-/* Adler.java - Computes Adler32 data checksum of a data stream
- Copyright (C) 1999, 2000 Free Software Foundation, Inc.
+/* Adler32.java - Computes Adler32 data checksum of a data stream
+ Copyright (C) 1999, 2000, 2001 Free Software Foundation, Inc.
This file is part of GNU Classpath.
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2, or (at your option)
any later version.
-
+
GNU Classpath is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
*/
/**
- * Computes Adler32 data checksum of a data stream.
- * The actual Adler32 algorithm is described in RFC 1950
- * (ZLIB Compressed Data Format Specification version 3.3).
- * Can be used to get the CRC32 over a stream if used with checked input/output
- * streams.
+ * Computes Adler32 checksum for a stream of data. An Adler32
+ * checksum is not as reliable as a CRC32 checksum, but a lot faster to
+ * compute.
+ *<p>
+ * The specification for Adler32 may be found in RFC 1950.
+ * (ZLIB Compressed Data Format Specification version 3.3)
+ *<p>
+ *<p>
+ * From that document:
+ *<p>
+ * "ADLER32 (Adler-32 checksum)
+ * This contains a checksum value of the uncompressed data
+ * (excluding any dictionary data) computed according to Adler-32
+ * algorithm. This algorithm is a 32-bit extension and improvement
+ * of the Fletcher algorithm, used in the ITU-T X.224 / ISO 8073
+ * standard.
+ *<p>
+ * Adler-32 is composed of two sums accumulated per byte: s1 is
+ * the sum of all bytes, s2 is the sum of all s1 values. Both sums
+ * are done modulo 65521. s1 is initialized to 1, s2 to zero. The
+ * Adler-32 checksum is stored as s2*65536 + s1 in most-
+ * significant-byte first (network) order."
+ *<p>
+ * "8.2. The Adler-32 algorithm
+ *<p>
+ * The Adler-32 algorithm is much faster than the CRC32 algorithm yet
+ * still provides an extremely low probability of undetected errors.
+ *<p>
+ * The modulo on unsigned long accumulators can be delayed for 5552
+ * bytes, so the modulo operation time is negligible. If the bytes
+ * are a, b, c, the second sum is 3a + 2b + c + 3, and so is position
+ * and order sensitive, unlike the first sum, which is just a
+ * checksum. That 65521 is prime is important to avoid a possible
+ * large class of two-byte errors that leave the check unchanged.
+ * (The Fletcher checksum uses 255, which is not prime and which also
+ * makes the Fletcher check insensitive to single byte changes 0 <->
+ * 255.)
+ *<p>
+ * The sum s1 is initialized to 1 instead of zero to make the length
+ * of the sequence part of s2, so that the length does not have to be
+ * checked separately. (Any sequence of zeroes has a Fletcher
+ * checksum of zero.)"
+ *
+ * @author John Leuner, Per Bothner
+ * @since JDK 1.1
*
* @see InflaterInputStream
* @see InflaterOutputStream
- *
- * @author Per Bothner
- * @date April 6, 1999.
*/
public class Adler32 implements Checksum
{
/** largest prime smaller than 65536 */
- private static int BASE = 65521;
-
- private int s1;
- private int s2;
+ private static final int BASE = 65521;
+
+ private int checksum; //we do all in int.
+
+ //Note that java doesn't have unsigned integers,
+ //so we have to be careful with what arithmetic
+ //we do. We return the checksum as a long to
+ //avoid sign confusion.
/**
- * Creates an Adler32 data checksum.
+ * Creates a new instance of the <code>Adler32</code> class.
+ * The checksum starts off with a value of 1.
*/
public Adler32 ()
{
}
/**
- * Resets the Adler32 data checksum as if no update was ever called.
+ * Resets the Adler32 checksum to the initial value.
*/
- public void reset () { s1 = 1; s2 = 0; }
+ public void reset ()
+ {
+ checksum = 1; //Initialize to 1
+ }
/**
- * Adds one byte to the data checksum.
+ * Updates the checksum with the byte b.
*
* @param bval the data value to add. The high byte of the int is ignored.
*/
public void update (int bval)
{
+ //We could make a length 1 byte array and call update again, but I
+ //would rather not have that overhead
+ int s1 = checksum & 0xffff;
+ int s2 = checksum >>> 16;
+
s1 = (s1 + (bval & 0xFF)) % BASE;
s2 = (s1 + s2) % BASE;
+
+ checksum = (s2 << 16) + s1;
}
/**
- * Adds the complete byte array to the data checksum.
+ * Updates the checksum with the bytes taken from the array.
+ *
+ * @param buffer an array of bytes
*/
public void update (byte[] buffer)
{
}
/**
- * Adds the byte array to the data checksum.
- *
- * @param buf the buffer which contains the data
- * @param off the offset in the buffer where the data starts
- * @param len the length of the data
+ * Updates the checksum with the bytes taken from the array.
+ *
+ * @param buf an array of bytes
+ * @param off the start of the data used for this update
+ * @param len the number of bytes to use for this update
*/
public void update (byte[] buf, int off, int len)
{
- int s1 = this.s1;
- int s2 = this.s2;
+ //(By Per Bothner)
+ int s1 = checksum & 0xffff;
+ int s2 = checksum >>> 16;
+
while (len > 0)
{
- // We can defer the modulo operation.
- int n = 4000;
+ // We can defer the modulo operation:
+ // s1 maximally grows from 65521 to 65521 + 255 * 3800
+ // s2 maximally grows by 3800 * median(s1) = 2090079800 < 2^31
+ int n = 3800;
if (n > len)
n = len;
len -= n;
s1 %= BASE;
s2 %= BASE;
}
- this.s1 = s1;
- this.s2 = s2;
+
+ /*Old implementation, borrowed from somewhere:
+ int n;
+
+ while (len-- > 0) {
+
+ s1 = (s1 + (bs[offset++] & 0xff)) % BASE;
+ s2 = (s2 + s1) % BASE;
+ }*/
+
+ checksum = (s2 << 16) | s1;
}
/**
*/
public long getValue()
{
- return ((long) s2 << 16) + s1;
+ return (long) checksum & 0xffffffffL;
}
}
projects / gcc.git / commitdiff