Question off webtest canoo

[Jose Antonio Avila Flores]

Hi.

How to continue running step when previous step fails

Thank.

_______________________________________________ WebTest mailing list WebTest@... http://lists.canoo.com/mailman/listinfo/webtest

[A Vander]

See the config element, you must set the haltonfailure attribute appropriately.

[Hernan Castagnola]

I am not sure but I think the haltonfailure/haltonerror attributes are related to the "webtest" task not the "steps" task. If you are running a test with multiple checking in a page and you want to check everything without stoping on the first failure I don't think this will solve your issue. There is another thread that shows a solution for telling webtest to continue with next step when one step fails
 <http://lists.canoo.com/pipermail/webtest/2009q2/012345.html>

haltonerror

Required? No, default is "true"

Determines if the execution of the current <webtest> stops if a program error (unhandled exception) is encountered in one of the test steps. An ANT BuildFailed exception will be raised if it is set to "Yes". This will terminate the execution of the ANT script.
Setting the option to "false" will stop the execution of the current <webtest> but continues with the next ANT task, e.g. another <webtest>.

haltonfailure

Required? No, default is "true"

Determines if the execution of the current test halts if a failure of one of the test steps is detected, e.g. a <verify...> test did not pass. Even when set to "false" all of the trailing <step>s of the current <webtest> will be skipped but processing will continue with the next <webtest>.

please let me know if that helped you.

Thanks,

Hernan

2009/8/27 A Vander <avander_be@...>

See the config element, you must set the haltonfailure attribute
appropriately.
--
View this message in context: http://www.nabble.com/Question-off-webtest-canoo-tp25163526p25170525.html
Sent from the WebTest mailing list archive at Nabble.com.

_______________________________________________
WebTest mailing list
WebTest@...
http://lists.canoo.com/mailman/listinfo/webtest

[Marc Guillemot]

Hi,

why do you want to continue if a failure occurs? Is it a failure or
isn't it a failure?

Cheers,
Marc.
--
Web: http://www.efficient-webtesting.com
Blog: http://mguillem.wordpress.com

Jose Antonio Avila Flores wrote:

> Hi.
>
> How to continue running step when previous step fails
>
> Thank.
>
>
>         Jose Antonio Avila
>         javila@... <mailto:javila@...>
>
>
> _______________________________________________ WebTest mailing list
> WebTest@... http://lists.canoo.com/mailman/listinfo/webtest

_______________________________________________
WebTest mailing list
WebTest@...
http://lists.canoo.com/mailman/listinfo/webtest

[Hernan Castagnola]

I think this depends on how you design your test case.

I think therer are at least two reason why I would like that:


1) Because of failure severity
Something that it's not expected but it's something "blocker" (you want to keep testing)
and some other failures that if that steps fails the you can't continue testing the next steps.

(But this depends on how your test cases are designed. )

Example to ilustrate that

//pseudocode
testSmokeTest(){
 Homepage.goto(context)
 Homepage.smokeTest()  ---> (When I see the report I just want to be see the failures according to the severity)
 Homepage.clickCategoryLink(15)
 Listing.smokeTest() ---->
}


2)Because of time spent on executions and analysis
Let's suppose you have a Page composed of modules

Homepage
   ->Module A
   ->Module B

Module A is composed by a list of links, some links should have some condition according to the context related to the Homepage you are testing.
So in some cases some links show have de "rel=nofollow" attribute for SEO reasons. This attribute is a dynamic and depends on the context of the page.

So I would like to create

a Class Homepage with the page logic
a Class Module A with the module logic
 on Module A a have a method that is in charged of testing the nofollowAttributes of the complete module (each link)

So once you create your ui model you can create a test like this

//pseudocode

testNoFollows(){
 Homepage.goto(context)
 Homepage.moduleA.testNoFollows()
}

It's very explicit and easy to understand what you are testing. And what your are expecting is not to stop the execution on the first failure. You would like to see all the links that are failing.

This way you run it once and then you analyse. You fix all the failures and then you run it again.

The oposite of this is running it once, detect one error,  fix that error, then running it another time fix it the second error, etc. It takes more time. 

Thanks,


Hernan

2009/9/4 Marc Guillemot <mguillemot@...>

Hi,

why do you want to continue if a failure occurs? Is it a failure or isn't it a failure?

Cheers,
Marc.
--
Web: http://www.efficient-webtesting.com
Blog: http://mguillem.wordpress.com

Jose Antonio Avila Flores wrote:

Hi.

How to continue running step when previous step fails

Thank.


       Jose Antonio Avila

       javila@... <mailto:javila@...>

_______________________________________________ WebTest mailing list WebTest@... http://lists.canoo.com/mailman/listinfo/webtest

_______________________________________________

WebTest mailing list
WebTest@...
http://lists.canoo.com/mailman/listinfo/webtest