Source Code Cross Referenced for Multiplication.java in  » Apache-Harmony-Java-SE » java-package » java » math » Java Source Code / Java DocumentationJava Source Code and Java Documentation

Java Source Code / Java Documentation
1. 6.0 JDK Core
2. 6.0 JDK Modules
3. 6.0 JDK Modules com.sun
4. 6.0 JDK Modules com.sun.java
5. 6.0 JDK Modules sun
6. 6.0 JDK Platform
7. Ajax
8. Apache Harmony Java SE
9. Aspect oriented
10. Authentication Authorization
11. Blogger System
12. Build
13. Byte Code
14. Cache
15. Chart
16. Chat
17. Code Analyzer
18. Collaboration
19. Content Management System
20. Database Client
21. Database DBMS
22. Database JDBC Connection Pool
23. Database ORM
24. Development
25. EJB Server geronimo
26. EJB Server GlassFish
27. EJB Server JBoss 4.2.1
28. EJB Server resin 3.1.5
29. ERP CRM Financial
30. ESB
31. Forum
32. GIS
33. Graphic Library
34. Groupware
35. HTML Parser
36. IDE
37. IDE Eclipse
38. IDE Netbeans
39. Installer
40. Internationalization Localization
41. Inversion of Control
42. Issue Tracking
43. J2EE
44. JBoss
45. JMS
46. JMX
47. Library
48. Mail Clients
49. Net
50. Parser
51. PDF
52. Portal
53. Profiler
54. Project Management
55. Report
56. RSS RDF
57. Rule Engine
58. Science
59. Scripting
60. Search Engine
61. Security
62. Sevlet Container
63. Source Control
64. Swing Library
65. Template Engine
66. Test Coverage
67. Testing
68. UML
69. Web Crawler
70. Web Framework
71. Web Mail
72. Web Server
73. Web Services
74. Web Services apache cxf 2.0.1
75. Web Services AXIS2
76. Wiki Engine
77. Workflow Engines
78. XML
79. XML UI
Java
Java Tutorial
Java Open Source
Jar File Download
Java Articles
Java Products
Java by API
Photoshop Tutorials
Maya Tutorials
Flash Tutorials
3ds-Max Tutorials
Illustrator Tutorials
GIMP Tutorials
C# / C Sharp
C# / CSharp Tutorial
C# / CSharp Open Source
ASP.Net
ASP.NET Tutorial
JavaScript DHTML
JavaScript Tutorial
JavaScript Reference
HTML / CSS
HTML CSS Reference
C / ANSI-C
C Tutorial
C++
C++ Tutorial
Ruby
PHP
Python
Python Tutorial
Python Open Source
SQL Server / T-SQL
SQL Server / T-SQL Tutorial
Oracle PL / SQL
Oracle PL/SQL Tutorial
PostgreSQL
SQL / MySQL
MySQL Tutorial
VB.Net
VB.Net Tutorial
Flash / Flex / ActionScript
VBA / Excel / Access / Word
XML
XML Tutorial
Microsoft Office PowerPoint 2007 Tutorial
Microsoft Office Excel 2007 Tutorial
Microsoft Office Word 2007 Tutorial
Java Source Code / Java Documentation » Apache Harmony Java SE » java package » java.math 
Source Cross Referenced  Class Diagram Java Document (Java Doc) 


001:        /*
002:         *  Licensed to the Apache Software Foundation (ASF) under one or more
003:         *  contributor license agreements.  See the NOTICE file distributed with
004:         *  this work for additional information regarding copyright ownership.
005:         *  The ASF licenses this file to You under the Apache License, Version 2.0
006:         *  (the "License"); you may not use this file except in compliance with
007:         *  the License.  You may obtain a copy of the License at
008:         *
009:         *     http://www.apache.org/licenses/LICENSE-2.0
010:         *
011:         *  Unless required by applicable law or agreed to in writing, software
012:         *  distributed under the License is distributed on an "AS IS" BASIS,
013:         *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
014:         *  See the License for the specific language governing permissions and
015:         *  limitations under the License.
016:         */
017:
018:        package java.math;
019:
020:        import org.apache.harmony.math.internal.nls.Messages;
021:
022:        /**
023:         * Static library that provides all multiplication of {@link BigInteger} methods.
024:         *
025:         * @author Intel Middleware Product Division
026:         * @author Instituto Tecnologico de Cordoba
027:         */
028:        class Multiplication {
029:
030:            /** Just to denote that this class can't be instantiated. */
031:            private Multiplication() {
032:            }
033:
034:            /**
035:             * Break point in digits (number of {@code int} elements)
036:             * between Karatsuba and Pencil and Paper multiply.
037:             */
038:            static final int whenUseKaratsuba = 63; // an heuristic value
039:
040:            /**
041:             * An array with powers of ten that fit in the type {@code int}.
042:             * ({@code 10^0,10^1,...,10^9})
043:             */
044:            static final int tenPows[] = { 1, 10, 100, 1000, 10000, 100000,
045:                    1000000, 10000000, 100000000, 1000000000 };
046:
047:            /**
048:             * An array with powers of five that fit in the type {@code int}.
049:             * ({@code 5^0,5^1,...,5^13})
050:             */
051:            static final int fivePows[] = { 1, 5, 25, 125, 625, 3125, 15625,
052:                    78125, 390625, 1953125, 9765625, 48828125, 244140625,
053:                    1220703125 };
054:
055:            /**
056:             * An array with the first powers of ten in {@code BigInteger} version.
057:             * ({@code 10^0,10^1,...,10^31})
058:             */
059:            static final BigInteger[] bigTenPows = new BigInteger[32];
060:
061:            /**
062:             * An array with the first powers of five in {@code BigInteger} version.
063:             * ({@code 5^0,5^1,...,5^31})
064:             */
065:            static final BigInteger bigFivePows[] = new BigInteger[32];
066:
067:            static {
068:                int i;
069:                long fivePow = 1L;
070:
071:                for (i = 0; i <= 18; i++) {
072:                    bigFivePows[i] = BigInteger.valueOf(fivePow);
073:                    bigTenPows[i] = BigInteger.valueOf(fivePow << i);
074:                    fivePow *= 5;
075:                }
076:                for (; i < bigTenPows.length; i++) {
077:                    bigFivePows[i] = bigFivePows[i - 1]
078:                            .multiply(bigFivePows[1]);
079:                    bigTenPows[i] = bigTenPows[i - 1].multiply(BigInteger.TEN);
080:                }
081:            }
082:
083:            /**
084:             * Performs a multiplication of two BigInteger and hides the algorithm used.
085:             * @see BigInteger#multiply(BigInteger)
086:             */
087:            static BigInteger multiply(BigInteger x, BigInteger y) {
088:                return karatsuba(x, y);
089:            }
090:
091:            /**
092:             * Performs the multiplication with the Karatsuba's algorithm.
093:             * <b>Karatsuba's algorithm:</b>
094:             *<tt>
095:             *             u = u<sub>1</sub> * B + u<sub>0</sub><br>
096:             *             v = v<sub>1</sub> * B + v<sub>0</sub><br>
097:             *
098:             *
099:             *  u*v = (u<sub>1</sub> * v<sub>1</sub>) * B<sub>2</sub> + ((u<sub>1</sub> - u<sub>0</sub>) * (v<sub>0</sub> - v<sub>1</sub>) + u<sub>1</sub> * v<sub>1</sub> +
100:             *  u<sub>0</sub> * v<sub>0</sub> ) * B + u<sub>0</sub> * v<sub>0</sub><br>
101:             *</tt>
102:             * @param op1 first factor of the product
103:             * @param op2 second factor of the product
104:             * @return {@code op1 * op2}
105:             * @see #multiply(BigInteger, BigInteger)
106:             */
107:            static BigInteger karatsuba(BigInteger op1, BigInteger op2) {
108:                BigInteger temp;
109:                if (op2.numberLength > op1.numberLength) {
110:                    temp = op1;
111:                    op1 = op2;
112:                    op2 = temp;
113:                }
114:                if (op2.numberLength < whenUseKaratsuba) {
115:                    return multiplyPAP(op1, op2);
116:                }
117:                /*  Karatsuba:  u = u1*B + u0
118:                 *              v = v1*B + v0
119:                 *  u*v = (u1*v1)*B^2 + ((u1-u0)*(v0-v1) + u1*v1 + u0*v0)*B + u0*v0
120:                 */
121:                // ndiv2 = (op1.numberLength / 2) * 32
122:                int ndiv2 = (op1.numberLength & 0xFFFFFFFE) << 4;
123:                BigInteger upperOp1 = op1.shiftRight(ndiv2);
124:                BigInteger upperOp2 = op2.shiftRight(ndiv2);
125:                BigInteger lowerOp1 = op1.subtract(upperOp1.shiftLeft(ndiv2));
126:                BigInteger lowerOp2 = op2.subtract(upperOp2.shiftLeft(ndiv2));
127:
128:                BigInteger upper = karatsuba(upperOp1, upperOp2);
129:                BigInteger lower = karatsuba(lowerOp1, lowerOp2);
130:                BigInteger middle = karatsuba(upperOp1.subtract(lowerOp1),
131:                        lowerOp2.subtract(upperOp2));
132:                middle = middle.add(upper).add(lower);
133:                middle = middle.shiftLeft(ndiv2);
134:                upper = upper.shiftLeft(ndiv2 << 1);
135:
136:                return upper.add(middle).add(lower);
137:            }
138:
139:            /**
140:             * Multiplies two BigIntegers.
141:             * Implements traditional scholar algorithm described by Knuth.
142:             *
143:             * <br><tt>
144:             *         <table border="0">
145:             * <tbody>
146:             *
147:             *
148:             * <tr>
149:             * <td align="center">A=</td>
150:             * <td>a<sub>3</sub></td>
151:             * <td>a<sub>2</sub></td>
152:             * <td>a<sub>1</sub></td>
153:             * <td>a<sub>0</sub></td>
154:             * <td></td>
155:             * <td></td>
156:             * </tr>
157:             *
158:             *<tr>
159:             * <td align="center">B=</td>
160:             * <td></td>
161:             * <td>b<sub>2</sub></td>
162:             * <td>b<sub>1</sub></td>
163:             * <td>b<sub>1</sub></td>
164:             * <td></td>
165:             * <td></td>
166:             * </tr>
167:             *
168:             * <tr>
169:             * <td></td>
170:             * <td></td>
171:             * <td></td>
172:             * <td>b<sub>0</sub>*a<sub>3</sub></td>
173:             * <td>b<sub>0</sub>*a<sub>2</sub></td>
174:             * <td>b<sub>0</sub>*a<sub>1</sub></td>
175:             * <td>b<sub>0</sub>*a<sub>0</sub></td>
176:             * </tr>
177:             *
178:             * <tr>
179:             * <td></td>
180:             * <td></td>
181:             * <td>b<sub>1</sub>*a<sub>3</sub></td>
182:             * <td>b<sub>1</sub>*a<sub>2</sub></td>
183:             * <td>b<sub>1</sub>*a1</td>
184:             * <td>b<sub>1</sub>*a0</td>
185:             * </tr>
186:             *
187:             * <tr>
188:             * <td>+</td>
189:             * <td>b<sub>2</sub>*a<sub>3</sub></td>
190:             * <td>b<sub>2</sub>*a<sub>2</sub></td>
191:             * <td>b<sub>2</sub>*a<sub>1</sub></td>
192:             * <td>b<sub>2</sub>*a<sub>0</sub></td>
193:             * </tr>
194:             *
195:             *<tr>
196:             * <td></td>
197:             *<td>______</td>
198:             * <td>______</td>
199:             * <td>______</td>
200:             * <td>______</td>
201:             * <td>______</td>
202:             * <td>______</td>
203:             *</tr>
204:             *
205:             * <tr>
206:             *
207:             * <td align="center">A*B=R=</td>
208:             * <td align="center">r<sub>5</sub></td>
209:             * <td align="center">r<sub>4</sub></td>
210:             * <td align="center">r<sub>3</sub></td>
211:             * <td align="center">r<sub>2</sub></td>
212:             * <td align="center">r<sub>1</sub></td>
213:             * <td align="center">r<sub>0</sub></td>
214:             * <td></td>
215:             * </tr>
216:             *
217:             * </tbody>
218:             * </table>
219:             *
220:             *</tt>
221:             *
222:             * @param op1 first factor of the multiplication {@code  op1 >= 0}
223:             * @param op2 second factor of the multiplication {@code  op2 >= 0}
224:             * @return a {@code BigInteger} of value {@code  op1 * op2}
225:             */
226:            static BigInteger multiplyPAP(BigInteger a, BigInteger b) {
227:                // PRE: a >= b
228:                int aLen = a.numberLength;
229:                int bLen = b.numberLength;
230:                int resLength = aLen + bLen;
231:                int resSign = (a.sign != b.sign) ? -1 : 1;
232:                // A special case when both numbers don't exceed int
233:                if (resLength == 2) {
234:                    long val = (a.digits[0] & 0xFFFFFFFFL)
235:                            * (b.digits[0] & 0xFFFFFFFFL);
236:                    int valueLo = (int) val;
237:                    int valueHi = (int) (val >>> 32);
238:                    return ((valueHi == 0) ? new BigInteger(resSign, valueLo)
239:                            : new BigInteger(resSign, 2, new int[] { valueLo,
240:                                    valueHi }));
241:                }
242:                int[] aDigits = a.digits;
243:                int[] bDigits = b.digits;
244:                int resDigits[] = new int[resLength];
245:                long carry;
246:                long bDigit;
247:                int i, j, m;
248:                // Common case
249:                for (j = 0; j < bLen; j++) {
250:                    carry = 0;
251:                    bDigit = (bDigits[j] & 0xFFFFFFFFL);
252:                    for (i = 0, m = j; i < aLen; i++, m++) {
253:                        carry += (aDigits[i] & 0xFFFFFFFFL) * bDigit
254:                                + (resDigits[m] & 0xFFFFFFFFL);
255:                        resDigits[m] = (int) carry;
256:                        carry >>>= 32;
257:                    }
258:                    resDigits[m] = (int) carry;
259:                }
260:                BigInteger result = new BigInteger(resSign, resLength,
261:                        resDigits);
262:                result.cutOffLeadingZeroes();
263:                return result;
264:            }
265:
266:            /**
267:             * Multiplies an array of integers by an integer value
268:             * and saves the result in {@code res}.
269:             * @param a the array of integers
270:             * @param aSize the number of elements of intArray to be multiplied
271:             * @param factor the multiplier
272:             * @return the top digit of production
273:             */
274:            private static int multiplyByInt(int res[], int a[],
275:                    final int aSize, final int factor) {
276:                long carry = 0;
277:
278:                for (int i = 0; i < aSize; i++) {
279:                    carry += (a[i] & 0xFFFFFFFFL) * (factor & 0xFFFFFFFFL);
280:                    res[i] = (int) carry;
281:                    carry >>>= 32;
282:                }
283:                return (int) carry;
284:            }
285:
286:            /**
287:             * Multiplies an array of integers by an integer value.
288:             * @param a the array of integers
289:             * @param aSize the number of elements of intArray to be multiplied
290:             * @param factor the multiplier
291:             * @return the top digit of production
292:             */
293:            static int multiplyByInt(int a[], final int aSize, final int factor) {
294:                return multiplyByInt(a, a, aSize, factor);
295:            }
296:
297:            /**
298:             * Multiplies a number by a positive integer.
299:             * @param val an arbitrary {@code BigInteger}
300:             * @param factor a positive {@code int} number
301:             * @return {@code val * factor}
302:             */
303:            static BigInteger multiplyByPositiveInt(BigInteger val, int factor) {
304:                int resSign = val.sign;
305:                if (resSign == 0) {
306:                    return BigInteger.ZERO;
307:                }
308:                int aNumberLength = val.numberLength;
309:                int[] aDigits = val.digits;
310:
311:                if (aNumberLength == 1) {
312:                    long res = (aDigits[0] & 0xFFFFFFFFL) * (factor);
313:                    int resLo = (int) res;
314:                    int resHi = (int) (res >>> 32);
315:                    return ((resHi == 0) ? new BigInteger(resSign, resLo)
316:                            : new BigInteger(resSign, 2, new int[] { resLo,
317:                                    resHi }));
318:                }
319:                // Common case
320:                int resLength = aNumberLength + 1;
321:                int resDigits[] = new int[resLength];
322:
323:                resDigits[aNumberLength] = multiplyByInt(resDigits, aDigits,
324:                        aNumberLength, factor);
325:                BigInteger result = new BigInteger(resSign, resLength,
326:                        resDigits);
327:                result.cutOffLeadingZeroes();
328:                return result;
329:            }
330:
331:            static BigInteger pow(BigInteger base, int exponent) {
332:                // PRE: exp > 0
333:                BigInteger res = BigInteger.ONE;
334:                BigInteger acc = base;
335:
336:                for (; exponent > 1; exponent >>= 1) {
337:                    if ((exponent & 1) != 0) {
338:                        // if odd, multiply one more time by acc
339:                        res = res.multiply(acc);
340:                    }
341:                    // acc = base^(2^i)
342:                    //a limit where karatsuba performs a faster square than the square algorithm
343:                    if (acc.numberLength == 1) {
344:                        acc = acc.multiply(acc); // square
345:                    } else {
346:                        acc = new BigInteger(1, square(acc.digits,
347:                                acc.numberLength));
348:                    }
349:                }
350:                // exponent == 1, multiply one more time
351:                res = res.multiply(acc);
352:                return res;
353:            }
354:
355:            /**
356:             *  Performs a<sup>2</sup>
357:             *  @param a The number to square.
358:             *  @param length The length of the number to square.
359:             */
360:            static int[] square(int[] a, int s) {
361:                int[] t = new int[s << 1];
362:                long cs;
363:                long aI;
364:                for (int i = 0; i < s; i++) {
365:                    cs = 0;
366:                    aI = (0xFFFFFFFFL & a[i]);
367:                    for (int j = i + 1; j < s; j++) {
368:                        cs += (0xFFFFFFFFL & t[i + j]) + aI
369:                                * (0xFFFFFFFFL & a[j]);
370:                        t[i + j] = (int) cs;
371:                        cs >>>= 32;
372:                    }
373:
374:                    t[i + s] = (int) cs;
375:                }
376:                BitLevel.shiftLeft(t, t, 0, 1);
377:                cs = 0;
378:
379:                for (int i = 0, index = 0; i < s; i++, index++) {
380:                    aI = (0xFFFFFFFFL & a[i]);
381:                    cs += aI * aI + (t[index] & 0xFFFFFFFFL);
382:                    t[index] = (int) cs;
383:                    cs >>>= 32;
384:                    index++;
385:                    cs += t[index] & 0xFFFFFFFFL;
386:                    t[index] = (int) cs;
387:                    cs >>>= 32;
388:                }
389:                return t;
390:            }
391:
392:            /**
393:             * Multiplies a number by a power of ten.
394:             * This method is used in {@code BigDecimal} class.
395:             * @param val the number to be multiplied
396:             * @param exp a positive {@code long} exponent
397:             * @return {@code val * 10<sup>exp</sup>}
398:             */
399:            static BigInteger multiplyByTenPow(BigInteger val, long exp) {
400:                // PRE: exp >= 0
401:                return ((exp < tenPows.length) ? multiplyByPositiveInt(val,
402:                        tenPows[(int) exp]) : val.multiply(powerOf10(exp)));
403:            }
404:
405:            /**
406:             * It calculates a power of ten, which exponent could be out of 32-bit range.
407:             * Note that internally this method will be used in the worst case with
408:             * an exponent equals to: {@code Integer.MAX_VALUE - Integer.MIN_VALUE}.
409:             * @param exp the exponent of power of ten, it must be positive.
410:             * @return a {@code BigInteger} with value {@code 10<sup>exp</sup>}.
411:             */
412:            static BigInteger powerOf10(long exp) {
413:                // PRE: exp >= 0
414:                int intExp = (int) exp;
415:                // "SMALL POWERS"
416:                if (exp < bigTenPows.length) {
417:                    // The largest power that fit in 'long' type
418:                    return bigTenPows[intExp];
419:                } else if (exp <= 50) {
420:                    // To calculate:    10^exp
421:                    return BigInteger.TEN.pow(intExp);
422:                } else if (exp <= 1000) {
423:                    // To calculate:    5^exp * 2^exp
424:                    return bigFivePows[1].pow(intExp).shiftLeft(intExp);
425:                }
426:                // "LARGE POWERS"
427:                /*
428:                 * To check if there is free memory to allocate a BigInteger of the
429:                 * estimated size, measured in bytes: 1 + [exp / log10(2)]
430:                 */
431:                long byteArraySize = 1 + (long) (exp / 2.4082399653118496);
432:
433:                if (byteArraySize > Runtime.getRuntime().freeMemory()) {
434:                    // math.01=power of ten too big
435:                    throw new OutOfMemoryError(Messages.getString("math.01")); //$NON-NLS-1$
436:                }
437:                if (exp <= Integer.MAX_VALUE) {
438:                    // To calculate:    5^exp * 2^exp
439:                    return bigFivePows[1].pow(intExp).shiftLeft(intExp);
440:                }
441:                /*
442:                 * "HUGE POWERS"
443:                 * 
444:                 * This branch probably won't be executed since the power of ten is too
445:                 * big.
446:                 */
447:                // To calculate:    5^exp
448:                BigInteger powerOfFive = bigFivePows[1].pow(Integer.MAX_VALUE);
449:                BigInteger res = powerOfFive;
450:                long longExp = exp - Integer.MAX_VALUE;
451:
452:                intExp = (int) (exp % Integer.MAX_VALUE);
453:                while (longExp > Integer.MAX_VALUE) {
454:                    res = res.multiply(powerOfFive);
455:                    longExp -= Integer.MAX_VALUE;
456:                }
457:                res = res.multiply(bigFivePows[1].pow(intExp));
458:                // To calculate:    5^exp << exp
459:                res = res.shiftLeft(Integer.MAX_VALUE);
460:                longExp = exp - Integer.MAX_VALUE;
461:                while (longExp > Integer.MAX_VALUE) {
462:                    res = res.shiftLeft(Integer.MAX_VALUE);
463:                    longExp -= Integer.MAX_VALUE;
464:                }
465:                res = res.shiftLeft(intExp);
466:                return res;
467:            }
468:
469:            /**
470:             * Multiplies a number by a power of five.
471:             * This method is used in {@code BigDecimal} class.
472:             * @param val the number to be multiplied
473:             * @param exp a positive {@code int} exponent
474:             * @return {@code val * 5<sup>exp</sup>}
475:             */
476:            static BigInteger multiplyByFivePow(BigInteger val, int exp) {
477:                // PRE: exp >= 0
478:                if (exp < fivePows.length) {
479:                    return multiplyByPositiveInt(val, fivePows[exp]);
480:                } else if (exp < bigFivePows.length) {
481:                    return val.multiply(bigFivePows[exp]);
482:                } else {// Large powers of five
483:                    return val.multiply(bigFivePows[1].pow(exp));
484:                }
485:            }
486:        }
www.java2java.com | Contact Us
Copyright 2009 - 12 Demo Source and Support. All rights reserved.
All other trademarks are property of their respective owners.