[Bioperl-l] possible bug printing GenBank feature qualfiers

[Torsten Seemann]

Hilmar Lapp wrote:
> The only problem with Heikki's version of the line is that if the
> value is undefined you get a (ugly) warning from perl stating that you
> printed an undefined value. Since normally tags should always have a
> value (even if an empty string) this is a rather theoretical issue.

I thought it would be worth mentioning here that tags like /pseudo,
which don't have a value, quoted or otherwise, can be set using the magic 
string value "_no_value", when using the older tag system, eg.

	$feature->add_tag_value('pseudo', '_no_value');

Perhaps the 'undef' value could be used to replace '_no_value' cf. NULL in SQL...

I'm not sure how this works with the new Bio::Annotation* approach, which uses 
typed objects rather than strings. And i'm also not sure if the "_no_value" 
business works with the GFF and other modules.

-- 
Torsten Seemann
Victorian Bioinformatics Consortium, Monash University, Australia
http://www.vicbioinformatics.com/