It seemed a bit odd to me to punish our users with multiple calls because of something we didn't design well in the first place. Typical case of where we should have provided an empty template method. It would have my preference to fix that rather than doing the isEnabled call twice.
Eelco On 3/29/07, Igor Vaynberg <[EMAIL PROTECTED]> wrote:
i think you should leave the isenabled check. isbehavioraccepted does it, sure. but as users override it, will they remember to call super? its not in javadoc and is not enforced. and if they do not call super they can override what is behavior's choice not the components. -igor On 3/29/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: > > Author: ehillenius > Date: Thu Mar 29 20:31:46 2007 > New Revision: 523920 > > URL: http://svn.apache.org/viewvc?view=rev&rev=523920 > Log: > header contributions test on isBehaviorAccepted > > Modified: > incubator/wicket/branches/wicket-1.x/jdk-1.4 > /wicket/src/main/java/wicket/Component.java > > Modified: incubator/wicket/branches/wicket-1.x/jdk-1.4 > /wicket/src/main/java/wicket/Component.java > URL: > http://svn.apache.org/viewvc/incubator/wicket/branches/wicket-1.x/jdk-1.4/wicket/src/main/java/wicket/Component.java?view=diff&rev=523920&r1=523919&r2=523920 > > ============================================================================== > --- incubator/wicket/branches/wicket-1.x/jdk-1.4/wicket/src/main/java/wicket/Component.java > (original) > +++ incubator/wicket/branches/wicket-1.x/jdk-1.4/wicket/src/main/java/wicket/Component.java > Thu Mar 29 20:31:46 2007 > @@ -1812,7 +1812,7 @@ > while (iter.hasNext()) > { > IBehavior behavior = > (IBehavior)iter.next(); > - if (behavior instanceof > IHeaderContributor && behavior.isEnabled(this)) > + if (behavior instanceof > IHeaderContributor && isBehaviorAccepted(behavior)) > { > > ((IHeaderContributor)behavior).renderHead( > container.getHeaderResponse()); > } > > >
