@Hamedz 如果使用较多的数据，则使用Jackson进行转换 光数据，使用apache… TestCase:

import java.lang.reflect.InvocationTargetException; import java.util.HashMap; import java.util.Map; import org.apache.commons.beanutils.BeanUtils; import com.fasterxml.jackson.databind.ObjectMapper; import lombok.AllArgsConstructor; import lombok.Data; import lombok.NoArgsConstructor; public class TestPerf { public static final int LOOP_MAX_COUNT = 1000; public static void main(String[] args) { Map<String, Object> map = new HashMap<>(); map.put("success", true); map.put("number", 1000); map.put("longer", 1000L); map.put("doubler", 1000D); map.put("data1", "testString"); map.put("data2", "testString"); map.put("data3", "testString"); map.put("data4", "testString"); map.put("data5", "testString"); map.put("data6", "testString"); map.put("data7", "testString"); map.put("data8", "testString"); map.put("data9", "testString"); map.put("data10", "testString"); runBeanUtilsPopulate(map); runJacksonMapper(map); } private static void runBeanUtilsPopulate(Map<String, Object> map) { long t1 = System.currentTimeMillis(); for (int i = 0; i < LOOP_MAX_COUNT; i++) { try { TestClass bean = new TestClass(); BeanUtils.populate(bean, map); } catch (IllegalAccessException e) { e.printStackTrace(); } catch (InvocationTargetException e) { e.printStackTrace(); } } long t2 = System.currentTimeMillis(); System.out.println("BeanUtils t2-t1 = " + String.valueOf(t2 - t1)); } private static void runJacksonMapper(Map<String, Object> map) { long t1 = System.currentTimeMillis(); for (int i = 0; i < LOOP_MAX_COUNT; i++) { ObjectMapper mapper = new ObjectMapper(); TestClass testClass = mapper.convertValue(map, TestClass.class); } long t2 = System.currentTimeMillis(); System.out.println("Jackson t2-t1 = " + String.valueOf(t2 - t1)); } @Data @AllArgsConstructor @NoArgsConstructor public static class TestClass { private Boolean success; private Integer number; private Long longer; private Double doubler; private String data1; private String data2; private String data3; private String data4; private String data5; private String data6; private String data7; private String data8; private String data9; private String data10; } }

如果你的类中有泛型类型，你应该使用TypeReference和convertValue()。

final ObjectMapper mapper = new ObjectMapper();
final MyPojo<MyGenericType> pojo = mapper.convertValue(map, new TypeReference<MyPojo<MyGenericType>>() {});

你也可以用它把pojo转换回java.util.Map。

final ObjectMapper mapper = new ObjectMapper();
final Map<String, Object> map = mapper.convertValue(pojo, new TypeReference<Map<String, Object>>() {});

我对Jackson和BeanUtils都进行了测试，发现BeanUtils要快得多。 在我的机器(Windows8.1, JDK1.7)中，我得到了这个结果。

BeanUtils t2-t1 = 286
Jackson t2-t1 = 2203

public class MainMapToPOJO {

public static final int LOOP_MAX_COUNT = 1000;

public static void main(String[] args) {
    Map<String, Object> map = new HashMap<>();
    map.put("success", true);
    map.put("data", "testString");

    runBeanUtilsPopulate(map);

    runJacksonMapper(map);
}

private static void runBeanUtilsPopulate(Map<String, Object> map) {
    long t1 = System.currentTimeMillis();
    for (int i = 0; i < LOOP_MAX_COUNT; i++) {
        try {
            TestClass bean = new TestClass();
            BeanUtils.populate(bean, map);
        } catch (IllegalAccessException e) {
            e.printStackTrace();
        } catch (InvocationTargetException e) {
            e.printStackTrace();
        }
    }
    long t2 = System.currentTimeMillis();
    System.out.println("BeanUtils t2-t1 = " + String.valueOf(t2 - t1));
}

private static void runJacksonMapper(Map<String, Object> map) {
    long t1 = System.currentTimeMillis();
    for (int i = 0; i < LOOP_MAX_COUNT; i++) {
        ObjectMapper mapper = new ObjectMapper();
        TestClass testClass = mapper.convertValue(map, TestClass.class);
    }
    long t2 = System.currentTimeMillis();
    System.out.println("Jackson t2-t1 = " + String.valueOf(t2 - t1));
}}

是的，它绝对可以避免到JSON的中间转换。使用像Dozer这样的深度复制工具，你可以直接将地图转换为POJO。下面是一个简单的例子:

示例POJO:

public class MyPojo implements Serializable {
    private static final long serialVersionUID = 1L;

    private String id;
    private String name;
    private Integer age;
    private Double savings;

    public MyPojo() {
        super();
    }

    // Getters/setters

    @Override
    public String toString() {
        return String.format(
                "MyPojo[id = %s, name = %s, age = %s, savings = %s]", getId(),
                getName(), getAge(), getSavings());
    }
}

转换代码示例:

public class CopyTest {
    @Test
    public void testCopyMapToPOJO() throws Exception {
        final Map<String, String> map = new HashMap<String, String>(4);
        map.put("id", "5");
        map.put("name", "Bob");
        map.put("age", "23");
        map.put("savings", "2500.39");
        map.put("extra", "foo");

        final DozerBeanMapper mapper = new DozerBeanMapper();
        final MyPojo pojo = mapper.map(map, MyPojo.class);
        System.out.println(pojo);
    }
}

输出:

MyPojo[id = 5，名字= Bob，年龄= 23，存款= 2500.39]

注意:如果你将源映射更改为map <String, Object>，那么你可以复制任意深度嵌套的属性(使用map <String, String>，你只能得到一层)。

杰克逊也能做到这一点。(而且自从你考虑使用jackson后，它似乎更舒服)。

使用ObjectMapper的convertValue方法:

final ObjectMapper mapper = new ObjectMapper(); // jackson's objectmapper
final MyPojo pojo = mapper.convertValue(map, MyPojo.class);

不需要转换成JSON字符串或其他东西;直接转换要快得多。

@Hamedz 如果使用较多的数据，则使用Jackson进行转换 光数据，使用apache… TestCase:

import java.lang.reflect.InvocationTargetException; import java.util.HashMap; import java.util.Map; import org.apache.commons.beanutils.BeanUtils; import com.fasterxml.jackson.databind.ObjectMapper; import lombok.AllArgsConstructor; import lombok.Data; import lombok.NoArgsConstructor; public class TestPerf { public static final int LOOP_MAX_COUNT = 1000; public static void main(String[] args) { Map<String, Object> map = new HashMap<>(); map.put("success", true); map.put("number", 1000); map.put("longer", 1000L); map.put("doubler", 1000D); map.put("data1", "testString"); map.put("data2", "testString"); map.put("data3", "testString"); map.put("data4", "testString"); map.put("data5", "testString"); map.put("data6", "testString"); map.put("data7", "testString"); map.put("data8", "testString"); map.put("data9", "testString"); map.put("data10", "testString"); runBeanUtilsPopulate(map); runJacksonMapper(map); } private static void runBeanUtilsPopulate(Map<String, Object> map) { long t1 = System.currentTimeMillis(); for (int i = 0; i < LOOP_MAX_COUNT; i++) { try { TestClass bean = new TestClass(); BeanUtils.populate(bean, map); } catch (IllegalAccessException e) { e.printStackTrace(); } catch (InvocationTargetException e) { e.printStackTrace(); } } long t2 = System.currentTimeMillis(); System.out.println("BeanUtils t2-t1 = " + String.valueOf(t2 - t1)); } private static void runJacksonMapper(Map<String, Object> map) { long t1 = System.currentTimeMillis(); for (int i = 0; i < LOOP_MAX_COUNT; i++) { ObjectMapper mapper = new ObjectMapper(); TestClass testClass = mapper.convertValue(map, TestClass.class); } long t2 = System.currentTimeMillis(); System.out.println("Jackson t2-t1 = " + String.valueOf(t2 - t1)); } @Data @AllArgsConstructor @NoArgsConstructor public static class TestClass { private Boolean success; private Integer number; private Long longer; private Double doubler; private String data1; private String data2; private String data3; private String data4; private String data5; private String data6; private String data7; private String data8; private String data9; private String data10; } }