On Dec 31, 2008, at 1:25 AM, Bertrand Delacretaz wrote:

On Wed, Dec 31, 2008 at 2:11 AM, Mark Lundquist <m...@wrinkledog.com> wrote:
...Cocoon is serving a web page with a bunch of occurrences of the "ndash" character (Unicode #8211). These displayed correctly with the old Saxon,
but now with the new version they instead look like this:

      â€"...

Did you try setting the encoding at the JVM level as well? (- Dfile.encoding=...)

Your problem might be related to some code using the platform's
default encoding in places, instead of the specified one.

Hi Bertrand,

Thanks but I don't think so :-)... no code within Saxon (the only thing that has changed) is reading any character data from outside the JVM, e.g. the offending characters are from the source document which has been parsed by Xalan, same as before, so the effect of the file.encoding property (or its absence) would not have changed.

I thought maybe somebody (Vadim?) w/ experience using Cocoon+Saxon might just be able to say "oh, that!" :-).

cheers, HNY,
—ml—

Attachment: smime.p7s
Description: S/MIME cryptographic signature

Reply via email to