I did try a few ways to see what the characters actually were. One was by pasting the results to Regex101, and I also did as you suggested to put that into Notepad++. The characters seemed to be "tabs", spaces, and potentially newline characters. The most confusing part is that the data had already gone through a data cleanser that specifically removed those characters not once, but twice, and it left these problem records. I ended up trying to break all of this apart using the "REGEX" tool in Alteryx to parse the field into three pieces. This was anything before what I wanted, the data I wanted, and then whatever came after. This allowed me to look again what and how many were failing. This code did clean all the records as I kept only the middle piece, removed the two additional columns and swapped this column out for the original. (^[\t\s]*)([\w].*[A-Za-z0-9.~\-`]*)([\s\t\n\r]*)$
When the same data was run through a Data Cleansing tool with the exact same configuration it worked just fine and stripped out the characters in a NEW workflow.
Apathetic - I hear you on the tool, but in many cases it gets the job done with smaller datasets without issues and faster than me writing all the cleansing. (unless it now doesn't work consistently).
I originally thought I had AMP turned on for either the calling workflow or the batch macro, but after review, it is turned on for neither. My hypothesis that this was somehow AMP related should not be true based on this unless the Data Cleanse tool has been modified and needs AMP turned on.
