Ignoring the return value of a function
Mark Hobley
int foobar(int n)
If I am not interested in the return value of the function, is it legal to
use the function in statement context as follows?
foobar(6); /* Is this legal? */
Or, is it necessary to create a dummy variable for the return value, and
disregard it, as follows?
int tmp;
tmp = foobar(6); /* variable tmp is not used anywhere else in the code */
Mark.
--
Mark Hobley
Linux User: #370818 http://markhobley.yi.org/
Ben Pfaff
> I have a function with a prototype as follows:
>
> int foobar(int n)
>
> If I am not interested in the return value of the function, is it legal to
> use the function in statement context as follows?
>
> foobar(6); /* Is this legal? */
Yes.
--
"Structure padding is the use of extraneous materials to
enhance the shape of a struct and make it more attractive to
members of the opposite struct. (See also "struct silicone.")"
--Eric Sosman
Andrew Poelstra
> markh...@hotpop.donottypethisbit.com (Mark Hobley) writes:
>
>> I have a function with a prototype as follows:
>>
>> int foobar(int n)
>>
>> If I am not interested in the return value of the function, is it legal to
>> use the function in statement context as follows?
>>
>> foobar(6); /* Is this legal? */
>
> Yes.
If you use lint, it might complain about it (since the
int return value is there for a reason, therefore you
should use it, goes the theory).
If you know the return value is irrelevant, you can do
(void) foobar(6);
if it makes you feel better.
--
Andrew Poelstra
http://www.wpsoftware.net/andrew
Paul N
wrote:
> I have a function with a prototype as follows:
>
> int foobar(int n)
>
> If I am not interested in the return value of the function, is it legal to
> use the function in statement context as follows?
>
> foobar(6); /* Is this legal? */
Yes, and you have probably already done so without realising. printf
returns an integer, which is nearly always ignored.
Ersek, Laszlo
> printf returns an integer, which is nearly always ignored.
An unacceptable practice, in my opinion. I can think of two excuses:
- You are calling multiple output operations on the same stdio stream in a
row, and you finish off that sequence with a call to ferror() or a checked
fflush(). (In the ferror() case with appropriately line-buffered or
unbuffered streams, no actual output need to have happened due to the
stream operations, so ferror() may not have a chance to signal any error,
but that is no problem, if the technique is used consistently, *and* there
is a final checked fflush() or fclose() on the stream.)
- You are writing an error message to stderr on your way to exit with a
non-zero exit status.
Sometimes, if logging doesn't work, nothing should work. (The expression
"logging works" can be defined with different guarantees.) Sometimes,
looking at an incomplete log file on a full disk, it is better to know
that nothing happened than not to know what happened.
Cheers,
lacos
Richard Bos
But please don't, because it will make most people who read your code
feel worse. Instead, get a better lint - these days, your compiler
should come with one, possibly built-in.
Richard
Nick
There are times when I do this - when I have a function whose return
value I usually use, but sometimes don't.
For example, the "create new stack frame" function in my interpreter
returns a pointer. This pointer can be used as a parameter to an "exit
from nested frames" to return early from a function (inside a loop,
(which creates its own frame) for example). Right at the start of
running a program I call the function to create a stack frame for the
entire program. I don't use that (early termination works entirely
differently) at all, so discard it with a (void) to mark that I know
there is a return value, and I know I don't want it.
--
Online waterways route planner | http://canalplan.eu
Plan trips, see photos, check facilities | http://canalplan.org.uk