Hello @soly 

Yu can use the transformer DuplicateFilter.

Thanks in Advance,

Danilo

These are the steps I used (FME 2021)

• AttributeSplitter on Column A using the hyphen as the delimiter (trimming whitespace)
• ListHistogrammer with the source being the list created by the AttributeSplitter
• 2 ListConcatenators, concatenating both the histogram count and value
• Matcher checking both the concatenated count and value attributes

This also works with more than 2 values in the cell.

An alternative method would be to use a ListSorter, instead of the ListHistogrammer, and follow that up with one ListConcatenator before going to the Matcher.

Both methods worked when testing with your sample data, however results varied once I made the values more complex. You could test out both methods with your actual dataset to determine which is preferred.