• To: <xml-dev@l...>
• Subject: Re: When to check entity WFness according to 4.3.2
• From: "Karl Waclawek" <karl@w...>
• Date: Tue, 26 Oct 2004 10:40:07 -0400
• References: <20041026142335.7309B190A44@m...>

----- Original Message ----- 
From: "Richard Tobin" <richard@i...>
To: <xml-dev@l...>
Sent: Tuesday, October 26, 2004 10:23 AM
Subject: Re:  When to check entity WFness according to 4.3.2


> >The only case when (c) does not catch the WF violation is when
> >the entity is never referenced. That is, Expat will not catch this:
> 
> ><!DOCTYPE doc [
> ><!ELEMENT doc (#PCDATA)>
> ><!ENTITY e "<foo>abc</not_foo>">
> >]>
> ><doc></doc>
> 
> Just to reiterate what I said earlier: expat is right not to report
> this as an error, because the document is well-formed, even though
> the entity "e" is not.

But the prolog is part of the document production, so if an entity declared
in the prolog is not well-formed, then should that not mean the document
is not well-formed? 

Btw, MSXML reports a WF violation for the document above.
Even if you are right, this at least indicates that the spec is
not very clear on that issue.

Karl

• Follow-Ups:
  • Re: When to check entity WFness according to 4.3.2
    • From: richard@i... (Richard Tobin)

• References:
  • Re: When to check entity WFness according to 4.3.2
    • From: richard@i... (Richard Tobin)

• Prev by Date: Re: RELAX NG pluggable type libraries
• Next by Date: RE: After XQuery, are we done?
• Previous by thread: Re: When to check entity WFness according to 4.3.2
• Next by thread: Re: When to check entity WFness according to 4.3.2
• Index(es):
  • Date
  • Thread