Curly braces

[Jorgen Andersson]

The first one wasn't invented for clarity or ease of reading, but for squeezing more text onto paper.

Wrong is evil and must be defeated. - Jeff Ello

[Jorgen Andersson]

That would depend on whether you consider the braces to be a part of the content or the header. If your answer is neither you should consider GNU style. ;P

Wrong is evil and must be defeated. - Jeff Ello

[den2k88]

I used to be in the first camp, I had to adapt to MISRA rules and now I prefer the second. It's way easier to find missing braces, move braced code and reindent it, even with powerful syntax highlighting - which is not a given in most IDEs for embedded development.

GCS/GE d--(d) s-/+ a C+++ U+++ P-- L+@ E-- W+++ N+ o+ K- w+++ O? M-- V? PS+ PE Y+ PGP t+ 5? X R+++ tv-- b+(+++) DI+++ D++ G e++ h--- r+++ y+++*      Weapons extension: ma- k++ F+2 X

[fgs1963]

I have no opinion about braces one way or the other but I have to ask - Why use different styles in different languages? Given it's a choice it seems your personal sense of style would prefer one style all the time. No?

[Jacquers]

I generally just use what's in the specific language's style guide. It makes it easier when different developers work on a project if we stick to those.

[MarkTJohnson]

A POX upon Tabs.

I’ve given up trying to be calm. However, I am open to feeling slightly less agitated.

[Single Step Debugger]

I'm in this camp

void AFunc()
{
}

And for short/inline statements in this one:

void AFunc() {}

I never mix them.

There is only one Vera Farmiga and Salma Hayek is her prophet! Advertise here – minimum three posts per day are guaranteed.

[Nelek]

exactly.

M.D.V. ;) If something has a solution... Why do we have to worry about?. If it has no solution... For what reason do we have to worry about? Help me to understand what I'm saying, and I'll explain it better to you Rating helpful answers is nice, but saying thanks can be even nicer.

[Nelek]

In the second one, even when an If has only one command to be executed.

M.D.V. ;) If something has a solution... Why do we have to worry about?. If it has no solution... For what reason do we have to worry about? Help me to understand what I'm saying, and I'll explain it better to you Rating helpful answers is nice, but saying thanks can be even nicer.

[godfetish]

The only sane format is:

void Afunc()
{
}

Why #1: Because half the non-curly scoped languages out there use words or other symbols for marking scope, and BEGIN just doesn't look right at the end of the line! Why #2: Java version 1.1 (the first language I used and learned that trailed the curly brace in books) was so horrible, it gave me one more thing to hate! Why #3: Other than using some automated reformatting that kicks lines to the right or left, I find it very difficult to find missing curly braces in some logic. Matched ones line up very clearly. If they trail the line, when things get a few indents deep, the code becomes hard to read and finding that one line you inserted outside of them takes more time than just begin clear the first time. So, always on its own line!

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119.

[Mircea Neacsu]

Please don't go there! One unwinnable debate is enough :) We all have our ways of setting braces and indenting which we know to be right and we'll quash the unbelievers who dare to have a different opinion. We've done that since the time of the first crusade :D

Mircea

[trønderen]

Jacquers wrote:

I generally just use what's in the specific language's style guide.

For C/C++ there is about 42 different bracing / indentation styles fighting for dominance, so referring to "the specific language's" style, as if it was unambiguous, is meaningless. From a logical point, I would prefer

void AFunc() {
}

but I have never seen that promoted in any style guide. The opening brace comes when the statement cannot be completed on the current line, indicating that a block is to follow. You should before you leave the line that you won't find a single one-statement line, but a block. I am one who think "if (day==sunday) weekend = true;" in one line is perfectly fine. "if (day==sunday) {" shows that the statement is not complete. The following block is indented. An indentation always follows a brace. No an indentation without an opening brace; no opening brace without an indentation.

"if (day==sunday)
{
}

starts the indentation before the brace - that is inconsistent! Undenting follows a closing brace. No undentation without a closing brace, no closing brace without an undentation.

"if (day==sunday)
{
// indented code block
}

undents before reaching the closing brace - that is inconsistent. Noone seems to agree with my logic. So I bow my head and follow whatever style guide is enforced upon me. Switch statements mess up indentation in C/C++. One common layout is:

switch (day) {
case (day == saturday):
case (day == sunday):
celebrate();
break;
default:
gotowork();
break;
}

The case alternatives are not blocks, so they don't need braces (that is also why they need the 'break'!), but they are indented! I certainly wish they were blocks, for consistency's sake (and I really dislike the 'default is fall through'), so I prefer to make them blocks, adding braces, to justify the indentation. Very few agree with this logic, too - they are so used to seeing indents without any braces justifying it that the inconsistency doesn't bother them.

[den2k88]

I f-f-fart in the general direction of spaces. I had to work on code where some dev used a single space, some two, some 3 and some idiot 8. As a result it was completely unindented. With tabs there wouldn't have been any issue.

GCS/GE d--(d) s-/+ a C+++ U+++ P-- L+@ E-- W+++ N+ o+ K- w+++ O? M-- V? PS+ PE Y+ PGP t+ 5? X R+++ tv-- b+(+++) DI+++ D++ G e++ h--- r+++ y+++*      Weapons extension: ma- k++ F+2 X

[Amarnath S]

The compiler doesn't care. Neither do I.

[den2k88]

You're not writing for the compiler.

GCS/GE d--(d) s-/+ a C+++ U+++ P-- L+@ E-- W+++ N+ o+ K- w+++ O? M-- V? PS+ PE Y+ PGP t+ 5? X R+++ tv-- b+(+++) DI+++ D++ G e++ h--- r+++ y+++*      Weapons extension: ma- k++ F+2 X

[Single Step Debugger]

Well, the compiler doesn't care if you use hungarian notation, but this is how office wars start.

There is only one Vera Farmiga and Salma Hayek is her prophet! Advertise here – minimum three posts per day are guaranteed.

[k5054]

I'd add that a number of us were influenced by K&R C, where you had

int foo(str, y)
char *str;
int y;
{
/* code */
}

So an opening curly bracket at the beginning of a function made sense. But then we have:

if(a == b ){
/* code */
}

ditto for while, for, etc. Yes, it is inconsistent between functions and every other compound block of code, but blessed by Kernighan and Ritchie, so who am I to argue. So maybe that's Why #4.

Keep Calm and Carry On

[trønderen]

godfetish wrote:

BEGIN just doesn't look right at the end of the line!

I really love 'It just ain't done!' arguments! BEGIN is perfectly fine at the end of the line. In my Pascal days, that was the established standard. Note that Pascal switch statements - CASE value OF - uses the OF keyword to start the case list. I never saw anyone put OF on a separate line. It is matched by an END, just like BEGIN. So why would there be a difference between BEGIN ... END and CASE OF ... END? Or would you write Pascal code as

CASE i
OF
0 : Write('zero');
1 : Write('one');
2 : Write('two');
3,4,5,6,7,8,9,10: Write('?')
END;

Note that if you put BEGIN and OF at separate lines, they are still at the end of the line :-)

[trønderen]

I usually do. The compiler is only one who really cares, and analyzes my writings thoroughly!

[honey the codewitch]

Why not just put blocks in if you prefer it that way? It also makes it so you can declare variables under the case without the compiler yelling at you.

case 1: {
// do work
}
break; // could be inside

To err is human. Fortune favors the monsters.