DSXchange

Can anyone please let me know how we deal with accented characters in DataStage. I am reading a .csv file (containing accented characters) in Sequential File stage with NLS_CODE set to UTF-8 and individual columns not having the extended 'Unicode' property.When I view the data I can see those accented characters.
Now I am using this file for lookup. In the output file I find those characters replaced by '?' etc.Please let me know why this is happening and which NLS_CODE should I select for Lookup stage.

Thanks in advance.