Extreme Programming - JUnit-diff-report

Author

JUnit-diff-report

Przemek Pokrywka

2005-10-25, 6:57 pm

I'd like to have a tool, that generates a report similar to maven's
JUnit report, but showing only the differences between current and last
JUnit report - which failing tests were fixed, which tests stopped to
pass, which tests were added and which removed.

I suspect, that I'd be able to write such a report generator with a
moderate effort, but I'm wondering if anybody ever heard of a tool like
that. I think it can be often useful to have only the changes
highlighted and not to have to scroll through the long pages of the
standard JUnit report.

Przemyslaw Pokrywka

Phlip

2005-10-25, 9:56 pm

Przemek Pokrywka wrote:

> I'd like to have a tool, that generates a report similar to maven's JUnit
> report, but showing only the differences between current and last JUnit
> report - which failing tests were fixed, which tests stopped to pass,
> which tests were added and which removed.

This is the XP newsgroup! Why are any tests, anywhere, failing??

Focus on the cause, not the symptom.

--
Phlip
[url]http://www.greencheese.org/Z

Land[/url] <-- NOT a blog!!!

David Carlton

2005-10-25, 9:56 pm

On Wed, 26 Oct 2005 01:21:11 GMT, "Phlip" <phlipcpp@yahoo.com> said:
> Przemek Pokrywka wrote:

[color=darkred]
> This is the XP newsgroup! Why are any tests, anywhere, failing??

Acceptance tests are allowed to fail, no?

Not that I think that's really the issue here: I agree with the
following:

> Focus on the cause, not the symptom.

David Carlton
carlton@bactrian.org

Phlip

2005-10-26, 3:59 am

David Carlton wrote:

> Acceptance tests are allowed to fail, no?

Any acceptance test that passes, it should trip a latch, and raise its
failure from a warning to error level. So for any commit of new code, all
unit tests must pass, and all acceptance tests that formerly passed must
still pass.

The strict verbiage of /Extreme Programming Explained/, 1st Edition, just
says "most acceptance tests must pass". This allows some of these tests to
be fuzzy, like tests on performance.

BTW I have played the "diff the bug reports" game many times, always in
legacy code...

--
Phlip
[url]http://www.greencheese.org/Z

Land[/url] <-- NOT a blog!!!

Robert C. Martin

2005-10-26, 6:57 pm

On Tue, 25 Oct 2005 19:48:27 -0700, David Carlton
<carlton@bactrian.org> wrote:

>On Wed, 26 Oct 2005 01:21:11 GMT, "Phlip" <phlipcpp@yahoo.com> said:
>
>
>
>Acceptance tests are allowed to fail, no?

Yes, but they latch. Once they pass, they are never allowed to fail
again.

-----
Robert C. Martin (Uncle Bob) | email: unclebob@objectmentor.com
Object Mentor Inc. | blog: www.butunclebob.com
The Agile Transition Experts | web: www.objectmentor.com
800-338-6716

"The aim of science is not to open the door to infinite wisdom,
but to set a limit to infinite error."
-- Bertolt Brecht, Life of Galileo

jeffgrigg@charter.net

2005-10-30, 6:59 pm

--- Przemek Pokrywka wrote:
> I'd like to have a tool, that generates a report similar to
> maven's JUnit report, but showing only the differences between
> current and last JUnit report
> - which failing tests were fixed,
> which tests stopped to pass,
> which tests were added and which removed.

It's a bad thing if some of your JUnit tests are failing. They should
all pass.

But given that some tests may fail...
"Cruise Control" is pretty good at reporting when JUnit tests are
failing, and when they get fixed. But it won't work "well" if some
tests keep failing for a long time and don't get fixed.

To do the report you suggest, particularly the added and removed tests,
I'd suggest using "ant", with the XML formatter. Then you can read,
interpret, and "diff" the XML files.

Or, you might direct your question to the "junit" Yahoo group:
http://groups.yahoo.com/group/junit/

I hope this helps!
- jeff

Sponsored Links