There are a number of typos in the 1992 version of Sedgewick,
"Algorithms in C++."  The ones that I (BFC) have found are listed
below.

 - p.26, line 10:  "delete stack" -> "delete [] stack"
   This code was correct in an earlier version of C++.

 - p. 77, lines 7, -8 & -1: "$C_1 = 0$" -> "$C_1 = 1$"

 - p. 78, line 12: "$C_1 = 0$" -> "$C_1 = 1$"

 - p.27, lines -11 & -10: "save" -> "acc"

 - p. 147, line 12: "delete a" --> "delete [] a"

 - p. 150, line -8: "a[k/2] <= v" --> a[k/2] < v"
                    This is not an error, just that no action is
                    necessary or desirable when equality holds.
 - p. 160, line 20: "delete a; delete p; delete info" -->
                    " delete [] a; delete [] p; delete [] info"

 - p. 195, line 17: "delete a" -> "delete [] a"

 - p. 279, line -16: "i -= j-1;" --> "i -= j;"

 - pp. 279-281:  There are several errors in the section on the
Brute-Force Algorithm and the beginning of the next section on the
Knuth-Morris-Pratt Algorithm (where the brute-force algorithm is
discussed more).  All of these errors seem to follow from the fact
that the code for the brute-force algorihm is not the code being
discussed.  Instead the book seems to be discussing code like the
following:

int brutesearch(char* const p, char* const a)
{
  int i = 0, j = 0,
      M = strlen(p),
      N = strlen(a);

  while ( j < M && i < N)
    if (a[i] == p[j])
      {i++; j++;}
    else
      {i -= j-1; j = 0;}

  if (j==M)
    return i-M;
  else
    return N;
}

 - p. 283, line 1: "initnext(char *p)" --> "void initnext(char *p)"

 - p. 283, initnext() function:  On the last past of the loop, in the
           increments, i is set to M and then next[M] is set to j, in error, as
           there is no Mth position in the array next.

 - p. 286, line -20: "next[2] is 7" --> "next[3] is 7"

 - p. 288, line 7:  In the for loop "j > 0" --> "j >= 0).
 - p. 288, line 15: "return n" --> "return ++i"

 - p. 294, line -7: (1 + 01)*(0 + 1) --> (1 + 01)* + (1 + 01)*0

 - p. 301:  The program on this page contains a number of errors.  Two
            examples: (1) 0 should be a correct answer indicating that
            a match of the empty string occurred.  Suggest returning
            -1 to indicate that no match occurred. (2) the program
            does not work for the simplest machine, i.e., one that matches a
            single character, say A.

	    The file
	    "composer:~caviness/cis320/95Fall/sedgewick-code/match.cc"
	    contains a correct version.  The file
	    "sedgewick-code/match-ling.cc" contains an analogous
	    version that matches the longest possible substring in the
	    input string.  The best way to understand all this code is
	    to start with a recursive version.  Such a program, that
	    makes matches of maximal length, is in the file
	    "sedgewick-code/findmatch.cc"

------------------------------------------------------------

This information is in the file 
composer:~caviness/cis320/95Fall/sedgewick-typos