[jdom-interest] DataConversionException javadoc and other issues

[Elliotte Rusty Harold]

The Javadoc for the DataConversionException constructor (see 
http://www.jdom.org/docs/apidocs/org/jdom/DataConversionException.html) 
is very misleading. It has nothing to do with missing attributes, and 
everything to do with a string not being able to be converted to a 
number.

I was just going to submit a patch, but then I began to wonder whther 
we still needed this exception class at all. It is only thrown, as 
near as I can see, by four methods in Attribute: getIntValue(), 
getFloatValue(), getDoubleValue(), getLongValue. If getIntValue() or 
getFloatValue(), etc. fail the natural response is a 
NumberFormatException. Would it be better to throw that and delete 
DataConversionException completely?

The big difference is that DataConversionException is checked and 
NumberFormatException is not. I think that this case  does satisfy 
the criterion for checked excepitons that, (following Bloch) " the 
exceptional condition cannot be prevented by proper use of the API 
/and/ the programmer using the API can take some useful action once 
confronted with the exception"


Last May, Jason objected that NumberFormatException was inappropriate 
for getBooleanValue. There's no BooleanFormatException. The 
java.lang.Boolean class interprets all non-true strings as false. I 
don't think we should go that far, but even if we keep 
DataConversionException the JavaDoc needs to be fixed.
-- 
Elliotte Rusty Harold
elharo at metalab.unc.edu
Processing XML with Java (Addison-Wesley, 2002)
   http://www.cafeconleche.org/books/xmljava
   http://www.amazon.com/exec/obidos/ISBN%3D0201771861/cafeaulaitA