Undefined behavior for of binding annotations with elementType=sliceOfStrings

[pmacik]

The binding annotation spec contains option sliceOfStrings for elementType in annotation.

However the spec does not tell how it should behave in various situations.
The following scenarios suggest the behavior:

Scenario 1: sliceOfStrings from array

Given a CRD is annotated with service binding annotation using the sliceOfStrings element type like in the following snippet:

...
metadata:
   annotations:
      service.binding/tags: path={.spec.tags},elementType=sliceOfStrings

and the respective CR contains the following spec:

...
kind: Backend
...
spec:
   tags:
      - knowledge
      - is
      - power

when a service binding is created to bind the service CR to an application the application is enriched by the binding with the following env variables:

BACKEND_TAGS_0="knowledge"
BACKEND_TAGS_1="is"
BACKEND_TAGS_2="power"

Scenario 2: sliceOfStrings from slice of maps

Given a CRD is annotated with service binding annotation using the sliceOfStrings element type like in the following snippet:

...
metadata:
   annotations:
      service.binding/url: path={.spec.connections},elementType=sliceOfStrings,sourceValue=url

and the respective CR contains the following spec:

...
kind: Backend
...
spec:
    connections:
      - type: primary
        url: primary.example.com
      - type: secondary
        url: secondary.example.com
      - type: '404'
        url: black-hole.example.com

when a service binding is created to bind the service CR to an application the application is enriched with the following env variables by the binding:

BACKEND_URL_0="primary.example.com"
BACKEND_URL_1="secondary.example.com"
BACKEND_URL_2="black-hole.example.com"

[sbose78]

This looks right to me. In scenario one, we have list of strings, and in the latter, we have a list of 'structs'.

@navidsh @Avni-Sharma Scenario 2 was in use by the strimzi proposal. Am I right?

[Avni-Sharma]

yes @sbose78 Scenario 2 is used in the case of strimzi. Referred redhat-developer/service-binding-operator#352 (comment) for cross-checking

[arthurdm]

Discussed in the hangout. @pmacik - thanks for bringing this up. Would you be able to please propose a PR for this one?

[pmacik]

Discussed in the hangout. @pmacik - thanks for bringing this up. Would you be able to please propose a PR for this one?

Sure, I'll send a PR for this