Some application code (such as parsers) may contain lots of string pointers. Currently, our analyses keeps all string pointers apart, potentially leading to large address sets.
To optimize our analysis runtime, address sets should be limited to only keep at most one distinct string pointer.