[Bioperl-l] VariationIO and Tempfile tests failing

[Jason Stajich]

On Wed, 25 Sep 2002, Hilmar Lapp wrote:

> Failed Test      Stat Wstat Total Fail  Failed  List of Failed
> -------------------------------------------------------------------------------
> t/Tempfile.t                    8    2  25.00%  3 6
> t/Variation_IO.t               25    3  12.00%  5 15 25
>
> I have no clue what's going wrong in Tempfile all of a sudden - did
> someone tamper with anything that may be involved?
>
(Pls test before checking in changes, especially when working with a
toplevel Bio::Root object!)


line 237 Bio::Root::Root was:
  return unless ref $self ne 'HASH';

however this will return false as ref($self) is 'Bio::Root::IO', instead,
should be:
  return unless ref $self && $self->isa('HASH');


The whole reason for this test is that Lincoln has
objects which use arrays underneath, just the implementation is flawed.
This works on perl 5.6.1 but we should test

my $hash = {};
print $hash->isa('HASH'), "\n";

on earlier perl versions.  can someone report back?

I am less worried about the underlying hash assumption now because
non-hash implmentations only inherit from Bio::Root::RootI and not
Bio::Root::Root.  Note to self/someone with more time: we need to include
documention of how to implement objects which inherit from
Bio::Root::RootI (in the RootI docs).  There are now about 5 methods you
have to implement (verbose,_cleanup_methods, _register_for_cleanup, etc)



> ChrisZ, I fixed Term/GOterm and the tests such that setting
> attributes to an empty string amounts to setting to undef. undef
> means 'not set' or 'not known', whereas an empty string may be a
> valid value (although it can't be sensibly most of the time). Also,
> MySQL takes empty strings as valid values and subsequently raises UK
> violations.
>
> 	-hilmar
> --
> -------------------------------------------------------------
> Hilmar Lapp                            email: lapp at gnf.org
> GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
> -------------------------------------------------------------
>
> _______________________________________________
> 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