#5 Allow multiple source files to form composite dataset

[Bjorn Beskow]

When specifying multiple sources, they are by default creating separate datasets, which are used in separate operations, one for each source file.

In many situations, it may however make sense to split one dataset into multiple smaller, separate files. It allows composition of different datasets from smaller bits and pieces, and hence promote reuse while reducing redundancy.

The CompositeDataSet already provides all needed functionality. It just needs to be exposed via the maven plugin and the underlying ant Operation.

An example usage might look like this:

      <execution>
        <id>clean-insert-from-multiple-files</id>
        <phase>test</phase>
        <goals>
          <goal>operation</goal>
        </goals>
        <configuration>
          <type>CLEAN_INSERT</type>
          <sources>
            <source>aSourceFile.xml</source>
            <source>anotherSourceFile.xml</source>
          </sources>
          <composite>true</composite>
          <combine>true</combine>
        </configuration>
      </execution>

[Jeff Jensen]

Ticket moved from /p/dbunit/feature-requests/245/

Can't be converted:

• _priority: 5
• _milestone: (future)

[Jeff Jensen]

Merged. Thank you for your work!