有四种序列化方式。
CamelCase策略,Java对象属性:personId,序列化后属性:persionId – 实际只改了首字母 大写变小写
PascalCase策略,Java对象属性:personId,序列化后属性:PersonId – 实际只改了首字母 小写变大写
SnakeCase策略,Java对象属性:personId,序列化后属性:person_id --大写字母前加下划线
KebabCase策略,Java对象属性:personId,序列化后属性:person-id -大写字母前加减号
public enum PropertyNamingStrategy {
CamelCase, //驼峰
PascalCase, //
SnakeCase, //大写字母前加下划线
KebabCase;
public String translate(String propertyName) {
switch (this) {
case SnakeCase: {
StringBuilder buf = new StringBuilder();
for (int i = 0; i < propertyName.length(); ++i) {
char ch = propertyName.charAt(i);
if (ch >= 'A' && ch <= 'Z') {
char ch_ucase = (char) (ch + 32);
if (i > 0) {
buf.append('_');
}
buf.append(ch_ucase);
} else {
buf.append(ch);
}
}
return buf.toString();
}
case KebabCase: {
StringBuilder buf = new StringBuilder();
for (int i = 0; i < propertyName.length(); ++i) {
char ch = propertyName.charAt(i);
if (ch >= 'A' && ch <= 'Z') {
char ch_ucase = (char) (ch + 32);
if (i > 0) {
buf.append('-');
}
buf.append(ch_ucase);
} else {
buf.append(ch);
}
}
return buf.toString();
}
case PascalCase: {
char ch = propertyName.charAt(0);
if (ch >= 'a' && ch <= 'z') {
char[] chars = propertyName.toCharArray();
chars[0] -= 32;
return new String(chars);
}
return propertyName;
}
case CamelCase: {
char ch = propertyName.charAt(0);
if (ch >= 'A' && ch <= 'Z') {
char[] chars = propertyName.toCharArray();
chars[0] += 32;
return new String(chars);
}
return propertyName;
}
default:
return propertyName;
}
}
发挥作用的是translate方法
了解了PropertyNamingStrategy后,看其是怎么发挥作用的,
阅读源码发现在buildBeanInfo时(注意是将bean转为json时构建json信息时,如果是map,JSONObject不会有这个转换)
if(propertyNamingStrategy != null && !fieldAnnotationAndNameExists){
propertyName = propertyNamingStrategy.translate(propertyName);
}
这里分别调用PropertyNamingStrategy对应的方法处理
常见误区
那么也就是说通过PropertyNamingStrategy的方式设置输出格式,只对javaBean有效,并且,至于转换结果,需要根据PropertyNamingStrategy#translate方法的内容具体分析
如果javaBean中的字段是用下划线间隔的,那么指定CamelCase进行序列化,也是无法转成驼峰的!
例如
Student student = new Student();
student.setTest_name("test");
SerializeConfig serializeConfig = new SerializeConfig();
serializeConfig.setPropertyNamingStrategy(PropertyNamingStrategy.CamelCase);
System.out.println(JSON.toJSONString(student,serializeConfig));
输出{test_name":“test”},因为执行 PropertyNamingStrategy#translate的CamelCase,仅仅只是,判断如果首字母大写转成小写。并不能完成,下划线到驼峰的转换
case CamelCase: {
char ch = propertyName.charAt(0);
if (ch >= 'A' && ch <= 'Z') {
char[] chars = propertyName.toCharArray();
chars[0] += 32;
return new String(chars);
}
return propertyName;
}
fastjson反序列化时,是能自动下划线转驼峰的。这点是很方便的。,在反序列化时无论采用那种形式都能匹配成功并设置值
String str = "{'user_name':123}";
User user = JSON.parseObject(str, User.class);
System.out.println(user);
输出{userName=‘123’}
fastjson在进行反序列化的时候,对每一个json字段的key值解析时,会调用
com.alibaba.fastjson.parser.deserializer.JavaBeanDeserializer#parseField
这个方法
以上面的例子为例,通过debug打个断点看一下解析user_id时的处理逻辑。
此时这个方法中的key为user_id,object为要反序列化的结果对象,这个例子中就是FastJsonTestMain.UserInfo
public boolean parseField(DefaultJSONParser parser, String key, Object object, Type objectType,
Map<String, Object> fieldValues, int[] setFlags) {
JSONLexer lexer = parser.lexer; // xxx
//是否禁用智能匹配;
final int disableFieldSmartMatchMask = Feature.DisableFieldSmartMatch.mask;
final int initStringFieldAsEmpty = Feature.InitStringFieldAsEmpty.mask;
FieldDeserializer fieldDeserializer;
if (lexer.isEnabled(disableFieldSmartMatchMask) || (this.beanInfo.parserFeatures & disableFieldSmartMatchMask) != 0) {
fieldDeserializer = getFieldDeserializer(key);
} else if (lexer.isEnabled(initStringFieldAsEmpty) || (this.beanInfo.parserFeatures & initStringFieldAsEmpty) != 0) {
fieldDeserializer = smartMatch(key);
} else {
//进行智能匹配
fieldDeserializer = smartMatch(key, setFlags);
}
***此处省略N多行***
}
再看下核心的代码,智能匹配smartMatch
public FieldDeserializer smartMatch(String key, int[] setFlags) {
if (key == null) {
return null;
}
FieldDeserializer fieldDeserializer = getFieldDeserializer(key, setFlags);
if (fieldDeserializer == null) {
if (this.smartMatchHashArray == null) {
long[] hashArray = new long[sortedFieldDeserializers.length];
for (int i = 0; i < sortedFieldDeserializers.length; i++) {
//java字段的nameHashCode,源码见下方
hashArray[i] = sortedFieldDeserializers[i].fieldInfo.nameHashCode;
}
//获取出反序列化目标对象的字段名称hashcode值,并进行排序
Arrays.sort(hashArray);
this.smartMatchHashArray = hashArray;
}
// smartMatchHashArrayMapping
long smartKeyHash = TypeUtils.fnv1a_64_lower(key);
//进行二分查找,判断是否找到
int pos = Arrays.binarySearch(smartMatchHashArray, smartKeyHash);
if (pos < 0) {
//原始字段没有匹配到,用fnv1a_64_extract处理一下再次匹配
long smartKeyHash1 = TypeUtils.fnv1a_64_extract(key);
pos = Arrays.binarySearch(smartMatchHashArray, smartKeyHash1);
}
boolean is = false;
if (pos < 0 && (is = key.startsWith("is"))) {
//上面的操作后仍然没有匹配到,把is去掉后再次进行匹配
smartKeyHash = TypeUtils.fnv1a_64_extract(key.substring(2));
pos = Arrays.binarySearch(smartMatchHashArray, smartKeyHash);
}
if (pos >= 0) {
//通过智能匹配字段匹配成功
if (smartMatchHashArrayMapping == null) {
short[] mapping = new short[smartMatchHashArray.length];
Arrays.fill(mapping, (short) -1);
for (int i = 0; i < sortedFieldDeserializers.length; i++) {
int p = Arrays.binarySearch(smartMatchHashArray, sortedFieldDeserializers[i].fieldInfo.nameHashCode);
if (p >= 0) {
mapping[p] = (short) i;
}
}
smartMatchHashArrayMapping = mapping;
}
int deserIndex = smartMatchHashArrayMapping[pos];
if (deserIndex != -1) {
if (!isSetFlag(deserIndex, setFlags)) {
fieldDeserializer = sortedFieldDeserializers[deserIndex];
}
}
}
if (fieldDeserializer != null) {
FieldInfo fieldInfo = fieldDeserializer.fieldInfo;
if ((fieldInfo.parserFeatures & Feature.DisableFieldSmartMatch.mask) != 0) {
return null;
}
Class fieldClass = fieldInfo.fieldClass;
if (is && (fieldClass != boolean.class && fieldClass != Boolean.class)) {
fieldDeserializer = null;
}
}
}
return fieldDeserializer;
}
通过上面的smartMatch方法可以看出,fastjson中之所以能做到下划线自动转驼峰,主要还是因为在进行字段对比时,使用了fnv1a_64_lower和fnv1a_64_extract方法进行了处理。
fnv1a_64_extract方法源码:
public static long fnv1a_64_extract(String key) {
long hashCode = fnv1a_64_magic_hashcode;
for (int i = 0; i < key.length(); ++i) {
char ch = key.charAt(i);
//去掉下划线和减号
if (ch == '_' || ch == '-') {
continue;
}
//大写转小写
if (ch >= 'A' && ch <= 'Z') {
ch = (char) (ch + 32);
}
hashCode ^= ch;
hashCode *= fnv1a_64_magic_prime;
}
return hashCode;
}
从源码可以看出,fnv1a_64_extract方法主要做了这个事:
去掉下划线、减号,并大写转小写
总结
fastjson中字段智能匹配的原理是在字段匹配时,使用了TypeUtils.fnv1a_64_lower方法对字段进行全体转小写处理。
之后再用TypeUtils.fnv1a_64_extract方法对json字段进行去掉"_“和”-"符号,再全体转小写处理。
如果上面的操作仍然没有匹配成功,会再进行一次去掉json字段中的is再次进行匹配。
如果上面的操作仍然没有匹配成功,会再进行一次去掉json字段中的is再次进行匹配。
智能匹配时默认开启的,需要手动关闭,看这个例子
String str = "{'user_name':123}";
ParserConfig parserConfig = new ParserConfig();
parserConfig.propertyNamingStrategy = PropertyNamingStrategy.SnakeCase;
User user = JSON.parseObject(str, User.class, parserConfig,Feature.DisableFieldSmartMatch);
System.out.println(user);
输出{userName=‘null’}
那么这种情况如何完成下划线到驼峰的转换
那么就需要使用parseConfig了
String str = "{'user_name':123}";
ParserConfig parserConfig = new ParserConfig();
parserConfig.propertyNamingStrategy = PropertyNamingStrategy.SnakeCase;
User user = JSON.parseObject(str, User.class,parserConfig,Feature.DisableFieldSmartMatch);
System.out.println(user);
那么此时PropertyNamingStrategy.SnakeCase又是如何发挥作用的?
断点PropertyNamingStrategy#translate方法
发现在构建JavaBeanDeserializer时
public JavaBeanDeserializer(ParserConfig config, Class<?> clazz, Type type){
this(config //
, JavaBeanInfo.build(clazz, type, config.propertyNamingStrategy, config.fieldBased, config.compatibleWithJavaBean, config.isJacksonCompatible())
);
}
if (propertyNamingStrategy != null) {
propertyName = propertyNamingStrategy.translate(propertyName);
}
add(fieldList, new FieldInfo(propertyName, method, field, clazz, type, ordinal, serialzeFeatures, parserFeatures,
annotation, fieldAnnotation, null, genericInfo));
会根据配置对propertyName进行translate。转换成对应格式的属性名称
常见误区:
与序列化误区相同,如果是map,JSONObject不会有这个转换,并且转换结果需要参照translate方方法逻辑来看
值的注意的是,JSONObject的toJavaObject方法,智能匹配会生效。可以放心得进行下划线和驼峰得互相转换
String str = "{'user_name':123}";
JSONObject object = (JSONObject) JSON.parse(str);
System.out.println(object);
User user = object.toJavaObject(User.class);
System.out.println(user);