Pitfalls and Solutions for Converting Java Beans to Maps

This article starts by describing a common scenario where Java beans need to be transformed into Map objects, but developers encounter subtle bugs such as type conversion (e.g., Long becoming Integer ) and property‑name mismatches caused by JSON libraries and the Java introspection API.

It first shows a simple FastJSON example that serializes a bean to JSON and parses it back to a map, highlighting how numeric values smaller than Integer.MAX_VALUE are demoted and how dates become timestamps.

Next, the article explores the BeanMap implementation from Apache Commons BeanUtils, revealing that the underlying java.beans.Introspector#decapitalize method incorrectly transforms getter names like getALong into the property name aLong , which leads to mismatched keys in the resulting map.

The same issue is reproduced with CGLIB's BeanMap , because its ReflectUtils#getBeanGetters also relies on the same decapitalization logic.

To solve these problems, the article proposes using Dubbo's PojoUtils.generalize method. This utility recursively converts beans, collections, arrays, enums, and primitive types while preserving the original class name and correctly handling property names. Sample Maven dependencies for Dubbo and a complete code demo are provided.

Key code snippets include the FastJSON conversion, the BeanUtils BeanMap usage, and the Dubbo solution:

import com.alibaba.fastjson.JSON;
import com.alibaba.fastjson.TypeReference;
Map
map = JSON.parseObject(JSON.toJSONString(bean), new TypeReference<Map<String, Object>>(){});

BeanMap beanMap = new BeanMap(bean);
System.out.println(beanMap);

Object generalized = PojoUtils.generalize(bean);
System.out.println(generalized);

The article concludes that developers should be aware of these hidden pitfalls, prefer robust utilities like Dubbo's PojoUtils , and always verify bean‑to‑map conversions with unit tests or debugging tools.