String的substring、split，外加StringTokenizer三者截取字符串性能测试

最终，StringTokenizer在截取字符串中效率最高，不论数据量大小，几乎持平。substring则要次之，数据量增加耗时也要随之增加。split则是表现最差劲的。

究其原因，split的实现方式是采用正则表达式实现，所以其性能会比较低。至于正则表达式为何低，还未去验证。split源码如下：

public String[] split(String regex, int limit) {
        return Pattern.compile(regex).split(this, limit);
    }

/***********************************************************************************************************************************************************************************/

package test.java.lang.ref;

import java.util.Random;
import java.util.StringTokenizer;

/**
 * String测试类
 *
 *
 */
public class StringTest {

    public static void main(String args[]) {
        String orginStr = getOriginStr(10);

        // ////////////String.splic()表现//////////////////////////////////////////////
        System.out.println("使用String.splic()的切分字符串");
        long st1 = System.nanoTime();
        String[] result = orginStr.split("\\.");
        System.out.println("String.splic()截取字符串用时：" + (System.nanoTime() - st1));
        System.out.println("String.splic()截取字符串结果个数：" + result.length);
        System.out.println();

        // ////////////StringTokenizer表现//////////////////////////////////////////////
        System.out.println("使用StringTokenizer的切分字符串");
        long st3 = System.nanoTime();
        StringTokenizer token = new StringTokenizer(orginStr, ".");
        System.out.println("StringTokenizer截取字符串用时:" + (System.nanoTime() - st3));
        System.out.println("StringTokenizer截取字符串结果个数：" + token.countTokens());
        System.out.println();

        // //////////////////String.substring()表现//////////////////////////////////////////

        long st5 = System.nanoTime();
        int len = orginStr.lastIndexOf(".");
        System.out.println("使用String.substring()切分字符串");
        int k = 0, count = 0;

        for (int i = 0; i <= len; i++) {
            if (orginStr.substring(i, i + 1).equals(".")) {
                if (count == 0) {
                    orginStr.substring(0, i);
                } else {
                    orginStr.substring(k + 1, i);
                    if (i == len) {
                        orginStr.substring(len + 1, orginStr.length());
                    }
                }
                k = i;
                count++;
            }
        }
        System.out.println("String.substring()截取字符串用时" + (System.nanoTime() - st5));
        System.out.println("String.substring()截取字符串结果个数：" + (count + 1));
    }

    /**
     * 构造目标字符串 eg：10.123.12.154.154
     *
     * @param len
     *            目标字符串组数(每组由3个随机数组成)
     * @return
     */
    private static String getOriginStr(int len) {

        StringBuffer sb = new StringBuffer();
        StringBuffer result = new StringBuffer();
        Random random = new Random();
        for (int i = 0; i < len; i++) {
            sb.append(random.nextInt(9)).append(random.nextInt(9)).append(random.nextInt(9));
            result.append(sb.toString());
            sb.delete(0, sb.length());
            if (i != len - 1)
                result.append(".");
        }

        return result.toString();
    }
}

 

http://my.oschina.net/kyo153/blog/41829