xslt-next-is-raw

["Voytenko, Dimitry" <DV...@SECTORBASE.COM>]

Hi,

This is what I found out and I'm not sure if it's ok. 
In Xalan J1 FormatterToDOM had processing instruction "xslt-next-is-raw" if
next node should be unescaped. Then TreeWalker could recognize this
processing instruction and output next text node without escaping. Now
TreeWalker has this implementation, but current DOMBuilder (replacement for
FormatterToDOM as I understood) can't create this processing instruction
because it doesn't implement org.apache.serialize.SerializerHandler (J1
checked directly output_handler if it's instance of FormatterToDOM).
Actually I don't know how it can affect transformation, I just accidentally
noticed that

Thanks