Question

I need to write a CVS export program which internally use UTF-8 encoding which originated from user input via web(so you can expect any characters). It’s Japanese system so I need to encode to Shift_JIS.

Now, when I change UTF-8 into Shift_JIS, I get errors like:

Encoding::UndefinedConversionError (U+7E6B from UTF-8 to Shift_JIS):

I want to either a) eliminate the character, or b) map the character to some other character
(or simply, to string ‘(U+7E6B)’)

It seems catch the exception and eliminate it as byte string but there must be easier way to do this.

What is the best way to do this conversion?

Answer 1

[Converting my follow-up comments to question to an answer]

I found encode has option and I can give encode with

:undef=>true, # for UndefinedConversionError :replace=>"?"

to have desired effect. can specify following also:

:invalid=>true, # for InvalidByteSequenceError