I would appreciate your insight in regards of the behavior of attribute rules and their validation. We are working on an (obviously branch versioned) Enterprise geodatabase.
I have a couple of rules which I am trying to test but having issues with the errors not being updated, meaning that a ones established error is not removed after the issue has been solved.
Example: One rule tests if a feature is in the country that it says in the corresponding attribute where it is supposed to be. The feature not being in the correct country is correctly flagged as error. After moving the point into the country boundary the validation status changes to No calculation required, validation required, has error(s) but after evaluating that point again, it only changes to No calculation required, no validation required, has error(s) but does not accept that the point is now in the correct location. Copying the point in the same location and evaluating again, it does not bring an error for that new point so I can confirm that the location is correct but the error not updated for the first point.
The same behavior can be seen in a rule that is testing against a regular expression for one of the attribute fields.
The behavior is as expected for another rule that checks that the type of the facility is in the name attribute: When I remove the type it lists the error, when I put the type back into the name, the error is removed from the list of errors.
Am I doing something wrong / expecting a behavior that is not as designed?
And another question: I have a rule that looks for duplicates: if the name of the facility in the same district appears again then it gives the error. Obviously this error shows for both features with the same name. Now when I change one of them, for this one the validation status is updated and the error removed when validating again. The problem is though that the second feature might stay untouched and thus the validation status is not updated and the error remains although there is no second feature with the same name again. Can I force a feature to be revalidated?
