The DECOMP modifier is used with the TOKENIZE routine to cause the routine to decompose variations of a character to a single character representation in the fields before comparison.
For example, an umlaut character "Ü" will be converted to the character "U". All accented variations of a capital 'A' as represented by the Unicode character values of 0x00C0 to 0x00C6 will be transformed to the unaccented capital 'A' as represented by the Unicode character value 0x0041. See Character Decomposition Mapping.