Skip to main content

Hello,

on the SchemaScanner transformer, in the data types section, there is a parameter to define the action to be taken for attributes with the value Missing/Null/Empty.

When ignore is selected, if an attribute has no entry, it is excluded from the schema. This seems correct for Missing. But is this the case for Null and Empty?

In the latter case, shouldn't the attribute remain in the data model?

If not, what other prossibility exists for retaining an attribute that has no value entry in the data model?

 

Best regards

Patrick

The NullAttributeMapper can map fields having null or empty values a set value. Maybe this can help you to get the Schema feature as you want. You also can map the attributes back to empty or null after the SchemaScanner, to get the output as you want (being null or empty).


The NullAttributeMapper can map fields having null or empty values a set value. Maybe this can help you to get the Schema feature as you want. You also can map the attributes back to empty or null after the SchemaScanner, to get the output as you want (being null or empty).

This is a pretty nice approach - that is, if you're not happy with the default fme_data_type fme chooses


The NullAttributeMapper can map fields having null or empty values a set value. Maybe this can help you to get the Schema feature as you want. You also can map the attributes back to empty or null after the SchemaScanner, to get the output as you want (being null or empty).

That did the trick. Thanks for your help!

With a large number of input paths, this means I have to define a null NullAttributeMapper for each of them before the schemas are merged, which might not be ideal.


That did the trick. Thanks for your help!

With a large number of input paths, this means I have to define a null NullAttributeMapper for each of them before the schemas are merged, which might not be ideal.

Cheers :)


Reply