multiplie grouping

[Dan Demidoff]

Hello All!

I recieve an xml from DBMS such as:

<ROWSET>
<ROW>
<Dimension1 id="...">...</Dimension1>
<Dimension2 id="...">...</Dimension2>
<Dimension3 id="...">...</Dimension3>
...
<Value1>Число</Value1>
<Value2>Число</Value2>
...
</ROW>
</ROWSET>

For example it may be a table "Good sales" laying out:
of countries (Dimension1);
of cities (Dimension2);
of consumer category (Dimension3)

Value1 reflectes a vegetables sales, Value2 reflectes fruites sales
and so on.

User may visually cut a dimension getting slices of remained
dimensions and SUMs of Value1, Value2 and so on.
For example client may cut all dimensions except country and he'll get
SUM of sales laying out of countries.

My task:
I need multiplie grouping laying out of remained dimensions (not
cutted) with XSLT.
It too simple with SQL:
For all dimensions: SELECT Dim1, Dim2, Dim3, ... SUM(1), SUM(Value2),
.... FROM TABLE GROUP BY Dim1, Dim2, Dim3
After Dim2 had cutted: SELECT Dim1, Dim3, ... SUM(Value1),
SUM(Value2), ... FROM TABLE GROUP BY Dim1, Dim2

How to do that with XSLT?
I suggest we need to define multiplie keys for all dimensions and
exclude dimensions from keys according to cutting (like in SQL above).
Hope for guru's advices.

Thanx.