Code Review Request for Bug #4802647

[Alan Bateman]

On 01/12/2011 22:42, Brandon Passanisi wrote:
> Hi Jason.  Thanks for your response.  I was thinking about how I can 
> improve the test using your suggestion. I could possibly do the 
> following:
>
>    1. Find all of the subclasses of AbstractCollection which override
>    removeAll(Collection<?>) and which also contain the spec language
>    which specifies that NullPointerException is thrown if the specified
>    collection is null.
>
>    2. Find all of the subclasses of AbstractCollection which override
>    retainAll(Collection<?>) and which also contain the spec language
>    which specifies that NullPointerException is thrown if the specified
>    collection is null.
>
>    3. Add the classes found in #1 and #2 to the test.
>
>    4. If any of the new classes added because of  #3 result in a test
>    failure, it might be a good idea to file a new bug as Bug #4802647
>    specifically mentions subclasses of AbstractCollection which do not
>    override remainAll, retainAll.
>
>    5. The public subclasses of AbstractCollection which do not override
>    removeAll, retainAll (probably) shouldn't be included in the test as
>    the currently existing NewAbstractCollection represents this scenario.
>
> What do you think?
>
Brandon - it's probably worth getting familiar with the existing tests, 
in in particular Martin's "Mother Of All Test" (Collection/MOAT.java).

-Alan.