Linit,
you need to take things step by step in order to find out where (or even if) the characters are being converted.
First off, do you actually have NLS activated in DataStage? Then do a simple read of your source file and write it out to another flat file. Do not use the "view data" but your favorite text editor that supports NLS to compare the two file's contents. Are they the same?
I see that you created your table with "CHARACTERSET US7ASCII
", not necessarily the correct setting for a table containing Hungarian characters.
</a>