[Patches] [ python-Patches-543867 ] test for patch #543865 & others
SourceForge.net
noreply at sourceforge.net
Tue Aug 5 10:02:03 EDT 2003
Patches item #543867, was opened at 2002-04-15 00:18
Message generated for change (Comment added) made by doerwalter
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=305470&aid=543867&group_id=5470
Category: Tests
Group: Python 2.4
>Status: Closed
>Resolution: Fixed
Priority: 5
Submitted By: Hernan Martinez Foffani (hfoffani)
Assigned to: Walter Dörwald (doerwalter)
Summary: test for patch #543865 & others
Initial Comment:
Here are 3 patches for:
- test_complex.py:
. add several checks to force execution of
unvisited
parts of complexobject.c code.
. add a test for complex floor division corresponding
bug #543387 and fix #543865
- test_complex_future.py
. add test for "future" true division.
(actually this is not a patch but the hole file)
- test_b1.py
. add test for bug #543840 and it's fix at patch
#543865
Regards,
-Hernan
----------------------------------------------------------------------
>Comment By: Walter Dörwald (doerwalter)
Date: 2003-08-05 18:01
Message:
Logged In: YES
user_id=89016
Checked in as:
Lib/test/test_complex.py 1.13+1.14
and
Lib/test/test_complex.py 1.12.6.1
----------------------------------------------------------------------
Comment By: Hernan Martinez Foffani (hfoffani)
Date: 2003-07-18 10:39
Message:
Logged In: YES
user_id=112690
thank you guys, for taking up these.
sadly, i don't have linux available anymore.
i must tell you that the patch was the result of a self-
excercise to learn about CPython. don't be surprise if you
don't see mayor improvements in code coverage percentages.
at the time I was very happy if I saw just 1 new LOC on a
branch tested: it meant that I understood Python's internals
involved.
sorry for the trouble.
regards,
-hernan
----------------------------------------------------------------------
Comment By: Raymond Hettinger (rhettinger)
Date: 2003-07-18 04:42
Message:
Logged In: YES
user_id=80475
Since __floordiv__ is deprecated for complex, it should be
left out of the test. Otherwise, it is ready to load.
Because they've already tagged the Py2.3 release candidate,
am marking this as a Py2.4 addition.
----------------------------------------------------------------------
Comment By: Walter Dörwald (doerwalter)
Date: 2003-07-17 13:25
Message:
Logged In: YES
user_id=89016
OK here is a patch (diff4.txt) that merges
test_complex_future.py into test_complex.py.
----------------------------------------------------------------------
Comment By: Raymond Hettinger (rhettinger)
Date: 2003-07-15 23:11
Message:
Logged In: YES
user_id=80475
That's a much better plan.
----------------------------------------------------------------------
Comment By: Walter Dörwald (doerwalter)
Date: 2003-07-15 20:58
Message:
Logged In: YES
user_id=89016
test_complex.py is checked in as rev 1.12.
test_complex_future.py looks good, although I don't exactly
understand its purpose, as there are tests from true
division and floor division in test_complex.py already. We
could extend check_div() in test_complex.py to try both
__truediv__ and __floordiv__. This would IMHO be better than
the code duplication in test_complex_future.py
----------------------------------------------------------------------
Comment By: Raymond Hettinger (rhettinger)
Date: 2003-07-15 08:58
Message:
Logged In: YES
user_id=80475
Converted test_complex_future.py to unittest format. See
attachment.
----------------------------------------------------------------------
Comment By: Raymond Hettinger (rhettinger)
Date: 2003-07-15 08:16
Message:
Logged In: YES
user_id=80475
diff3.txt looks good and runs fine on my machine. I'm
surprised that it did not increase code coverage. The
comparisons of a ** 105 to itself should be commented as a
whitebox test of a special case code path.
----------------------------------------------------------------------
Comment By: Walter Dörwald (doerwalter)
Date: 2003-07-14 23:08
Message:
Logged In: YES
user_id=89016
Reworking test_complex.py.diff3 into the attached diff3.txt
I don't see any increase in code coverage for
complexobject.c (it stays at 91.84%) except for the file
writing test which brings the coverage up to 92.95%. The
current tests in test_complex.py seem to cover most test
cases from test_complex.py.diff3 (except that
test_complex.py.diff3 uses ** and % and the current
test_complex.py uses pow and __mod__). So I'd say we add the
file writing test and drop the rest.
I'll look at test_complex_future.py tomorrow.
----------------------------------------------------------------------
Comment By: Walter Dörwald (doerwalter)
Date: 2003-07-14 10:53
Message:
Logged In: YES
user_id=89016
OK, I'll see if I can look at the patches later today.
----------------------------------------------------------------------
Comment By: Raymond Hettinger (rhettinger)
Date: 2003-07-14 09:49
Message:
Logged In: YES
user_id=80475
Walter, do you care to add these in unittest form?
Also, since the tests are to validate bug fixes, they are
appropriate to go into Py2.3.
----------------------------------------------------------------------
Comment By: Brett Cannon (bcannon)
Date: 2003-07-06 21:58
Message:
Logged In: YES
user_id=357491
The patch is now severely outdated since test_complex.py has
been converted to PyUnit, but it might be pertinent to go through
the patch and see if any tests are there that could be added.
----------------------------------------------------------------------
Comment By: Hernan Martinez Foffani (hfoffani)
Date: 2002-04-22 20:38
Message:
Logged In: YES
user_id=112690
Regarding "the error msg for complex pow says "remainder";
it shouldn't" you are correct, the exception string has a
bad wording.
----------------------------------------------------------------------
Comment By: Hernan Martinez Foffani (hfoffani)
Date: 2002-04-22 20:31
Message:
Logged In: YES
user_id=112690
On:
vereq(a ** 105, a ** 105) ... etc ...
The c code in complexobject.c has special cases when the
exponent is > 100, < than -100, and in-between.
I didn't want to test for equality with constants to avoid
messing up with floating point issues.
----------------------------------------------------------------------
Comment By: Tim Peters (tim_one)
Date: 2002-04-22 20:21
Message:
Logged In: YES
user_id=31435
I'm not sure what lines like
vereq(a ** 105, a ** 105)
vereq(b ** -105, b ** -105)
vereq(b ** -30, b ** -30)
are trying to test. That we get the same answer when we do
exactly the same thing twice?
Note that complex % has been deprecated: no point adding a
test for a deprecated feature.
The error msg for complex pow says "remainder"; it
shouldn't.
----------------------------------------------------------------------
Comment By: Guido van Rossum (gvanrossum)
Date: 2002-04-22 20:09
Message:
Logged In: YES
user_id=6380
OK, I've deleted them for you. Who do you expect to review
this?
----------------------------------------------------------------------
Comment By: Hernan Martinez Foffani (hfoffani)
Date: 2002-04-22 20:02
Message:
Logged In: YES
user_id=112690
Yes to both questions. I'm withdrawing test_complex.py
and test_b1.py.
I can't delete them and I double checked that I were
correctly logged in as hfoffani.
SourceForge error:
File Delete: ArtifactFile: Permission Denied
----------------------------------------------------------------------
Comment By: Guido van Rossum (gvanrossum)
Date: 2002-04-22 19:51
Message:
Logged In: YES
user_id=6380
I don't understand your comment. Are you withdrawing the
files test_complex.py and test_b1.py? Have you uploaded
these to separate patch issues? You should be able to delete
them as the original submitter; ifthis doesn't work, let me
know and I'll do it.
----------------------------------------------------------------------
Comment By: Hernan Martinez Foffani (hfoffani)
Date: 2002-04-15 01:48
Message:
Logged In: YES
user_id=112690
Following Tim's advise to group together bug/fix/test, I'll
leave this patch entry for improvements in the tests of
complex numbers.
Then the valid files are:
21173: test_complex_future.py
and
21180: test_complex.diff3
----------------------------------------------------------------------
Comment By: Hernan Martinez Foffani (hfoffani)
Date: 2002-04-15 01:47
Message:
Logged In: YES
user_id=112690
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=305470&aid=543867&group_id=5470
More information about the Patches
mailing list