Does Java 1.7 use a different character encoding?

Question

We are migrating our application from Java 1.6 to Java 1.7. We recompiled the code using Java 1.7 and received an error while compiling which was due to a character (an Ó).

Was there a change in Java 1.7 related to characters? Our application does a lot of processing of incoming files to then load them into a database and I want to ensure that when we upgrade to Java 1.7 that the reading of a file from java and the writing to the database of that content wont result in some odd character conversions.

Do I need to be concerned at all when upgrading to 1.7? If so, how to I get the same encoding that we had in Java 1.6?

Answer 1

The error occurs because you've told the Java compiler that your source is UTF-8 encoded, but it still contains some ISO-8859-1 extended characters. I recently had to fix similar errors in a codebase that was migrated from 1.5 to 1.6. I believe that Java 7 is much stricter about UTF-8 encoding than previous versions and will issue errors where previously the incorrect encodings were silently accepted.

You will need to make sure that your source code is "Unicode-clean", that is, you must replace any extended ISO-8859-1 characters with their Unicode equivalents.