[kaffe] ByteToCharEUCJP/CharToByteEUCJP

[Atsushi Nemoto]

I found CharToByteSHIFT_JIS use CharToByteIconv but CharToByteEUCJP
does not.  Is there any reason?  Extending CharToByteIconv is better
than table-based implementation, isn't it?

OTOH CharToByteEUCJP have a typo (203E vs. 20E3).

		    // 0x7E -> U+203E OVERLINE
...
	    if (!US_ASCII && (chr == 0x20E3)) {

Please fix this (203E is correct) if CharToByteEUCJP will not be
reimplemented.

---
Atsushi Nemoto