资料来源 : Free On-Line Dictionary of Computing

regular expression
     
        1.  (regexp, RE) One of the {wild
        card} patterns used by {Unix} utilities such as {grep}, {sed}
        and {awk} and editors such as {vi} and {Emacs}.  These use
        conventions similar to but more elaborate than those described
        under {glob}.  A regular expression is a sequence of
        characters with the following meanings:
     
        An ordinary character (not one of the special characters
        discussed below) matches that character.
     
        A backslash (\) followed by any special character matches the
        special character itself.  The special characters are:
     
        "." matches any character except NEWLINE; "RE*" (where
        the "*" is called the "{Kleene star}") matches zero
        or more occurrences of RE.  If there is any choice, the
        longest leftmost matching string is chosen, in most
        regexp {flavour}s.
     
        "^" at the beginning of an RE matches the start of a line and
        "$" at the end of an RE matches the end of a line.
     
        [string] matches any one character in that string.  If the
        first character of the string is a "^" it matches
        any character (except NEWLINE, in most regexp {flavour}s)
        and the remaining characters in the string.  "-" may be used
        to indicate a range of consecutive ASCII characters.
     
        \( RE \) matches whatever RE matches and \n, where n is a
        digit, matches whatever was matched by the RE between the nth
        \( and its corresponding \) earlier in the same RE.  In
        many flavours ( RE ) is used instead of \( RE \)
     
        The concatenation of REs is a RE that matches the
        concatenation of the strings matched by each RE.
     
        \< matches the beginning of a word and \> matches the end of a
        word.  In many flavours of regexp, \> and \< are replaced by
        "\b", the special character for "word boundary".
     
        RE\{m\} matches m occurences of RE.  RE\{m,\} matches m or
        more occurences of RE.  RE\{m,n\} matches between m and n
        occurences.
     
        The exact details of how regexp will work in a given
        application vary greatly from flavour to flavour.  A comprehensive
        survey of regexp flavours is found in Friedl 1997 (see below).
     
        [Jeffrey E.F. Friedl, "{Mastering Regular
        Expressions(http://enterprise.ic.gc.ca/~jfriedl/regex/index.html)},
        O'Reilly, 1997.]
     
        2. Any description of a {pattern} composed from combinations
        of {symbols} and the three {operators}:
     
        Concatenation - pattern A concatenated with B matches a match
        for A followed by a match for B.
     
        Or - pattern A-or-B matches either a match for A or a match
        for B.
     
        Closure - zero or more matches for a pattern.
     
        The earliest form of regular expressions (and the term itself)
        were invented by mathematician {Stephen Cole Kleene} in the
        mid-1950s, as a notation to easily manipulate "regular sets",
        formal descriptions of the behaviour of {finite state
        machines}, in {regular algebra}.
     
        [S.C. Kleene, "Representation of events in nerve nets and
        finite automata", 1956, Automata Studies. Princeton].
     
        [J.H. Conway, "Regular algebra and finite machines", 1971, Eds
        Chapman & Hall].
     
        [Sedgewick, "Algorithms in C", page 294].
     
        (1997-08-03)