[Bioperl-l] return undef

Chris Fields cjfields at uiuc.edu
Mon Nov 6 12:58:27 EST 2006


>> On a similar note, is it safe to simply substitute "use vars" for  
>> "our"
>> and correctly commenting the list of variables if qw() was  
>> previously used?
>
> Should be. But note that we didn't officially agree on having the Run
> package require Perl 5.6... Is there a particular person who maintains
> the Run package, or does that fall to the pumpkin of Core?

I believe Albert Vilella was taking care of bioperl-run, but I'm not  
absolutely sure.

I think we can assume that, if core rel. 1.5.2  requires perl 5.6.1,  
anything that requires bioperl rel. 1.5.2 should likewise require  
perl 5.6.1 as well.  Probably should make that explicitly stated  
somewhere.

>> If so, there are 340+ "use vars" in the bioperl-run......any wise  
>> ideas
>> on a generic script that can do an intelligent search and replace.
>> Bearing in mind that these lists can span multiple lines?
>
> It'll need a little futzing, just be very careful it is working
> perfectly before applying it to everything. All of Core and the other
> packages need the same work done on them...

I have gone through a few and manually switched 'use vars' over to  
'our'.  You can always try making the necessary changes globally but  
I think you'll see a lot of failed tests.

>> I set up the following page a couple of days ago to house some  
>> scripts
>> that can be used to do audit checks of code:
>> http://www.bioperl.org/wiki/Auditing
>>
>> If anyone has any scripts, Perl, shell or otherwise fell free to add
>> them! :-)
>
> I created a one-off for the conversion to 'use base', but its a bit of
> delirium-induced hack. I'll add it when I get the chance; its
> essentially a more complex version of what you want to do. It also  
> ought
> to be applied to all the other Bioperl packages...
>
> (You should make mention of the existing audit scripts in CVS for
> bioperl-live in maintenance/)

Not sure how much more global code auditing should be done before the  
1.5.2 final release, but if it passes tests then I don't have a problem.

Chris

Christopher Fields
Postdoctoral Researcher
Lab of Dr. Robert Switzer
Dept of Biochemistry
University of Illinois Urbana-Champaign





More information about the Bioperl-l mailing list