Skip to Content.
Sympa Menu

shibboleth-dev - RE: IQ-sites.xml regular expression

Subject: Shibboleth Developers

List archive

RE: IQ-sites.xml regular expression


Chronological Thread 
  • From: "Scott Cantor" <>
  • To: "'Ian Young'" <>
  • Cc: <>
  • Subject: RE: IQ-sites.xml regular expression
  • Date: Tue, 9 Nov 2004 18:05:26 -0500
  • Organization: The Ohio State University

> It sort of made sense, as after all the domain is modelled on a DNS
> name, and DNS names aren't case sensitive. Regular expression systems
> often have a flag you can use to get that behaviour.

David Champion and I talked about it a bit. This particular one (the one
supplied in the XML parser) doesn't, one reason being it's Unicode-aware by
necesssity, and Unicode makes case folding hard.

There appear to be some Unicode-aware options, like ICU, but it introduces
"yet another dependency(tm)" that historically made using Xerces attractive.

The thing you're looking at may resemble DNS but it's not. It's attribute
"scope", or "domain of assertion", if you will. In practice, it's not useful
or valuable to be sending mixed case in that.

> This on the other hand is a better explanation all round. I guess I'd
> better bear this in mind the next time I'm looking for an example of
> something :-)

Always feel free to tell somebody, just take it with a grain of salt.

-- Scott




Archive powered by MHonArc 2.6.16.

Top of Page