W3C home > Mailing lists > Public > xsl-editors@w3.org > October to December 1999

Re: Output to external general parsed entity

From: James Clark <jjc@jclark.com>
Date: Mon, 06 Dec 1999 09:00:24 +0700
Message-ID: <384B18B8.7EA034FF@jclark.com>
To: mhkay <mhkay@iclway.co.uk>
CC: xsl-editors@w3.org
mhkay wrote:
> 
> The first sentence of 16.1 states that the output is a well-formed XML
> external general parsed entity (EGPE).
> 
> But surely if the xml/text declaration includes standalone="yes" or "no",
> then it isn't an EGPE?
> 
> Equally, if it includes a document type declaration, then it isn't an EGPE?

I think the spec covers the DTD case, when it says "If the XSLT
processor generated a document type declaration because of the
doctype-system attribute, then the above requirements apply to the
entity with the generated document type declaration removed".

It should probably say something similar for a SDD.

> That suggests that the first sentence of 16.1 should read "... is either a
> well-formed XML external general parsed entity or a well-formed XML document
> or both".
> 
> It also implies that the requirements for conformance with the namespaces
> recommendation should be phrased differently, since as written they don't
> work when the output isn't an EGPE.
> 
> This would also imply in turn that it is an error to request the output of a
> standalone pseudo-attribute or a document type declaration if the children
> of the root node include text nodes, or include no element node, or include
> more than one element node: presumably an error that the XSLT processor
> should report or recover from.

I agree this should be an error (I can't think of any useful recovery
behaviour).

James
Received on Sunday, 5 December 1999 22:41:49 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:59:49 GMT