I'm ok with changing the semantics & reflecting that in the docs. As long as we are careful not to create objects unnecessarily (if we return a constant that should be fine).
Sent from my iPhone > On 2014/05/17, at 3:20, Gary Gregory <garydgreg...@gmail.com> wrote: > > Hm... the contract for LogEvent.getContextMap() says it can return null, but > our implementation never returns null and in fact uses > java.util.Collections.EMPTY_ > MAP. > > What about changing the LogEvent.getContextMap() Javadoc to require that a > map always be returned? > > Gary > > >> On Fri, May 16, 2014 at 2:18 PM, Gary Gregory <garydgreg...@gmail.com> wrote: >> LogEvent.getContextMap() can return null if there is no map. >> >> It seems that a default of a read-only empty Map >> (java.util.Collections.EMPTY_MAP) would be safer since it would avoid any >> NPEs from our own call sites and users' as well. >> >> Thoughts? >> >> Gary >> >> -- >> E-Mail: garydgreg...@gmail.com | ggreg...@apache.org >> Java Persistence with Hibernate, Second Edition >> JUnit in Action, Second Edition >> Spring Batch in Action >> Blog: http://garygregory.wordpress.com >> Home: http://garygregory.com/ >> Tweet! http://twitter.com/GaryGregory > > > > -- > E-Mail: garydgreg...@gmail.com | ggreg...@apache.org > Java Persistence with Hibernate, Second Edition > JUnit in Action, Second Edition > Spring Batch in Action > Blog: http://garygregory.wordpress.com > Home: http://garygregory.com/ > Tweet! http://twitter.com/GaryGregory
