Thank you all. Indeed, the component was not 'immediate', and that's
why it wasn't set up.

My task of the moment it to add pax-exam tests for some individual
components; in the past, I only tested this stuff with broad
functional tests. This has led me, on this particular occasion,  end
up trying to test this not-immediate component.

How do other people create pax-exam tests for @Components that are not
immediate? By adding DS metadata to pax-exam bundles?






g! scr:info com.basistech.ws.worker.core.Worker
*** Bundle: rosapi-worker-core (59)
Component Description:
  Name: com.basistech.ws.worker.core.Worker
  Implementation Class: com.basistech.ws.worker.core.Worker
  Default State: enabled
  Activation: immediate
  Configuration Policy: ignore
  Activate Method: activate
  Deactivate Method: deactivate
  Modified Method: -
  Configuration Pid: [com.basistech.ws.worker.core.Worker]
  Services:
    com.basistech.ws.worker.api.WorkerInterface
  Service Scope: singleton
  Reference: BusService
    Interface Name: com.basistech.ws.worker.api.WorkerBusService
    Cardinality: 1..1
    Policy: static
    Policy option: reluctant
    Reference Scope: bundle
  Reference: WorkerComponentServices
    Interface Name: com.basistech.ws.worker.core.WorkerComponentServices
    Cardinality: 1..1
    Policy: static
    Policy option: reluctant
    Reference Scope: bundle
  Reference: WorkerConfiguration
    Interface Name: com.basistech.ws.worker.core.config.WorkerConfiguration
    Cardinality: 1..1
    Policy: static
    Policy option: reluctant
    Reference Scope: bundle
  Component Description Properties:
  Component Configuration:
    ComponentId: 5
    State: active
    SatisfiedReference: BusService
      Target: null
      Bound to:        39
      Reference Properties:
          component.id = 3
          component.name = com.basistech.ws.worker.bus.impl.BusService
          licensePathname =
/Users/benson/x/rosapi1.5/worker/core/../../assemblies/common-configuration/src/main/resources/etc/rosapi/rlp-license.xml
          objectClass = [com.basistech.ws.worker.api.WorkerBusService]
          service.bundleid = 24
          service.id = 39
          service.pid = com.basistech.ws.bus
          service.scope = bundle
    SatisfiedReference: WorkerComponentServices
      Target: null
      Bound to:        56
      Reference Properties:
          objectClass = [com.basistech.ws.worker.core.WorkerComponentServices]
          service.bundleid = 59
          service.id = 56
          service.scope = singleton
    SatisfiedReference: WorkerConfiguration
      Target: null
      Bound to:        40
      Reference Properties:
          component.id = 6
          component.name =
com.basistech.ws.worker.core.config.WorkerConfigurationImpl
          configurationFilePathname =
/Users/benson/x/rosapi1.5/worker/core/src/test/config/path-subst-worker-config.yaml
          endpointsPathname =
/Users/benson/x/rosapi1.5/worker/core/src/test/config/all-endpoints-no-dte.yaml
          native-root =
/var/folders/80/5l86669j3278_x4hlpntlz380000gp/T/nativeroot6966159272930866297
          objectClass =
[com.basistech.ws.worker.core.config.WorkerConfiguration]
          service.bundleid = 59
          service.id = 40
          service.pid = com.basistech.ws.worker
          service.scope = bundle
          subst = tsbus
    Component Configuration Properties:
        component.id = 5
        component.name = com.basistech.ws.worker.core.Worker

On Fri, Aug 12, 2016 at 6:05 AM, Ferry Huberts <[email protected]> wrote:
>
>
> On 12/08/16 12:02, Carsten Ziegeler wrote:
>>>
>>> Thanks Carsten, that was my expectation but I was waiting for the
>>> “inspect cap service” output to confirm.
>>>
>>> One thing that can be quite confusing is that the service
>>> references, with cardinality of 1..1, are described as both
>>> satisfied and unbound. This sounds like a contradiction until you
>>> realise the component has not yet been instantiated. I assume that
>>> SCR at this point does know which services it *will* bind when the
>>> component is activated… maybe these could be shown here?
>
>
> I think that would be misleading because when the bundle composition changes
> before activation is needed then a different resolution can/will be
> activated.
>
>>>
>> Hmm, yeah we could do that - on the other hand "unbound" means it's
>> satisfied, otherwise it would state the reference as "unsatisfied"
>> :) Maybe, it's more a wording problem?
>>
>> Carsten
>>
>
> --
> Ferry Huberts
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to