Re: [Jmri-developers] [Bulk] Properties in NamedBeans

[Randall W. <ran...@al...>]

On 27 Nov 2011, at 11:43, Bob Jacobsen wrote:

> 
> On Nov 27, 2011, at 8:10 AM, Randall Wood wrote:
> 
>> In NamedBean, the key parameter for getProperty and setProperty is an Object instead of a String. Does anyone recall why? Is there a reason not to change property keys to be Strings?
>> 
>> I'm asking because the JavaBeans introspection and property manipulation methods all use Strings for property names, and wonder if the NamedBean should match the JavaBeans design patterns.
> 
> 
> I don't recall in detail, but I suspect it was a gesture toward generality.
> 
> It only requires changing the interface and a few implementation files to convert the type:
> 
> M       java/src/jmri/NamedBean.java
> M       java/src/jmri/implementation/AbstractNamedBean.java
> M       java/src/jmri/jmrit/catalog/CatalogTreeIndex.java
> M       java/src/jmri/jmrit/catalog/CatalogTreeFS.java
> 
> along with two using files:
> M       java/src/jmri/managers/configurexml/AbstractNamedBeanManagerConfigXML.java
> M       java/test/jmri/implementation/NamedBeanTest.java

Using the "Find Usages" tool in NetBeans, I can't find any use of the NamedBean.getProperty() and NamedBean.setProperty() methods outside the above code, but I did find a three implementations of getProperty() and setProperty(). Each implementation uses a "parameters" HashMap to story properties set using setProperty(), however, I cannot find any use of that HashMap outside getProperty(), setProperty(), and getPropertyKeys(), which suggests that the entire NamedBean getProperty/setProperty mechanism is essentially unused (except in tests). Can someone please verify this, or point to where I missed this?

Randall Wood
Alexandria Software
202.683.8604
ran...@al...
http://www.alexandriasoftware.com