There are no receivers for the messages because they're functional rather
than OO. When it comes to unit tests, though, I feel that remaining
strictly OO makes them awkward.


On 5 January 2014 08:34, Gary Gregory <[email protected]> wrote:

> I've not been a Hamcrest fan. It feels to much like a DL that makes for
> funky looking tests. In general I do not like static imports for
> non-constants because it is not OO, You cannot tell who the receiver of a
> message is.
>
> Gary
>
>
> -------- Original message --------
> From: Ralph Goers
> Date:01/05/2014 04:49 (GMT-05:00)
> To: Log4J Developers List
> Cc: [email protected]
> Subject: Re: Question/suggestion regarding use of assertions in unit
> tests.
>
> The tests weren't migrated from junit 3, although I spent many years
> working with it.  Frankly, until I saw your patch today I was unaware of
> directly using Hamcrest. Since junit uses it I really have no problem if we
> do if it makes the tests more readable.
>
> Ralph
>
> On Jan 5, 2014, at 12:38 AM, Matt Sicker <[email protected]> wrote:
>
> Hi all, newcomer here. I'm wondering what your opinions are on using
> Hamcrest matchers in unit tests. That is, using the assertThat() methods
> and the matchers that go with it. It makes many otherwise hard to read
> assertions far more literate, plus it provides some nice error messages
> explaining why an assertion failed (unlike the native assert keyword and
> certain related methods in org.junit.Assert).
>
> I don't know if the tests were migrated from JUnit 3 or anything, but I do
> believe it's the preferred way of asserting things in JUnit.
>
> And before anyone says something like "patches welcome", I'd be glad to
> help update unit tests for such a thing. :)
>
> --
> Matt Sicker <[email protected]>
>
>


-- 
Matt Sicker <[email protected]>

Reply via email to