I have quite a complex xslt based app that I will soon be
releasing to
the world.
It's a multiuser webapp, based on libxslt. Users can plugin
their own
xslts to a framework allowing all sorts of web2.0-y type
websites:
blogs, planets, etc...
One of the areas where my app doesn't work well right now is
error
handling. That's because the only error handling I know
about with
libxslt is the libxslt error handler.
There are two reasons why the current error handler is now
really good
enough for me:
- it hardly ever provides enough information about the
error
- it is not process specific - there is only one error log
and no
contextual way to separate one thread's errors from
another thread's
I'd like to do something about this in the near term.
Three solutions immediately suggest themselves:
- allow an error handler to be set against a particular
stylesheet and
somehow alter the xslt engine code to use that error
handler
- allow a per-thread context to be passed to the error
handler, say a
transform context; you'd have to register the error
handler in a
special way to get the extra-argument,; in python say:
def error_handler(ctx, pctx, error_message):
global error_messages
error_messages[pctx].add(error_message)
libxslt.registerContextErrorHandler(error_handler_func,
"")
- allow real per-thread error handlers, setting an error
handler in
the per-thread data space and checking for it's existance
before
logging an error
Anybody care to shoot holes in these ideas?
--
Nic Ferrier
http://www.tapsellfer
rier.co.uk for all your tapsell ferrier needs
_______________________________________________
xml mailing list, project page http://xmlsoft.org/
xml gnome.org
http://mai
l.gnome.org/mailman/listinfo/xml
|