XOM API for Processing XML with Java

["David Bullock" <db AT dawnbreaks.net>]

On 10/14/06, John Cowan <cowan AT ccil.org> wrote:
> Elliotte Harold scripsit:
>
> > The problem is that a Writer doesn't give you enough information to tell
> > which characters need to be escaped and which don't. JDOM floundered on
> > this shoal too, though they went the other way; serializing to writers
> > but potentially producing malformed output. I'm not sure there's a
> > good option here. :-(
>
> If the Writer is an instance of StringWriter or CharArrayWriter, you
> can safely use the full Unicode repertoire.  Otherwise, assume the ASCII
> repertoire (even EBCDIC OutputStreamWriters implement that much).




OK, I see what TextWriter is doing.  Well, given that the configurable state
of an nu.xom.TextWriter is simply a Writer and an encoding, what's to
prevent a constructor like:

 Serializer(Writer out, String encoding)

... if I happen to lie about the encoding, then that's my own fault?  In the
specific JSP use-case I'm discussing, I can reliably
pageContext.getResponse().getCharacterEncoding()
... so it's not an issue to actually supply the 'meta info' about the Writer
externally,  if the Writer interface won't itself provide it.


Just a thought.

cheers,
David.