➜
This issue tracker has been migrated to GitHub,
and is currently read-only.
For more information,
see the GitHub FAQs in the Python's Developer Guide.
Created on 2013-01-11 14:13 by stephane.poirier, last changed 2022-04-11 14:57 by admin. This issue is now closed.
| Files | ||||
|---|---|---|---|---|
| File name | Uploaded | Description | Edit | |
| testQSort33.py | stephane.poirier, 2013-01-11 14:13 | |||
| Messages (3) | |||
|---|---|---|---|
| msg179680 - (view) | Author: Stephane Poirier (stephane.poirier) | Date: 2013-01-11 14:13 | |
qsort doesn't sort arrays that are defined with double elements. You can find attached a script that shows this function working well for int (as described in the ctypes documentation). If int array is replaced by double array qsort doesn't sort any more. Even worse, with python 2.6, the difference between a[0] and b[0] may be nan. The script can be launch by python.exe testQSort33.py |
|||
| msg179683 - (view) | Author: Mark Dickinson (mark.dickinson) *  |
Date: 2013-01-11 14:53 | |
What happens if you replace the line:
CMPFUNC = CFUNCTYPE(c_double, POINTER(c_double), POINTER(c_double))
in your example code with
CMPFUNC = CFUNCTYPE(c_int, POINTER(c_double), POINTER(c_double))
?
|
|||
| msg179685 - (view) | Author: Mark Dickinson (mark.dickinson) * |
Date: 2013-01-11 14:57 | |
Closing as invalid: the issue is not in Python but with the CMPFUNC line. Your py_cmp_func should return an integer rather than a float to be compatible with libc.qsort. |
|||
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2022-04-11 14:57:40 | admin | set | github: 61138 |
| 2013-01-11 14:57:43 | mark.dickinson | set | status: open -> closed resolution: not a bug messages: + msg179685 |
| 2013-01-11 14:53:53 | mark.dickinson | set | nosy:
+ mark.dickinson messages: + msg179683 |
| 2013-01-11 14:24:18 | serhiy.storchaka | set | nosy:
+ amaury.forgeotdarc, belopolsky, meador.inge |
| 2013-01-11 14:13:56 | stephane.poirier | create | |