Comment on style

[pocmatos]

Hi all,

While I was programming 5 minutes ago a recurring issue came up and
this time I'd like to hear some opinions on style. Although they are
usually personal I do think that in this case as also to do with making
the code easier to read.

Imagine a function returning void (for example) and it's body is a big
if with lots of special cases:

void foo(const MyClass &c) {
if(c.bar()) {
x();
} else if(c.foobar()) {
y();
} else if(c.mybar()) {
z();
}
}

The question is if you prefer to see like I've wrote it or with a
return; after a call to x(), and another one after a call to y(), etc
etc... ?

Cheers,

Paulo Matos

[Ian Collins]

po******@gmail. com wrote:

Hi all,

While I was programming 5 minutes ago a recurring issue came up and
this time I'd like to hear some opinions on style. Although they are
usually personal I do think that in this case as also to do with making
the code easier to read.

Imagine a function returning void (for example) and it's body is a big
if with lots of special cases:

void foo(const MyClass &c) {
if(c.bar()) {
x();
} else if(c.foobar()) {
y();
} else if(c.mybar()) {
z();
}
}

The question is if you prefer to see like I've wrote it or with a
return; after a call to x(), and another one after a call to y(), etc
etc... ?

What purpose would the returns serve? You are already in an if else
structure.

--
Ian Collins.

[Alf P. Steinbach]

* po******@gmail. com:

Imagine a function returning void (for example) and it's body is a big
if with lots of special cases:

void foo(const MyClass &c) {
if(c.bar()) {
x();
} else if(c.foobar()) {
y();
} else if(c.mybar()) {
z();
}
}

The question is if you prefer to see like I've wrote it
No, you should always use proper indentation, like

void foo(const MyClass &c) {
if(c.bar()) {
x();
} else if(c.foobar()) {
y();
} else if(c.mybar()) {
z();
}
}
or with a
return; after a call to x(), and another one after a call to y(), etc
etc... ?

With exception safe code it doesn't matter technically. But with
respect to clarity it's proper only for precondition checking. Within
the meat of the function it can hide the structure of the code (of
course that assumes proper indentation to begin with), and it can make
maintenance more difficult, for example where you want to add some
common action at the end, so there it's a judgement call.

For the C language the story is different, and I think the best advice I
can give to a novice here is: if you do not understand the difference
between C and C++ wrt. early returns, then absolutely don't use them;
otherwise, use clarity of code as the main guideline.

--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?

[Pep]

po******@gmail. com wrote:

Hi all,

While I was programming 5 minutes ago a recurring issue came up and
this time I'd like to hear some opinions on style. Although they are
usually personal I do think that in this case as also to do with making
the code easier to read.

Imagine a function returning void (for example) and it's body is a big
if with lots of special cases:

void foo(const MyClass &c) {
if(c.bar()) {
x();
} else if(c.foobar()) {
y();
} else if(c.mybar()) {
z();
}
}

The question is if you prefer to see like I've wrote it or with a
return; after a call to x(), and another one after a call to y(), etc
etc... ?

Cheers,

Paulo Matos

Yep, style is definitely personal.

Ian Collins has already asked what purpose would the returns serve as you
are already in an if/else block. I agree with his question and would point
out that you have said this a big function, so you need to consider whether
it would be helpful for another developer to come at a later time and have
to deal with multiple exit points when trying to solve a bug in the
routine, which I would say it is not.
Alf P. Steinbach has stated that you should use proper indentation which
again I agree with but as to whether proper indentation demands 2 spaces, 4
spaces or tabs is open to personal interpretation. Myself, I prefer tabs.

Now I differ from most people that follow the K&R style because I would
prefer the code writtens as

void foo(const MyClass &c)
{

if(c.bar())
{
x();
}
else if(c.foobar())
{
y();
}
else if(c.mybar())
{
z();
}

}

Which I find more readable foe one thing and find it easier to comment out a
block of code because

} else if(c.foobar()) {
y();
} else if(c.mybar()) {

requires you to edit the "} else if(c.foobar()) {" line to place the "else"
statement on a separate line in order to comment out the line or insert a
"/*" in the middle of the else line, whereas

}
else if(c.foobar())
{
y();
}
else if(c.mybar())

simply requires you to place a /* ahead of the "else if(c.foobar())" line
and a "*/" after the closing brace.

Similarly I prefer to see

else if(c.foobar())
{
y();
}

as opposed to

else if(c.foobar())
y();

As you say, style is personal.

Finally I hate single character variable names, even in example code, even
though I sometimes (rarely) use them myself :)

Cheers,
Pep.

[marco]

po******@gmail. com ha scritto:

Hi all,

While I was programming 5 minutes ago a recurring issue came up and
this time I'd like to hear some opinions on style. Although they are
usually personal I do think that in this case as also to do with making
the code easier to read.

Imagine a function returning void (for example) and it's body is a big
if with lots of special cases:

void foo(const MyClass &c) {
if(c.bar()) {
x();
} else if(c.foobar()) {
y();
} else if(c.mybar()) {
z();
}
}

The question is if you prefer to see like I've wrote it or with a
return; after a call to x(), and another one after a call to y(), etc
etc... ?

Cheers,

Paulo Matos

From your post I understand you want to know the difference between the
following two cases:

case 1:
void foo(const MyClass &c) {
if(c.bar()) {
x();
} else if(c.foobar()) {
y();
} else if(c.mybar()) {
z();
}
}

case 2:
void foo(const MyClass &c) {
if(c.bar()) {
x();
return;
} else if(c.foobar()) {
y();
return;
} else if(c.mybar()) {
z();
return;
}
}

Am I write or not?

[marco]

marco ha scritto:

po******@gmail. com ha scritto:

The question is if you prefer to see like I've wrote it or with a
return; after a call to x(), and another one after a call to y(), etc
etc... ?

Cheers,

Paulo Matos

From your post I understand you want to know the difference between the
following two cases:

case 1:
void foo(const MyClass &c) {
if(c.bar()) {
x();
} else if(c.foobar()) {
y();
} else if(c.mybar()) {
z();
}
}

case 2:
void foo(const MyClass &c) {
if(c.bar()) {
x();
return;
} else if(c.foobar()) {
y();
return;
} else if(c.mybar()) {
z();
return;
}
}

Am I write or not?

of course in case #2 else "must" be removed!

[benben]

> Now I differ from most people that follow the K&R style because I would

prefer the code writtens as

void foo(const MyClass &c)
{

if(c.bar())
{
x();
}
else if(c.foobar())
{
y();
}
else if(c.mybar())
{
z();
}

}

I'd agree and that's what I would do...except when there's only one
statement in the block (and it'll remain in one statement in the near
future.) I prefer:

if (c.bar()) x();
else if (c.foobar()) y();
else if (c.mybar()) z();

Regards,
Ben

[Pep]

benben wrote:

Now I differ from most people that follow the K&R style because I would
prefer the code writtens as

void foo(const MyClass &c)
{

if(c.bar())
{
x();
}
else if(c.foobar())
{
y();
}
else if(c.mybar())
{
z();
}

}

I'd agree and that's what I would do...except when there's only one
statement in the block (and it'll remain in one statement in the near
future.) I prefer:

if (c.bar()) x();
else if (c.foobar()) y();
else if (c.mybar()) z();

Regards,
Ben

Yep, I can go along with that though myself I would still write it as a
brace'd block of code :)

Cheers,
Pep.

[Howard]

"benben" <be******@yahoo .com.au> wrote in message
news:43******** *************** @news.optusnet. com.au...

Now I differ from most people that follow the K&R style because I would
prefer the code writtens as void foo(const MyClass &c)
{

if(c.bar())
{
x();
}
else if(c.foobar())
{
y();
}
else if(c.mybar())
{
z();
}

}

I'd agree and that's what I would do...except when there's only one
statement in the block (and it'll remain in one statement in the near
future.) I prefer:

if (c.bar()) x();
else if (c.foobar()) y();
else if (c.mybar()) z();

The problem with that style is that many IDE's debuggers have trouble with
allowing you to step through that code nicely. I often set a breakpoint on
the action following the if, so that I can quickly see when the condition
succeeds. That's not possible if I can't set a breakpoint on it because
it's on the same line as the if. So... I'd put the action on the line
following the if, like this:

if (a)
doa();
else if b
do(b);

(Also, sometimes I'm not positive that I'll never add a second line of
action, or I get lost in what else goes with what if, so I end up adding
back those curly braces anyway. :-))

-Howard

[Aaron Graham]

> Which I find more readable foe one thing and find it easier to comment out a

block of code because

} else if(c.foobar()) {
y();
} else if(c.mybar()) {

requires you to edit the "} else if(c.foobar()) {" line to place the "else"
statement on a separate line in order to comment out the line or insert a
"/*" in the middle of the else line, whereas

}
else if(c.foobar())
{
y();
}
else if(c.mybar())

simply requires you to place a /* ahead of the "else if(c.foobar())" line
and a "*/" after the closing brace.
Or...you could just do this:

// } else if(c.foobar()) {
// y();
} else if(c.mybar()) {

The problem with the "every brace gets its own line" philosophy is that
reading the code requires a lot more scrolling. It's extremely
aggrivating trying to figure out a piece of code where half of all the
lines are just braces, because there's never very much information on
one page.

I find that It's easy to tell how often someone has to read/debug other
peoples' code by how he structures his his own code. If he uses a
separate line for every brace, tabs all over the place (or worse yet,
spaces mixed with tabs), hungarian notation, etc., then he probably
hasn't spent much time looking at someone else's code.

And, no offense, but generally, people who code in a vacuum like that
don't often write good code. In fact, the best coders I know don't get
to write their own code very often. They spend most of their time
fixing someone else's bugs, since they're the only ones who can.
Finally I hate single character variable names, even in example code, even
though I sometimes (rarely) use them myself :)

There's nothing wrong with it if its scope is very localized. If I can
see what 'x' is simply by looking at the line of code above where it's
used, there's no reason to give 'x' a long name; it's just clutter.

Aaron