记录一次并没有什么用的对比测试

一、假设有一个这样的数据对象需要不断的产生，销毁

public class TestData {
　　private String id;
　　private String val;

　　public String getId() {
　　　　return id;
　　}

　　public void setId(String id) {
　　　　this.id = id;
　　}

　　public String getVal() {
　　　　return val;
　　}

　　public void setVal(String val) {
　　　　this.val = val;
　　}
}

二、构造一些那样的对象，分别存放在map和list中

private ConcurrentMap<String, TestData> makeTestMap(int count) {
    ConcurrentMap<String, TestData> map = new ConcurrentHashMap<String, TestData>();
    for (int i = 0; i < count; i++) {
        TestData data = new TestData();
        data.setId(String.valueOf(i));
        data.setVal(UUID.randomUUID().toString());
        map.put(String.valueOf(i), data);
    }
    return map;
}

private List<TestData> makeTestList(int count) {
    List<TestData> list = new ArrayList<TestData>();
    for (int i = 0; i < count; i++) {
        TestData data = new TestData();
        data.setId(String.valueOf(i));
        data.setVal(UUID.randomUUID().toString());
        list.add(data);
    }
    return list;
}

三、从map和List中随机的取一个，并返回花费时间

public long getMapTime(String posStr, Map<String, TestData> map) {
    long t = System.nanoTime();
    map.get(posStr).getVal();
    // System.out.println("map get!!! " + map.get(posStr).getVal());
    return System.nanoTime() - t;
}

public long getListTime(String posStr, List<TestData> list) {
    long t = System.nanoTime();
    for (TestData td : list) {
        if (posStr.equals(td.getId())) {
        td.getVal();
        // System.out.println("list get!!! " + td.getVal());
        break;
        }
    }
    return System.nanoTime() - t;
}

四、随机从map和List中分别取一百次测试数据的值，计算平均花费时长

@Test
public void testPerformance() {
    int CAPACITY = 50000;
    int TEST_COUNT = 100;
    ConcurrentMap<String, TestData> map = makeTestMap(CAPACITY);
    List<TestData> list = makeTestList(CAPACITY);
    long mapSum = 0L;
    long listSum = 0L;
    for (int i = 0; i < TEST_COUNT; i++) {
        int pos = new Random().nextInt(CAPACITY);
        String posStr = String.valueOf(pos);
        // System.out.println(i+".("+posStr+")");
        // System.out.println("MAP take - " + testMap(posStr, map) + " ns");
        // System.out.println("LIST take - " + testList(posStr, list) + " ns");
        mapSum += getMapTime(posStr, map);
        listSum += getListTime(posStr,list);
    }
    System.out.println("map average take "+ mapSum/100 + " ns");
    System.out.println("list average take "+ listSum/100 + " ns");
}

五、一些测试数据

// 容器中共500条数据
// map average take 2011 ns
// list average take 109312 ns

// 容器中共5000条数据
// map average take 3020 ns
// list average take 408248 ns

// 容器中共50000条数据
// map average take 3917 ns
// list average take 1452222 ns

// 容器中共500000条数据
// map average take 7548 ns
// list average take 9193437 ns