• To: <michael.h.kay@n...>,"'Larry Fillion'" <larry.fillion@c...>,<xml-dev@l...>
• Subject: Re: need a [ in an ATTLIST
• From: "Rob Lugt" <roblugt@e...>
• Date: Tue, 2 Apr 2002 17:53:28 +0100
• References: <000e01c1da60$00b3bf20$6501a8c0@pcukmka>

Michael Kay wrote:-
> >
> > As for the parser allowing a default value containing the "["
> > - this is a bug.
> >
>
> Are you sure?
>
> I read "Validity Constraint: Attribute Value Legal. The declared default
> value must meet the lexical constraints of the declared attribute type".
>
> Firstly, this is a VC not a WFC.

Quite true.  Therefore probably not a bug unless the parser is being used in
validating mode.

>
> Secondly, it's not obvious that the list of enumerated values is a
"lexical
> constraint", though it would seem reasonable to read it that way.
>

I think it is.  Richard Tobin may correct me (again) on this, but I think
the word "lexical" is employed here to differentiate between VCs that check
for legal tokens and VCs that check for valid document content such as valid
IDREFs.

> But in any case, it seems to me that a non-validating parser should allow
> <!ATTLIST myelement myattribute1 (value1|value2) "value[">
> through, even though it's nonsense.
>

Agreed.  Sometimes I forget there are non-validating parsers around ;-))

~Rob

• Follow-Ups:
  • Re: need a [ in an ATTLIST
    • From: Richard Tobin <richard@c...>

• References:
  • RE: need a [ in an ATTLIST
    • From: "Michael Kay" <michael.h.kay@n...>

• Prev by Date: RE: need a [ in an ATTLIST
• Next by Date: SQL DTD and related questions
• Previous by thread: RE: need a [ in an ATTLIST
• Next by thread: Re: need a [ in an ATTLIST
• Index(es):
  • Date
  • Thread