Hi @bradosav​ ,

I'd suggest the SchemaScanner for this as you can use it to detect schema drift (last section in that article). In your case, since you have a text file that is line delimited, you can simply read in the text file and perform the join to detect any changes between what attributes the table has and what your expected schema is.

A few things to keep in mind, you may want to use Group By fme_feature_type in the SchemaScanner to ensure you are getting a single feature per feature type (table). Also, make note of the Ignore Attributes Containing to get rid of the format/feature attributes that aren't relevant and whether you want empty attributes to be included or not (in this case you likely want to set it to Use Default Data Type depending on what your tables look like).

If you are using a version of FME that doesn't have the SchemaScanner, this webinar covers a few solutions that should work in it's place.

Thanks a lot! I didn't need grouping, but keeping empty features, yes! Solved!