| [birt-news] XML dataset aggregation |
|
I am using BIRT 2.2 and trying to create a dataset based on
an XML data source. The XML file contains the top <projects> element,
which encapsulates a list of <project> elements each of which contains one
<projectName> sub-element a list of <member> sub-elements. Here is
a generic structure of the XML file: <projects> <project> <projectName>P-1<projectName> <members> <member>Alice</member> <member>Bob</member> <member>Dave</member> …. </members> </project> <project> … </project> … </projects> I am using XPath expressions to define a data set with two
columns. One column should list projectName values and the other column should
list the count of <member> sub-elements for the corresponding project. I defined Table Mapping in “Edit Data Set”
dialog using “/projects/*” XPath _expression_. I also defined the “Column
Mapping” for the first column as “projectName” XPath _expression_.
This works fine and I see a list of project names in “Preview Results”,
However, when I define the second column as “count(members/member)”
the column contains empty cells. Is there a way to define XML aggregate operations in BIRT
and how to do it correctly? Thanks. Alec |