From the paper, we originally had:
class FilterOperator {
// ...
onNewInstance(instance) {
trigger(aexpr(this.expression, instance))
.onBecomeTrue(() => this.add(instance))
.onBecomeFalse(() => this.remove(instance));
}
}
However, by reformulating the paramerizable aexpr to a call with that exact parameter we get:
trigger(aexpr(() => this.expression(instance)))
Can our chosen implementation strategy handle this case?
This should be true for ticking and rewriting, but not for interpretation, if the this.expression refers to locally scoped variables, as this.expression is now not called directly in the interpreter anymore! (which it was before, so it could handle this scenario due to the explicitly provided local scope object)