[Bioperl-l] Tests

Chris Fields cjfields at uiuc.edu
Mon Jun 26 11:45:51 EDT 2006


My opinion: tests should cover methods and expected results and are based on
what the module actually accomplishes.  Some classes (like SeqIO, SearchIO)
are normally relatively easy to build tests for b/c the expected results are
in the file being parsed.  Tests which check calculated results from modules
(Bio::Align::DNAStatictics for instance) I would think are trickier since
you should confirm the calculations are correct through independent means.

Links:

http://www.bioperl.org/wiki/Advanced_BioPerl#Designing_Good_Tests

http://search.cpan.org/~mschwern/Test-Simple-0.62/lib/Test/Tutorial.pod

The link above uses Test::Simple or Test::More; we use Test (but have
considered moving to Test::More using Devel::Cover).

My 2c

Chris

> -----Original Message-----
> From: bioperl-l-bounces at lists.open-bio.org [mailto:bioperl-l-
> bounces at lists.open-bio.org] On Behalf Of Sendu Bala
> Sent: Monday, June 26, 2006 9:45 AM
> To: bioperl-l at lists.open-bio.org
> Subject: [Bioperl-l] Tests
> 
> What level of testing is expected to be done in a test file? Is there
> such a thing as too many tests? Tests for every possible (documented)
> way of achieving a result with a module's method? Tests for every
> conceivable way of misusing a method?
> 
> If I come across a test for a module that doesn't test for everything
> the module can do, should I add tests as a matter of course? Would this
> be beneficial, or a waste of time (given that the module probably is
> bug-free already)?
> 
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/bioperl-l



More information about the Bioperl-l mailing list