fix an error message and a comment in _testcapimodule.c #392
Conversation
|
thanks |
| @@ -1205,7 +1205,8 @@ test_k_code(PyObject *self) | |||
| value = PyLong_AsUnsignedLongMask(num); | |||
| if (value != (unsigned long)-0x42) | |||
| return raiseTestError("test_k_code", | |||
| "PyLong_AsUnsignedLongMask() returned wrong value for long 0xFFF...FFF"); | |||
| "PyLong_AsUnsignedLongMask() returned wrong " | |||
| "value for long -0xFFF..000042"); | |||
There was a problem hiding this comment.
0xFFF...000042 looks strange. Is this really the right value? (And what does the ... mean here?)
There was a problem hiding this comment.
first we have:
num = PyLong_FromString("-FFFFFFFF000000000000000042", NULL, 16);
and then the check is whether PyLong_AsUnsignedLongMask returned -0x42,
i.e. truncated num as expected.
so yes, IMHO -0xFFF..000042 is the right value.
and I guess the .. means a big enough amount of zeros, such that the
truncation would indeed result in -0x42.
ISTM the original author felt this would be more elegant than writing
-0xFFFFFFFF000000000000000042 again...
There was a problem hiding this comment.
Fair enough. Given that it's just an error message, and that that error message is at least unique enough to identify exactly what's failing, it's probably good enough. Sorry for the bikeshedding.
fix two mistakes in Modules/_testcapimodule.c:
(note that I mentioned these mistakes in bpo-27298, but only as a side note.)
Just in case, I ran the test module on my Windows 10, and as expected, it seems that this minor patch doesn't break anything.