Using Knime on a Mac with csv files, I often encounter that the column names become garbled when adding them to nodes (java snippet, rule based row filter, etc). See below:
Thanks for the screenshot. I've not seen that kind of behaviour before, but I have a suspicion that it is linked to the File Encoding used by the File Reader or the CSV Reader.
To confirm, you could run the following in the terminal:
file -I filename.csv
filename.csv: text/plain; charset=utf-8
In the reader node 'Advanced' Tab you can set the 'Character decoding' to match the file type. Does that help?
I had already tried UTF8, UTF16, and ISO 8891 to see if it helped in the file reader but it just changed the symbols at the beginning of the first column name. Then when I write the data to a new CSV the column names have spaces between each letter.
Running that file enconding checker in terminal says the file is UTF-8.
File came from Tableau server (download data). And opens just fine in tableau and excel.