Re: Re: Umlaute

On Tue, Jun 03, 2003 at 04:15:11AM +0100, Matthias Radestock wrote:
> [email protected] wrote:
> >With my character set I can't see an umlaut, in the define statement.
> >Are there two characters in the result of evaluating a?  What are they?  
> >Is a's string-length 1?  
> 
> (equal? "?" "\u00e4")
>          ^----weird character.
> returns #t. The latter is the unicode point for the former. The string 
> length is 1. So it looks like input is ok and that the problem is 
> output. Also note that read/write invariance via with-output-to-string 
> is preserved, but via files is *not*.

The problem is not in any SISC code.  We use System.out, which is a 
PrintStream, for the default REPL output.  PrintStreams write bytes in 
the system's default character set, which results in the two character 
representation.

I changed DynamicEnvironment to default to PrintWriter and 
correspondingly InputStreamReader, so we now read and write characters 
correctly.  

	Scott

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)

iD8DBQE+3CRir9IW4v3mHtQRArZYAJ4t0A+KRY5SeiApHPRJfA867IL6wQCeM6jf
sRo4Vojcq6AvF4lkvVIeWhc=
=4zRb
-----END PGP SIGNATURE-----