Transformator to show all attributes without values/with only empty fields

I don't think there is anything 'off-the-shelf', but this could be one way to tackle it.

• AttributeExposer exposing fme_feature_type
• PythonCaller to create an attribute containing comma-separated attributes with null values (code below)
• Aggregator with Group By set to fme_feature_type, creating a list of null values
• ListDuplicateRemover to remove duplicates
• ListConcatenator resulting in one feature per feature type, with an attribute listing all nulled attributes.

From there you could probably do your filtering.

import fmeobjects
 
def processFeature(feature):
    
    null_lst=[]
    
    for tmpAttrName in feature.getAllAttributeNames():
        tmpAttrVal = feature.getAttribute(tmpAttrName)
        if tmpAttrVal == '':
            null_lst.append(tmpAttrName)
            
    str = ','.join(null_lst)
    feature.setAttribute("null_list",str)
 

One possible way using the custom transformer NullAttributeCounter, the feature merger tests whether the count of null attributes equals the number of records in the workspace and returns the attribute name where that case is met

Prior to this, you may need to use a null attribute mapper to make sure empty and missing attributes are null