forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jeff Turner <>
Subject Re: Forrest on OS/390
Date Tue, 18 Mar 2003 13:12:00 GMT
On Tue, Mar 18, 2003 at 10:39:13AM +0100, Oliver Alth wrote:
> Hello Jeff,
> I replaced the Xerces with 2.3 version, but it didn't help.
> I am not converting those xml files, i just unpack them on USS.
> Just the script files forrest, ant, and so on have to be converted to
> EBCDIC, otherwise they cannot be executed. So the files are ASCII. I
> just checked the xconf File with a hex editor. I can't find the x85
> (NEL) nor the 2028 unicode.
> Setup...FATAL_E 2003-03-18 09:38:14.352 [        ] (): Exception caught
> org.xml.sax.SAXParseException: Content is not allowed in trailing section.
>         at
> org.apache.xerces.parsers.AbstractSAXParser.parse(
>         at
>         at
>         at org.apache.cocoon.Main.main(

That bit of code is reading an XML file, WEB-INF/logkit.xconf

I think the problem is that your default JVM encoding is set to EBCDIC,
and somewhere in the software stack (Xerces, Avalon, Cocoon), the byte ->
char encoding isn't explicitly set to UTF-8, so it defaults to EBCDIC and
breaks.  Here is a good page describing this stuff:

So it depends on how badly you want to get Cocoon (Forrest is just a
layer on top) working on OS/390 :)

If you simply want to get some HTML from XML, and don't have time to
fight Avalon/Cocoon encoding issues, you might want to look at simpler
tools like Anakia (, which
are currently building

If you really want to use Cocoon and Forrest, and don't mind being the
Cocoon-OS/390 guinea pig, I'm very happy to help debug these issues and
commit fixes.  We should move over to the cocoon-users list then, as it's
not Forrest-specific.


> Have you got another idea?
> Thanks,
> Oliver

View raw message