Usually, our developers will run the QUnit tests directly through the
browser. This applies to our Continuous Integration build
The test case source does the following:
- Using Watin, navigate to the Url with the QUnit tests.
- Execute each test and parse the results in to a data structure
- There is a single test in this suite that looks like this:
[Test, TestCaseSource("RunTests")]
public void Run(QUnitTestResult result)
{
if (!result.IsSuccess)
{
// ReSharper runner clears the output window between each
call to Run() _allFailedTestsOutput.AppendLine(result.Description);
} Assert.IsTrue(result.IsSuccess, result.Description);
}
If TestCaseSource executed only when it required (like the "delayed
evaluation" in LINQ), it would open up many other possible uses.
2011/1/21 Stein Jakob Nordbø <email address hidden>
> Hi, hope you don't mind that I jump in on this one,
>
> We're also running into the same problem; our current work-around is to
> run all QUnit tests within a single NUnit test. The advantage of running
> QUnit tests in a source is to be able to translate 1:1 to NUnit tests so
> that our build system reports QUnit test failures individually rather
> than in summary, and that our test counts include both NUnit and QUnit
> tests. Naturally, running browser-based tests is quite time-consuming in
> the normal flow of development, and since we don't want our back-end
> developers to be waiting for the full suite of javascript tests when
> working in another area of the system, we prefer to defer running of
> these until the build on the build server (hence Explicit+Category); UI
> devs run the JS tests more regularily, but manually.
>
>
> Stein Jakob
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/538070
>
> Title:
> Need a way to generate data when test is executed
>
> Status in NUnit Test Framework:
> Confirmed
>
> Bug description:
> NUnit 2.5.2 and 2.5.3 run every method targeted by TestCaseSource,
> even if the test targeting it is Explicit or the test fixture is
> Explicit.
>
> The behaviour I expect is: NUnit shouldn't run a test's TestCaseSource
> if the test itself won't be run.
>
> Detail:
>
> I have an expensive TestCaseSource generating a few hundred thousand
> permutations and combinations. I've applied Explicit to both the test
> targeting the source and the fixture surrounding both test and source.
>
> NUnit console spends ten minutes needlessly constructing test cases
> from the TestCaseSource before ignoring them and proceeding with the
> non-Explicit tests.
>
> If I throw NotImplementedException from the first line of the
> expensive TestCaseSource. NUnit runs the non-Explicit tests
> immediately. If I put a MessageBox.Show call in the first line, I see
> the message box.
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/nunit-3.0/+bug/538070/+subscribe
>
Usually, our developers will run the QUnit tests directly through the
browser. This applies to our Continuous Integration build
The test case source does the following:
- Using Watin, navigate to the Url with the QUnit tests.
- Execute each test and parse the results in to a data structure
- There is a single test in this suite that looks like this:
[Test, TestCaseSource( "RunTests" )]
_allFailedTes tsOutput. AppendLine( result. Description) ;
Assert. IsTrue( result. IsSuccess, result. Description) ;
public void Run(QUnitTestResult result)
{
if (!result.IsSuccess)
{
// ReSharper runner clears the output window between each
call to Run()
}
}
If TestCaseSource executed only when it required (like the "delayed
evaluation" in LINQ), it would open up many other possible uses.
2011/1/21 Stein Jakob Nordbø <email address hidden>
> Hi, hope you don't mind that I jump in on this one, /bugs.launchpad .net/bugs/ 538070 xception from the first line of the /bugs.launchpad .net/nunit- 3.0/+bug/ 538070/ +subscribe
>
> We're also running into the same problem; our current work-around is to
> run all QUnit tests within a single NUnit test. The advantage of running
> QUnit tests in a source is to be able to translate 1:1 to NUnit tests so
> that our build system reports QUnit test failures individually rather
> than in summary, and that our test counts include both NUnit and QUnit
> tests. Naturally, running browser-based tests is quite time-consuming in
> the normal flow of development, and since we don't want our back-end
> developers to be waiting for the full suite of javascript tests when
> working in another area of the system, we prefer to defer running of
> these until the build on the build server (hence Explicit+Category); UI
> devs run the JS tests more regularily, but manually.
>
>
> Stein Jakob
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https:/
>
> Title:
> Need a way to generate data when test is executed
>
> Status in NUnit Test Framework:
> Confirmed
>
> Bug description:
> NUnit 2.5.2 and 2.5.3 run every method targeted by TestCaseSource,
> even if the test targeting it is Explicit or the test fixture is
> Explicit.
>
> The behaviour I expect is: NUnit shouldn't run a test's TestCaseSource
> if the test itself won't be run.
>
> Detail:
>
> I have an expensive TestCaseSource generating a few hundred thousand
> permutations and combinations. I've applied Explicit to both the test
> targeting the source and the fixture surrounding both test and source.
>
> NUnit console spends ten minutes needlessly constructing test cases
> from the TestCaseSource before ignoring them and proceeding with the
> non-Explicit tests.
>
> If I throw NotImplementedE
> expensive TestCaseSource. NUnit runs the non-Explicit tests
> immediately. If I put a MessageBox.Show call in the first line, I see
> the message box.
>
> To unsubscribe from this bug, go to:
> https:/
>