[ansi-test-devel] Wrong test?

[Ville Voutilainen]

On 26 December 2011 18:36, Juan Jose Garcia-Ripoll
<juanjose.garciaripoll at googlemail.com> wrote:
> The test is named CTYPECASE.14 in the test suite, so I doubt that it is
> actually checking MACROLET, but rather how CTYPECASE treats the evaluated
> form, and which is the active environment in which it is evaluated.
> Otherwise the test could have been simply a TYPECASE with an otherwise
> clause.

Perhaps so; I don't know the history and evolution of the tests, so I
can't say what that
test is actually trying to accomplish. Your explanation makes sense,
perhaps it's there
to ascertain that CTYPECASE performs no funny business with a macro, that is, it
checks that the expansion works correctly and that CTYPECASE uses the current
environment instead of some other environment.

I know for a fact that we had errors in abcl related to such
environment handling
in macros, and the ansi-tests helped weed out such problems. To me, it
still looks
like this case doesn't even purport to raise a correctable error, but
the intent is
that the CTYPECASE evaluates to :GOOD if the implementation does everything
properly.