if(a == b == c)

[ahmd]

Just confirming, is comparison 1 evaluating to comparison 2 or 3?

	int a = 0, b = 0, c = 0, d = 0;

	//Comparison 1
	if(a == b == c)
	{
		d++;
	}

	//Comparison 2
	bool res = a == b;
	if(res == c)
	{
		d++;
	}

	//Comparison 3
	if(a == b &&
		b == c)
	{
		d++;
	}

[Eri523]

[...] is comparison 1 evaluating to comparison 2 or 3?

First off, comparisons don't evaluate to one another, they evaluate to a bool result. So I suppose you're simply asking whether they all have the same result.

In #1 I'm not 100% sure about the evaluation sequence mandated by the standard, but that doesn't matter either: In the end you're comparing the result of an equality comparison between two equal ints (i.e. true) with a third int that's 0. Since true is converted to a non-zero integer (usually 1, but I don't think that's mandatory, but any non-zero value would lead to the same result) the final result is true.

In #2 you're enforcing a certain evaluation sequence by assigning an intermediate result to a temporary variable, but aside from this it essentially is the same as #1.

#3 is obvious, isn't it? You're comparing 0 and 0 for equality twice. The comparison evaluates to true in both cases, so the && eventually will evaluate to true as well.

However, I don't see what that has to do with Win API. Could it be that you were one off when clicking the section you wanted to post that in?

[ahmd]

It said, "C++ and WinAPI", so I'm taking it as C++ (since there's no C section here). If it's a wrong section, I'll let moderator move it.

Going back to my question. I'm not asking about all three situations. I'm asking about comparison #1 and what it evaluates to. And no, your assumption is not correct. The first if statement will result in a false.

[Paul McKenzie]

It said, "C++ and WinAPI", so I'm taking it as C++ (since there's no C section here).

C++ precedence rules aren't Windows API issues.

http://www.cppreference.com/wiki/lan...tor_precedence

The first if statement will result in a false.

a == b is true. Then

true == c

is false, since c is 0, so the entire expression is false.

Regards,

Paul McKenzie

[Eri523]

And no, your assumption is not correct. The first if statement will result in a false.

a == b is true. Then

true == c

is fakse, since c is 0, so the entire expression is false.

D'oh! That's what happens when I make conclusions too quickly...

But, ahmd, if you already knew that, why then did you ask? Did you intend to test us?

[ahmd]

I'm trying to save myself coding time. If you look at the mathematical sense, a == b == c should evaluate to "if 'a' equals 'b' and 'b' equals 'c'", which is not the case for if(a == b == c) statement.

[Eri523]

Hmmm... Didn't you learn something from all the threads you probably have read here during the last two years? How many times have you seen something like this in a beginner's program:

if (answer == 'y' || 'n')

This obviously won't work: Programming languages (maybe except some experimental ones) don't understand everyday's language used in inter-human communication.

What you are trying may be seen as a bit less daring since you're "just" trying to make C++ understand some common math jargon idioms. Eventually, however, this won't work either, simply because that's not part of the C++ language. (The language does have these "chained assignments" which look pretty similar but actually are something completely different.)

Chances are that you actually may find a language that understands this, but of course this will not be C++ and so you have to abandon C++ and make a migration. Also, perhaps this language understands your simplified syntax but produces slower code. Are you really willing to save a few keystrokes at the expense of the efficiency of your final program?

[Igor Vartanov]

I'm trying to save myself coding time. If you look at the mathematical sense, a == b == c should evaluate to "if 'a' equals 'b' and 'b' equals 'c'", which is not the case for if(a == b == c) statement.

C is not a mathematician's language, that's what you should keep in mind at coding time. This habit will save your debugging time which cost is much bigger than coding time.

[ahmd]

Wow, thank you for your condescending replies.

I'm not trying to learn the C++ language, I was simply asking about that specific construct. What shook me is that normally VC compiler gives you a warning if you write something like if(a = 1) but not in this case...

[MrViggy]

Not mine.

if (a = 1)

Is a perfectly valid statement. There are many times I do something like:

if (bQuit = SomeFuncCall())
{
   // Do more stuff here
}

Also,

if (a == b == c)

is valid as well. As already stated, this will be evaluated left to right, returning "false" in your example.

Viggy

[Igor Vartanov]

Wow, thank you for your condescending replies.

I always get surprised how other people are able to find condescension in a trivial reply to a trivial question with an obvious answer. Does it sound condescending to say 'think in French when you write in French'?

I'm not trying to learn the C++ language, I was simply asking about that specific construct. What shook me is that normally VC compiler gives you a warning if you write something like if(a = 1) but not in this case...

If you had a warning thing as a main point, why did you talk about mathematical notation? Don't you find this a bit confusing?

BTW, it shake me every time how easily you get excited by the things I'd never even notice, or paid really less attention.

[ahmd]

You might be right. I think I had a bad day yesterday. It all actually started in another thread when people were treating me like some noob programmer... I guess there are quite a few of them here.

[Igor Vartanov]

Well, you're definitely not a noob, so why to bother about what others say or think of you. Being positive about yourself is a great thing, almost the same great as being demanding to yourself and forgiving to others.

[ahmd]

Wow, I didn't know I'd get this kinda encouragement on this site It was just too many things going wrong at once... I'm sure you've experienced that... anyway, thanks.

[ovidiucucu]

If it's a wrong section, [...]

[ Moved thread ]