On Mar 20, 2010, at 11:42 AM, Junio C Hamano wrote:
Brian Gernhardt [off-list ref] writes:
quoted
Several tests did not use test_expect_success for their setup
commands. Putting these start commands into the testing framework
means both that errors during setup will be caught quickly and that
non-error text will be suppressed without -v.
Nice. Should the one in 4103 be three separate tests? They seem to be
preparing the set of test input files and are not expected to fail (unless
of course you are futzing with diff-tree, but for that we have t4013).
I don't think they need to be. t4103 was the first one I did and I did it as three tests since it was three separate "blocks". I had also tried moving the diff-trees closer to where their result is used until I saw how much they were reused.
~~ Brian