Subject: Re: meaning of "contain" in the XSLT spec From: James Clark <jjc@xxxxxxxxxx> Date: Sun, 03 Oct 1999 09:17:27 +0700 |
Mike Brown wrote: > > Michael Kay wrote: > > "Containing" surely allows an ancestor as well as a parent > > [...] > > In SAXON I allow xsl:attribute to add an attribute to an > > element created using an ancestor literal result element > > James Clark wrote: > > That's the intended interpretation. You've edited out the important bit of my message. The complete message was: > > Oh! Is that a definitive interpretation? (a) "Containing" surely allows an > > ancestor as well as a parent, (b) I thought it was talking about the result > > tree not the stylesheet tree. > > > > In SAXON I allow xsl:attribute to add an attribute to an element created > > using an ancestor literal result element in the source stylesheet, or even > > an element created using a different template (i.e. you can do an > > xsl:call-template to a template that contains the xsl:attribute). The only > > constraint is that the element in the result tree must be open and must not > > have had any character content added to it. > > That's the intended interpretation. "containing" means "parent", but the key point is that it's talking about the *result* tree not the stylesheet tree (which should be clear from the fact that it says "result element node"). Suppose that if the xsl:attribute element had been an xsl:element it would have added an element as a child of a particular element X. What the spec is trying to saying is that xsl:attribute element adds an attribute to element X. James XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
RE: meaning of "contain" in the XSL, Mike Brown | Thread | RE: meaning of "contain" in the XSL, Mike Brown |
Re: first-of-type, Clark C. Evans | Date | decode( source,'if-one','1', ... , , Clark C. Evans |
Month |