Hello good people,
Thank you Andreas, J.Pietschmann, and Riz for helping me out. Your
suggestions got me started on the right track. For your edification, here
is what I came up with that works:
xsl:for-each select=MUInventory/Category !-- Outer Loop --
xsl:variable name=lastCatxsl:value-of select=
normalize-space(preceding-sibling::node()[2]/id)//xsl:variable
xsl:variable name=lastTypexsl:value-of select=
normalize-space(preceding-sibling::node()[2]/Type)//xsl:variable
xsl:if test=concat(id,Type) != concat($lastCat,$lastType)
fo:table-row span=all
fo:table-cell width=3in
fo:block
xsl:value-of select=concat(./id,' - ', ./Type)/ !--
last = xsl:value-of select=$lastCat/ - xsl:value-of
select=$lastType/
Current NODE: xsl:value-of select=./node/ and
Preceding NODE: xsl:value-of select=preceding-sibling::node()[2]/node/
--
/fo:block
/fo:table-cell
/fo:table-row
/xsl:if
...next
First, I got rid of the text nodes just to minimize ambiguous behaviors.
Then I went about testing your suggestions and found that when using
xsl:if test=self::node()[not(id=preceding-sibling::node()[2]/id)] and
self::node()[not(Type=preceding-sibling::node()[2]/Type)] instead of
xsl:if test=concat(id,Type) != concat($lastCat,$lastType), it would
evaluate true for the first node and the last node and give me a header for
Motorcycles - Cruisers and Motorcycles - Standard, leaving out the
others. I had to use [2] instead of [1] in the path to get the correct
node. I have no solid reason as to why, but maybe it's the manner in which
the nodes are being numbered internally on the object stack. As you can
see, I had troubleshooting code in the output. If I used the expressions
self::node()[not(id=preceding-sibling::node()[2]/id)] and
self::node()[not(Type=preceding-sibling::node()[2]/Type)] on separate
xsl:if... statements, it still would exibit the same same wrong behavior
as witnessed when chaining them together with boolean logic. So much for
elegance, I concatenated and made a direct comparism and viola`, it now
gives me the table below:
My XML file will always be in category order, so there is no sorting or
other problem areas to anticipate.
Motorcycles - Cruisers
2001 Harley Davidson XL1200 Sportster
2002 Harley Davidson 883 Sportster
2003 Honda VT750 Shadow Spirit
2000 Yamaha Royal Star Venture
2002 Yamaha V Star 650 Custom
2004 Yamaha V Star 650 Silverado Classic
2004 Yamaha V Star 650 Classic
2005 Yamaha Royal Star Tour Deluxe
Motorcycles - Off-Road
2002 Suzuki DR200
Motorcycles - Scooters
2003 Aprilia Rally 50
2003 Aprilia SR 50
Motorcycles - Sport
2004 Suzuki Katana 600
ATV - Sport
2001 Yamaha Banshee
Motorcycles - Standard
2004 Suzuki SV650
Thanks again!
Curtis Fisher
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Hi,
I am attempting to compare the values in one node with the preceding
node.
When using the preceding-sibling on the first node, it retuns a
null (which is good), but thereafter returns the value of the first
node (only) as the tree is processed.
Not completely sure, but the problem might be that:
preceding-sibling::node()
returns the first node on the preceding-sibling axis, but 'first' here
being
'in document order'... (as in: take the set of all preceding-siblings and
of
that set, return the one that appears first in the source document)
If I use the axisName following:: or following-sibling:: then it works
fine.
Yes, because here the immediately following (sibling) node is always also
first one following in the document as well.
Try:
self::node()[not(text()=preceding-sibling::node()[1]/text())]
IIC, the addition of '[1]' should make sure you return the immediately
preceding-sibling.
HTH!
Greetz,
Andreas
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
