RFC: Character set converters...

[David Daney <ddaney at avtrex dot com>]

The problem at hand:

Run jCIFS (http://jcifs.samba.org) code in the libgcj runtime.

One problem we have is the need for a "UnicodeLittleUnmarked" character
set encoder for String.getBytes(String enc).  This is supplied in Sun's
jre, but not libgcj.

My first attempt was to create the class:

gnu.gcj.convert.Output_UnicodeLittleUnmarked.java

This works, but I was thinking that perhaps a better way would be to add
a general purpose adapter into gnu.gcj.convert.UnicodeToBytes so that it
could use the java.nio.charset converters.  Thus eliminating duplicated
code.

Does this make any sense?

I might be willing to either commit my Output_UnicodeLittleUnmarked.java
or implement the adapter scheme, and put the converter in java.nio.charset.

Comments?

David Daney.