home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Chip 1998 November
/
Chip_1998-11_cd.bin
/
tema
/
Cafe
/
main.bin
/
BigInteger.java
< prev
next >
Wrap
Text File
|
1998-01-23
|
50KB
|
1,445 lines
/*
* 97/02/24, @(#)BigInteger.java 1.7
*
* Copyright (c) 1996, 1997 Sun Microsystems, Inc. All Rights Reserved.
*
* This software is the confidential and proprietary information of Sun
* Microsystems, Inc. ("Confidential Information"). You shall not
* disclose such Confidential Information and shall use it only in
* accordance with the terms of the license agreement you entered into
* with Sun.
*
* SUN MAKES NO REPRESENTATIONS OR WARRANTIES ABOUT THE SUITABILITY OF THE
* SOFTWARE, EITHER EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE
* IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR
* PURPOSE, OR NON-INFRINGEMENT. SUN SHALL NOT BE LIABLE FOR ANY DAMAGES
* SUFFERED BY LICENSEE AS A RESULT OF USING, MODIFYING OR DISTRIBUTING
* THIS SOFTWARE OR ITS DERIVATIVES.
*
* CopyrightVersion 1.1_beta
*
*/
package java.math;
import java.util.Random;
/**
* Immutable arbitrary-precision integers. All operations behave as if
* BigIntegers were represented in two's-complement notation (like Java's
* primitive integer types). BigIntegers provide analogues to all of Java's
* primitive integer operators, and all relevant static methods from
* java.lang.Math. Additionally, BigIntegers provide operations for modular
* arithmetic, GCD calculation, primality testing, prime generation,
* single-bit manipulation, and a few other odds and ends.
*
* <P>Semantics of arithmetic operations exactly mimic those of java's integer
* arithmetic operators, as defined in The Java Language Specification. For
* example, division by zero throws an ArithmeticException, and division of
* a negative by a positive yields a negative (or zero) remainder. All of
* the details in the Spec concerning overflow are ignored, as BigIntegers
* are made as large as necessary to accommodate the results of an operation.
*
* <P>Semantics of shift operations extend those of Java's shift operators
* to allow for negative shift distances. A right-shift with a negative
* shift distance results in a left shift, and vice-versa. The unsigned
* right shift operator (>>>) is omitted, as this operation makes little
* sense in combination with the "infinite word size" abstraction provided
* by this class.
*
* <P>Semantics of bitwise logical operations are are exactly mimic those of
* Java's bitwise integer operators. The Binary operators (and, or, xor)
* implicitly perform sign extension on the shorter of the two operands
* prior to performing the operation.
*
* <P>Comparison operations perform signed integer comparisons, analogous to
* those performed by java's relational and equality operators.
*
* <P>Modular arithmetic operations are provided to compute residues, perform
* exponentiation, and compute multiplicative inverses. These methods always
* return a non-negative result, between 0 and (modulus - 1), inclusive.
*
* <P>Single-bit operations operate on a single bit of the two's-complement
* representation of their operand. If necessary, the operand is sign
* extended so that it contains the designated bit. None of the single-bit
* operations can produce a number with a different sign from the the
* BigInteger being operated on, as they affect only a single bit, and the
* "infinite word size" abstraction provided by this class ensures that there
* are infinitely many "virtual sign bits" preceding each BigInteger.
*
*
* @see BigDecimal
* @version 1.7, 97/11/25
* @author Josh Bloch
*/
public class BigInteger extends Number {
/*
* The number is internally stored in "minimal" sign-magnitude format
* (i.e., no BigIntegers have a leading zero byte in their magnitudes).
* Zero is represented with a signum of 0 (and a zero-length magnitude).
* Thus, there is exactly one representation for each value.
*/
private int signum;
private byte[] magnitude;
/*
* These "redundant fields" are initialized with recognizable nonsense
* values, and cached the first time they are needed (or never, if they
* aren't needed).
*/
private int bitCount = -1;
private int bitLength = -1;
private int firstNonzeroByteNum = -2; /* Only used for negative numbers */
private int lowestSetBit = -2;
//Constructors
/**
* Translates a byte array containing the two's-complement representation
* of a (signed) integer into a BigInteger. The input array is assumed to
* be big-endian (i.e., the most significant byte is in the [0] position).
* (The most significant bit of the most significant byte is the sign bit.)
* The array must contain at least one byte or a NumberFormatException
* will be thrown.
*/
public BigInteger(byte[] val) throws NumberFormatException{
if (val.length == 0)
throw new NumberFormatException("Zero length BigInteger");
if (val[0] < 0) {
magnitude = makePositive(val);
signum = -1;
} else {
magnitude = stripLeadingZeroBytes(val);
signum = (magnitude.length == 0 ? 0 : 1);
}
}
/**
* Translates the sign-magnitude representation of an integer into a
* BigInteger. The sign is represented as an integer signum value (-1 for
* negative, 0 for zero, 1 for positive). The magnitude is represented
* as a big-endian byte array (i.e., the most significant byte is in the
* [0] position). An invalid signum value or a 0 signum value coupled
* with a nonzero magnitude will result in a NumberFormatException.
* A zero length magnitude array is permissible, and will result in
* in a value of 0 (irrespective of the given signum value).
*/
public BigInteger(int signum, byte[] magnitude)
throws NumberFormatException{
this.magnitude = stripLeadingZeroBytes(magnitude);
if (signum < -1 || signum > 1)
throw(new NumberFormatException("Invalid signum value"));
if (this.magnitude.length==0) {
this.signum = 0;
} else {
if (signum == 0)
throw(new NumberFormatException("signum-magnitude mismatch"));
this.signum = signum;
}
}
/**
* Translates a string containing an optional minus sign followed by a
* sequence of one or more digits in the specified radix into a BigInteger.
* The character-to-digit mapping is provided by Character.digit.
* Any extraneous characters (including whitespace), or a radix outside
* the range from Character.MIN_RADIX(2) to Character.MAX_RADIX(36),
* inclusive, will result in a NumberFormatException.
*/
public BigInteger(String val, int radix) throws NumberFormatException {
int cursor = 0, numDigits;
if (radix < Character.MIN_RADIX || radix > Character.MAX_RADIX)
throw new NumberFormatException("Radix out of range");
if (val.length() == 0)
throw new NumberFormatException("Zero length BigInteger");
/* Check for leading minus sign */
signum = 1;
if (val.charAt(0) == '-') {
if (val.length() == 1)
throw new NumberFormatException("Zero length BigInteger");
signum = -1;
cursor = 1;
}
/* Skip leading zeros and compute number of digits in magnitude */
while (cursor<val.length() && val.charAt(cursor)==ZERO_CHAR)
cursor++;
if (cursor==val.length()) {
signum = 0;
magnitude = new byte[0];
return;
} else {
numDigits = val.length() - cursor;
}
/* Process first (potentially short) digit group, if necessary */
int firstGroupLen = numDigits % digitsPerLong[radix];
if (firstGroupLen == 0)
firstGroupLen = digitsPerLong[radix];
String group = val.substring(cursor, cursor += firstGroupLen);
BigInteger tmp = valueOf(Long.parseLong(group, radix));
/* Process remaining digit groups */
while (cursor < val.length()) {
group = val.substring(cursor, cursor += digitsPerLong[radix]);
long groupVal = Long.parseLong(group, radix);
if (groupVal <0)
throw new NumberFormatException("Illegal digit");
tmp = tmp.multiply(longRadix[radix]).add(valueOf(groupVal));
}
magnitude = tmp.magnitude;
}
/**
* Translates a string containing an optional minus sign followed by a
* sequence of one or more decimal digits into a BigInteger. The
* character-to-digit mapping is provided by Character.digit.
* Any extraneous characters (including whitespace) will result in a
* NumberFormatException.
*/
public BigInteger(String val) throws NumberFormatException {
this(val, 10);
}
/**
* Returns a random number uniformly distributed on [0, 2**numBits - 1]
* (assuming a fair source of random bits is provided in rndSrc).
* Note that this constructor always returns a non-negative BigInteger.
* Throws an IllegalArgumentException if numBits < 0.
*/
public BigInteger(int numBits, Random rndSrc)
throws IllegalArgumentException {
this(1, randomBits(numBits, rndSrc));
}
private static byte[] randomBits(int numBits, Random rndSrc)
throws IllegalArgumentException {
if (numBits < 0)
throw new IllegalArgumentException("numBits must be non-negative");
int numBytes = (numBits+7)/8;
byte[] randomBits = new byte[numBytes];
/* Generate random bytes and mask out any excess bits */
if (numBytes > 0) {
rndSrc.nextBytes(randomBits);
int excessBits = 8*numBytes - numBits;
randomBits[0] &= (1 << (8-excessBits)) - 1;
}
return randomBits;
}
/**
* Returns a randomly selected BigInteger with the specified bitLength
* that is probably prime. The certainty parameter is a measure of
* the uncertainty that the caller is willing to tolerate: the probability
* that the number is prime will exceed 1 - 1/2**certainty. The execution
* time is proportional to the value of the certainty parameter. The
* given random number generator is used to select candidates to be
* tested for primality. Throws an ArithmeticException if bitLength < 2.
*/
public BigInteger(int bitLength, int certainty, Random rnd) {
if (bitLength < 2)
throw new ArithmeticException("bitLength < 2");
BigInteger p;
do {
/*
* Select a candidate of exactly the right length. Note that
* Plumb's generator doesn't handle bitLength<=16 properly.
*/
do {
p = new BigInteger(bitLength-1, rnd).setBit(bitLength-1);
p = (bitLength <= 16
? (bitLength > 2 ? p.setBit(0) : p)
: new BigInteger(plumbGeneratePrime(p.magnitude), 1));
} while (p.bitLength() != bitLength);
} while (!p.isProbablePrime(certainty));
signum = 1;
magnitude = p.magnitude;
}
/**
* This private constructor differs from its public cousin
* with the arguments reversed in two ways: it assumes that its
* arguments are correct, and it doesn't copy the magnitude array.
*/
private BigInteger(byte[] magnitude, int signum) {
this.signum = (magnitude.length==0 ? 0 : signum);
this.magnitude = magnitude;
}
//Static Factory Methods
/**
* Returns a BigInteger with the specified value. This factory is provided
* in preference to a (long) constructor because it allows for reuse
* of frequently used BigIntegers (like 0 and 1), obviating the need for
* exported constants.
*/
public static BigInteger valueOf(long val) {
/* If -MAX_CONSTANT < val < MAX_CONSTANT, return stashed constant */
if (val == 0)
return ZERO;
if (val > 0 && val <= MAX_CONSTANT)
return posConst[(int) val];
else if (val < 0 && val >= -MAX_CONSTANT)
return negConst[(int) -val];
/* Dump two's complement binary into valArray */
byte valArray[] = new byte[8];
for (int i=0; i<8; i++, val >>= 8)
valArray[7-i] = (byte)val;
return new BigInteger(valArray);
}
private final static BigInteger ZERO = new BigInteger(new byte[0], 0);
/**
* Initialize static constant array when class is loaded.
*/
private final static int MAX_CONSTANT = 16;
private static BigInteger posConst[] = new BigInteger[MAX_CONSTANT+1];
private static BigInteger negConst[] = new BigInteger[MAX_CONSTANT+1];
static {
for (int i = 1; i <= MAX_CONSTANT; i++) {
byte[] magnitude = new byte[1];
magnitude[0] = (byte) i;
posConst[i] = new BigInteger(magnitude, 1);
negConst[i] = new BigInteger(magnitude, -1);
}
}
/**
* Returns a BigInteger with the given two's complement representation.
* Assumes that the input array will not be modified (the returned
* BigInteger will reference the input array if feasible).
*/
private static BigInteger valueOf(byte val[]) {
return (val[0]>0 ? new BigInteger(val, 1) : new BigInteger(val));
}
// Arithmetic Operations
/**
* Returns a BigInteger whose value is (this + val).
*/
public BigInteger add(BigInteger val) throws ArithmeticException {
if (val.signum == 0)
return this;
else if (this.signum == 0)
return val;
else if (val.signum == signum)
return new BigInteger(plumbAdd(magnitude, val.magnitude), signum);
else if (this.signum < 0)
return plumbSubtract(val.magnitude, magnitude);
else /* val.signum < 0 */
return plumbSubtract(magnitude, val.magnitude);
}
/**
* Returns a BigInteger whose value is (this - val).
*/
public BigInteger subtract(BigInteger val) {
return add(new BigInteger(val.magnitude, -val.signum));
}
/**
* Returns a BigInteger whose value is (this * val).
*/
public BigInteger multiply(BigInteger val) {
if (val.signum == 0 || this.signum==0)
return ZERO;
else
return new BigInteger(plumbMultiply(magnitude, val.magnitude),
signum * val.signum);
}
/**
* Returns a BigInteger whose value is (this / val). Throws an
* ArithmeticException if val == 0.
*/
public BigInteger divide(BigInteger val) throws ArithmeticException {
if (val.signum == 0)
throw new ArithmeticException("BigInteger divide by zero");
else if (this.signum == 0)
return ZERO;
else
return new BigInteger(plumbDivide(magnitude, val.magnitude),
signum * val.signum);
}
/**
* Returns a BigInteger whose value is (this % val). Throws an
* ArithmeticException if val == 0.
*/
public BigInteger remainder(BigInteger val) throws ArithmeticException {
if (val.signum == 0)
throw new ArithmeticException("BigInteger divide by zero");
else if (this.signum == 0)
return ZERO;
else if (this.magnitude.length < val.magnitude.length)
return this; /*** WORKAROUND FOR BUG IN R1.1 OF PLUMB'S PKG ***/
else
return new BigInteger(plumbRemainder(magnitude,val.magnitude),
signum);
}
/**
* Returns an array of two BigIntegers. The first ([0]) element of
* the return value is the quotient (this / val), and the second ([1])
* element is the remainder (this % val). Throws an ArithmeticException
* if val == 0.
*/
public BigInteger[] divideAndRemainder(BigInteger val)
throws ArithmeticException {
BigInteger result[] = new BigInteger[2];
if (val.signum == 0) {
throw new ArithmeticException("BigInteger divide by zero");
} else if (this.signum == 0) {
result[0] = result[1] = ZERO;
} else if (this.magnitude.length < val.magnitude.length) {
/*** WORKAROUND FOR A BUG IN R1.1 OF PLUMB'S PACKAGE ***/
result[0] = ZERO;
result[1] = this;
} else {
byte resultMagnitude[][] =
plumbDivideAndRemainder(magnitude, val.magnitude);
result[0] = new BigInteger(resultMagnitude[0], signum*val.signum);
result[1] = new BigInteger(resultMagnitude[1], signum);
}
return result;
}
/**
* Returns a BigInteger whose value is (this ** exponent). Throws
* an ArithmeticException if exponent < 0 (as the operation would yield
* a non-integer value). Note that exponent is an integer rather than
* a BigInteger.
*/
public BigInteger pow(int exponent) throws ArithmeticException {
if (exponent < 0)
throw new ArithmeticException("Negative exponent");
if (signum==0)
return (exponent==0 ? ONE : this);
/* Perform exponetiation using repeated squaring trick */
BigInteger result = valueOf(exponent<0 && (exponent&1)==1 ? -1 : 1);
BigInteger baseToPow2 = this;
while (exponent != 0) {
if ((exponent & 1)==1)
result = result.multiply(baseToPow2);
if ((exponent >>= 1) != 0)
baseToPow2 = new BigInteger(
plumbSquare(baseToPow2.magnitude), 1);
}
return result;
}
/**
* Returns a BigInteger whose value is the greatest common denominator
* of abs(this) and abs(val). Returns 0 if this == 0 && val == 0.
*/
public BigInteger gcd(BigInteger val) {
if (val.signum == 0)
return this.abs();
else if (this.signum == 0)
return val.abs();
else
return new BigInteger(plumbGcd(magnitude, val.magnitude), 1);
}
/**
* Returns a BigInteger whose value is the absolute value of this
* number.
*/
public BigInteger abs() {
return (signum >= 0 ? this : this.negate());
}
/**
* Returns a BigInteger whose value is (-1 * this).
*/
public BigInteger negate() {
return new BigInteger(this.magnitude, -this.signum);
}
/**
* Returns the signum function of this number (i.e., -1, 0 or 1 as
* the value of this number is negative, zero or positive).
*/
public int signum() {
return this.signum;
}
// Modular Arithmetic Operations
/**
* Returns a BigInteger whose value is this mod m. Throws an
* ArithmeticException if m <= 0.
*/
public BigInteger mod(BigInteger m) {
if (m.signum <= 0)
throw new ArithmeticException("BigInteger: modulus not positive");
BigInteger result = this.remainder(m);
return (result.signum >= 0 ? result : result.add(m));
}
/**
* Returns a BigInteger whose value is (this ** exponent) mod m. (If
* exponent == 1, the returned value is (this mod m). If exponent < 0,
* the returned value is the modular multiplicative inverse of
* (this ** -exponent).) Throws an ArithmeticException if m <= 0.
*/
public BigInteger modPow(BigInteger exponent, BigInteger m) {
if (m.signum <= 0)
throw new ArithmeticException("BigInteger: modulus not positive");
/* Workaround for a bug in Plumb: x^0 (y) dumps core for x != 0 */
if (exponent.signum == 0)
return ONE;
boolean invertResult;
if ((invertResult = (exponent.signum < 0)))
exponent = exponent.negate();
BigInteger base = (this.signum < 0 || this.compareTo(m) >= 0
? this.mod(m) : this);
BigInteger result;
if (m.testBit(0)) { /* Odd modulus: just pass it on to Plumb */
result = new BigInteger
(plumbModPow(base.magnitude, exponent.magnitude, m.magnitude), 1);
} else {
/*
* Even modulus. Plumb only supports odd, so tear it into
* "odd part" (m1) and power of two (m2), use Plumb to exponentiate
* mod m1, manually exponentiate mod m2, and use Chinese Remainder
* Theorem to combine results.
*/
/* Tear m apart into odd part (m1) and power of 2 (m2) */
int p = m.getLowestSetBit(); /* Max pow of 2 that divides m */
BigInteger m1 = m.shiftRight(p); /* m/2**p */
BigInteger m2 = ONE.shiftLeft(p); /* 2**p */
/* Caculate (base ** exponent) mod m1 */
BigInteger a1 = new BigInteger
(plumbModPow(base.magnitude, exponent.magnitude, m1.magnitude),1);
/* Caculate (this ** exponent) mod m2 */
BigInteger a2 = base.modPow2(exponent, p);
/* Combine results using Chinese Remainder Theorem */
BigInteger y1 = m2.modInverse(m1);
BigInteger y2 = m1.modInverse(m2);
result = a1.multiply(m2).multiply(y1).add
(a2.multiply(m1).multiply(y2)).mod(m);
}
return (invertResult ? result.modInverse(m) : result);
}
/**
* Returns (this ** exponent) mod(2**p)
*/
private BigInteger modPow2(BigInteger exponent, int p) {
/*
* Perform exponetiation using repeated squaring trick, chopping off
* high order bits as indicated by modulus.
*/
BigInteger result = valueOf(1);
BigInteger baseToPow2 = this.mod2(p);
while (exponent.signum != 0) {
if (exponent.testBit(0))
result = result.multiply(baseToPow2).mod2(p);
exponent = exponent.shiftRight(1);
if (exponent.signum != 0)
baseToPow2 = new BigInteger(
plumbSquare(baseToPow2.magnitude), 1).mod2(p);
}
return result;
}
/**
* Returns this mod(2**p). Assumes that this BigInteger >= 0 and p > 0.
*/
private BigInteger mod2(int p) {
if (bitLength() <= p)
return this;
/* Copy remaining bytes of magnitude */
int numBytes = (p+7)/8;
byte[] mag = new byte[numBytes];
for (int i=0; i<numBytes; i++)
mag[i] = magnitude[i + (magnitude.length - numBytes)];
/* Mask out any excess bits */
int excessBits = 8*numBytes - p;
mag[0] &= (1 << (8-excessBits)) - 1;
return (mag[0]==0 ? new BigInteger(1, mag) : new BigInteger(mag, 1));
}
/**
* Returns modular multiplicative inverse of this, mod m. Throws an
* ArithmeticException if m <= 0 or this has no multiplicative inverse
* mod m (i.e., gcd(this, m) != 1).
*/
public BigInteger modInverse(BigInteger m) throws ArithmeticException {
if (m.signum != 1)
throw new ArithmeticException("BigInteger: modulus not positive");
/* Calculate (this mod m) */
BigInteger modVal = this.remainder(m);
if (modVal.signum < 0)
modVal = modVal.add(m);
if (!modVal.gcd(m).equals(ONE))
throw new ArithmeticException("BigInteger not invertible");
return new BigInteger(plumbModInverse(modVal.magnitude,m.magnitude),1);
}
// Shift Operations
/**
* Returns a BigInteger whose value is (this << n). (Computes
* floor(this * 2**n).)
*/
public BigInteger shiftLeft(int n) {
if (n==0)
return this;
if (n<0)
return shiftRight(-n);
int nBytes = n/8;
int nBits = n%8;
byte result[] = new byte[(bitLength()+n)/8+1];
if (nBits == 0) {
for (int i=nBytes; i<result.length; i++)
result[result.length-1-i] = getByte(i-nBytes);
} else {
for (int i=nBytes; i<result.length; i++)
result[result.length-1-i] = (byte)
((getByte(i-nBytes) << nBits)
| (i==nBytes ? 0
: ((getByte(i-nBytes-1)&0xff) >> (8-nBits))));
}
return valueOf(result);
}
/**
* Returns a BigInteger whose value is (this >> n). Sign extension is
* performed. (Computes floor(this / 2**n).)
*/
public BigInteger shiftRight(int n) {
if (n==0)
return this;
if (n<0)
return shiftLeft(-n);
if (n >= bitLength())
return (signum<0 ? valueOf(-1) : ZERO);
int nBytes = n/8;
int nBits = n%8;
byte result[] = new byte[(bitLength-n)/8+1];
if (nBits == 0) {
for (int i=0; i<result.length; i++)
result[result.length-1-i] = getByte(nBytes+i);
} else {
for (int i=0; i<result.length; i++)
result[result.length-1-i] = (byte)
((getByte(nBytes+i+1)<<8 | (getByte(nBytes+i)&0xff)) >> nBits);
}
return valueOf(result);
}
// Bitwise Operations
/**
* Returns a BigInteger whose value is (this & val). (This method
* returns a negative number iff this and val are both negative.)
*/
public BigInteger and(BigInteger val) {
byte[] result = new byte[Math.max(byteLength(), val.byteLength())];
for (int i=0; i<result.length; i++)
result[i] = (byte) (getByte(result.length-i-1)
& val.getByte(result.length-i-1));
return valueOf(result);
}
/**
* Returns a BigInteger whose value is (this | val). (This method
* returns a negative number iff either this or val is negative.)
*/
public BigInteger or(BigInteger val) {
byte[] result = new byte[Math.max(byteLength(), val.byteLength())];
for (int i=0; i<result.length; i++)
result[i] = (byte) (getByte(result.length-i-1)
| val.getByte(result.length-i-1));
return valueOf(result);
}
/**
* Returns a BigInteger whose value is (this ^ val). (This method
* returns a negative number iff exactly one of this and val are
* negative.)
*/
public BigInteger xor(BigInteger val) {
byte[] result = new byte[Math.max(byteLength(), val.byteLength())];
for (int i=0; i<result.length; i++)
result[i] = (byte) (getByte(result.length-i-1)
^ val.getByte(result.length-i-1));
return valueOf(result);
}
/**
* Returns a BigInteger whose value is (~this). (This method returns
* a negative value iff this number is non-negative.)
*/
public BigInteger not() {
byte[] result = new byte[byteLength()];
for (int i=0; i<result.length; i++)
result[i] = (byte) ~getByte(result.length-i-1);
return valueOf(result);
}
/**
* Returns a BigInteger whose value is (this & ~val). This method,
* which is equivalent to and(val.not()), is provided as a convenience
* for masking operations. (This method returns a negative number iff
* this is negative and val is positive.)
*/
public BigInteger andNot(BigInteger val) {
byte[] result = new byte[Math.max(byteLength(), val.byteLength())];
for (int i=0; i<result.length; i++)
result[i] = (byte) (getByte(result.length-i-1)
& ~val.getByte(result.length-i-1));
return valueOf(result);
}
// Single Bit Operations
/**
* Returns true iff the designated bit is set. (Computes
* ((this & (1<<n)) != 0).) Throws an ArithmeticException if n < 0.
*/
public boolean testBit(int n) throws ArithmeticException {
if (n<0)
throw new ArithmeticException("Negative bit address");
return (getByte(n/8) & (1 << (n%8))) != 0;
}
/**
* Returns a BigInteger whose value is equivalent to this number
* with the designated bit set. (Computes (this | (1<<n)).)
* Throws an ArithmeticException if n < 0.
*/
public BigInteger setBit(int n) throws ArithmeticException {
if (n<0)
throw new ArithmeticException("Negative bit address");
int byteNum = n/8;
byte[] result = new byte[Math.max(byteLength(), byteNum+2)];
for (int i=0; i<result.length; i++)
result[result.length-i-1] = getByte(i);
result[result.length-byteNum-1] |= (1 << (n%8));
return valueOf(result);
}
/**
* Returns a BigInteger whose value is equivalent to this number
* with the designated bit cleared. (Computes (this & ~(1<<n)).)
* Throws an ArithmeticException if n < 0.
*/
public BigInteger clearBit(int n) throws ArithmeticException {
if (n<0)
throw new ArithmeticException("Negative bit address");
int byteNum = n/8;
byte[] result = new byte[Math.max(byteLength(), (n+1)/8+1)];
for (int i=0; i<result.length; i++)
result[result.length-i-1] = getByte(i);
result[result.length-byteNum-1] &= ~(1 << (n%8));
return valueOf(result);
}
/**
* Returns a BigInteger whose value is equivalent to this number
* with the designated bit flipped. (Computes (this ^ (1<<n)).)
* Throws an ArithmeticException if n < 0.
*/
public BigInteger flipBit(int n) throws ArithmeticException {
if (n<0)
throw new ArithmeticException("Negative bit address");
int byteNum = n/8;
byte[] result = new byte[Math.max(byteLength(), byteNum+2)];
for (int i=0; i<result.length; i++)
result[result.length-i-1] = getByte(i);
result[result.length-byteNum-1] ^= (1 << (n%8));
return valueOf(result);
}
/**
* Returns the index of the rightmost (lowest-order) one bit in this
* number (i.e., the number of zero bits to the right of the rightmost
* one bit). Returns -1 if this number contains no one bits.
* (Computes (this==0? -1 : log2(this & -this)).)
*/
public int getLowestSetBit() {
/*
* Initialize lowestSetBit field the first time this method is
* executed. This method depends on the atomicity of int modifies;
* without this guarantee, it would have to be synchronized.
*/
if (lowestSetBit == -2) {
if (signum == 0) {
lowestSetBit = -1;
} else {
/* Search for lowest order nonzero byte */
int i;
byte b;
for (i=0; (b = getByte(i))==0; i++)
;
lowestSetBit = 8*i + trailingZeroCnt[b & 0xFF];
}
}
return lowestSetBit;
}
// Miscellaneous Bit Operations
/**
* Returns the number of bits in the minimal two's-complement
* representation of this number, *excluding* a sign bit, i.e.,
* (ceil(log2(this < 0 ? -this : this + 1))). (For positive
* numbers, this is equivalent to the number of bits in the
* ordinary binary representation.)
*/
public int bitLength() {
/*
* Initialize bitLength field the first time this method is executed.
* This method depends on the atomicity of int modifies; without
* this guarantee, it would have to be synchronized.
*/
if (bitLength == -1) {
if (signum == 0) {
bitLength = 0;
} else {
/* Calculate the bit length of the magnitude */
int magBitLength = 8*(magnitude.length-1)
+ bitLen[magnitude[0] & 0xff];
if (signum < 0) {
/* Check if magnitude is a power of two */
boolean pow2 = (bitCnt[magnitude[0]&0xff] == 1);
for(int i=1; i<magnitude.length && pow2; i++)
pow2 = (magnitude[i]==0);
bitLength = (pow2 ? magBitLength-1 : magBitLength);
} else {
bitLength = magBitLength;
}
}
}
return bitLength;
}
/*
* bitLen[i] is the number of bits in the binary representaion of i.
*/
private final static byte bitLen[] = {
0, 1, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4,
5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5,
6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6,
6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6,
7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7,
7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7,
7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7,
7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7,
8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8,
8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8,
8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8,
8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8,
8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8,
8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8,
8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8,
8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8};
/**
* Returns the number of bits in the two's complement representation
* of this number that differ from its sign bit. This method is
* useful when implementing bit-vector style sets atop BigIntegers.
*/
public int bitCount() {
/*
* Initialize bitCount field the first time this method is executed.
* This method depends on the atomicity of int modifies; without
* this guarantee, it would have to be synchronized.
*/
if (bitCount == -1) {
/* Count the bits in the magnitude */
int magBitCount = 0;
for (int i=0; i<magnitude.length; i++)
magBitCount += bitCnt[magnitude[i] & 0xff];
if (signum < 0) {
/* Count the trailing zeros in the magnitude */
int magTrailingZeroCount = 0, j;
for (j=magnitude.length-1; magnitude[j]==0; j--)
magTrailingZeroCount += 8;
magTrailingZeroCount += trailingZeroCnt[magnitude[j] & 0xff];
bitCount = magBitCount + magTrailingZeroCount - 1;
} else {
bitCount = magBitCount;
}
}
return bitCount;
}
/*
* bitCnt[i] is the number of 1 bits in the binary representation of i.
*/
private final static byte bitCnt[] = {
0, 1, 1, 2, 1, 2, 2, 3, 1, 2, 2, 3, 2, 3, 3, 4,
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
4, 5, 5, 6, 5, 6, 6, 7, 5, 6, 6, 7, 6, 7, 7, 8};
/*
* trailingZeroCnt[i] is the number of trailing zero bits in the binary
* representaion of i.
*/
private final static byte trailingZeroCnt[] = {
8, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
5, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
6, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
5, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
7, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
5, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
6, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
5, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0,
4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0};
// Primality Testing
/**
* Returns true if this BigInteger is probably prime, false if it's
* definitely composite. The certainty parameter is a measure
* of the uncertainty that the caller is willing to tolerate:
* the method returns true if the probability that this number is
* is prime exceeds 1 - 1/2**certainty. The execution time is
* proportional to the value of the certainty parameter.
*/
public boolean isProbablePrime(int certainty) {
/*
* This test is taken from the DSA spec.
*/
int n = certainty/2;
if (n <= 0)
return true;
BigInteger w = this.abs();
if (w.equals(TWO))
return true;
if (!w.testBit(0) || w.equals(ONE))
return false;
/* Find a and m such that m is odd and w == 1 + 2**a * m */
BigInteger m = w.subtract(ONE);
int a = m.getLowestSetBit();
m = m.shiftRight(a);
/* Do the tests */
Random rnd = new Random();
for(int i=0; i<n; i++) {
/* Generate a uniform random on (1, w) */
BigInteger b;
do {
b = new BigInteger(w.bitLength(), rnd);
} while (b.compareTo(ONE) <= 0 || b.compareTo(w) >= 0);
int j = 0;
BigInteger z = b.modPow(m, w);
while(!((j==0 && z.equals(ONE)) || z.equals(w.subtract(ONE)))) {
if (j>0 && z.equals(ONE) || ++j==a)
return false;
z = z.modPow(TWO, w);
}
}
return true;
}
// Comparison Operations
/**
* Returns -1, 0 or 1 as this number is less than, equal to, or
* greater than val. This method is provided in preference to
* individual methods for each of the six boolean comparison operators
* (<, ==, >, >=, !=, <=). The suggested idiom for performing these
* comparisons is: (x.compareTo(y) <op> 0), where <op> is one of the
* six comparison operators.
*/
public int compareTo(BigInteger val) {
return (signum==val.signum
? signum*byteArrayCmp(magnitude, val.magnitude)
: (signum>val.signum ? 1 : -1));
}
/*
* Returns -1, 0 or +1 as big-endian unsigned byte array arg1 is
* <, == or > arg2.
*/
private static int byteArrayCmp(byte[] arg1, byte[] arg2) {
if (arg1.length < arg2.length)
return -1;
if (arg1.length > arg2.length)
return 1;
/* Argument lengths are equal; compare the values */
for (int i=0; i<arg1.length; i++) {
int b1 = arg1[i] & 0xff;
int b2 = arg2[i] & 0xff;
if (b1 < b2)
return -1;
if (b1 > b2)
return 1;
}
return 0;
}
/**
* Returns true iff x is a BigInteger whose value is equal to this number.
* This method is provided so that BigIntegers can be used as hash keys.
*/
public boolean equals(Object x) {
if (!(x instanceof BigInteger))
return false;
BigInteger xInt = (BigInteger) x;
if (xInt.signum != signum || xInt.magnitude.length != magnitude.length)
return false;
/* This test is just an optimization, which may or may not help */
if (xInt == this)
return true;
for (int i=0; i<magnitude.length; i++)
if (xInt.magnitude[i] != magnitude[i])
return false;
return true;
}
/**
* Returns the BigInteger whose value is the lesser of this and val.
* If the values are equal, either may be returned.
*/
public BigInteger min(BigInteger val) {
return (compareTo(val)<0 ? this : val);
}
/**
* Returns the BigInteger whose value is the greater of this and val.
* If the values are equal, either may be returned.
*/
public BigInteger max(BigInteger val) {
return (compareTo(val)>0 ? this : val);
}
// Hash Function
/**
* Computes a hash code for this object.
*/
public int hashCode() {
int hashCode = 0;
for (int i=0; i<magnitude.length; i++)
hashCode = 37*hashCode + (magnitude[i] & 0xff);
return hashCode * signum;
}
// Format Converters
/**
* Returns the string representation of this number in the given radix.
* If the radix is outside the range from Character.MIN_RADIX(2) to
* Character.MAX_RADIX(36) inclusive, it will default to 10 (as is the
* case for Integer.toString). The digit-to-character mapping provided
* by Character.forDigit is used, and a minus sign is prepended if
* appropriate. (This representation is compatible with the (String, int)
* constructor.)
*/
public String toString(int radix) {
if (signum == 0)
return "0";
if (radix < Character.MIN_RADIX || radix > Character.MAX_RADIX)
radix = 10;
/* Compute upper bound on number of digit groups and allocate space */
int maxNumDigitGroups = (magnitude.length + 6)/7;
String digitGroup[] = new String[maxNumDigitGroups];
/* Translate number to string, a digit group at a time */
BigInteger tmp = this.abs();
int numGroups = 0;
while (tmp.signum != 0) {
BigInteger b[] = tmp.divideAndRemainder(longRadix[radix]);
digitGroup[numGroups++] = Long.toString(b[1].longValue(), radix);
tmp = b[0];
}
/* Put sign (if any) and first digit group into result buffer */
StringBuffer buf = new StringBuffer(numGroups*digitsPerLong[radix]+1);
if (signum<0)
buf.append('-');
buf.append(digitGroup[numGroups-1]);
/* Append remaining digit groups padded with leading zeros */
for (int i=numGroups-2; i>=0; i--) {
/* Prepend (any) leading zeros for this digit group */
int numLeadingZeros = digitsPerLong[radix]-digitGroup[i].length();
if (numLeadingZeros != 0)
buf.append(zeros[numLeadingZeros]);
buf.append(digitGroup[i]);
}
return buf.toString();
}
/* zero[i] is a string of i consecutive zeros. */
private static String zeros[] = new String[64];
static {
zeros[63] =
"000000000000000000000000000000000000000000000000000000000000000";
for (int i=0; i<63; i++)
zeros[i] = zeros[63].substring(0, i);
}
/**
* Returns the string representation of this number, radix 10. The
* digit-to-character mapping provided by Character.forDigit is used,
* and a minus sign is prepended if appropriate. (This representation
* is compatible with the (String) constructor, and allows for string
* concatenation with Java's + operator.)
*/
public String toString() {
return toString(10);
}
/**
* Returns the two's-complement representation of this number. The array
* is big-endian (i.e., the most significant byte is in the [0] position).
* The array contains the minimum number of bytes required to represent
* the number (ceil((this.bitLength() + 1)/8)). (This representation is
* compatible with the (byte[]) constructor.)
*/
public byte[] toByteArray() {
byte[] result = new byte[byteLength()];
for(int i=0; i<result.length; i++)
result[i] = getByte(result.length-i-1);
return result;
}
/**
* Converts this number to an int. Standard narrowing primitive conversion
* as per The Java Language Specification.
*/
public int intValue() {
int result = 0;
for (int i=3; i>=0; i--)
result = (result << 8) + (getByte(i) & 0xff);
return result;
}
/**
* Converts this number to a long. Standard narrowing primitive conversion
* as per The Java Language Specification.
*/
public long longValue() {
long result = 0;
for (int i=7; i>=0; i--)
result = (result << 8) + (getByte(i) & 0xff);
return result;
}
/**
* Converts this number to a float. Similar to the double-to-float
* narrowing primitive conversion defined in The Java Language
* Specification: if the number has too great a magnitude to represent
* as a float, it will be converted to infinity or negative infinity,
* as appropriate.
*/
public float floatValue() {
/* Somewhat inefficient, but guaranteed to work. */
return Float.valueOf(this.toString()).floatValue();
}
/**
* Converts the number to a double. Similar to the double-to-float
* narrowing primitive conversion defined in The Java Language
* Specification: if the number has too great a magnitude to represent
* as a double, it will be converted to infinity or negative infinity,
* as appropriate.
*/
public double doubleValue() {
/* Somewhat inefficient, but guaranteed to work. */
return Double.valueOf(this.toString()).doubleValue();
}
static {
System.loadLibrary("math");
plumbInit();
}
private final static BigInteger ONE = valueOf(1);
private final static BigInteger TWO = valueOf(2);
private final static char ZERO_CHAR = Character.forDigit(0, 2);
/**
* Returns a copy of the input array stripped of any leading zero bytes.
*/
static private byte[] stripLeadingZeroBytes(byte a[]) {
int keep;
/* Find first nonzero byte */
for (keep=0; keep<a.length && a[keep]==0; keep++)
;
/* Allocate new array and copy relevant part of input array */
byte result[] = new byte[a.length - keep];
for (int i = keep; i<a.length; i++)
result[i - keep] = a[i];
return result;
}
/**
* Takes an array a representing a negative 2's-complement number and
* returns the minimal (no leading zero bytes) unsigned whose value is -a.
*/
static private byte[] makePositive(byte a[]) {
int keep, j;
/* Find first non-sign (0xff) byte of input */
for (keep=0; keep<a.length && a[keep]==-1; keep++)
;
/* Allocate output array. If all non-sign bytes are 0x00, we must
* allocate space for one extra output byte. */
for (j=keep; j<a.length && a[j]==0; j++)
;
int extraByte = (j==a.length ? 1 : 0);
byte result[] = new byte[a.length - keep + extraByte];
/* Copy one's complement of input into into output, leaving extra
* byte (if it exists) == 0x00 */
for (int i = keep; i<a.length; i++)
result[i - keep + extraByte] = (byte) ~a[i];
/* Add one to one's complement to generate two's complement */
for (int i=result.length-1; ++result[i]==0; i--)
;
return result;
}
/*
* The following two arrays are used for fast String conversions. Both
* are indexed by radix. The first is the number of digits of the given
* radix that can fit in a Java long without "going negative", i.e., the
* highest integer n such that radix ** n < 2 ** 63. The second is the
* "long radix" that tears each number into "long digits", each of which
* consists of the number of digits in the corresponding element in
* digitsPerLong (longRadix[i] = i ** digitPerLong[i]). Both arrays have
* nonsense values in their 0 and 1 elements, as radixes 0 and 1 are not
* used.
*/
private static int digitsPerLong[] = {0, 0,
62, 39, 31, 27, 24, 22, 20, 19, 18, 18, 17, 17, 16, 16, 15, 15, 15, 14,
14, 14, 14, 13, 13, 13, 13, 13, 13, 12, 12, 12, 12, 12, 12, 12, 12};
private static BigInteger longRadix[] = {null, null,
valueOf(0x4000000000000000L), valueOf(0x383d9170b85ff80bL),
valueOf(0x4000000000000000L), valueOf(0x6765c793fa10079dL),
valueOf(0x41c21cb8e1000000L), valueOf(0x3642798750226111L),
valueOf(0x1000000000000000L), valueOf(0x12bf307ae81ffd59L),
valueOf( 0xde0b6b3a7640000L), valueOf(0x4d28cb56c33fa539L),
valueOf(0x1eca170c00000000L), valueOf(0x780c7372621bd74dL),
valueOf(0x1e39a5057d810000L), valueOf(0x5b27ac993df97701L),
valueOf(0x1000000000000000L), valueOf(0x27b95e997e21d9f1L),
valueOf(0x5da0e1e53c5c8000L), valueOf( 0xb16a458ef403f19L),
valueOf(0x16bcc41e90000000L), valueOf(0x2d04b7fdd9c0ef49L),
valueOf(0x5658597bcaa24000L), valueOf( 0x6feb266931a75b7L),
valueOf( 0xc29e98000000000L), valueOf(0x14adf4b7320334b9L),
valueOf(0x226ed36478bfa000L), valueOf(0x383d9170b85ff80bL),
valueOf(0x5a3c23e39c000000L), valueOf( 0x4e900abb53e6b71L),
valueOf( 0x7600ec618141000L), valueOf( 0xaee5720ee830681L),
valueOf(0x1000000000000000L), valueOf(0x172588ad4f5f0981L),
valueOf(0x211e44f7d02c1000L), valueOf(0x2ee56725f06e5c71L),
valueOf(0x41c21cb8e1000000L)};
/**
* These routines provide access to the two's complement representation
* of BigIntegers.
*/
/**
* Returns the length of the two's complement representation in bytes,
* including space for at least one sign bit,
*/
private int byteLength() {
return bitLength()/8 + 1;
}
/* Returns sign bit */
private int signBit() {
return (signum < 0 ? 1 : 0);
}
/* Returns a byte of sign bits */
private byte signByte() {
return (byte) (signum < 0 ? -1 : 0);
}
/**
* Returns the specified byte of the little-endian two's complement
* representation (byte 0 is the LSB). The byte number can be arbitrarily
* high (values are logically preceded by infinitely many sign bytes).
*/
private byte getByte(int n) {
if (n >= magnitude.length)
return signByte();
byte magByte = magnitude[magnitude.length-n-1];
return (byte) (signum >= 0 ? magByte :
(n <= firstNonzeroByteNum() ? -magByte : ~magByte));
}
/**
* Returns the index of the first nonzero byte in the little-endian
* binary representation of the magnitude (byte 0 is the LSB). If
* the magnitude is zero, return value is undefined.
*/
private int firstNonzeroByteNum() {
/*
* Initialize bitCount field the first time this method is executed.
* This method depends on the atomicity of int modifies; without
* this guarantee, it would have to be synchronized.
*/
if (firstNonzeroByteNum == -2) {
/* Search for the first nonzero byte */
int i;
for (i=magnitude.length-1; i>=0 && magnitude[i]==0; i--)
;
firstNonzeroByteNum = magnitude.length-i-1;
}
return firstNonzeroByteNum;
}
/**
* Native method wrappers for Colin Plumb's big positive integer package.
* Each of these wrappers (except for plumbInit) behaves as follows:
*
* 1) Translate input arguments from java byte arrays into plumbNums.
*
* 2) Perform the requested computation.
*
* 3) Translate result(s) into Java byte array(s). (The subtract
* operation translates its result into a BigInteger, as its result
* is, effectively, signed.)
*
* 4) Deallocate all of the plumbNums.
*
* 5) Return the resulting Java array(s) (or, in the case of subtract,
* BigInteger).
*/
private static native void plumbInit();
private static native byte[] plumbAdd(byte[] a, byte[] b);
private static native BigInteger plumbSubtract(byte[] a, byte[] b);
private static native byte[] plumbMultiply(byte[] a, byte[] b);
private static native byte[] plumbDivide(byte[] a, byte[] b);
private static native byte[] plumbRemainder(byte[] a, byte[] b);
private static native byte[][] plumbDivideAndRemainder(byte[] a, byte[] b);
private static native byte[] plumbGcd(byte[] a, byte[] b);
private static native byte[] plumbModPow(byte[] a, byte[] b, byte[] m);
private static native byte[] plumbModInverse(byte[] a, byte[] m);
private static native byte[] plumbSquare(byte[] a);
private static native byte[] plumbGeneratePrime(byte[] a);
}