The art of not being one with the universe

I need to use MySQL in one of the projects at work. Fair enough, databases are useful little thingies. Big thingies. Frustrations mount quickly, though.

Example.

1. Dataset in text file contains N unique rows which I may want to update now and then.
2. N is large.
3. No column value is unique to each row, so there is no natural primary key.
4. Every row is guaranteed to be unique.
5. Does MySQL have some operator, procedure, function or command I can use to avoid duplicate rows when updating database from new text source? Can’t find any.