[Q21.5] Handle UTF-8 more robustly; pass through information about incorrect sequences

[Stephen J. Turnbull]

Aidan Kehoe writes:

 > I’m not sure what you mean by that. With this patch, for every string X,
 > (encode-coding-string (decode-coding-string X 'utf-8) 'utf-8) is identical
 > to X, independent of whether X is valid UTF-8.

What I want documented is what happens in the case of

  (encode-coding-string (snafu (decode-coding-string X 'utf-8) 'utf-8))

for arbitrary #'snafu that returns a string.  Eg,

  (defun snafu (dx)
    "Delete every other character from DX."
    (concat (let ((dxl (append dx nil)))
              (prog1 dxl
                (while (cdr dxl)
                  (setcdr dxl (cddr dxl))
                  (setq dxl (cdr dxl)))))))