]. So as the <Path> declaration is a global the
validator has it available and should validate against it.

If you want to just ignore the layout of everything inside PluggableType
consider changing to processContents="skip" which just checks for
well-formed XML.

[1] http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/#Wildcard_details


For my personal edification, why are you against local elements?

dc

On Fri, Oct 23, 2009 at 12:16:01PM -0400, Scott Cantor wrote:
> Any chance you're using a separate tool to validate the file and it's what's
> reporting the violation?
>
> The Xerces validator doesn't notice this because it's laxly validating the
> wildcard content of CredentialResolver and thus skipping it. That's
> technically correct behavior, and is one of the reasons I don't require
> schemas and validation on the content models of the plugins. That's all
> handled manually.
>
> So on one level your tool (if I'm guessing correctly) may have a bug, and I
> would note that the concern about what's "legal" inside the plugins is
> really beside the point because anything is legal there. It's an open
> content model with a wildcard. You can stick anything in, it just won't load
> the plugin (and it will tell you that).
>
> The schema's not as consistent as I would like in that it does define a
> number of local elements, so I'm inclined to consider revising it, but I
> have to be very careful not to invalidate any existing instances, so it has
> to be done very deliberately.
>
> Renaming an element (without removing the original) is a way of avoiding
> that testing because I can guarantee nothing new will break and the old
> stuff that worked already keeps working.
>
> -- Scott
>
>



Archive powered by MHonArc 2.6.16.

Top of Page