10.12.2021 17:25, Kevin Wolf wrote:
Am 06.12.2021 um 18:59 hat John Snow geschrieben:
On Fri, Dec 3, 2021 at 7:22 AM Vladimir Sementsov-Ogievskiy <
vsement...@virtuozzo.com> wrote:

We are going to use do_run_test() in multiprocessing environment, where
we'll not be able to change original runner object.

Happily, the only thing we change is that last_elapsed and it's simple
to do it in run_tests() instead. All other accesses to self in
do_runt_test() and in run_test() are read-only.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsement...@virtuozzo.com>
---
  tests/qemu-iotests/testrunner.py | 4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/tests/qemu-iotests/testrunner.py
b/tests/qemu-iotests/testrunner.py
index fa842252d3..a9f2feb58c 100644
--- a/tests/qemu-iotests/testrunner.py
+++ b/tests/qemu-iotests/testrunner.py
@@ -287,7 +287,6 @@ def do_run_test(self, test: str) -> TestResult:
                                diff=diff, casenotrun=casenotrun)
          else:
              f_bad.unlink()
-            self.last_elapsed.update(test, elapsed)
              return TestResult(status='pass', elapsed=elapsed,
                                casenotrun=casenotrun)

@@ -353,6 +352,9 @@ def run_tests(self, tests: List[str]) -> bool:
                      print('\n'.join(res.diff))
              elif res.status == 'not run':
                  notrun.append(name)
+            elif res.status == 'pass':
+                assert res.elapsed is not None
+                self.last_elapsed.update(t, res.elapsed)

              sys.stdout.flush()
              if res.interrupted:
--
2.31.1


(I continue to be annoyed by the "None" problem in mypy, but I suppose it
really can't be helped. Nothing for you to change with this patch or
series. I just wish we didn't need so many assertions ...)

I'm inclined to say it's a None problem in our code, not in mypy.
Essentially it comes from the fact that we're abusing a string
(res.status) and None values to distinguish different types of results
that have a different set of valid attributes.

Of course, Python already provides a language feature to distinguish
different types of results that have a different set of attributes and
that wouldn't run into this problem: subclasses.


Agree, you are right. I'll look if it is simple to refactor.

--
Best regards,
Vladimir

Reply via email to