[
https://issues.apache.org/jira/browse/TUSCANY-3606?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12881721#action_12881721
]
Simon Laws commented on TUSCANY-3606:
-------------------------------------
There's a good question here about whether we should be removing intents from
the places they don't belong all the way through the structural hierarchy or
only checking for exclusivity in the places where intents apply.
Off the top of my head it's say the latter as they are only attached to the
higher levels as a way of getting them flowed down the the lower leves. There
doesn't seem to be much point in doing validity processing at the higher level.
I do though notice that the CompositePolicyBuilder does call resolveAndCheck()
explicitly at the higher levels and the removeConstrainedIntents() is called
from various places so we need to look to see if we can rationalize it.
> POL_4007 Problems with mutually exclusive intent checking
> ---------------------------------------------------------
>
> Key: TUSCANY-3606
> URL: https://issues.apache.org/jira/browse/TUSCANY-3606
> Project: Tuscany
> Issue Type: Bug
> Components: Java SCA Policy
> Reporter: Brent Daniel
>
> POL_4007 specifies two mutually exclusive intents at the Composite level. The
> intents are constrained to sca:binding, so we should only be checking for
> exclusivity at the binding level. Right now, we are checking (and failing)
> the intents in this test case at the Component level.
> There is some code in ComponentPolicyBuilderImpl.removeConstrainedIntents()
> that attempts to remove intents from any policy subject to which they do not
> apply. However, it does this based on the ExtensionType for the subject,
> which seems to be null for all SCA constructs other than bindings and
> implementations. It looks like it may be null for the default binding as well.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
