Hello it is true what to support unicode source files only StringBuffer implementation must be changed? if so, then task can be simplified by converting any files read by hGetStringBuffer to UTF-32 (PackedString) representation and putting in memory array in this form. After this, we must change indexing of ByteArray to indexing of Array Int Char, and somewhat replace call to mkFastSubStringBA#. btw, why in FastString module unicode strings are saved as [Int], not as String itself? -- Best regards, Bulat mailto:bulatz@HotPOP.com