Hi Ed and Manfred,
I have committed the code, I need note something here to avoid surprising:
1. For the fixing code in class NavigationHandlerImpl, I only keep the
code that handle case that start from method call node and removed the
code that handles the case start from switch node due to it fails the
cases which starting from switch node both legacy and newly-added;
2. I have do the Hudson testing against weblogic and the failed tests are:
/./javaee6web/flowTraversalCombinations/target/failsafe-reports/com.sun.faces.test.javaee6web.flowtraversalcombinations.ReturnNaviToOtherNodesIT.txt:Tests
run: 5, Failures: 0, Errors: 1, Skipped: 0, Time elapsed: 0.263 sec <<<
FAILURE! - in
com.sun.faces.test.javaee6web.flowtraversalcombinations.ReturnNaviToOtherNodesIT
./javaee6web/flowTraversalCombinations/target/failsafe-reports/com.sun.faces.test.javaee6web.flowtraversalcombinations.SwitchNaviToOtherNodesIT.txt:Tests
run: 5, Failures: 4, Errors: 0, Skipped: 0, Time elapsed: 0.167 sec <<<
FAILURE! - in
com.sun.faces.test.javaee6web.flowtraversalcombinations.SwitchNaviToOtherNodesIT
./javaee7/cdiBeanValidator/target/failsafe-reports/com.sun.faces.test.javaee7.cdibeanvalidator.Issue3014IT.txt:Tests
run: 1, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 1.179 sec <<<
FAILURE! - in com.sun.faces.test.javaee7.cdibeanvalidator.Issue3014IT
/
3. I have do the Hudson testing against glassfish and the failed tests are:
/./javaee6web/flowTraversalCombinations/target/failsafe-reports/com.sun.faces.test.javaee6web.flowtraversalcombinations.ReturnNaviToOtherNodesIT.txt:Tests
run: 5, Failures: 0, Errors: 1, Skipped: 0, Time elapsed: 0.231 sec <<<
FAILURE! - in
com.sun.faces.test.javaee6web.flowtraversalcombinations.ReturnNaviToOtherNodesIT
./javaee6web/flowTraversalCombinations/target/failsafe-reports/com.sun.faces.test.javaee6web.flowtraversalcombinations.SwitchNaviToOtherNodesIT.txt:Tests
run: 5, Failures: 4, Errors: 0, Skipped: 0, Time elapsed: 0.166 sec <<<
FAILURE! - in
com.sun.faces.test.javaee6web.flowtraversalcombinations.SwitchNaviToOtherNodesIT/
4. Cause I didn't find failed test cases starting from method call node,
so I set the bug #3593 as resolved;
5. I filed jira bugs for the failed Tests:
JAVASERVERFACES-3694
<
https://java.net/jira/browse/JAVASERVERFACES-3694>: for
com.sun.faces.test.javaee6web.flowtraversalcombinations.ReturnNaviToOtherNodesIT.testReturnNaviToFlowCallNode
JAVASERVERFACES-3695
<
https://java.net/jira/browse/JAVASERVERFACES-3695>: for tests in
com.sun.faces.test.javaee6web.flowtraversalcombinations.SwitchNaviToOtherNodesIT.txt/
/6. about com.sun.faces.test.javaee7.cdibeanvalidator.Issue3014IT.txt
for Weblogic, I will observe the result in Hudson today, if still
failed, I will file a bug also.
BR,
Zhijun
On 1/16/15, 6:56, Edward Burns wrote:
>>>>>> On Thu, 15 Jan 2015 14:22:46 +0800, zhijun Ren<ren.zhijun_at_oracle.com> said:
> RZ> I have attached change bundle to the jira issue. The change is big
> RZ> and including 25 cases to cover all the navigation possibilities. I
> RZ> have discarded your original tests because of the complete coverage.
>
> RZ> I noticed that the flow can be defined via xml file and java
> RZ> annotation, so I don't know whether need add more cases to test the
> RZ> combinations of the navigation test with different flow definition
> RZ> ways.
>
> There is clearly a lot of work here! Don't bother with the Java Defined
> flow, since whether it is XML or Java, the data structure should end up
> being the same.
>
> RZ> Please help to review due to the big code size and I am looking
> RZ> forward your comments.
>
> I have attached them to the bug.
>
> RZ> I have tested all the 25 cases against GF and get the following
> RZ> results and I think the failed tests may be caused by the Mojarra
> RZ> implementation code and I am checking into it.
>
> Does this mean that you want to check in what you have now and address
> this failure in another commit? I am ok with that as long as you put
> the commit output of all the commits related to 3593 on the work log of
> the JIRA. This makes it easier to the port to other code lines.
>
> This is what I said earlier:
>
> EB> Yes, that's understandable. It might be better to do the work in
> EB> stages and commit the parts that currently do work, then file one or
> EB> more separate bugs for the remainder.
>
> Thanks for tackling this!
>
> Ed
>