Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I am a U.S. citizen who recently moved to Japan. When I entered the country, my first and middle names were combined to create a 20 character given name that appears on my Japan-issued ID.

Ignoring the more common issues of computer systems not accepting non-Japanese names, it has been a bureaucratic nightmare, struggling with applying for credit cards, opening banking and investment accounts, and linking things together, as each system has different lengths limits for names, and in some systems my name is input as Unicode "double width romaji", which in some systems will not match the otherwise identical ASCII version of the name.

I truly wish I was offered the ability during immigration to just drop the middle name, but even then my first name is longer than some systems' 5 character limit, where I am often put in as "Cortl".



> "double width romaji"

The double-width latin characters are the bane of every data analysis using Asian data. It's a roll of the dice whether any given Latin character will be in typical ASCII form or double-width form.

I'm sure there's some "seriously insensitive character comparison" library out there, but it's not built into my SQL server's "LIKE" operator, that's for sure.


The seriously insensitive character comparison you're looking for is NFKC normalization (see http://en.wikipedia.org/wiki/Unicode_equivalence).

You would probably be able to normalize all your strings before putting them into your database. You'd lose no functionality unless at some point later you really did need to distinguish "C" from "C".




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: