XOM API for Processing XML with Java

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

Elliotte Rusty Harold writes:

> I think I agree with this, though it may be uglier than the current 
> default output. Nothing will be automatically converted. Conversion 
> will take place if and only if the user tells the serializer to 
> change the white space in some way.

But then aren't you using setLineSeperator (or another method) for a
dual function or to cause a side-effect, one to set the line seperator
and second turning on/off automatic conversion or some other feature?

Would this be reasonable?

   1. Always apply John's conversion alg. to attribute values and text nodes
   2. add/subtract any WS need for pretty printing
   3. send to output stream

And of course document this and that any \r will be escaped and if you
really want a line break always use a \n.

I think this will preserve roundtripping in the two cases of

    XOM -> serialize -> text -> parse -> XOM
    text -> parse -> XOM -> serialize -> text (assuming line seperator is 
same)

But as often the case, I could be wrong.

Brad