Constructing, not selecting, XSL node set variable - xslt

I wish to construct an XSL node set variable using a contained for-each loop. It is important that the constructed node set is the original (a selected) node set, not a copy.
Here is a much simplified version of my problem (which could of course be solved with a select, but that's not the point of the question). I've used the <name> node to test that the constructed node set variable is in fact in the original tree and not a copy.
XSL version 1.0, processor is msxsl.
Non-working XSL:
<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:output method="text" encoding="iso-8859-1" omit-xml-declaration="yes" />
<xsl:template match="/">
<xsl:variable name="entries">
<xsl:for-each select="//entry">
<xsl:copy-of select="."/>
</xsl:for-each>
</xsl:variable>
<xsl:variable name="entryNodes" select="msxsl:node-set($entries)"/>
<xsl:for-each select="$entryNodes">
<xsl:value-of select="/root/name"/>
<xsl:value-of select="."/>
</xsl:for-each>
</xsl:template>
</xsl:stylesheet>
XML input:
<?xml version="1.0" encoding="UTF-8"?>
<root>
<name>X</name>
<entry>1</entry>
<entry>2</entry>
</root>
Wanted output:
X1X2
Actual output:
12
Of course the (or a) problem is the copy-of, but I can't work out a way around this.

There isn't a "way around it" in XSLT 1.0 - it's exactly how this is supposed to work. When you have a variable that is declared with content rather than with a select then that content is a result tree fragment consisting of newly-created nodes (even if those nodes are a copy of nodes from the original tree). If you want to refer to the original nodes attached to the original tree then you must declare the variable using select. A better question would be to detail the actual problem and ask how you could write a suitable select expression to find the nodes you want without needing to use for-each - most uses of xsl:if or xsl:choose can be replaced with suitably constructed predicates, maybe involving judicious use of xsl:key, etc.
In XSLT 2.0 it's much more flexible. There's no distinction between node sets and result tree fragments, and the content of an xsl:variable is treated as a generic "sequence constructor" which can give you new nodes if you construct or copy them:
<xsl:variable name="example" as="node()*">
<xsl:copy-of select="//entry" />
</xsl:variable>
or the original nodes if you use xsl:sequence:
<xsl:variable name="example" as="node()*">
<xsl:sequence select="//entry" />
</xsl:variable>

I wish to construct an XSL node set variable using a contained
for-each loop.
I have no idea what that means.
It is important that the constructed node set is the original (a
selected) node set, not a copy.
This part I think I understand a little better. It seems you need to replace:
<xsl:variable name="entries">
<xsl:for-each select="//entry">
<xsl:copy-of select="."/>
</xsl:for-each>
</xsl:variable>
with:
<xsl:variable name="entries" select="//entry"/>
or, preferably:
<xsl:variable name="entries" select="root/entry"/>
The resulting variable is a node-set of the original entry nodes, so you can do simply:
<xsl:for-each select="$entries">
<xsl:value-of select="/root/name"/>
<xsl:value-of select="."/>
</xsl:for-each>
to get your expected result.
Of course, you could do the same thing by operating directly on the original nodes, in their original context - without requiring the variable.
In response to the comments you've made:
We obviously need a better example here, but I think I am getting a vague idea of where you want to go with this. But there are a few things you must understand first:
1.
In order to construct a variable which contains a node-set of nodes in their original context, you must use select. This does not place any limits whatsoever on what you can select. You can do your selection all at once, or in stages, or even in a loop (here I mean a real loop). You can combine the intermediate selections you have made in any way sets can be combined: union, intersection, or difference. But you must use select in all these steps, otherwise you will end up with a set of new nodes, no longer having the context they did in the source tree.
IOW, the only difference between using copy and select is that the former creates new nodes, which is precisely what you wish to avoid.
2.
xsl:for-each is not a loop. It has no hierarchy or chronology. All the nodes are processed in parallel, and there is no way to use the result of previous iteration in the current one - because no iteration is "previous" to another.
If you try to use xsl:for-each in order to add each of n processed nodes to a pre-existing node-set, you will end up with n results, each containing the pre-existing node-set joined with one of the processed nodes.
3.
I think you'll find the XPath language is quite powerful, and allows you to select the nodes you want without having to go through the complicated loops you hint at.

It might help if you showed us a problem that can't be trivially solved in XSLT 1.0. You can't solve your problem the way you are asking for: there is no equivalent of xsl:sequence in XSLT 1.0. But the problem you have shown us can be solved without such a construct. So please explain why you need what you are asking for.

Related

Is there a way to count the elements generated by an XSL within the same XSL?

If I have an XSL that creates output like this simple/rough example:
<Parent1>
<ABC><xsl:value-of select="SomeValue1"/></ABC>
<DEF><xsl:value-of select="SomeValue2"/></DEF>
<GHI><xsl:value-of select="SomeValue3"/></GHI>
...
<YZ><xsl:value-of select="SomeValue9"/></YZ>
</Parent1>
... within this same XSL, how can I count how many children the XSL will produce?
You can generate your content into a variable, count the children in the variable, and then emit the content of the variable:
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="2.0">
<xsl:template match="/">
<xsl:variable name="temp-results">
<Parent1>
<ABC><xsl:value-of select="SomeValue1"/></ABC>
<DEF><xsl:value-of select="SomeValue2"/></DEF>
<GHI><xsl:value-of select="SomeValue3"/></GHI>
...
<YZ><xsl:value-of select="SomeValue9"/></YZ>
</Parent1>
</xsl:variable>
<xsl:text>Number of children:</xsl:text>
<xsl:value-of select="count($temp-results/Parent1/*)"/>
<xsl:sequence select="$temp-results"/>
</xsl:template>
</xsl:stylesheet>
One possibility is wrapping the whole output process in a variable and then count its descendants.
So, for example, you can use the following XSLT code
<xsl:template match="/Parent1">
<xsl:variable name="output">
<ABC><xsl:value-of select="SomeValue1"/><ZZZ>Some Grandchild</ZZZ></ABC>
<DEF><xsl:value-of select="SomeValue2"/></DEF>
<GHI><xsl:value-of select="SomeValue3"/></GHI>
...
<YZ><xsl:value-of select="SomeValue9"/></YZ>
</xsl:variable>
<xsl:value-of select="concat('Outputting ', count($output/descendant::*), ' elements.
')" />
<xsl:copy-of select="$output" />
</xsl:template>
Its output is
Outputting 5 elements.
<ABC>
<ZZZ>Some Grandchild</ZZZ>
</ABC>
<DEF/>
<GHI/>
...
<YZ/>
This code accomplishes three things:
First it generates the result and puts it into the variable
It counts all the
children (child::* axis) or
descendants (descendant::* axis) as in the example above
of the elements in the variable
It copies the variable to the output stream
This approach can even be nested - meaning that it can be applied several times, one after another.
Your choices are:
(a) find a way of computing the result as a function of the input
(b) capture the output in a variable and run a second phase of processing against that variable.
(c) a blend of the above: compute some intermediate result in a variable, and use that variable as input to both processes.
In the example you've given, the first approach works perfectly well; but I guess your real problem is more complex than that, otherwise you wouldn't be asking.

XSLT: for-each loop with key not returning all nodes

I am a novice XSLT developer. I have been asked to fix an issue on a project where the original developer is no longer with us. In the XSLT, there is a for-each loop using a key and a count
<xsl:for-each select="ns0:BOM[count(. | key('subsat', ns0:BomText01)[1]) = 1][ns0:BomText01]">
...
This is the key:
<xsl:key name="subsat" match="ns0:Parts/ns0:BOM[ns0:FindNum!='0']" use="ns0:BomText01" />
In the XML file being transformed, there are two sibling nodes that represent sub-parts:
<ns0:BOM referentId="10000:65091335:65359080">
<ns0:BomText01>3069260-303-SUB0027</ns0:BomText01>
<ns0:ItemNumber>My_part_1</ns0:ItemNumber>
<ns0:ItemType>Part</ns0:ItemType>
<ns0:Qty>67</ns0:Qty>
</ns0:BOM>
<ns0:BOM referentId="10000:65102551:86713230">
<ns0:BomText01>3069260-303-SUB0027</ns0:BomText01>
<ns0:ItemNumber>My_part_2</ns0:ItemNumber>
<ns0:ItemType>Part</ns0:ItemType>
<ns0:Qty>67</ns0:Qty>
</ns0:BOM>
However, the loop is only picking up the first node (My_part_1). I suspect it's because of the count=1 but I really don't know. And I don't know how to modify it. Ideas? If I need to include more data, let me know.
Assuming that the relevant part of your XSLT looks something like this:
<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:ns0="ns0" version="1.0">
<xsl:key name="subsat" match="ns0:BOM[ns0:FindNum!='0']" use="ns0:BomText01"/>
<xsl:template match="ns0:Parts">
<xsl:for-each
select="ns0:BOM[count(. | key('subsat', ns0:BomText01)[1]) = 1][ns0:BomText01]">
<xsl:value-of select="."/>
</xsl:for-each>
</xsl:template>
</xsl:stylesheet>
It will only print the first of the elements because it is selecting the BOM elements which have an unique BomText01 value. That's the expected result.
If the BomText01 is an ID field (as it seems it is) and you expected to get both result (perhaps, because their ItemNumber contains different values), the error is possibly in your source (which assigned equal IDs when it should not have done so).
If you change one of those values in the source, you should be able to select both and verify this.

XSLT/Xpath -- sum function performance

I often use this xpath sum(preceding::*/string-length())
It does what I need it to do (provides a character count of all text up to this context in the XML file).
Problem: it is slow.
Is there a different built in function that I should be using instead? Or an extension?
UPDATE:
Based on Michael Kay's comment, I explored XSLT 3.0 <accumulator>. It was my first try with 3.0 (I had to update OxygenXML to make it work). I haven't fully adapted it to my needs, but initial test below shows promise.
<xsl:output method="xml" />
<xsl:accumulator
name="f:string-summ"
post-descent="f:accum-string-length"
as="xs:integer"
initial-value="0">
<xsl:accumulator-rule
match="text/*"
new-value="$value + string-length()"/>
</xsl:accumulator>
<xsl:template match="text/*">
<xsl:value-of select="f:accum-string-length()" />
</xsl:template>
Off topic: Stack Overflow needs an "XSLT-3.0" tag.
If you call this function on every node, then you stylesheet performance will be O(n^2) in the number of nodes.
The function is incorrect anyway. The preceding axis gives you your parent's preceding siblings and also the children of your parent's preceding siblings, so the string length of your cousins is being counted more than once.
Try defining a memo function something like this:
<xsl:function name="f:preceding-string-length" saxon:memo-function="yes">
<xsl:param name="n" as="element()"/>
<xsl:sequence select="sum(ancestor::*/preceding-sibling::*[1]/(f:preceding-string-length(.) + string-length(.)))"/>
</xsl:function>
Or use an XSLT 3.0 accumulator, which amounts to much the same thing.
I don't think the sum function is slow, the navigation to all preceding elements and the computation of the string length of all contents is expensive. As for optimizing it, which XSLT 2.0 processor do you use?

adding multiple filters

As I have mentioned in this post:
dynamic multiple filters in xsl
Basically, I want to apply multiple filters to my xml using "for loop" and these filters are dynamic which are coming from some other xml
sth like this:
foreach(list/field[#ProgramCategory=$Country][not(contain(#Program,$State1][not(contain(#Program,$State2][not(contain(#Program,$State3][not(contain(#Program,$Staten])
The problem is that I can get n no. of states which I am getting through for loop of other xml.
I cannot use document() function as suggested by Dimitre so I was thinking of achieving it by:
<xsl:variable name="allprograms">
<xsl:for-each select="/list2/field2">
<xsl:text disable-output-escaping="yes">[not(contains(#Program,'</xsl:text><xsl:value-of select="#ProgramID"></xsl:value-of><xsl:text disable-output-escaping="yes">'))]</xsl:text>
</xsl:for-each>
</xsl:variable>
gives me something like this:
[not(contains(#Program,'Virginia'))][not(contains(#Program,'Texas'))][not(contains(#Program,'Florida'))]
I want to use this above value as a filter in the for loop below and I am not sure how to achieve that
<xsl:for-each="list/field[not(contains(#Program,'Virginia'))][not(contains(#Program,'Texas'))][not(contains(#Program,'Florida'))]">
Before this I also have a for loop to filter United States
xsl:for-each="list/field $allprograms">
<xsl:value-of select="#ows_ID" />
</xsl:for-each>
I want my answer to be 1082, 1088..
I can add the xml here too if there is any confusion..
Jack,
From the previous solution you just need to add to this:
<xsl:param name="pFilteredStates">
<state>Virginia</state>
<state>Texas</state>
<state>Florida</state>
</xsl:param>
the following (changing the current variable definition that relies on the document() function):
<xsl:variable name="vFiltered" select=
"ext:node-set($pFilteredStates)/*
"/>
Where the "ext:" prefix needs to be bound to this namespace (this is the EXSLT namespace -- if your XSLT processor doesn't implement exslt:node-set() then you need to find what xxx:node-set() extension it implements, or tell us what is your XSLT processor and people will provide this information):
"http://exslt.org/common"
So, your <xsl:stylesheet> may look like the following:
<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
xmlns:ext="http://exslt.org/common" exclude-result-prefixes="ext">
I still recommend that the $pFilteredStates parameter should be passed by the initiator of the transformation -- in which case you can delete the definition of $vFiltered and replace every reference to it with $pFilteredStates` and the transformation should work OK.

XSL Reuse? YES! But: Element must not contain an xsl:import element! :-(

I am using a heavy stylesheet with a lot of recurring transformations, so I thought it would be smart to reuse the same chunks of code, so I would not need to make the same changes at a bunch of different places. So I discovered , but -alas- it won't allow me to do it. When trying to run it in Sonic Workbench I get the following error:
An xsl:for-each element must not contain an xsl:import element
This is my stylesheet code:
<xsl:template match="/">
<InboundFargoMessage>
<EdiSender>
<xsl:value-of select="TransportInformationMessage/SenderId"/>
</EdiSender>
<EdiReceiver>
<xsl:value-of select="TransportInformationMessage/RecipientId"/>
</EdiReceiver>
<EdiSource>PORLOGIS</EdiSource>
<EdiDestination>FARGO</EdiDestination>
<Transportations>
<xsl:for-each select="TransportInformationMessage/TransportUnits/TransportUnit">
<xsl:import href="TransportCDMtoFDM_V0.6.xsl"/>
</xsl:for-each>
<xsl:for-each select="TransportInformationMessage/Waybill/TransportUnits/TransportUnit">
<xsl:import href="TransportCDMtoFDM_V0.6.xsl"/>
</xsl:for-each>
</Transportations>
</InboundFargoMessage>
</xsl:template>
</xsl:stylesheet>
I will leave out the child xsl-sheets for now, as the problem appears to be happening at the base.
If I cannot use xsl:import, is there any option of reuse?
If I cannot use xsl:import, is there
any option of reuse?
You can use <xsl:import>.
All <xsl:import> elements must be the first element children of <xsl:stylesheet>
As an alternative, an <xsl:include> element has to be globally defined (a child of <xsl:stylesheet>) but can be preceded by any other xslt instruction that can be placed globally.
You need to be aware of and understand well the rules of using these two XSLT instructions. I'd recommend reading a good book on XSLT.
The main unit of reusability in XSLT is the template (<xsl:template>).
The importing stylesheet can use (via <xsl:call-template> or <xsl:apply-templates>) any template that is defined in any imported stylesheet.
Each of included XSL files should contain a template(s).
The main file includes the others in the beginning and then calls the templates with call-template or apply-templates from various places.
Thanks for all the suggestions, which were somewhat helpful, but allow me formulate a complete answer. As suggested, the answer to the question of re-use lies in the xsl:templates. Templates can be defined by enclosing them within . Then, whereever necessary, they can be summoned by adding a element. Also, they can be put into separate xsl sheets, as long as they are imported at the top of the parent xsl sheet.
Thus, the solution to my questions looks as follows:
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" xmlns:far="http://www.itella.com/fargo/fargogate/" xmlns:a="http://tempuri.org/XMLSchema.xsd" xmlns:p="http://tempuri.org/XMLSchema.xsd">
<xsl:import href="TransportCDMtoFDM_V0.6.xsl"/>
<xsl:template match="/">
<InboundFargoMessage>
<EdiSender>
<xsl:value-of select="TransportInformationMessage/SenderId"/>
</EdiSender>
<EdiReceiver>
<xsl:value-of select="TransportInformationMessage/RecipientId"/>
</EdiReceiver>
<EdiSource>PORLOGIS</EdiSource>
<EdiDestination>FARGO</EdiDestination>
<Transportations>
<xsl:for-each select="TransportInformationMessage/TransportUnits/TransportUnit">
<xsl:call-template name="transport"/>
</xsl:for-each>
<xsl:for-each select="TransportInformationMessage/Waybill/TransportUnits/TransportUnit">
<xsl:call-template name="transport"/>
</xsl:for-each>
</Transportations>
</InboundFargoMessage>
</xsl:template>
Where the file 'TransportCDMtoFDM_V0.6.xsl' contains the template called "transport".
There is just one problem left: Using templates, all the nodes mentioned within the template are used, even if they are empty. So the remaining question is how to leave out the empty nodes?