Feed AttributeValidator with current data schema and cached data

Hi!

As the attribute types are available more and more along the workspace, it would be nice to have ways to feed the AttributeValidator with those (type: number, string, char(20) etc..)
For example :
Attribute1 is defined as a string
import type to Validator
Attribute1 Type String
This would allow to identify outliers in datasets early.
A bit harder but i guess possible, use the cached data to generate rules (min,max, categories..).

I currently use EDA tools for this as it is too slow to go through all attributes manually with FME.