Here’s an idiom that I think would be useful in C#: “property methods” in the style of jQuery.
There are many methods in jQ – html(), val(), width() – that are getters when they are called without an argument, and setters when they are. This makes for some very nice fluent code:
$(“#someElement”).width(“100px”).html(“hello there”);
window.alert(“The width of #someElement is ” + $(“#someElement”).width());
I’d love to be able to do this in C#, to bring in some of that fluency and terseness. For example, I currently have to write:
myPerson.FirstName = “Joe”;
myPerson.LastName = “Smith”;
myPerson.ZipCode = “12345”;
The code smell here is that I refer to myPerson 3 times. Don’t you just wanna say “Yeah, I get it, myPerson, sheesh.” Wouldn’t it be better to write:
myPerson.FirstName(“Joe”).LastName(“Smith”).ZipCode(“12345”);
Of course I could implement the above with overloads and private fields. But I shouldn’t need to do that much notation.
I think it can be done in C# without new keywords. Simply allow methods to have get and set bodies:
public class Person
{
public string FirstName() { get; set ; }
public string LastName () { get; set ; }
public string ZipCode () { get; set ; }
}
The compiler would interpret these much like regular properties, backed by (automagical) private fields, with the caveat that a passed value invokes set. The setter returns the parent object, giving us the fluency.
And as with properties, you could add bodies to get and set as you desire, where the value argument is implied for set.
Overall, this is the equivalent of:
public class Person
{
public string FirstName() // getter
{
return _FirstName;
}
public Person FirstName(string value) // setter
{
_FirstName = value;
return this; // the fluent bit
}
private string _FirstName; // backing store
//…and repeat for each property
}
Again, I could write my classes this way. But that would be so much boilerplate, uninformative code. Please compiler, do it for me.