Re: XSLT Subtract function?

--
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

Yesterday I was tasked to produce a list of records from an Oracle database where the result set contained rows where the value in one of the columns was not one of a long list. (Bear with me, please, we'll get to XSLT in a moment.)

The problem was that the list of values to omit was simply too long for SQL*Plus. So I went to plan B. I selected all the rows from the table in question into an XML document. I prepared a second XML document that contained an element for each of the values to be omitted.

I then wrote a stylesheet with this template:

<xsl:template match="ROWSET">
 <ROWSET>
   <xsl:apply-templates select="ROW[not(CODE = $active-codes/ROWSET/ROW/CODE)]" />
 </ROWSET>
</xsl:template>

So the output contained all the ROW elements in the original document whose CODE child string value didn't match any of the CODE string values in the document containing the values I wanted to omit from the result.

In short, I wanted to subtract the set of values contained in the second document from the first, resulting in a document containing only the non-matching values.

After I finished, it occurred to me that this probably wasn't so unusual a need, and that probably someone had written a general-purpose function for this. What I had in mind is a function that would take four arguments:

1) A base document (base document) [$base]
2) A document containing the values to subtract from the base document (exemplar document) [$exemplar]
3) The XPath to the elment in the base document holding the values of interest. [$base-element]
4) The XPath to the element in the exemplar document holding the values of interest. [$list-element]

The result would be a document with the same structure as the base document containing only those elements which didn't have a matching value.

Here is an example of two documents:

The base document:

<?xml version="1.0"?>
<grocery-list>
 <item>apple</item>
 <item>milk</item>
 <item>cheese</item>
 <item>coffee</item>
 <item>asparagus</item>
</grocery-list>

The exemplar document:

<?xml version="1.0"?>
<inventory>
 <sku>apple</sku>
 <sku>pear</sku>
 <sku>grapes</sku>
</inventory>

What I would like is a general-purpose function that would subtract those elements from the first document whose values matched the elements in the second.

<grocery-list>
 <item>milk</item>
 <item>cheese</item>
 <item>coffee</item>
 <item>asparagus</item>
</grocery-list>

I've poked around on Google, but haven't found a useful was to compose the search. Does anyone on the list know of such a function or have you written one? I am reluctant to re-invent the wheel, particularly when mine is not likely to be as round and smooth as an existing one.

Thanks.

--
Charles Knell
cknell@xxxxxxxxxx - email