Sun 03 Mar 2013 08:18:55 PM UTC, comment #10:
Also, if you run test stemleaf
from command, it works correctly.
Only, make check fails
|
Sun 03 Mar 2013 08:02:54 PM UTC, comment #9:
Yes,
stemleaf ([0 0.5 1], "")
generates the warning and then does the right
thing. So, I think you are right.
|
Sun 03 Mar 2013 07:59:05 PM UTC, comment #8:
I think the problem is with the warning system. See bug #35850. I tried to fix 35850 and was finding that the %!warning tests would very often print out results of lastwarn() rather than the current warning. This screws up the test system which in this case is expecting "X truncated to integer values" but is getting the string from the last time warning() was called which was probably from a print routine.
Just to be certain that the code itself is working one should try
from the Octave prompt and verify that it generates a warning.
|
Sun 03 Mar 2013 05:42:13 PM UTC, comment #7:
I built the latest dev with updated
stemleaf. make check fails as Philip reported.
But, if I run ./run-octave --no-gui
test stemleaf
it passes all 17 tests.
Any ideas as to what is wrong?
|
Sun 03 Mar 2013 03:26:30 PM UTC, comment #6:
Rik,
This all looks great. Thanks, again, for the
cleanup of the code. As usual, I missed a number of
items.
Any idea what causes the "epstool" error reported
by Philip below?
|
Sun 03 Mar 2013 02:55:48 PM UTC, comment #5:
I applied your changeset here (http://hg.savannah.gnu.org/hgweb/octave/rev/3e6d15a2a50b). I also did some cleanup for Octave coding conventions, like space between function name and '(', and removed unused variables left over from previous versions and checked that change in here (http://hg.savannah.gnu.org/hgweb/octave/rev/0f0e970723ec).
|
Sun 03 Mar 2013 02:49:58 PM UTC, comment #4:
This is a mystery. The problem does not come
directly from stemleaf. There is nothing in stemleaf
that leads to a reference to epstool. Also, I could
not find anything in the assert code that causes this
(but, I suspect that it is there). I removed epstool
from my system, but this does not induce the error.
Since Rik provided that test, maybe he can provide a
clue.
If you have a minute you might try removing the "tmp ="
from the !warning... at line 598. This is nothing more than
a guess.
|
Sun 03 Mar 2013 10:32:52 AM UTC, comment #3:
[philip@deskprn octave]$ hg summary
parent: 16180:3e6d15a2a50b tip
Update stemleaf to provide new features and minor fix to printd
Mageia 2 Linux
make check gives:
fntests.log:
Could probing for epstool help with polishing stemleaf's output?
|
Sun 03 Mar 2013 03:21:43 AM UTC, comment #2:
To clarify: file #27554 replaces #27553
|
Sat 02 Mar 2013 11:30:26 PM UTC, comment #1:
Minor typo in patch, correction below.
Sorry,
Michael
(file #27554)
|
Sat 02 Mar 2013 11:11:56 PM UTC, original submission:
After several interchanges with Rik I and he have implemented
many enhancements and corrections to stemleaf.
I have also included in this patch minor changes to printd
which improve the documentation and prevent the tests
for printd from writing output.
|