You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by kf...@collab.net on 2004/08/06 15:12:05 UTC
Re: Test suite setting exit code to non-zero on skipped test
"Brian W. Fitzpatrick" <fi...@red-bean.com> writes:
> =========
> Proposal:
>
> Is there any reason why we should return non-zero just for skipping a
> test? I can't see any reason why we need to error if we skip something,
> but perhaps someone here has a good reason (svn blame tells me that
> kfogel wrote the line that sets skips to non-zero).
The situation is more complex than that. The _Predicate class in
testcase.py uses the error number as an index into an array of action
strings (see _Predicate.run_text(), and see the r8649 diff).
This can be fixed, but the fix is more than just a one-line change...
-Karl
> My suggested patch is:
>
> ---------------------------------------------------------------
> Index: testcase.py
> ===================================================================
> --- testcase.py (revision 10504)
> +++ testcase.py (working copy)
> @@ -107,7 +107,7 @@
> print ex.__doc__
> sys.exit(255)
> except svntest.Skip, ex:
> - error = 2
> + error = 0
> except svntest.Failure, ex:
> error = 1
> # We captured Failure and its subclasses. We don't want to print
> ---------------------------------------------------------------
>
> Thotz?
> -Fitz
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
> For additional commands, e-mail: dev-help@subversion.tigris.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Test suite setting exit code to non-zero on skipped test
Posted by "Brian W. Fitzpatrick" <fi...@red-bean.com>.
On Fri, 2004-08-06 at 10:12, kfogel@collab.net wrote:
> "Brian W. Fitzpatrick" <fi...@red-bean.com> writes:
> > =========
> > Proposal:
> >
> > Is there any reason why we should return non-zero just for skipping a
> > test? I can't see any reason why we need to error if we skip something,
> > but perhaps someone here has a good reason (svn blame tells me that
> > kfogel wrote the line that sets skips to non-zero).
>
> The situation is more complex than that. The _Predicate class in
> testcase.py uses the error number as an index into an array of action
> strings (see _Predicate.run_text(), and see the r8649 diff).
>
> This can be fixed, but the fix is more than just a one-line change...
Ook. Well in that case, Christian's patch should do the trick:
==============
diff -Naur cvs2svn-1.0.0-rc4-orig/svntest/main.py
cvs2svn-1.0.0-rc4/svntest/main.py
--- cvs2svn-1.0.0-rc4-orig/svntest/main.py 2004-06-23
06:43:38.000000000 +0200
+++ cvs2svn-1.0.0-rc4/svntest/main.py 2004-08-06 11:57:04.000000000
+0200
@@ -516,7 +516,10 @@
if testnum is None:
for n in range(1, len(test_list)):
- if run_one_test(n, test_list):
+ test_exit_code = run_one_test(n, test_list)
+ if test_exit_code == 2:
+ test_exit_code = 0 # ignore skipped test cases
+ if test_exit_code:
exit_code = 1
else:
exit_code = run_one_test(testnum, test_list)
==============
-Fitz
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org