[Scilab-Dev] Bug No- 5466 Advice needed

[Vincent COUVERT]

Hi,

You are right.

I fixed this bug a few weeks ago in the 5.2 branch. I have just added a 
non-regression test and add an item in the changes file.

Best regards.

On 28/03/2010 14:01, shankhs ch wrote:
> HI
> I reproduced the bug using the program given in the 5466 bug page 
> http://bugzilla.scilab.org/show_bug.cgi?id=5466 :
> A = sprand(400,400,0.01);
> B = sprand(100,100,0.1);
> unix('rm test_matfile.mat'); // be sure not file like this one exists
> savematfile('test_matfile.mat','A','B');
> clear
> loadmatfile('test_matfile.mat','A','B');
> size(A)
> size(B)
> clear
> A = rand(10,10);
> savematfile('test_matfile.mat','A');
> clear
> loadmatfile('test_matfile.mat','A');
> size(A) // It should display a size like 400 x 400 instead of 10 x 10
> The output is indeed 400. 400. But the logic says that it should be 
> 10. 10. isn't it?
> I think I have identified the problem which is :
> "deletion"/"creation of another file with the same name( 
> test_matfile.mat )" fails once the file is "created"/"if present in 
> the directory as a result of previous execution" using the *unix('rm 
> test_matfile.mat'); *command.
> Moreover the file is getting appended instead of getting created fresh 
> everytime which is causing the complained behavior.To check this, run 
> the above program and execute a number of times , check the filesize 
> after every execution.
> I wanted to write this one in the bug page but since this is my first 
> time and I am new to scilab so I thought its better to ask you guys.
> Any thoughts and advices?
> Thanks and regards
> shankhs

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.scilab.org/pipermail/dev/attachments/20100329/1c10bbb7/attachment.htm>