[Bioperl-l] Please test

[Jason Stajich]

On Thu, 10 Oct 2002, Shailesh L Mistry wrote:

>
>
>
> Failed Test     Stat Wstat Total Fail  Failed  List of Failed
> -------------------------------------------------------------
> t\BioGraphics.t               14    3  21.43%  1-3
> t\RootIO.t                    14    2  14.29%  13-14
> t\SearchIO.t                 658    4   0.61%  19 317 342 347
> t\consed.t                    15    2  13.33%  12 14
> t\scf.t          255 65280    15   14  93.33%  2-15
> 4 subtests skipped.
>
>
> 1) t\BioGraphics................FAILED tests 1-3
>
> inside biographics.t, function do_compare is checking the header
> of the incoming files. According to the PNG spec, the header
> should be \211PNG\r\n\032\n. I am using ptkdb to debug and I can
> not find a way to check that the value in $test_data matches
> this. However, I can hexdump the png files in the data folder and
> it has a header of \211PNG\r\r\n\032\r\n.
>
> Can someone please confirm the header of the png files in the data
> folder on unix. Does anyone know how to get ptkdb/some other debugger
> to show the hex value in $test_data?
>

This is windows linefeeds causing problems because the files may not have
been checked in/out as binary mode.

Lincoln - when we add binary files we need to do

% cvs add -kb filename.png

I've back added this tag by doing

% cvs admin -kb t1 t2 t3

Can you try a fresh checkout and see how this look.  Make sure that the
option has been applied by doing

[jason@sonogno biographics]$ cvs status t3/version3.png
===================================================================
File: version3.png      Status: Up-to-date

   Working revision:    1.1
   Repository revision: 1.1
/home/repository/bioperl/bioperl-live/t/data/biographics/t3/version3.png,v
   Sticky Tag:          (none)
   Sticky Date:         (none)
   Sticky Options:      -kb
                        ^^^
                        Make sure this is there, it it isn't the anon cvs
hasn't synced or my cvs admin cmd didn't work.


>
>
> 2) t\RootIO.....................FAILED tests 13-14
> The value in $iotest->mode is 'r' but we test for 'w'.
> This is just failing now so something may have changed recently.
> I will dig deeper later.
>
>
> 3) t\SearchIO...................FAILED tests 19, 317, 342, 347
> These fail because we 'got:1.46134e-90' but 'expected:1.46134e-090'.
> This could be a WinNT vs unix problem or even something deeper.
> However, the answer is still correct and we could make the test check
> for both values. Any thoughts....?

This is just activestate perl/windows being stupid about how it pads
numbers. We'll have to figure out a better way to generate the numbers but
this just a perl problem not in our code since we're getting the right
number.  can't test for numeric equality without using Math::BigFloat
which I don't want to introduce here for something so simple.

>
>
> 4) t\consed.....................FAILED tests 12, 14
> Test 12 'got:114' but 'expected:179'
> Test 14 'got:179' but 'expected:114'
> I need to dig into this deeper, more news later.
>

I think at one point chad never reported to have t/consed work on non-unix
- I may have fixed the OS level grep he was using a while ago, may be more
things to check.

>
> 5) t\scf........................dubious DIED. FAILED tests 2-15
> I need to dig into this deeper, more news later.

Same here.  Could be some dynamic loading going on too, can't remember how
this test was setup.

> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l@bioperl.org
> http://bioperl.org/mailman/listinfo/bioperl-l
>

-- 
Jason Stajich
Duke University
jason at cgt.mc.duke.edu