[MOBY-dev] findService and the hierarchy

[Edward Kawas]

Currently, the flag only works on inputs.

Eddie

-----Original Message-----
From: moby-dev-bounces at lists.open-bio.org
[mailto:moby-dev-bounces at lists.open-bio.org] On Behalf Of Andreas Groscurth
Sent: August-05-08 6:07 AM
To: Core developer announcements
Subject: Re: [MOBY-dev] findService and the hierarchy

Hi,

oh my god... i mixed inputs and outputs... yes there is no service which 
consumes Annotated_GO_Term... there are only services which returns it.

So the find service call only checks for the inputs then ? Or is it 
possible to find a service which consumes an object with a specific 
namespace and returns an Annotated_GO_Term object or its parent?
currently i ask for each output definition by its own.

Sorry again for the confusion ....

Andreas

Edward Kawas wrote:
> Hi Andreas,
>
> I believe that the results are correct.
>
> There are no services in the default registry that consume
> Annotated_GO_Term. 1 service that consumes its parent (GO_Term) and 689
> services that strictly consume Object.
>
> When you set the flag, includeParentDatatype, the behaviour is like doing
a
> findservice for services that consume Annotated_GO_Term or any parent type
> up to and including Object with includeParentDatatype=false. 
>
> Does this make sense?
>
> Eddie
>
> -----Original Message-----
> From: moby-dev-bounces at lists.open-bio.org
> [mailto:moby-dev-bounces at lists.open-bio.org] On Behalf Of Andreas
Groscurth
> Sent: August-03-08 11:33 PM
> To: Core developer announcements
> Subject: Re: [MOBY-dev] findService and the hierarchy
>
> Martin Senger wrote:
>
>   
>> I posted my results - now it is up to Eddie or Mark to draw
>> conclusion from that results... 
>>     
>
>
> could any of you two state something about that ?
>
> Thanks
> Andreas
>
> Andreas Groscurth wrote:
>   
>> i like to bring this up for discussion again ;-)
>>
>> thanks
>>
>> Martin Senger wrote:
>>     
>>>> In my opinion when searching for services consuming 
>>>> Annotated_GO_Term via
>>>> the findService method i should also include the service which 
>>>> consumes a
>>>> GO_Term datatype, as this is the parent datatype (per default
>>>> 'includeParentDatatype' is true)....
>>>>
>>>>     
>>>>         
>>> I assume that by "I should also include" you actually meant "the 
>>> registry
>>> response should also include". Is my understanding correct?
>>>
>>> If so, I am afraid that it is not the issue of the jMoby library but the
>>> registry itself. Eddie or Mark can confirm or reject this hypothesis.
>>>
>>> The BioMoby API allows to specify two XML tags when calling findService
>>> method: the
>>>
>>> <expandObjects>X</expandObjects>
>>> and
>>> <expandServices>X</expandServices>
>>>
>>> Where X can be either 1 or 0.
>>>
>>> The jMoby's findService() method sets these two tags according to the 
>>> two
>>> boolean parameters that the method can get:
>>>
>>> boolean includeChildrenServiceTypes
>>> and
>>> boolean includeParentDataTypes
>>>
>>> I have tried to call the findService, with the Annotated_GO_Term input,
>>> using all combinations of these two tags (you can do it using the Moby
>>> command-line client with the optioins -nc and/or -np). The results are
>>> either no service or 713 services:
>>>
>>> expandObjects     expandServices    # of found services
>>> 1                 1                 713
>>> 0                 0                 0
>>> 1                 0                 713
>>> 0                 0                 0
>>>
>>> I have feeling that this is not a correct result. Could Eddie or Mark
>>> perhaps comment on it?
>>>
>>> [I vaguely remember that we had this conversation about these tags 
>>> with Mark
>>> years(!) ago - but I do not recall what was the conclusion. I only, and
>>> again vaguely, remember to have this issue in my to-do list - which 
>>> would
>>> indicate that something could be done in the jMoby library.]
>>>
>>> Cheers,
>>> Martin
>>>
>>>   
>>>       
>>     
>
>
>   


-- 
=====================================================
  Dipl. Bioinf. Andreas Groscurth
  Bioinformatics Software Developer
  Plant Computational Biology group
  Max-Planck Institute for plant breeding research
  Carl-von-Linne Weg 10
  50829 Cologne
  Germany
  +49(0) 221 5062449
=====================================================

_______________________________________________
MOBY-dev mailing list
MOBY-dev at lists.open-bio.org
http://lists.open-bio.org/mailman/listinfo/moby-dev