[jira] [Created] (COLLECTIONS-536) (Code style) map.size() call in MapUtils.putAll()

["Tagir Valeev (JIRA)" <ji...@apache.org>]

Tagir Valeev created COLLECTIONS-536:
----------------------------------------

             Summary: (Code style) map.size() call in MapUtils.putAll()
                 Key: COLLECTIONS-536
                 URL: https://issues.apache.org/jira/browse/COLLECTIONS-536
             Project: Commons Collections
          Issue Type: Bug
          Components: Map
    Affects Versions: 4.0, 3.2.1
         Environment: Any
            Reporter: Tagir Valeev
            Priority: Trivial


In class org.apache.commons.collections(4).MapUtils there's a method putAll(final Map<K, V> map, final Object[] array) which starts with 
map.size();  // force NPE

Actually map.size() is not that harmless call for any map. For example, consider java.util.concurrent.ConcurrentHashMap size() implementation: depending on the circumstances it may take even more time than the rest of the putAll method (at least prior to JDK 8). I'd suggest to replace it with more conventional check like:

if(map == null) {
    throw new NullPointerException();
}

If you still want to save bytes, you may use map.getClass(). It's final, so it will never be overridden to do something strange and it's even used by JavaC for the same purpose (to trigger NPE). For example, if you compile and disassemble this code:
public class Outer {
    public class Inner {}
    public void test(Outer n) { n.new Inner(); }
}
You will see a getClass() call which is used to trigger NPE.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)