Re: DecoderUS_ASCII

[Roman Kennke]

Am Fr, den 16.07.2004 schrieb Tom Tromey um 0:48:
> >>>>> "Roman" == Roman Kennke <address@hidden> writes:
> 
> Roman> 2004-07-15  Roman Kennke  <address@hidden>
> Roman>        * gnu/java/io/decode/DecoderUS_ASCII.java:
> Roman>        Added decoder for ASCI decoding of Unicode.
> Roman>        * java/lang/System.java:
> Roman>        Registered new US_ASCII decoder.
> 
> Roman>      // XXX FIXME - Cheat a little for ASCII.
> Roman>      // Remove when we get a real "ASCII En/Decoder"
> 
> You might as well remove this FIXME...

well, I left it in, since there is no encoder there yet.

> Also, do we need a corresponding ASCII encoder?
> 
> A couple nits about the implementation: it doesn't really follow the
> current coding standard.  Also, having a table for the identity
> function seems a little excessive, but really I don't care much.

I only copied the most part from 8859-1. US_ASCII is a subset of this,
so this seemed ok here. But of course: you are right.

> One bigger problem is that, from what I recall, the JDK treats ASCII
> just like 8859-1 -- and that Java programs rely on this.  Perhaps the
> alias really is the best thing to do... ?

It would not really hurt IMO. So we leave it as it is and only take out
the FIXME?

/Roman

signature.asc
Description: Dies ist ein digital signierter Nachrichtenteil