Text nodes hold the non-markup, non-Entity content of
an Element or Attribute.
When a document is first made available to the DOM, there is only
one Text object for each block of adjacent plain-text. Users (ie,
applications) may create multiple adjacent Texts during editing --
see
Element.normalize for discussion.
Note that CDATASection is a subclass of Text. This is conceptually
valid, since they're really just two different ways of quoting
characters when they're written out as part of an XML stream.
since: PR-DOM-Level-1-19980818.