I'm working on normalizing a legacy database for an international company and came across a weird problem. Someone decided that contacts for a company can just have a "name" field. It's not split up by first and last name, so we can't conveniently sort by last name. I'm aware that the family name frequently comes first for some Asians. What the heck do I do there? I need to move duplicate information from the office contacts and administrative users into a new table, but I don't know how to split up the names. Some of them have clearly "Westernized" their names, but who? Further, if I figure out which name is which, do I add some sort of boolean field to the table to allow the two names to be displayed in reverse? These are issues get more subtle than I ever realized and I don't have any convenient way to wade through this other than the anguished screams of users seeing their names displayed wrong. I feel sorry for the internationalization folks. I'm sure that they constantly have issues like this.
Following the XP "simplest thing that could possibly work" philosophy, I'm considering splitting the names into first and last and not trying to do any fancy tricks with the ordering. If someone's family name is their first name, it will still display as such. But what happens when they enter new information? I suppose we can just tell them that they input the data incorrectly, but no client ever likes to hear that. Hmm...
Re:Names in schemas
jdavidb on 2002-09-06T15:51:11
"Firstname" "initial" breaks down horribly for "J. David Blackstone."
On the other hand, I always know it's a telemarketer when they ask for "David J. Blackstone."
I've come to believe that most people are pretty understanding when it comes to "computers" screwing up their names. It's when the situtation can't be corrected that they get irrate.
For example, many names of the form "X Y Z" are first name X, middle name Y, last name Z (like Sean Michael Burke)
But many X Y Z names (typically in Latin America) are first name X, primary family name Y, and second (mother's) family name Z, like Mario Vargas Llosa -- he's Mr. Vargas Llosa (his father was Ernesto J. Vargas Maldonado, his mother was
Dora Llosa Ureta). So:
Sean Michael Burke -> Burke, Sean Michael
but:
Mario Vargas Llosa -> Vargas Llosa, Mario
But wait, there's more! I've heard that when alphabetizing Icelanders' names, you alphabetize by the first name. That is, "Björk Guðmundsdóttir" is sorted under B -- but I think she's still Ms. Guðmundsdóttir, not Ms. Björk.
An adaptation of what librarians do when cataloguing, would be for you to add a "_" before the start of the last name so you know where the sortable stuff starts. But I don't think that tells you how to turn the name into [Title] [Lastnamepart]. I.e., if you tag Wen Ho Lee as "_Wen Ho Lee", that doesn't tell you whether it's Mr. Wen or Mr. Wen Ho -- it just tells you to sort under W. God only knows what we'd do if he had a son with the same name: Wen Ho, Jr., Lee?
RUN AWAY!
