Re: I am missing Expression architecture...

From: Giulio Cesare Solaroli (slrgcs..bn-italy.com)
Date: Fri Oct 03 2003 - 04:38:33 EDT

  • Next message: Giulio Cesare Solaroli: "Re: [Still LONG] Re: [LONG] reasons why "advanced" expressions are needed"

    On Thursday, Oct 2, 2003, at 11:14 Europe/Rome, Dirk Olmes wrote:

    >> I think it would be excellent to start all this discussion writing
    >> down
    >> a few test cases that describe exactly the problem we are trying to
    >> solve, how we want to write the application code to use express these
    >> feature, and then add the code to Cayenne until all the new tests
    >> pass.
    >
    > very XP style, I like that :-)

    That's what I have done while implementing these same feature for EOF,
    and it was really helpful.

    >> I have tried to look for Cayenne and DBUnit, but did not find anything
    >> relevant.
    >>
    >> Is there anybody doing any work to integrate Cayenne with a DB testing
    >> framework like DBUnit or something like that?
    >
    > I investigated into a cayenne-DBUnit cooperation some time ago,
    > especially
    > since it seems to be the perfect match: let cayenne do its work and
    > check
    > with DBUnit whether the database is in expected state afterwards.

    I didn't look at the Cayenne source tree very carefully, and missed the
    test package.
    I am going to take a closer look.

    > [...]
    > I think it will take some effort to integrate DBUnit into the cayenne
    > test
    > cases. In that process there's still a number of decisions to make:
    > should
    > DBUnit fill the database upon each test run? each TestCase? only once?
    > I've used DBUnit successfully in conjunction with EOF and TopLink in
    > the
    > past and can share what I found out, even some code though it may be
    > only of
    > limited use.

    We have written the DBTestingKit, and extension to OCUnit written by
    Sente (www.sente.ch).
    Our approach was to allow every testCase if its tests where read-only
    or not.
    If the tests where read-only, you could also specify if the DB had to
    be created before running the tests, or whether the tests should use an
    already available DB (as the tests were read-only, you could safely run
    them against an already running DB).

    Anyway, these are only test performance tuning; it's true you should
    run these tests often, but even if they are not optimally tuned, I
    would be much more interested in great framework code!! ;-]

    Anyway, I think that all experiences on these topics should be welcome.

    Giulio Cesare



    This archive was generated by hypermail 2.0.0 : Fri Oct 03 2003 - 04:38:27 EDT