[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: CDATA strangeness

• To: xml-dev@l...
• Subject: Re: CDATA strangeness
• From: Richard Tobin <richard@c...>
• Date: Mon, 5 Aug 2002 23:04:21 +0100 (BST)
• Cc:
• In-reply-to: <NDBBLNCOMCPLKABNJBLKOEBAGDAA.ramin@w...>
• Organization: HCRC, University of Edinburgh

In article <NDBBLNCOMCPLKABNJBLKOEBAGDAA.ramin@w...> you write:

><!--[if IE]><script language=javascript>ie5=1;</script><![endif]-->

That bit's OK, it's just a comment.

><![if !IE]><script language=javascript>ie5=0;</script><![endif]>

But that isn't.

>I thought a CDATA section *had* to start with '<![CDATA[' 

That's right.

>and plain '<![' should be ignored.

But that isn't.

<![ followed by anything except CDATA is illegal in an XML instance
(in the DTD it could be a conditional section).  It's just a syntax
error.  Less-than must be escaped when it isn't markup - XML doesn't
just ignore it.

>If you pass this through the tidy/xerces-C combo, the parser complains that
>it expects to see a CDATA section start around the '<![if' section. 

It may not be the best error message in the world - it's presumably
just assuming that you must have meant CDATA because that's the only
thing that's legal after <![.

>Who's right and who's wrong here? The web-site or tidy/xerces?

Well if it's *meant* to be XML, then the web-site is just wrong.
But is it meant to be XML?  I'm not sure what its status is as HTML
(ie SGML).  

-- Richard

• References:
  • CDATA strangeness
    • From: "Ramin Firoozye" <ramin@w...>

• Prev by Date: Re: CDATA strangeness
• Next by Date: Re: XMLSchema.dtd not fully defined?
• Previous by thread: Re: CDATA strangeness
• Next by thread: Re: CDATA strangeness
• Index(es):
  • Date
  • Thread