In postgres there is a simple string function SPLIT_PART to get the n-th part of a given string that is splitted on defined delimiters. A simple example would be:
SELECT SPLIT_PART('A,B,C,D', ',', 2);
split_part
-----------
B
So the function takes an input string, splits it at the given delimiter and returns the n-th substring.
Of course it is not a very complex task to get this functionality in Oracle (using a combination of instr, substr; maybe embetted in a user-defined function) - but I think it would be nice to get this task done with a single built-in function. This would make some common csv-import tasks very convenient.
As Frank Kulash mentioned in his comment: an (optional) enclosing character would make such a function even more valuable.