Re: Encoding and unescaped-data

[Luis Casillas <luis at casillas dot org>]

$LANG is an OS-specific way of telling Java which encoding to use, so  
I'm not surprised you run into trouble.  What do you get when you run  
the following Kawa code in your Windows laptop?

   (<java.lang.System>:getProperty "file.encoding")

If you have control over what arguments are passed to the Java  
process that's running your code, you may want to add a "- 
Dfile.encoding=UTF-8" argument.

Other questions:

1. How is the port that you're using to output the data being opened?

2. When are you setting the port-char-encoding, relative to when the  
port is being opened?

On Aug 22, 2007, at 2:39 PM, Daniel Terhorst wrote:

> I tried running the code to my Linux machine and it actually worked
> just fine over there, save that I needed to add <?xml version="1.0"
> encoding="UTF-8"?> to the output file. The $LANG environment variable
> was en_US.UTF-8.
>
> Still no luck on my Windows laptop, however. I tried changing the LANG
> system and user environment variables to en_US.UTF-8 and UTF-8, but
> that didn't work. They were originally C.
>
> I also tried setting the port-char-encoding in my example:
>
> (set! port "UTF-8")
> (set! port-char-encoding "UTF-8")
>
> but that didn't seem to make a difference either.
>
>
>
>
> On 8/22/07, Luis Casillas <casillas@mercedsystems.com> wrote:
> > Are you somewhere specifying that you want to output data in  
> > UTF-8?  If no
> > output encoding is specified, Java will pick one based on your  
> > locale; if
> > that's not UTF-8, then unencodable characters will be output as  
> > "?".  If you
> > are on a Unix variant, what's the value of your $LANG environment  
> > variable?
> >
> > (I once ran into a problem where Kawa will not compile source  
> > files with
> > UTF-8 characters correctly if the locale wasn't set to be UTF-8;  
> > this is
> > probably related.)