(
Jump to the original submission)
Thu 28 Feb 2013 04:42:36 PM UTC, comment #10:
Thanks, I merged your addendum:
http://hg.savannah.gnu.org/hgweb/octave/graph/f279fd604c7e
|
Thu 28 Feb 2013 08:03:10 AM UTC, comment #9:
Oops, sorry... here is the patch
(file #27541)
|
Thu 28 Feb 2013 08:02:42 AM UTC, comment #8:
Here is a patch for the bug number in classes.tst.
|
Wed 27 Feb 2013 08:20:51 PM UTC, comment #7:
I have pushed and merged your change:
http://hg.savannah.gnu.org/hgweb/octave/graph/c6b2409672e9
I noticed your commit message doesn't mention the bug number. I had to hunt around to find it. :-/
The bug number you refer to in classes.tst (38170) also seems to be wrong and is pointing at an unrelated problem.
Can you please fix that?
|
Tue 26 Feb 2013 01:45:03 PM UTC, comment #6:
Here is a minor update of my patch (rebased and tested).
(file #27533)
|
Thu 21 Feb 2013 03:56:07 PM UTC, comment #5:
I just found out that the same problem was already exhibited (although not advertised in the comments) by the example file #24630 attached to bug #35118.
|
Mon 11 Feb 2013 07:57:06 PM UTC, comment #4:
Here is the same patch with test_classes.m renamed to classes.tst.
(file #27444)
|
Fri 08 Feb 2013 07:57:27 AM UTC, comment #3:
A few tests concerning classes and concatenation are located test/class-concat (added by JWE in Nov 2011).
In the patch that I just posted, I have added my own tests in test/classes/test_classes.m.
Should I move my concatenation tests into test/class-concat, or rather merge the old tests from test/class-concat into test/classes/test_classes.m ?
[Note: this was originally discussed on the maintainer mailing-list, http://octave.1599824.n4.nabble.com/Tracking-bug-38128-Overloaded-vertcat-not-called-in-some-cases-td4649193.html]
|
Fri 08 Feb 2013 07:51:49 AM UTC, comment #2:
The attached patch seems to fix the problem.
I have added a few tests in test/classes/test_classes.m, to check that vertical and horizontal concatenation still work in the case where nothing is overloaded.
[Note: this was initially posted on the patch tracker, see patch #7935. Mea culpa.]
(file #27422)
|
Sat 26 Jan 2013 01:14:17 PM UTC, comment #1:
horzcat() is affected by the exact same problem
|
Sun 20 Jan 2013 09:35:29 PM UTC, original submission:
Let us consider a class @A, with the following constructor:
and the following overload vertcat function:
The following script illustrates the problem:
It seems that the overloaded vertcat function is only called when both arguments are class A objects.
I would expect it to be called as soon as one of the two arguments is of class A (this is what happens, for instance, with an overloaded plus() function).
The problem has been observed in Octave 3.6.2 (Ubuntu 12.10 install) and 3.6.4-rc1 (built from branch stable, revision 7ad3eea8a3af).
Attachment: test script + @A class directory.
|
(Note: upload size limit is set to 16384 kB, after insertion of the required
escape characters.)
Attach File(s):
Comment:
Depends on the following items: None found
Items that depend on this one: None found
Follow 8 latest changes.
| Date |
Changed By |
Updated Field |
Previous Value |
=> |
Replaced By |
| Thu 28 Feb 2013 08:03:10 AM UTC | jbect | Attached File | - |  | Added jbFixObjConcatProblem3Addendum.patch, #27541 |
| Wed 27 Feb 2013 08:20:51 PM UTC | jordigh | Status | Patch Submitted |  | Fixed |
| | | Open/Closed | Open |  | Closed |
| Tue 26 Feb 2013 01:45:03 PM UTC | jbect | Attached File | - |  | Added jbFixObjConcatProblem3.patch, #27533 |
| Mon 11 Feb 2013 07:57:06 PM UTC | jbect | Attached File | - |  | Added jbFixObjConcatProblem2.patch, #27444 |
| Fri 08 Feb 2013 02:09:32 PM UTC | jordigh | Status | None |  | Patch Submitted |
| Fri 08 Feb 2013 07:51:49 AM UTC | jbect | Attached File | - |  | Added jbFixObjConcatProblem.patch, #27422 |
| Sun 20 Jan 2013 09:35:29 PM UTC | None | Attached File | - |  | Added octaveVertcatIssue.zip, #27305 |