Different kind of Angular Directives patterns

Question

angular.directive('ppd:reset', function(e1,e2) {
    return function(linkElement) {      
        linkElement.wrap()....
        });
    };
});

And

angular.directive('ppd:reset', [function() {
        return function(scope, elm, attrs) {   } 

}]);

What is the difference between these two directives pattern ?

Answer 1

If you declare your factories with the bracket notation, you can avoid problems when minifying your code:

angular.directive('ppd:reset', ["$compile", function(compile) {
    return function(scope, elm, attrs) {   } 

}]);

The injector looks at your function parameter names in order to know what to inject. If a minification process renames those, The injector doesn't know what to do anymore. Minification will of course not touch string values which is why the array notation works fine.

Answer 2

The difference is that version #1 is a simple way that angular does support for writing directives that don't require any injectable modules. The version #2 is for having injectables. So, let's say your directive relied on the $timeout service, then you would have a definition like below. For me, its easier to not think and just use the array syntax even if there are no injectables.

angular.directive('ppd:reset', ['$timeout', function($timeout) { 
       return function(scope, elm, attrs) {   } 
}]);