[Bioperl-l] Wrapper definitions format standard (was: Bio::Tools::Run::WrapperBase: Capture the command thatis executed?)

[Mark A. Jensen]

Sounds like both Galaxy and ACD converters to the underlying CommandExts config 
format (built on exported global vars) would be useful and make the system very 
flexible; could then import any existing wrapper definitions, and use emboss 
acds to make quick BP wrappers on the fly for any emboss program, producing 
wrappers with the unified api that CommandExts provides.

----- Original Message ----- 
From: "Chris Fields" <cjfields at illinois.edu>
To: "Peter" <biopython at maubp.freeserve.co.uk>
Cc: "Mark A. Jensen" <maj at fortinbras.us>; "bioperl-l" 
<bioperl-l at lists.open-bio.org>; "Ben Bimber" <bimber at wisc.edu>
Sent: Monday, April 19, 2010 10:30 AM
Subject: Re: [Bioperl-l] Bio::Tools::Run::WrapperBase: Capture the command 
thatis executed?


On Apr 19, 2010, at 9:09 AM, Peter wrote:

> On Mon, Apr 19, 2010 at 2:03 PM, Mark A. Jensen <maj at fortinbras.us> wrote:
>> Hey Ben,
>> Yes, the new stuff is in the bioperl-dev/trunk
>> svn://code.open-bio.org/bioperl/bioperl-dev/trunk/Bio/Tools/Run
>> (code.open-bio.org was working as of this morning...)
>>
>> This is all pretty experimental, so use at own risk. The XML-based
>> config system is experimental too at the moment (tho' all code
>> is available, in bioperl-dev/trunk/Bio/Tools). I've been getting
>> good feedback from other Bio* peeps, who are advocating using an
>> existing config standard like emboss ACD or the Galaxy format.
>
> The EMBOSS ACD format is text based and as such rather niche.
> I was more suggesting the ability to parse the EMBOSS ACD files
> into the chosen XML format would be a neat way to automatically
> generate wrappers for all the EMBOSS tools at once.
>
> The Galaxy workflow command line tool description is XML and,
> might be suitably general. There are likely others out there too
> (no sense reinventing the wheel if possible).

...

The discussion we've been having with SoC on this (I think Brad Chapman's 
comments in particular) make me think this isn't a completely resolved issue, 
possibly b/c deriving a generic system is either extremely hard, not 
interesting, or simply that no one has thought of it yet.  Galaxy XML is the 
closest I can think of to a standard that might work.  I agree completely re: 
comments on ACD.

chris