[Previously raised here:https://connect.microsoft.com/SQLServer/feedback/details/127230/token-string-tokeniser-string-occurence-function but closed without explanation so I'm re-raising]
Taking an incoming string value and tokenising it (i.e. treating as a delimited string of smaller strings) is something we do very often in the Derived Column component. Typical use case is when someone doesn't wish to use the Flat File Connection Manager to parse a text file because the parsed coluns are not needed until a later execution tree.
However its very very hard to do the parsing in a Derived COlumn component - it basically takes a combination of FINDSTRING & SUBSTRING function calls which ends up looking very messy.