Just FYI,

The getPropertyDescriptors method you describe simply delegates much of it's
job to the java.beans.Introspector. (This bug, in particular, describes an
issue which exists in the output from the Introspector.) Further, the
approach of BeanUtils to this point in time has been to avoid rebuilding or
redoing anything currently found in Introspector - this means the bad
(Introspector bugs) comes with the good.

My guess will be that this issue should and will be marked "WONTFIX", though
I'll leave it to those actively working with BeanUtils to make the final
decision. The easiest workaround, of course, is to choose a slightly more
verbose name

The capitalization rule of PropertyDescriptor#getName() is as follows:

If the leading two characters following set(get) prefix are capitalized,
the leading character of a property name is also capitalized.

Otherwise, the leading character of a property name is not capitalized.

examples:

set(get)AA --> AA (aA is expected)
set(get)NAmount --> NAmount (nAmount is expected)
set(get)Aa -> aa
set(get)VcName -> vcName

I don't know whether this rule is designated for acronyms or not. Method names
conforming to Java's naming rule would not encounter this bug. But, if database
people prefer to another naming rule that needs type prefix preceding nouns,
workarounds are required.

This is my workaround. See commented lines:

import java.beans.BeanInfo;
import java.beans.Introspector;
import java.beans.PropertyDescriptor;
public class TestCapitalization {

private String aA;
private String aa;
private String nAmount;
private String vcName;

public String getAA()

public String getAa()

public String getNAmount()

public String getVcName()

public void setAA(String aA)

public void setAa(String aa)

public void setNAmount(String nAmount)

public void setVcName(String vcName)

/*
workaround: The leading character not to be capitalized.

public String getaA()

public void setaA(String aA)

*/

public static void main(String[] args) throws Exception {
BeanInfo beanInfo = Introspector.getBeanInfo(TestCapitalization.class);
PropertyDescriptor[] props = beanInfo.getPropertyDescriptors();
for (int i=0; i < props.length; ++i)

}

}

BeanUtils simply uses the standard introspection functionality of the underlying
JDK. It does not make it's own determination of what is a property or not, nor
can it influence the behavior of the underlying Introspector class.

The workaround is for your class to follow the property name design patterns
described in the JavaBeans Specification if you want BeanUtils, or any other
software that deals with JavaBeans properties, to find them.

I believe that this issue could (and should) be fixed. The standard JDK provides methods for dealing with property names as described in this issue. Please have a look at: http://java.sun.com/j2se/1.4.2/docs/api/java/beans/Introspector.html#decapitalize%28java.lang.String%29

I believe that the problem is in the and more specifically in this part of the code:

PropertyDescriptor[] descriptors = getPropertyDescriptors(bean);
if (descriptors != null) {

for (int i = 0; i < descriptors.length; i++) {
if (name.equals(descriptors[i].getName()))

}
}

That should be changed to use decapitalize method, so:

name.equals(Introspector.html#decapitalize(descriptors[i].getName()))

Could you please re-consider opening / fixing this issue?

@Pawel

Craig pointed out that BeanUtils is following the JavaBean specification wrt to property names and I see nothing in your post that persuades me to change that decision. If we implemented your suggestion then anyone who had use the correct name in this circumstance would no longer work