Rewrite FixNesting implementation to be tree-based.

This mega-patch rips out the FixNesting implementation and the related
ChildDef components.  The primary algorithmic change is to convert from
use of tokens to tree nodes, which are far more amenable to the style
of processing that FixNesting uses.  Additionally, FixNesting has been
changed to go bottom-up rather than top-down, in order to avoid needing
to implement backtracking.

This patch simplifies a good deal of the relevant logic, since we no
longer need to continually recalculate the nesting structure when
processing things.  However, the conversion to the alternate format
incurs some overhead, so for small inputs these changes are not a win.
One possibility to greatly reduce the constant factors here is to switch
to entirely using libxml's representation, and never serializing tokens;
this would require one to rewrite injectors, however.

The iterative post-order traversal in FixNesting is a bit subtle, but
we have essentially reified the stack and continuations.

We've removed support for %Core.EscapeInvalidChildren.

Signed-off-by: Edward Z. Yang <ezyang@mit.edu>

library/HTMLPurifier/ConfigSchema/schema/Core.EscapeInvalidChildren.txt

@@ -2,9 +2,11 @@ Core.EscapeInvalidChildren
 TYPE: bool
 DEFAULT: false
 --DESCRIPTION--
-When true, a child is found that is not allowed in the context of the
+<p><strong>Warning:</strong> this configuration option is no longer does anything as of 4.6.0.</p>
+
+<p>When true, a child is found that is not allowed in the context of the
 parent element will be transformed into text as if it were ASCII. When
 false, that element and all internal tags will be dropped, though text will
 be preserved.  There is no option for dropping the element but preserving
-child nodes.
+child nodes.</p>
 --# vim: et sw=4 sts=4