Re: Html escaping characters inside ElementProcessor
Actually, that javadoc comment is very confusing. What I meant when I wrote it is that since 2.1.3 the "content" of a Text node is no longer the result of explicitly escaping whatever came through the constructor, which was the previous behaviour.
Now Text contents are kept just as they are specified in their constructors, but the template writer can ask Text nodes to provide an escaped version of their contents when writing the results (not during processing). This saves quite a lot of unescape+escape operations in template processing, because all texts directly coming from the input template and going unmodified to the template output are not touched at all.