[MOBY-l] Services accepting multiple inputs
Ola Spjuth
ola.spjuth at farmbio.uu.se
Thu Apr 3 12:23:49 UTC 2008
Andreas,
Thanks a lot for the explanation.
Cheers,
/Ola
On Apr 3, 2008, at 12:47 , Andreas Groscurth wrote:
> Ola Spjuth wrote:
>> I suppose it would be possible to set up another version of the
>> service that accepted a MobyDataComposite containing an arbitrary
>> number of wrapped MobyDataObject, and delivers a MobyDataComposite
>> as output, containing an arbitrary number of objects?
> no - because MobyDataComposite is the represent of complex data
> types, NOT for an arbtrary number of inputs.
>
> A service can accept one MobyDataObject (a single Object), one
> MobyDataComposite (a single complex type, like AminoAcidSequence)
> or more of them or a combination of them... but its not for using a
> collection of intputs just because it could be convenient to call a
> service with several different inputs !
>
> just use for each input object a differnent cal - that would
> mobyrequest have also to do... becaus a service which defines one
> input can not take more inputs...
>
>> This means you need 2 different services to accommodate for the
>> case of one input object and another for many inputs, and the
>> situation is left to service providers and not handled by Moby. Is
>> this a correct observation?
> Why 2 different services ? Either your service takes e.g. one GI to
> do something, or takes a collection... Of course its up to the
> service provider to know what his services does and therefore to
> offer one of either possibilities. And if the services takes a
> collection of input - of course you can set more than one input
> (like 3 GIs), BUT if the service is registered with only one input
> - you have to stick to its rules.
>
> Best
> andreas
>>
>> On Apr 3, 2008, at 10:47 , Andreas Groscurth wrote:
>>
>>
>>> [snip]
>>> maybe there is a way, as Mark apparently indicates, but i cant
>>> see how shall work....
>>>
>>
>> Yes, I am also curious to hear about that!
>>
>> Cheers,
>>
>> .../Ola
>>
>
More information about the moby-l
mailing list