XOM API for Processing XML with Java

[John Cowan <jcowan AT reutershealth.com>]

Bradley S. Huffman scripsit:

> > On the other hand, if the user does set the line separator to a 
> > linefeed (for example) then I think I'm justified in changing all 
> > line breaks including those in attribute values to line feeds.  

I don't think so.  Suppose I'm running on a Unix system, where \n is the
line separator and \r is an obscure, usually useless control character.  If
I have an \r in character content or attribute value somewhere, it
*almost certainly* means that I read a document, either native or from
some foreign system, that had an explicit 
 in it.  Whoever wrote that
meant what he said, and it should be properly re-created on output.

> Look at that, 3 paragraphs to try and explain what to do with two
> little characters :-) There's got to be a easier way that we're missing.

I think this covers all cases:

1.  Output every \n as the current line separator.
2.  Output every \r as a character reference.
3.  Output every \t as a character reference in attribute values only.
4.  Output every character not represented in the output encoding as
    a character reference.
5.  Output every space character using the current output encoding,
    unless you need to break up a line, in which case output it as the
    current line separator.
6.  Output every other character using the current output encoding.

-- 
Dream projects long deferred            John Cowan <jcowan AT reutershealth.com>
usually bite the wax tadpole.            http://www.ccil.org/~cowan
        --James Lileks                  http://www.reutershealth.com