oasp/oasp4j#207: added example for solution 3 (CSV based test with se…

[Max-Goebel]

…renity)

There are some problems with dependencies, so this change should not be accepted without correction. I have just created it in order to discuss advantages and disadvantages of data driven test with serenity.

This are the two problems with the dependencies:

• Guava must be upgraded to version 18.0. I don't know which side effects this may have. Please review and test it.
• A new dependency "xml-apis" must be created. Unfortunately, this dependency causes the AccessControlSchemaXmlValidationTest to fail. I have no idea why and how to fix it.

Furthermore, I think I found a little bug in UcManageOrderPositionImpl and fixed it.

[oasp-ci]

Can one of the admins verify this patch?

[hohwille]

build error looks like dependency(-management) problem that I can not easily track down with the diffs. In general your PR looks right. So 👍 already for your input.
For the test-cases can you follow our own conventions and use AssertJ assertions instead of JUnit-Standard-Asserts. Simply extends one of the four Base-Classes that we provide for each test-level such as ComponentTest. Further replace TODO with 1-2 sentences JavaDoc what is the goal of the test.
Cool is especially that before your change we were checking currentOrderPosition against itself.
To see the actual value of serenity it would be nice to observer the reporting if the test fails.
I suppose that this is where it gets interesting. Everything else could IMHO be done with plain JUnit and AssertJ about as simple.

[Max-Goebel]

• AssertJ assertions: Done
• Replace TODO: Done
• Build error: Not done. I could not reproduce it. I've just got the failing testcase AccessControlSchemaXmlValidationTest, but there are no build errors.
• Test reports: See ...\oasp4j\oasp4j-samples\oasp4j-sample-core\target\site\serenity

[oasp-ci]

Can one of the admins verify this patch?

[Devonfw-ci]

Can one of the admins verify this patch?

[hohwille]

We need some testing guys to have a look. @llaszkie any feedback?
This PR is more than overdue and effort for resolving conflicts, etc. rises if we do not merge for ages.
To me this looks like a nice demonstration of combination testing. However, restaurant sample is going to be deleted soon. How should we proceed?

[hohwille]

I have liked this as a task for #632 so we can utilize code from this PR for MTS. As the sample has been deleted, I am going to close this PR as it can not be merged anymore.