[Bioperl-l] Re: bl2seq reader chokes when no results (PR#927)

Hilmar Lapp lapp@gnf.org
Wed, 25 Apr 2001 09:50:00 -0700


Jeffrey Chang wrote:
>
> However, I
> believe parsing code should be separate from code that diagnoses
> unspecified faults in the input file.  One of the reasons is because there
> is only one interpretation of a correct file, but possibly many
> interpretations of ambiguously incorrect ones.  This allows people to use
> different interpretations.
> 
> Keeping things separate does impose some overhead than if it were tightly
> coupled.  But, you always have the option of not using the extra code if
> you're not worried about faults in the output.  If the diagnostic code
> were integrated, you wouldn't have that option.
> 

I definitely agree. Sane but still powerful approach. Would cause some
extra work in module design though, but definitely worth it.

	Hilmar
-- 
-------------------------------------------------------------
Hilmar Lapp                            email: lapp@gnf.org
GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
-------------------------------------------------------------