Re: [xsl] Can someone help me understand why this isn't working?

Subject: Re: [xsl] Can someone help me understand why this isn't working?
From: "Luke Shannon" <lshannon@xxxxxxxxxxxxxxx>
Date: Thu, 20 Jan 2005 10:09:24 -0500
Responses below.

----- Original Message ----- 
From: "Wendell Piez" <wapiez@xxxxxxxxxxxxxxxx>
To: <xsl-list@xxxxxxxxxxxxxxxxxxxxxx>
Sent: Wednesday, January 19, 2005 7:33 PM
Subject: Re: [xsl] Can someone help me understand why this isn't working?

> Hey Luke,
> At 06:08 PM 1/19/2005, you wrote:
> >I have a basic example I am trying to get working to convert ;&lt;P&gt;
> ><P>.
> Is this by design? Do you have the alternative of getting a different form
> of input?

Not at present. This is what I have to work with at this time.

> >Here is the XML:
> >
> ><?xml version="1.0" encoding="iso-8859-1"?>
> ><?xml-stylesheet type="text/xsl" href="hello.xsl"?>
> ><greeting>&lt;P&gt;Hello, world!&lt;/P&gt;</greeting>
> >
> >Here is the XSL:
> >
> ><?xml version="1.0" encoding="iso-8859-1"?>
> ><xsl:stylesheet version="1.0"
> >xmlns:xsl="";>
> ><xsl:output method="html"/>
> ><xsl:template match="/">
> ><html>
> ><head>
> ><title>Today's greeting</title>
> ></head>
> ><body>
> ><xsl:apply-templates select="greeting"/>
> ></body>
> ></html>
> ></xsl:template>
> ><xsl:template match="P">
> ><em><xsl:apply-templates/></em>
> ></xsl:template>
> ></xsl:stylesheet>
> >
> >I am guessing that the value of greeting selected in the apply-templates
> >call contains &lt;P&gt;Hello, world!&lt;/P&gt;
> The string value of the greeting element in the source (as of the text
> child of that element) is, yes, the sequence of characters you are
> (Keeping in mind that in that sequence &lt; is < etc. - as nodes in the
> tree, the character references have been resolved. That is, the string
> value is actually "<P>Hello world!</P>" because that's what you had in
> input once it was parsed into the tree.)
> This is a far cry from a p element node. The only element is the greeting
> element, whose text value this is. No template you can write will pick up
> an element that doesn't exist. :->

I can see that now :-)
> >I was hoping because the output type was HTML a result tree containing
> ><p>Hello, world!</p> would be created that my match="P" would process (my
> >apologies if my terminology is not correct, still getting up to speed
> >this).
> The terminology is fine, but alas XSLT is not designed to work this way.
> The processor does not re-parse strings on the fly into more nodes in the
> tree. (Presumably you hide the XML with entities because you don't want it
> processed.) There is nothing here to be matched: the apply-templates will
> only find the text node. To handle it requires some kind of second parse.
> There seems to be an emerging industry in handling such "pseudo-XML".
> Generally this practice is frowned upon in polite circles, because it
> violates the spirit of XSLT node-think, tempts one to the heresy of
> tag-writing (much as I might indulge in that vice in private, it's not
> something I'd announce to the world), and is just generally nasty work.
> (Expect things to break frequently whenever things fail to parse, which
> be often in such uncontrolled environments.) Stay clean, we urge: process
> nodes and don't try to reparse strings.
> If you have to do it, there are generally two approaches:
> 1. Use an XSLT processor that supports disable-output-escaping, and use it
> to write this content directly to files, where you can try to parse it
> succeed if you are lucky). If you are smart you might even do useful
> transformations at that stage. But you will probably have to clean up by

This is the approach I have gone with using regular expressions in Java. I
don't like the solution but I feel more confident in it now having talked to
some XSL experts. :-)
> 2. Use an extension function such as Saxon's, and try to parse those
> strings in place.
> >Am I thinking about this the wrong way? Any advice would help.
> Optimally, you'd probably avoid pseudo-markup and find a way to enjoy the
> full benefits of XML.
> If that is not an option, work the data over in two passes (or more) as
> described above.
> Good luck,
> Wendell

Thank you.

> ___&&__&_&___&_&__&&&__&_&__&__&&____&&_&___&__&_&&_____&__&__&&_____&_&&_
>      "Thus I make my own use of the telegraph, without consulting
>       the directors, like the sparrows, which I perceive use it
>       extensively for a perch." -- Thoreau

Current Thread