- From: Kay, Michael <Michael.Kay@softwareag.com>
- Date: Tue, 26 Aug 2003 14:45:50 +0200
- To: Per Bothner <per@bothner.com>, public-qt-comments@w3.org
- Message-ID: <DFF2AC9E3583D511A21F0008C7E62106073DD0AE@daemsg02.software-ag.de>
XQuery does not provide the namespace axis, and does not provide any other construct that allows you to get your hands on a namespace node. Therefore, the sequence that provides the content for a new element can never contain namespace nodes, except in the very limited and controlled way allowed by computed namespace constructors. Michael Kay > -----Original Message----- > From: Per Bothner [mailto:per@bothner.com] > Sent: 26 August 2003 03:25 > To: public-qt-comments@w3.org > Subject: namespace nodes in element content > > > > Section 3.7.1.3 "Content" of the new XQuery draft does not say what > happens to namespace nodes in the content of a direct element > constructor. There are two choices: > > (1) They are allowed, and treated as in the context expression of a > Computed Element Constructor. I.e. they must precede > attributes which > must precede other content. > > (2) They are not allowed, in which case the specification should say > that, as it does for document nodes. > > I think (1) is more consistent. Otherwise, why can the content of a > direct element constructor allow attributes but not namespaces, while > the content of a computed constructor allows both? > > An alternative is to dis-allow *both* attributes and > namespace nodes in > direct element constructors, since you can create attributes > using both > literal attributes as well as attribute nodes in enclosed > content, and > that's arguably a bit of a weird hack. Since we have > computed element > constructors, we don't need this special handling. > -- > --Per Bothner > per@bothner.com http://per.bothner.com/ > > >
Received on Tuesday, 26 August 2003 08:48:34 UTC