The old and new strings in the 's' and 'x' commands ('s/<old>/<new>/' and
'x/<old>/<new>/') are defined as follows:

<old> is a regular expression that is searched for in each line in the range
	(use 'h reg' for more information on regular expressions (RE's)).
	The characters '^', '$', '.', '*' and '[' do not stand for themselves
	unless they are preceded by a '`.
	In addition to the normal RE notation, <old> may contain RE's
	delimited by '\\(' and '\\)'. These define sub-strings that can be
	referenced in <new>.

<new> is not a RE, so '^', '$', '.', '*' and '[' may be used as normal
	characters, but '&' denotes the whole string matched by <old>, and
	'\1','\2', ..., denote the strings matched by the first, second, ...,
	RE's delimited by '\\(' and '\\)' in <old>.
Note that 's' replaces only the first instance of <old> in each line unless
the command is postfixed by a 'g', whereas 'x' displays all instances of <old>.
