The purpose of nativetoascii is to let you write
code using native character sets, then translate it to ASCII for compilation. The "javac" compiler needs non-ASCII characters to be escaped.
But
Java's I/O system can read the native characters directly (otherwise nativetoascii, written in Java itself, wouldn't work, right?) Data files don't need to be run through nativetoascii, and in fact, the \u0000 escape sequences don't mean anything to Java's I/O system, only to the compiler.
So don't translate your file before reading it. If Java can't seem to read it correctly unchanged, then you may need to fiddle with character set encodings, but you shouldn't need to modify the data file itself.