I've tried splitting the results (packages) up into diff. xml files, but
it still collates them under (root).
Am I doing something wrong?
E.g.:
<?xml version="1.0" encoding="UTF-8"?>
<testsuites>
<testsuite id="0" package="MyPackage" name="SomeName"
timestamp="2011-10-12T11:46:46" hostname="host" tests="8" time="123"
errors="0" failures="0">
...
(with multiple testsuites for differing packages)
--
[neil@fnx ~]# rm -f .signature
[neil@fnx ~]# ls -l .signature
ls: .signature: No such file or directory
[neil@fnx ~]# exit
to my knowledge, there's no official specification of the JUnit XML file
format - so unfortunately, every tool seems to talk a dialect of its
own... ;O(
In case of Jenkins, an attribute 'package' of the 'testsuite' element is
ignored as far as I know. Instead you can use a fully qualified class
name in the 'classname' attribute of the embedded 'testcase' element,
e.g. classname="MyPackage.SomeName". Jenkins will process anything left
from the last dot as a package prefix.
Does this work for you?
Cheers,
Simon.
--
Excellent, yes, that does the trick! Thanks!