XOM API for Processing XML with Java

["Bradley S. Huffman" <hip AT cs.okstate.edu>]

John Cowan writes:

> People who synthesize XOM objects as opposed to building them
> should only be inserting \n characters, because that's what XML
> uses internally as the line-end.  If you look at the contents of a
> Text node representing the character content of
> 
>       <foo>This is a record
>       all about grownups</foo>
> 
> it will always be "This is a record\n\tall about grownups", based on
> the XML Recommendation, no matter what line terminator was used in
> the document.
> 
> So in either case the presence of \r at the XOM level represents an
> anomaly that should force the creation of an escape sequence.

Hmmm, I could easily see users, for portablity, doing something like

    String line_seperator = System.getProperty("line.seperator");
             ...
    attribute.setValue("one and two" + line_sperator + "and three");
             ...
    serializer.setLineSeperator(line_seperator);

and then wondering why the string is one char. longer on some systems then
on others.

Brad