Re: omitted test scripts?

From: Daniel Barkalow <barkalow@iabervon.org>
Date: Sep 19, 2005, 17:01 UTC
Message-ID: <Pine.LNX.4.63.0509191239090.23242@iabervon.org>
In-Reply-To: <7v1x3lrpxi.fsf@assigned-by-dhcp.cox.net>

On Mon, 19 Sep 2005, Junio C Hamano wrote:

Show 11 quoted lines

> Chuck Lever <cel@citi.umich.edu> writes:
> 
> > i noticed while testing the new cache API that there are no tests under 
> > t/ for git-merge-index.
> 
> In the past, when I did any major butchering of existing code or
> made nontrivial additions, I wrote test scripts to cover what
> should happen (and what to be preserved) first to make sure the
> changes or additions would not introduce regression.  I think we
> have been reasonably successful (Daniel could fill us in with
> experiences with read-tree tests).

I think the only thing missing from the read-tree tests was for refusing to create directory/file conflicts which would otherwise be permitted when not using the --emu23 flag. Other than that, everything I neglected to handle caused something to fail. That reminds me that I still owe you a replacement test for this case.

> I have been expecting people to follow suit, without explicitly
> asking them to do, so when I took over the project.  Further I
> got sloppy when accepting any sizable patches.  Sorry about
> that.

I think I also owe you some tests for fetch. Any suggestions on writing unit tests? (I'd like to test that fetch.c actually calls functions in the right order and such, even in complicated cases)

	-Daniel
*This .sig left intentionally blank*