I've got a function in an object like this:
arrayofElements: function(item) {
var result = [];
if (item.isA) {
result.push(new Div('aClass', labels['A']));
}
if (item.isC) {
result.push(new Div('cClass', labels['C']));
}
if (item.isD) {
result.push(new Div('dClass', labels['D']));
}
return result;
},
How can this be refactored? I dislike having to push() each item conditionally.
You could move the is* properties to a sub-object so as to isolate them, and loop over the sub-object's properties
item.is={A:'aClass', C:'cClass'};
...
arrayofElements: function(item) {
var result = [];
for (p in item.is) {
result.push(new Div(item.is[p], labels[p]));
}
return result;
},
The values in item.is could be the classes (as shown), or the values in item.is could be objects with a class property, or you could use the property name p as an index in another object to get the classes. Which depends largely on what the items represent and what the element classes are most closely associated with.
arrayofElements: function(item) {
var result = [],
items = ['A', 'C', 'D'];
while(var i=0; i < items.length; i++) {
if(item["is" + items[i]]) {
result.push(new Div(items[i] + 'Class', labels[items[i]]))
}
}
return result;
},
Remember that obj.foo is the same as obj['foo'] in Javascript
A variant of Larry's solution:
arrayofElements: function(item) {
var result = [], letters = ['A','C','D'], ltr, i, len;
len = letters.length;
for (i = 0; i < len; i++) {
ltr = letters[i];
if (item["is"+ltr]) {
result.push(new Div(ltr.toLowerCase()+"Class", labels[ltr]));
}
}
return result;
},
This function avoids using push() and is functionally identical to yours:
arrayofElements: function(item) {
if (item.isA) {
return [new Div('aClass', labels['A'])];
}
if (item.isC) {
return [new Div('cClass', labels['C'])];
}
if (item.isD) {
return [new Div('dClass', labels['D'])];
}
return [];
},
{isA:true,isC:true} it will return [new Div('aClass', labels['A']),new Div('cClass',labels['C'])] Array#push modifies the array it is called on.
itemhave?