RFR: 8164479: Update JAX-WS RI integration to latest version

[Mandy Chung]

> On Nov 8, 2016, at 5:18 AM, Roman Grigoriadi <roman.grigoriadi at oracle.com> wrote:
> 
> Thanks for suggestions Mandy. xjc -m option is used to pass module name if module-info generation is desired. Here is updated webrev with all suggestions included:
> 
> http://cr.openjdk.java.net/~aefimov/8164479/04/
> 


+1.


One javadoc comment:

  90      * If {@code java.beans.Beans} is visible then it's {@link Beans#instantiate}

I don’t see an import java.beans.Beans and so does javadoc issue warning for {@link Beans#instantiate}?

You can run make docs to check for any javadoc warnings.

No need to generate a new webrev.

Mandy

> Roman
> 
> 
> On 11/03/2016 09:37 PM, Mandy Chung wrote:
>>> On Nov 3, 2016, at 4:53 AM, Roman Grigoriadi <roman.grigoriadi at oracle.com> wrote:
>>> 
>>> CCC for CommandInfo changes is now approved (http://ccc.us.oracle.com/8047773), you can find new webrev here:
>>> 
>>> http://cr.openjdk.java.net/~aefimov/8164479/03/
>>> 
>> javax/activation/CommandInfo.java
>> 
>> 92      * When java.beans.Beans is not visible (when java.desktop module is not
>> 
>> Formatting nit: “java.beans.Beans” and “java.desktop” can be wrapped with {@code…}
>> 
>>  218                 Class<?> beanClass = Class.forName(cn, true, loader);
>> 
>> I suggest to not to initialize the class in the Class::forName call (i.e. s/true/false)
>> 
>>  220                     return beanClass.newInstance();
>> Class::newInstance is deprecated in JDK 9.  Can you replace it with Class::getDeclaredConstructor and Constructor::newInstance
>> 
>> 
>> What does wsimport and xjc -m option do?
>> 
>>   67     private static final String JAXWS_PACKAGE = "java.xml.ws”;
>> 
>> - Is it a module name, not package name?
>> 
>> Does java.xml.ws module still require java.rmi?  Should it be updated?
>> 
>> Otherwise, looks okay to me.
>> 
>> Mandy
>