xsl-list
[Top] [All Lists]

Re: XSL pattern needed for begin/end elements

2004-07-07 13:09:24
Tracy.

I haven't worked through this too carefully - but here is a pseudo-code method that might work in the sibling case. It is based on the idea of selecting all following nodes for processing based on their next <hyperlink_end> element having matching attributes to the current hyperlink-begin.
(which looks like what you were intending)

<xsl:template match='hyperlink-begin'>
<HyperLink xlink:href="{concat(locator_url/@protocol,'/',locator_url/@host_name)}"> <xsl:apply-templates select='following-sibling::*[following-sibling::hyperlink_end[1][(_at_)id=current()/@end]' mode='INLINK'/>
</HyperLink>
</xsl:template>

<xsl:template match='text_run' mode='INLINK'>
<xsl:choose>
<xsl:when test='@emphaisis="bold"'>
<b><xsl:value-of select='.'/></b>
</xsl:when>
<xsl:otherwise>
<xsl:value-of select='.'/>
</xsl:otherwise>
</xsl:choose>
</xsl:template>

There are other ways of doing the 'INLINK' mode processing - depending on hwo complex it gets.
E.g. - you could have separate templates matching 'text_run[(_at_)emphasis]' 
etc.

You may need an additional template
e.g.
<xsl:template match='*' mode='INLINE'>
   <xsl:apply-templates/> <!-- or whatever else you want to do -->
</xsl:template>
if you need to process non-sibling intervening elements.

I think something close to the above should work.

HTH.


Tracy Atteberry wrote:

Mike,

The current project is a demo for something that will eventually be
written in C/C++.  Then as you say, we can then walk the DOM tree and
maintain a separate context stack to help solve the problem.

For now, we can definitely assume that these elements are siblings.  In
fact, for most real source documents this will be the case.  Given that
assumption, I would love to know the not-too-difficult solution, as this
is my immediate problem.

As for the more general case, a hyperlink may in some cases overlap text
runs.  For example:

<doc>
 <p>
<text_run emphasis="bold">Click <hyperlink_begin id="111" end="222">
       <locator_url protocol="http" host_name="www.sf.net"/>
     </hyperlink_begin>
     here
   </text_run>
   <text_run> to download.</text_run>
   <hyperlink_end id="222" begin="111"/>
 </p>
</doc>

In fact, hyperlinks can overlap paragraphs and other document elements
though this is rarely seen in practice.

-Tracy


-----Original Message-----
From: Mike Trotman [mailto:mike(_dot_)trotman(_at_)datalucid(_dot_)com] Sent: Wednesday, July 07, 2004 1:26 PM
To: xsl-list(_at_)lists(_dot_)mulberrytech(_dot_)com
Subject: Re: [xsl] XSL pattern needed for begin/end elements


If the begin and end elements are siblings at the same level then the problem is tractable and probably not too difficult to solve.

However if they can occur at different levels then this means that one of them is enclosed inside an element that excludes the other (I think).

Can you give any example of a case where the begin and end elements are not siblings at the same level?
I ask because:
a) I can't picture how this would make sense given the information that you require them to contain
b) If one of them does occur inside an element that excludes the other
- what would you want to to with the excluded part of this elements content / tree?
   - If you start closing all the parent elements etc (and opening them

again to match the orphaned end tags)
then you are destroying the structure and meaning of the XML data which XSLT is designed to help preserve.

I.e. if they are not siblings at the same level then the XML data 'structure' is totally inappropriate for XSLT
and the 1st thing you should do is process it using something else.

I have documents like this - and I process them by walking the DOM tree and maintaining a separate STACK of whatever I consider my current context to be. (I am doing this to detect overlap between different document layers marked in exactly the way you describe.)

Tracy Atteberry wrote:

Hi all,

I'm looking for an XSL pattern to solve the problem of going from XML that has separate begin and end elements to one that does not.

Please, please note that I do not control either the source or target XML formats. If I did, this would be much easier.

Source XML snip:

<doc>
<hyperlink_begin id=3D"111" end=3D"222">
  <locator_url protocol=3D"http" host_name=3D"www.sf.net"/>
</hyperlink_begin>
<text_run>Click</text_run>
<text_run emphasis=3D"bold">here.</text_run>
<hyperlink_end id=3D"222" begin=3D"111"/>
</doc>

Target XML example:

<cod>
<HyperLink xlink:href=3D"http://www.sf.net";>
  Click <b>here.</b>
</HyperLink>
</cod>

In my case I can assume that associated begin and end hyperlink tags will occur as siblings -- though generally this is not the case and in fact, this is the reason the begin and end tags are unique elements.

I have a template that /almost/ works so feel free to let me know why it fails OR suggest a completely different solution.

Current XSL template snip:

<xsl:template match=3D"//hyperlink_begin">
  <xsl:variable name=3D"linkUrl">
      <xsl:value-of select=3D"locator_url/@protocol"/>
      <xsl:text>://</xsl:text>
      <xsl:value-of select=3D"locator_url/@host_name"/>
  </xsl:variable>
  <xsl:variable name=3D"endID" select=3D"@end"/>
  <xsl:element name=3D"HyperLink">
<xsl:attribute name=3D"xlink:href"><xsl:value-of select=3D"$linkUrl"/></xsl:attribute> <xsl:apply-templates select=3D"(following-sibling::*) except (following-sibling::hyperlink_end[(_at_)id=3D$endID]/following-sibling::*)"/

  </xsl:element>
</xsl:template>

This produces the correct hyperlink but the template for text_run elements gets called twice this way -- once inside the hyperlink, then again as templates continue to be applied.

Any help would be greatly appreciated.  Thanks!

Tracy Atteberry

PS. I'm using Saxon 8

--+------------------------------------------------------------------
XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list
To unsubscribe, go to: http://lists.mulberrytech.com/xsl-list/
or e-mail: <mailto:xsl-list-unsubscribe(_at_)lists(_dot_)mulberrytech(_dot_)com>
--+--




--
Datalucid Limited
8 Eileen Road
South Norwood
London SE25 5EJ
United Kingdom

/
tel :0208-239-6810
mob: 0794-725-9760
email: mike(_dot_)trotman(_at_)datalucid(_dot_)com

UK Co. Reg:   4383635
VAT Reg.:   798 7531 60

/