testrunner option to disable -1

[Fred Drake]

When used with zc.buildout and the zc.recipe.testrunner recipe, it's generally reasonable to include the "-1" option in the "defaults" setting. It would be good to have an option that turns off the "-1" effects, so that the few times when re-enabling the extra reporting makes sense, it can be done, and treated as the special case that it is.

[Benji York (benji)]

If it were a lower case letter option; I'd suggest upper case to turn it off. By that line of reasoning -! would be my suggestion, but that's kind of silly.

Perhaps instead we add --first-failure as an alias of -1 and --no-first-failure or --all-failures as its opposite.

[Fred Drake (fdrake)]

On 5/18/07, Benji York <email address hidden> wrote:
> Perhaps instead we add --first-failure as an alias of -1 and --no-first-
> failure or --all-failures as its opposite.

Something like that would work for me, though we should be especially
careful about the names. There are enough people that find the
semantics of -1 confusing as it stands; it's very specific to doctest
behavior.

[Benji York (benji)]

Fred Drake wrote:
> On 5/18/07, Benji York <email address hidden> wrote:
>> Perhaps instead we add --first-failure as an alias of -1 and --no-first-
>> failure or --all-failures as its opposite.
>
> Something like that would work for me, though we should be especially
> careful about the names. There are enough people that find the
> semantics of -1 confusing as it stands; it's very specific to doctest
> behavior.

How about --first-doctest-failure and --all-doctest-failures; kinda long
though.
--
Benji York
Senior Software Engineer
Zope Corporation

[Fred Drake (fdrake)]

On 5/18/07, Benji York <email address hidden> wrote:
> How about --first-doctest-failure and --all-doctest-failures; kinda long
> though.

Kinda long, and not completely clear still. Since this is about
reporting, and non-doctest tests already do what this is about making
doctest tests do, perhaps

    --hide-secondary-failures

and

    --show-secondary-failures

would be good? Still long, but optparse only requires enough of that
to be unambiguously identifiable.

  -Fred

--
Fred L. Drake, Jr. <fdrake at gmail.com>
"Chaos is the score upon which reality is written." --Henry Miller

[Benji York (benji)]

--hide-secondary-failures and --show-secondary-failures sound good to me. As an upside, non doctest tests that have yet to be invented could use those options as well.

[Fred Drake (fdrake)]

Added support for --hide-secondary-failures and --show-secondary-failures in zope.testing trunk, revision 75845.

[Wolfgang Schnerring (wosc)]

has long been released