EasyExcel之动态表头导出不生效

今天接到一个优化需求，表格导出后的表头顺序和页面不一致，要优化成一致的。根据传入的字段，动态导出数据，并保证顺序。
我看到导出的实体类都有@ExcelProperty注解，同时也在官网查看了这注解的含义和使用。

@ExcelProperty有两个属性可以帮我们排序：index和order，所以我就想每次在去写excel的时候，对映射类字段的index去动态排序。
注意index的使用

• index的值相同时会抛出异常
• index的值不连续时会插入空白列

然后想直接用反射动态修改index

    /**** @param headList 前端上送的动态表头* @param clazz 导出实体类的class对象*/private static void handleExcelHead(List<String> headList, Class<?> clazz) {try {for (String excelHead : headList) {Field declaredField = clazz.getDeclaredField(excelHead);ExcelProperty annotation = declaredField.getAnnotation(ExcelProperty.class);InvocationHandler invocationHandler = Proxy.getInvocationHandler(annotation);Field field = invocationHandler.getClass().getDeclaredField("memberValues");field.setAccessible(true);Map<String, Object> memberValues = (Map<String, Object>) field.get(invocationHandler);memberValues.put("index", headList.indexOf(excelHead));}}catch (Exception e){e.printStackTrace();}}

完整代码如下：

public class Test {public static void main(String[] args) throws Exception {List<String> headList = new ArrayList<>();headList.add("name");headList.add("sex");headList.add("age");// 动态处理表头handleExcelHead(headList,User.class);EasyExcel.write("test.xlsx").head(User.class).sheet("test").includeColumnFiledNames(headList).doWrite(init());}/**** @param headList 前端上送的动态表头* @param clazz 导出实体类的class对象*/private static void handleExcelHead(List<String> headList, Class<?> clazz) {try {for (String excelHead : headList) {Field declaredField = clazz.getDeclaredField(excelHead);ExcelProperty annotation = declaredField.getAnnotation(ExcelProperty.class);InvocationHandler invocationHandler = Proxy.getInvocationHandler(annotation);Field field = invocationHandler.getClass().getDeclaredField("memberValues");field.setAccessible(true);Map<String, Object> memberValues = (Map<String, Object>) field.get(invocationHandler);memberValues.put("index", headList.indexOf(excelHead));}}catch (Exception e){e.printStackTrace();}}private static List<User> init() {List<User> userList = new ArrayList<>();for (int i = 0; i < 10; i++) {User user = new User();user.setName("test"+i);user.setSex("男");user.setAge(20 + i);userList.add(user);}return userList;}
}

代码运行后发现和预期结果一样。但是当我则调整表头顺序并增加表头后发现不生效了。后来debug发现由于使用的是class对象，只要这个jvm不重启或者这个对象不被回收修改的index就一直存在，所以我们每次应该把index恢复成默认值。但是发现还是不行。后来在想这框架应该是使用了缓存吧，毕竟导出实体是固定的。果然：

ClassUtils 这里使用了缓存

    public static final Map<Class<?>, FieldCache> FIELD_CACHE = new ConcurrentHashMap<>();private static FieldCache declaredFields(Class<?> clazz) {if (clazz == null) {return null;}return FIELD_CACHE.computeIfAbsent(clazz, key -> {List<Field> tempFieldList = new ArrayList<>();Class<?> tempClass = clazz;// When the parent class is null, it indicates that the parent class (Object class) has reached the top// level.while (tempClass != null) {Collections.addAll(tempFieldList, tempClass.getDeclaredFields());// Get the parent class and give it to yourselftempClass = tempClass.getSuperclass();}// Screening of fieldMap<Integer, List<Field>> orderFieldMap = new TreeMap<Integer, List<Field>>();Map<Integer, Field> indexFieldMap = new TreeMap<Integer, Field>();Map<String, Field> ignoreMap = new HashMap<String, Field>(16);ExcelIgnoreUnannotated excelIgnoreUnannotated = clazz.getAnnotation(ExcelIgnoreUnannotated.class);for (Field field : tempFieldList) {declaredOneField(field, orderFieldMap, indexFieldMap, ignoreMap, excelIgnoreUnannotated);}return new FieldCache(buildSortedAllFieldMap(orderFieldMap, indexFieldMap), indexFieldMap, ignoreMap);});}

这下难住我了，这个缓存也不刷新。后来想他把缓存放在map里，key是Class，我直接每次都给他remove不就可以了。加了一行代码放在处理表头前边，果然可行！

ClassUtils.FIELD_CACHE.remove(User.class);