The CASE expression (sorry about that, couldn't resist) -or rather the nested CASE expressions- are, as already mentioned, overly complicated.

Basically, it makes use of LEFT(@Text, Number) to get the first Number characters of the string, then uses CASE to decide if there is a @StringDelimiter somewhere in that last part - if there is, it takes the rightmost bit until the last @SDtringDelimiter, otherwise it takes the whole part.

The nested CASE are required to prevent out-of-bound errors in the SUBSTRING functions.