- * First calculate number width and decimal place (last + 1 for
- * non-decimal numbers). If the stored decimal is subsequent to
- * ours, make our size longer by that difference
- * (right-"shifting"); similarly, if ours is subsequent the
- * stored, then extend the stored size by the difference.
- * Finally, re-assign the stored values.
+ * Find the last digit and
+ * the last decimal point that is adjacent to a digit.
+ * The alignment indicator "\&" overrides everything.