Jump to content

  • Log In with Google      Sign In   
  • Create Account

We need your feedback on a survey! Each completed response supports our community and gives you a chance to win a $25 Amazon gift card!


C# null testing


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • You cannot reply to this topic
9 replies to this topic

#1 ChaosEngine   Crossbones+   -  Reputation: 2509

Like
0Likes
Like

Posted 27 May 2008 - 12:57 PM

Sometimes it's the little things that really get to you. Right now, I'm getting really sick of typing " != null" in C#. This seems to be a very common test in C# (checking if someone has subscribed to an event, testing for a SelectedItem, the result of a find), so couldn't they have let you do a C++ style
if (someObj)
and cast null to false? I don't see why this would be a problem. I'm a relative noob with C#, so feel free to tell if I'm not using the correct C# idioms (in fact, please tell me there's a better way!) and while I'm ranting, I want specialised generics damnit!! [grin]
if you think programming is like sex, you probably haven't done much of either.-------------- - capn_midnight

Sponsor:

#2 TheUnbeliever   Members   -  Reputation: 961

Like
0Likes
Like

Posted 27 May 2008 - 01:05 PM

Potentially of value may be the ?? null coalescing operator.

EDIT: Actually, I see this won't be particularly useful in the examples you give, but it's still handy to have.

#3 ChaosEngine   Crossbones+   -  Reputation: 2509

Like
0Likes
Like

Posted 27 May 2008 - 01:10 PM

Thanks, Unbeliever, but I'm afraid that doesn't help.

The ?? null coalescing operator is basically a shortcut for the ?: ternary op, which means that both operands must be of the same type, i.e. you can't do

void SomeMethod()
{}

someObject ?? SomeMethod();



anymore than you can do

someObject != null ? SomeMethod() : SomeOtherMethod()




#4 Nypyren   Crossbones+   -  Reputation: 4826

Like
0Likes
Like

Posted 27 May 2008 - 01:50 PM

In C#, it's that way because of a case like this:


object a;
object b;

// something to one or both of the objects

if (a == b)

// now what happens if you have a typo:

if (a = b)



Semantically, the expression with a typo will reduce to "if (object)", exactly like your example for testing against null. Since C# requires that the expression inside an "if" reduce to a boolean, instead of getting a runtime error, you get a compile-time error.

You could try bypassing this in C# 3.0 with an implicit typecast operator as an extension method, but I suspect those are specifically disallowed.

#5 Spodi   Members   -  Reputation: 642

Like
0Likes
Like

Posted 27 May 2008 - 02:53 PM

Sorry, think you'll just have to get used to it. I have never found a way around it, which I think is good. Trust me, its not that bad - at least its specific on what you are checking. C# likes to make you say what you want, not just guess it and hope its right, and its worth it. Think of it this way - you may waste a little time writing out != or == null, but at least you'll save a lot more using the .NET framework and the GC. ^_^
NetGore - Open source multiplayer RPG engine

#6 Telastyn   Crossbones+   -  Reputation: 3730

Like
0Likes
Like

Posted 27 May 2008 - 04:05 PM

If it's a user made class, and it makes sense, you can make an implicit casting operator to bool so if(foo){} works.

That said, you shouldn't. That behavior is specific to C, carried forward for compatibility reasons into C++ and promptly abandoned by anything since. nulls are far less common in a language like C# that forces initialization before use and promotes exceptions for exceptional circumstances (and null coalescing where appropriate).

Though yeah, invoking an unsubscribed/null event should be a no-op...

#7 TheTroll   Members   -  Reputation: 882

Like
0Likes
Like

Posted 27 May 2008 - 06:27 PM

I am trying to why you are having to do so much testing for null. I very rarely have to do this. So I am thinking there might be something about your design that is causing the need for this.

theTroll

#8 Cygon   Crossbones+   -  Reputation: 1157

Like
0Likes
Like

Posted 27 May 2008 - 08:01 PM

You have a null test whenever you implement your own object.Equals() override. It's not that uncommon.

I don't know of a shorter test, but I'm often even using object.ReferenceEquals(x, null) instead of x != null in order to avoid calling the operator != override. I guess that's just something you'll have to accept when programming in C#.

#9 Koobazaur   Members   -  Reputation: 691

Like
0Likes
Like

Posted 27 May 2008 - 09:25 PM

Quote:
Original post by ChaosEngine
if (someObj)
and cast null to false?

I don't see why this would be a problem.


What if someObj is a bool? now you need extra checks to see if it is either null or false ( if (someObj != true) if (someObj != null) [false] else [null])

#10 Telastyn   Crossbones+   -  Reputation: 3730

Like
0Likes
Like

Posted 28 May 2008 - 02:40 AM

Quote:
Original post by Koobazaur
Quote:
Original post by ChaosEngine
if (someObj)
and cast null to false?

I don't see why this would be a problem.


What if someObj is a bool? now you need extra checks to see if it is either null or false ( if (someObj != true) if (someObj != null) [false] else [null])


bools are not nullable.

A class which is implicitly convertible to bool will handle the null case.




Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS