Is it safe to check a pointer to not being NULL
by writing simply if(pointer)
or do I have to use if(pointer != NULL)
?
14 Answers
You can; the null pointer is implicitly converted into boolean false while non-null pointers are converted into true. From the C++11 standard, section on Boolean Conversions:
A prvalue of arithmetic, unscoped enumeration, pointer, or pointer to member type can be converted to a prvalue of type
bool
. A zero value, null pointer value, or null member pointer value is converted tofalse
; any other value is converted totrue
. A prvalue of typestd::nullptr_t
can be converted to a prvalue of typebool
; the resulting value isfalse
.
Yes, you could.
- A null pointer is converted to false implicitly
- a non-null pointer is converted to true.
This is part of the C++ standard conversion, which falls in Boolean conversion clause:
§ 4.12 Boolean conversions
A prvalue of arithmetic, unscoped enumeration, pointer, or pointer to member type can be converted to a prvalue of type bool. A zero value, null pointer value, or null member pointer value is converted to false; any other value is converted to true. A prvalue of type std::nullptr_t can be converted to a prvalue of type bool; the resulting value is false.
Yes, you can. In fact, I prefer to use if(pointer)
because it's simpler to read and write once you get used to it.
Also note that C++11 introduced nullptr
which is preferred over NULL
.
-
10A pointer isn't a boolean expression. It's converted implicitly. If it's better to read when you have to remember this conversion to understand is your opinion. It's just one kind of coding style.– harperCommented Jul 21, 2013 at 12:17
-
7@harper You can say it's a coding style. But you can apply the same logic to
if(som_integer)
vsif(some_integer != 0)
because integers are also not booleans, right? I prefer to avoid0
orNULL
in an if-statement.– Yu HaoCommented Jul 21, 2013 at 12:33 -
13I would agree it's simply a matter of coding style. I have come to prefer
if (pointer)
myself, butif (ptr != nullptr)
seems perfectly legitimate to me. On the other hand, if I saw someone on my team who wroteif (some_integer)
I would make them change it toif (some_integer != 0)
. However, I won't pretend that's not a relatively arbitrary preference on my part - I simply prefer not to treat pointers and integers the same.– JoelCommented Jul 21, 2013 at 13:48 -
5@franji1 Then how about
if(isReady)
if(filePtr)
if(serviceNo)
? Making bad variable names on purpose doesn't mean much in this case. Anyway I already got your point and understood it, but I can insist using my own coding style myself, OK?– Yu HaoCommented Jul 22, 2013 at 13:13 -
2@YuHao - I only meant to obfuscate the type on purpose to emphasize my point. I decided to generate an "answer" that better gets my point across. (Sadly, I tend to write code using both styles, but tend to prefer seeing the expanded version when doing maintenance, but tend to prefer the shorter version shen creating it).– franji1Commented Jul 22, 2013 at 13:20
Question is answered, but I would like to add my points.
I will always prefer if(pointer)
instead of if(pointer != NULL)
and if(!pointer)
instead of if(pointer == NULL)
:
- It is simple, small
Less chances to write a buggy code, suppose if I misspelled equality check operator
==
with=
if(pointer == NULL)
can be misspelledif(pointer = NULL)
So I will avoid it, best is justif(pointer)
.
(I also suggested some Yoda condition in one answer, but that is diffrent matter)Similarly for
while (node != NULL && node->data == key)
, I will simply writewhile (node && node->data == key)
that is more obvious to me (shows that using short-circuit).- (may be stupid reason) Because NULL is a macro, if suppose some one redefine by mistake with other value.
-
8Using = instead of == almost always generates a compiler warning, in the days when it didn't people would use if ( NULL == ptr )– paulmCommented Jul 21, 2013 at 17:25
-
2@paulm that I just added this point its called Yoda Condition some people don't like it as its less readable. Commented Jul 21, 2013 at 17:26
-
3
(boolean expression)? true : false
is completely pointless. The expression evaluates either totrue
or tofalse
; what you say is "if it's true, give me true, if it's false, give me false". In short: It's completely equivalent to the boolean expression itself. Note thatnode == NULL
is a boolean expression. BTW, your two implementations return exactly the opposite of each other. Either you want!=
in the first, or only one!
in the second.– celtschkCommented Jul 21, 2013 at 18:26 -
BTW, one possible protection against
=
instead of==
is to make your variablesconst
whenever possible. For example, you could define your function asisEmnpy(node* const head) { ... }
, and then the compiler would refuse to compile it if you accidentally wrotenode = NULL
instead ofnode == NULL
. Of course that only works for variables which you really don't need to change.– celtschkCommented Jul 21, 2013 at 18:32 -
2Because the smart pointer classes have
T* get() const
instead ofoperator T*() const
to avoid implicit conversions. They do however have anoperator bool() const
. Commented Jul 24, 2013 at 8:03
Explicitly checking for NULL could provide a hint to the compiler on what you are trying to do, ergo leading to being less error-prone.
Yes, you can. The ability to compare values to zeros implicitly has been inherited from C, and is there in all versions of C++. You can also use if (!pointer)
to check pointers for NULL.
The relevant use cases for null pointers are
- Redirection to something like a deeper tree node, which may not exist or has not been linked yet. That's something you should always keep closely encapsulated in a dedicated class, so readability or conciseness isn't that much of an issue here.
Dynamic casts. Casting a base-class pointer to a particular derived-class one (something you should again try to avoid, but may at times find necessary) always succeeds, but results in a null pointer if the derived class doesn't match. One way to check this is
Derived* derived_ptr = dynamic_cast<Derived*>(base_ptr); if(derived_ptr != nullptr) { ... }
(or, preferrably,
auto derived_ptr = ...
). Now, this is bad, because it leaves the (possibly invalid, i.e. null) derived pointer outside of the safety-guardingif
block's scope. This isn't necessary, as C++ allows you to introduce boolean-convertable variables inside anif
-condition:if(auto derived_ptr = dynamic_cast<Derived*>(base_ptr)) { ... }
which is not only shorter and scope-safe, it's also much more clear in its intend: when you check for null in a separate if-condition, the reader wonders "ok, so
derived_ptr
must not be null here... well, why would it be null?" Whereas the one-line version says very plainly "if you can safely castbase_ptr
toDerived*
, then use it for...".The same works just as well for any other possible-failure operation that returns a pointer, though IMO you should generally avoid this: it's better to use something like
boost::optional
as the "container" for results of possibly failing operations, rather than pointers.
So, if the main use case for null pointers should always be written in a variation of the implicit-cast-style, I'd say it's good for consistency reasons to always use this style, i.e. I'd advocate for if(ptr)
over if(ptr!=nullptr)
.
I'm afraid I have to end with an advert: the if(auto bla = ...)
syntax is actually just a slightly cumbersome approximation to the real solution to such problems: pattern matching. Why would you first force some action (like casting a pointer) and then consider that there might be a failure... I mean, it's ridiculous, isn't it? It's like, you have some foodstuff and want to make soup. You hand it to your assistant with the task to extract the juice, if it happens to be a soft vegetable. You don't first look it at it. When you have a potato, you still give it to your assistant but they slap it back in your face with a failure note. Ah, imperative programming!
Much better: consider right away all the cases you might encounter. Then act accordingly. Haskell:
makeSoupOf :: Foodstuff -> Liquid
makeSoupOf p@(Potato{..}) = mash (boil p) <> water
makeSoupOf vegetable
| isSoft vegetable = squeeze vegetable <> salt
makeSoupOf stuff = boil (throwIn (water<>salt) stuff)
Haskell also has special tools for when there is really a serious possibility of failure (as well as for a whole bunch of other stuff): monads. But this isn't the place for explaining those.
⟨/advert⟩
-
1I can only see one sentence in this interminable screed that actually answers the question. Commented Jul 21, 2013 at 22:18
-
@EJP: if you take the question literally ("can I use"), then it's not answered explicitly at all (the answer is simply "yes"). I tried to give proper reasons for why the OP should in fact use
if(ptr)
rather thanif(ptr != nullptr)
, to which there is quite a bit more to say. Commented Jul 22, 2013 at 1:52
Yes, Both are functionally the same thing. But in C++ you should switch to nullptr in the place of NULL;
Yes. In fact you should. If you're wondering if it creates a segmentation fault, it doesn't.
yes, of course! in fact, writing if(pointer) is a more convenient way of writing rather than if(pointer != NULL) because: 1. it is easy to debug 2. easy to understand 3. if accidently, the value of NULL is defined, then also the code will not crash
As others already answered well, they both are interchangeable.
Nonetheless, it's worth mentioning that there could be a case where you may want to use the explicit statement, i.e. pointer != NULL
.
Yes, you can always do this as 'IF' condition evaluates only when the condition inside it goes true. C does not have a boolean return type and thus returns a non-zero value when the condition is true while returns 0 whenever the condition in 'IF' turns out to be false. The non zero value returned by default is 1. Thus, both ways of writing the code are correct while I will always prefer the second one.
-
1The non-zero value by default is undefined if I remember correctly. Commented Jul 21, 2013 at 22:19
I think as a rule of thumb, if your if-expression can be re-written as
const bool local_predicate = *if-expression*;
if (local_predicate) ...
such that it causes NO WARNINGS, then THAT should be the preferred style for the if-expression. (I know I get warnings when I assign an old C BOOL
(#define BOOL int
) to a C++ bool
, let alone pointers.)
"Is it safe..?" is a question about the language standard and the generated code.
"Is is a good practice?" is a question about how well the statement is understood by any arbitrary human reader of the statement. If you are asking this question, it suggests that the "safe" version is less clear to future readers and writers.
-
My intention was to ask if it is safe. Thefore I used that wording. However, what you wrote here is not an answer to the question. Instead, it should be a comment under the question. You can delete the answer and add a comment under the question then.– danijarCommented Jul 24, 2013 at 11:29
-
@danijar Don't you remember when you was new to StackOverflow and searched for the 'Comment' section without success? Someone with 7 reputation can't do that.– BroxzierCommented Jul 24, 2013 at 13:20
-
@JimBalter Which is very confusing, since you can see others do so. When I was new to SO someone blamed me for doing that.– BroxzierCommented Aug 26, 2013 at 9:09
-
@JimBalter I am not murdering and stealing. I was telling danijar that Fred Mitchell was a new user and could not post comments.– BroxzierCommented Aug 26, 2013 at 9:14
-
@JimBalter Which you started today. Also you are the one not understanding instead. That comment is only supporting the confusing of this.– BroxzierCommented Aug 26, 2013 at 9:47
0
ornullptr
. (NULL
is a C'ism, and requires including a header file.)NULL
in C++ from hereon becauseNULL
is implementation dependent macro which might gives you ambiguous behaviors.