转载请注明出处:http://blog.csdn.net/xiaojimanman/article/details/38924981
近期在项目工作中有一个关于文本对照的需求,经过这段时间的学习,总结了这篇博客内容:求两个字符串的最大公共子串。
算法思想:基于图计算两字符串的公共子串。详细算法思想參照下图:
输入字符串S1:achmacmh 输入字符串S2:macham
1)第a步,是将字符串s1,s2分别按字节拆分,构成一个二维数组;
2)二维数组中的值如b所看到的,比方第一行第一列的值表示字符串s2和s1的第一个字节是否相等,若相等就是1,否则就是0,终于产生b所看到的的二维数组;
3)分别求二维数组中斜线上的公共因子(斜线为元素a右下角值,即a[i][j]的下一个元素是a[i+1][j+1];公共因子为1所在的位置构成的字符串);
4)对全部公共因子排序,返回最大的公共因子的值。
详细的实现代码例如以下所看到的:
package cn.lulei.compare; import java.util.ArrayList; import java.util.Collections; import java.util.Comparator; import java.util.List; public class StringCompare { private int a; private int b; public String getMaxLengthCommonString(String s1, String s2) { if (s1 == null || s2 == null) { return null; } a = s1.length();//s1长度做行 b = s2.length();//s2长度做列 if(a== 0 || b == 0) { return ""; } //设置匹配矩阵 boolean [][] array = new boolean[a][b]; for (int i = 0; i < a; i++) { char c1 = s1.charAt(i); for (int j = 0; j < b; j++) { char c2 = s2.charAt(j); if (c1 == c2) { array[i][j] = true; } else { array[i][j] = false; } } } //求全部公因子字符串,保存信息为相对第二个字符串的起始位置和长度 List<ChildString> childStrings = new ArrayList<ChildString>(); for (int i = 0; i < a; i++) { getMaxSort(i, 0, array, childStrings); } for (int i = 1; i < b; i++) { getMaxSort(0, i, array, childStrings); } //排序 sort(childStrings); if (childStrings.size() < 1) { return ""; } //返回最大公因子字符串 int max = childStrings.get(0).maxLength; StringBuffer sb = new StringBuffer(); for (ChildString s: childStrings) { if (max != s.maxLength) { break; } sb.append(s2.substring(s.maxStart, s.maxStart + s.maxLength)); sb.append("\n"); } return sb.toString(); } //排序,倒叙 private void sort(List<ChildString> list) { Collections.sort(list, new Comparator<ChildString>(){ public int compare(ChildString o1, ChildString o2) { return o2.maxLength - o1.maxLength; } }); } //求一条斜线上的公因子字符串 private void getMaxSort(int i, int j, boolean [][] array, List<ChildString> sortBean) { int length = 0; int start = j; for (; i < a && j < b; i++,j++) { if (array[i][j]) { length++; } else { sortBean.add(new ChildString(length, start)); length = 0; start = j + 1; } if (i == a-1 || j == b-1) { sortBean.add(new ChildString(length, start)); } } } //公因子类 class ChildString { int maxLength; int maxStart; ChildString(int maxLength, int maxStart){ this.maxLength = maxLength; this.maxStart = maxStart; } } /** * @param args */ public static void main(String[] args) { // TODO Auto-generated method stub System.out.println(new StringCompare().getMaxLengthCommonString("achmacmh", "macham")); } }
程序终于运行结果是:
对于两个文件的比对个人觉得能够參照这样的算法思想(自己如今并为实现),在日后的博客中将会写到。
上述实现过程中,用数组保存了全部的公共子串信息,然后排序取最大的子串,这样的做法假设仅仅是求最大子串的话,算法就不是非常合理,因此做了例如以下改动,List仅仅保存当前计算中最大的子串,详细实现例如以下:
/** *@Description: 字符串比較 */ package com.lulei.test; import java.util.ArrayList; import java.util.List; public class StringCompare { private int a; private int b; private int maxLength = -1; public String getMaxLengthCommonString(String s1, String s2) { if (s1 == null || s2 == null) { return null; } a = s1.length();//s1长度做行 b = s2.length();//s2长度做列 if(a== 0 || b == 0) { return ""; } //设置匹配矩阵 boolean [][] array = new boolean[a][b]; for (int i = 0; i < a; i++) { char c1 = s1.charAt(i); for (int j = 0; j < b; j++) { char c2 = s2.charAt(j); if (c1 == c2) { array[i][j] = true; } else { array[i][j] = false; } } } //求全部公因子字符串,保存信息为相对第二个字符串的起始位置和长度 List<ChildString> childStrings = new ArrayList<ChildString>(); for (int i = 0; i < a; i++) { getMaxSort(i, 0, array, childStrings); } for (int i = 1; i < b; i++) { getMaxSort(0, i, array, childStrings); } StringBuffer sb = new StringBuffer(); for (ChildString s: childStrings) { sb.append(s2.substring(s.maxStart, s.maxStart + s.maxLength)); sb.append("\n"); } return sb.toString(); } //求一条斜线上的公因子字符串 private void getMaxSort(int i, int j, boolean [][] array, List<ChildString> sortBean) { int length = 0; int start = j; for (; i < a && j < b; i++,j++) { if (array[i][j]) { length++; } else { //直接add,保存全部子串,以下的推断,仅仅保存当前最大的子串 //sortBean.add(new ChildString(length, start)); if (length == maxLength) { sortBean.add(new ChildString(length, start)); } else if (length > maxLength) { sortBean.clear(); maxLength = length; sortBean.add(new ChildString(length, start)); } length = 0; start = j + 1; } if (i == a-1 || j == b-1) { //直接add,保存全部子串,以下的推断,仅仅保存当前最大的子串 //sortBean.add(new ChildString(length, start)); if (length == maxLength) { sortBean.add(new ChildString(length, start)); } else if (length > maxLength) { sortBean.clear(); maxLength = length; sortBean.add(new ChildString(length, start)); } } } } //公因子类 class ChildString { int maxLength; int maxStart; ChildString(int maxLength, int maxStart){ this.maxLength = maxLength; this.maxStart = maxStart; } } /** * @param args */ public static void main(String[] args) { // TODO Auto-generated method stub System.out.println(new StringCompare().getMaxLengthCommonString("abcdef", "defabc")); } }