Re: 5. xml:id error processing being a "must"

On Wed, 5 Jan 2005, Norman Walsh wrote:
> |
> | Should the spec say that xml:id errors MAY simply be ignored? In
> | effect, what's the point of the error checking? The following
> | document:
> |
> |    <test xml:id="!%&(@!"/>
> |
> | ...is well-formed, and MUST result in a document representation where
> | the element has the ID "!%&(@!", despite it being technically in
> | xml:id error.
> 
> I'm reluctant to allow general purpose xml:id processors, like my own
> SAX filter, to always ignore all errors and claim to be conformant.

Fair enough. I retract my comment.

I agree with this resolution.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Received on Thursday, 20 January 2005 13:00:49 UTC