xslt challenging situation related to iterations

[hilz]

Hi all
I have this situation where I have an xml file similar to this:

<Root>
<MyElement year="2004"><Amount>10</Amount></MyElement>
<MyElement year="2004"><Amount>11</Amount></MyElement>
<MyElement year="2005"><Amount>15</Amount></MyElement>
<MyElement year="2006"><Amount>4</Amount></MyElement>
<MyElement year="2006"><Amount>7</Amount></MyElement>
<MyElement year="2004"><Amount>20</Amount></MyElement>
...
...
</Root>
Now I want to summarize this by transforming it into something like this

<Root>
<MyElement year="2004">
<Amounnt>10</Amount>
<Amounnt>11</Amount>
<Amounnt>20</Amount>
</MyElement>
<MyElement year="2005">
<Amounnt>15</Amount>
</MyElement>
<MyElement year="2006">
<Amounnt>4</Amount>
<Amounnt>7</Amount>
</MyElement>
</Root>
The problem I am having is how to loop through the MyElement elements
and get all amounts for a given unique year, then go to the next year.
In other words, I want to be able to use the year as a unique key for my
iterations.

Any help is greatly appreciated.
Thanks

[Martin Honnen]

hilz wrote:

<Root>
<MyElement year="2004"><Amount>10</Amount></MyElement>
<MyElement year="2004"><Amount>11</Amount></MyElement>
<MyElement year="2005"><Amount>15</Amount></MyElement> <Root>
<MyElement year="2004">
<Amounnt>10</Amount>
<Amounnt>11</Amount>
<Amounnt>20</Amount>
The problem I am having is how to loop through the MyElement elements
and get all amounts for a given unique year, then go to the next year.
In other words, I want to be able to use the year as a unique key for my
iterations.

Then define a key and use if for grouping:

<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet
xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
version="1.0">

<xsl:output method="xml" indent="yes" />

<xsl:key name="elementByYear" match="MyElement" use="@year" />

<xsl:template match="Root">
<xsl:copy>
<xsl:apply-templates select="MyElement[generate-id() =
generate-id(key('elementByYear', @year)[1])]" />
</xsl:copy>
</xsl:template>

<xsl:template match="MyElement">
<xsl:copy>
<xsl:apply-templates select="@*" />
<xsl:apply-templates select="key('elementByYear', @year)/Amount" />
</xsl:copy>
</xsl:template>

<xsl:template match="@* | node()">
<xsl:copy>
<xsl:apply-templates select="@* | node()" />
</xsl:copy>
</xsl:template>

</xsl:stylesheet>

--

Martin Honnen
http://JavaScript.FAQTs.com/

[Joe Kesselman]

> <xsl:apply-templates select="MyElement[generate-id() =

generate-id(key('elementByYear', @year)[1])]" />

For those who haven't seen this trick before, it may be worth explaining...

The expression "generate-id(node1)=generate-id(node2): is a way of
testing whether node1 and node2 are the same node... often a useful
test, but one that accidentally got let out of XSLT 1.0.

In this case, we're checking it against the list of nodes keyed by the
same @year value -- specifically, against the first node in that list.
The result is that we process all the possible key values, but only the
first node for each key. This gives us a way to enumerate the keys,
which also got left out of XSLT 1.0.

Then, in the template for that node, we can retrieve and process
everything associated with that key.

Clever solution, and it has become a standard XSLT idiom... but it's
definitely not obvious to a beginner!
--
() ASCII Ribbon Campaign | Joe Kesselman
/\ Stamp out HTML e-mail! | System architexture and kinetic poetry

[hilz]

Joe Kesselman wrote:

<xsl:apply-templates select="MyElement[generate-id() =
generate-id(key('elementByYear', @year)[1])]" />

For those who haven't seen this trick before, it may be worth explaining...

The expression "generate-id(node1)=generate-id(node2): is a way of
testing whether node1 and node2 are the same node... often a useful
test, but one that accidentally got let out of XSLT 1.0.

In this case, we're checking it against the list of nodes keyed by the
same @year value -- specifically, against the first node in that list.
The result is that we process all the possible key values, but only the
first node for each key. This gives us a way to enumerate the keys,
which also got left out of XSLT 1.0.

Then, in the template for that node, we can retrieve and process
everything associated with that key.

Clever solution, and it has become a standard XSLT idiom... but it's
definitely not obvious to a beginner!

Thank you Joe and Martin.
You've given me what is necessary for me to proceed.
I also found this example while searching. I guess it is the same as
what Martin suggested:

http://www.jenitennison.com/xslt/gro...muenchian.html
thanks

[Andy Dingley]

On Mon, 13 Feb 2006 21:45:53 -0500, Joe Kesselman
<ke************@comcast.net> wrote:

Clever solution, and it has become a standard XSLT idiom... but it's
definitely not obvious to a beginner!

Is _anything_ in XSLT obvious to a beginner? I've never known a
language like it for needing a "cookbook" approach.

OTOH, I'm just starting to learn Scheme.... 8-)

[Joe Kesselman]

Andy Dingley wrote:

Is _anything_ in XSLT obvious to a beginner? I've never known a
language like it for needing a "cookbook" approach.

Depends on what languages the beginner has been exposed to in the past.
Most folks have only been taught procedural programming, and thinking in
terms of pattern-matching and recursion comes hard until you get used to it.
--
() ASCII Ribbon Campaign | Joe Kesselman
/\ Stamp out HTML e-mail! | System architexture and kinetic poetry