[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: AW: Re: octave-forge loadpath
From: |
David Bateman |
Subject: |
Re: AW: Re: octave-forge loadpath |
Date: |
Fri, 12 Jan 2007 23:16:09 +0100 |
User-agent: |
Thunderbird 1.5.0.7 (X11/20060921) |
address@hidden wrote:
> Hi all again
>
> Thanks David for your help, the loadpath problem is solved.
>
> But the real reason for this mail is a problem with the "test"
> function
> In the octave documentation is written, that "test name" will take
> the first function named "name" which is found in the octave path
> ...
> I wrote some tests in bar.m (OctPlot - bar.m).
>
> the second entry in "path" is (the first one is ".")
> /usr/local/libexec/octave/2.9.9/site/octplot-0.3.99
> this is a directory which holds the m-files of octplot like bar.m
> but if I enter at the octave prompt "test bar" following error
> message appears:
> ????? /usr/local/share/octave/2.9.9/m/plot/bar.m has no tests
> available
> octave:8>
>
> of course, octplot is enabled ...
> Michael
>
The code segment in test.m that resolves the function name to a file is
__file = file_in_loadpath (__name);
if (isempty (__file))
__file = file_in_loadpath ([__name, ".m"]);
endif
if (isempty (__file))
__file = file_in_loadpath ([__name, ".cc"]);
endif
So it appears that the problem is in fact with the file isn't found in
the load path correctly. At first I didn't see this issue and then I
remove the recent fixes to parse.y(load_fcn_from_file) and the problem
came back. So I think this is a bug that would go away with a rebuild
from the CVS. I really hope its not too long till a 2.9.10 release now
as there are many changes since 2.9.9 that would be good to see released..
Cheers
David