Pedro,
Thank you for your reply. I should have mentioned that I do not want the build
to fail within the same target that the tests are run. After my
run.server.tests has been run, there's another target invoked that undeploys
the cactified application. If I fail the build after the tests are run, this
undeploy target is never invoked and the application remains on the server.
The simple solution, of course, is to have all the tasks done within the same
target (i.e. within your run target, have the undeploy task written right
before where fail appears). This just seems contrary to how I think cactus
and ANT should be working together. Namely, properties that are set in
runservertest's testtarget should persist throughout the rest of the build.
Thanks again for your time.
Keith
Quoting Pedro Nevado [EMAIL PROTECTED]:
Hi there,
This works for me. I hope it helps.
Pedro
target name=run
mkdir dir=${tests.reports.dir}/
junit printsummary=false haltonerror=false
haltonfailure=false fork=${junit.fork}
errorproperty=test.failed failureproperty=test.failed
formatter type=plain usefile=false/
formatter type=xml/
classpath
//.
/classpath
batchtest todir=${tests.reports.dir} unless=testClass
fileset dir=${tests.class.dir}
includes=**/*Test.class/
/batchtest
/junit
junitreport todir=${tests.reports.dir}
fileset dir=${tests.reports.dir} includes=TEST-*.xml/
report format=frames todir=${tests.reports.dir}/html/
/junitreport
fail if=test.failed message=Tests failed after making report./
/target
-Mensaje original-
De: Keith Collison [mailto:[EMAIL PROTECTED]
Enviado el: viernes, 01 de julio de 2005 22:30
Para: cactus-user@jakarta.apache.org
Asunto: Problem using runservertests failureproperty
Hello,
In my current environment, I manually create a cactified war file and
use runservertests to execute the tests. Everything happens as
expected, with one important hitch -- my failureproperty is not
getting set as expected.
Here are my two relevant targets:
target name=run.server.tests depends=init
runservertests
testurl=${cactus.test.url}
starttarget=start.tomcat
stoptarget=stop.cactus
testtarget=test.server
/
echo message=within run.server.tests, test
failed=[${test.failed}]/
/target
target name=test.server
description=Runs junit on test cases.
delete dir=${test.result.dir}/
mkdir dir=${test.result.dir}/
property name=test value=Test*/
junit printsummary=true
haltonerror=false
haltonfailure=false
failureproperty=test.failed
classpath
path refid=cactus.classpath/
pathelement location=${cactus.classes.dir}/
/classpath
formatter type=plain usefile=false/
formatter type=xml/
batchtest todir=${test.result.dir}
fileset dir=${test.web.dir}
include name=**/${test}.java/
exclude name=**/Base*.java/
exclude name=**Abstract**/
/fileset
/batchtest
/junit
echo message=test failed:[${test.failed}]/
/target
Note the echo tasks I have. When I run the tests and there are test
failures, my output is:
[echo] within test.server target, test failed=[true]
[echo] within run.server.tests, test failed=[${test.failed}]
Obviously the failure property is not being permanently set by the
test.server target Am I misunderstanding how this is supposed to work?
Are there other ways of accomplishing the same goal (without use of the
cactus task)?
Thank you in advance.
Keith
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
