Subject: Re: Maximum "child-depth" of current node?
From: "Dimitre Novatchev" <dnovatchev@xxxxxxxxx>
Date: Tue, 16 Oct 2007 10:02:55 -0700
|
Do have a look here:
http://www.dpawson.co.uk/xsl/sect2/N2193.html
The most up-to-date XSLT 2.0 code is:
<xsl:stylesheet version="2.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:xs="http://www.w3.org/2001/XMLSchema"
xmlns:f="http://fxsl.sf.net/"
exclude-result-prefixes="xs f"
>
<xsl:import href="../f/func-map.xsl"/>
<xsl:function name="f:maxDepth" as="xs:integer">
<xsl:param name="pNode" as="node()"/>
<xsl:sequence select=
"if(not($pNode/node())) then 0
else
max(f:map(f:maxDepth(), $pNode/node() ) ) + 1
"
/>
</xsl:function>
<xsl:function name="f:maxDepth" as="element()">
<f:maxDepth/>
</xsl:function>
<xsl:template match="f:maxDepth" mode="f:FXSL" as="xs:integer">
<xsl:param name="arg1" as="node()"/>
<xsl:sequence select="f:maxDepth($arg1)"/>
</xsl:template>
</xsl:stylesheet>
and the logic is concentrated just in these three lines:
if(not($pNode/node())) then 0
else
max(f:map(f:maxDepth(), $pNode/node() ) ) + 1
--
Cheers,
Dimitre Novatchev
---------------------------------------
Truly great madness cannot be achieved without significant intelligence.
---------------------------------------
To invent, you need a good imagination and a pile of junk
-------------------------------------
You've achieved success in your field when you don't know whether what
you're doing is work or play
On 10/16/07, Scott Trenda <Scott.Trenda@xxxxxxxx> wrote:
> Hey XSL-List. Got a bugger of a problem that started to drive me nuts
> yesterday, figured I'd ask here first thing today to avoid pulling my
> hair out over this. :P
>
> I just found out yesterday that the nest-hierarchy system I set up for a
> recent big project has to essentially be done in _reverse_ for a
> different format, but this one isn't as simple. I'm basically making a
> list header hierarchy into a nested HTML table header in one format, and
> I have to make it into similar-looking CSV in the other format. Take
> this example data:
>
> <x>
> <c>col1</c>
> <c>col2</c>
> <g n='grp1'>
> <c>col3</c>
> <c>col4</c>
> <c>col5</c>
> </g>
> <g n='grp2'>
> <c>col6</c>
> <g n='grp3'>
> <c>col7</c>
> <c>col8</c>
> </g>
> <c>col9</c>
> </g>
> <c>col10</c>
> <c>col11</c>
> </x>
>
> HTML output:
> ------------------------------------------------------------------------
> --------
> | | | grp1 | grp2 | |
> |
> | | |--------------------|---------------------------| |
> |
> | | | | | | | grp3 | | |
> |
> | | | | | | |-------------| | |
> |
> | col1 | col2 | col3 | col4 | col5 | col6 | col7 | col8 | col9 | col10 |
> col11 |
> ------------------------------------------------------------------------
> --------
>
> CSV output:
> , , , , ,grp2, , , , ,
> , ,grp1, , , ,grp3, , , ,
> col1,col2,col3,col4,col5,col6,col7,col8,col9,col10,col11
>
>
> For HTML (this part is done already), I have a key that gets the <c> or
> <g> elements at a specified depth - since col1, col2, col10, and col11
> actually exist in the first <tr> of the table, they belong with grp1 and
> grp2 at the top level. But those cells are bottom-valigned because
> there's a data table beneath it, and it makes sense to have the label
> sitting directly above it. In the CSV output, I need to alter the
> structure so they actually appear there in the result document.
>
> With that, here's the key I'm using for the HTML version:
> <key name="cols-at-depth" match="c|g[.//c]" use="count(ancestor::g)
> + 1"/>
>
> Later in the stylesheet, I find the max column depth
> ($total-header-levels), start processing with key('cols-at-depth', 1)
> and loop until I'm at key('cols-at-depth', $total-header-levels). But
> for the CSV version, I essentially need to go the opposite way - rather
> than counting the node's depth from its farthest <g> ancestor, I need to
> count the depth of its deepest child branch. If I could do this with a
> key, it would definitely be best, but just finding the algorithm to get
> it in the first place would be a good start. My strategy from there is
> to do a loop similar to the HTML cols-at-depth algorithm above, but the
> CSV version would hold off on making the parent group entries until
> absolutely necessary (at the point where the output is on the
> nth-to-last output row, and parent group has at least one branch n
> levels deep). Any ideas on XPath trickery I could use here?
>
> I've included a trimmed-down version of the HTML-output stylesheet at
> the bottom.
>
>
> ... and I know we've had this discussion before - I'm stuck with XSLT
> 1.0. Thanks in advance!
>
> ~ Scott
>
>
>
> <xsl:stylesheet version="1.0"
> xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
>
> <xsl:output method="html" encoding="utf-8"/>
>
> <xsl:key name="columns-at-depth" match="c|g[.//c]"
> use="count(ancestor::g) + 1"/>
>
> <xsl:variable name="total-header-levels">
> <xsl:for-each select="//c">
> <xsl:sort select="count(ancestor::g)" data-type="number"/>
> <xsl:if test="position() = last()">
> <xsl:value-of select="count(ancestor::g) + 1"/>
> </xsl:if>
> </xsl:for-each>
> </xsl:variable>
>
>
> <xsl:template match="/">
> <table>
> <xsl:call-template name="loop-rows"/>
> </table>
> </xsl:template>
>
> <xsl:template name="loop-rows">
> <xsl:param name="row" select="1"/>
> <tr>
> <xsl:apply-templates select="key('columns-at-depth',
> $row)"/>
> </tr>
> <xsl:if test="$row < $total-header-levels">
> <xsl:call-templates name="loop-rows">
> <xsl:with-param name="row" select="$row + 1"/>
> </xsl:apply-templates>
> </xsl:if>
> </xsl:template>
>
>
> <xsl:template match="c">
> <th rowspan="{$total-header-levels - count(ancestor::g)}">
> <xsl:value-of select="."/>
> </th>
> </xsl:template>
>
> <xsl:template match="g">
> <th colspan="{count(.//c)}">
> <xsl:value-of select="@n"/>
> </th>
> </xsl:template>
>
>
> </xsl:stylesheet>
|