-
Notifications
You must be signed in to change notification settings - Fork 4
Collecting multiple feature parsing errors #13
Description
Huh, a good question.
I assume, arguing from the POV of an end user, that it might be a better way to return the errors for all affected features, even if it affects all features.
On the other hand, e.g. for 831 building block features or even more raster cell features, the returned error list would explode quickly. I am not sure if that is the best option.
Of course, one might think about implementing some additional logic to detect whether all features produce the same error(s) and only return those error messages bundled for a pseudonym like "'ALL {{numberOfFeatures}} FEATURES' have error(s) ...". But that would require additional logic and I actually would not really prefer such an option.
A feature-wise decleration of errors is more precise, even though the error-list might grow big.
A fast fail (implemented currently) lacks the information of how many features share the same error(s). E.g. if 100 of 831 features lack the property, an error-list of 100 items would point the user to 100 problems. A fast fail will only show one problematic feature, which might be misleading from the POV of the end user.
So, if it does not take too much time, a full list of errors for all problematic features is desirable.
Regards,
Christian
Originally posted by @cDanowski in https://github.com/SebaDro/kommonitor-importer/issues/5#issuecomment-594162634