In what situations Argument Dependent name Look-up (ADL) kicks in? - c++

In Wikipedia article below quote is mentioned:
ADL only occurs if the normal lookup of an unqualified name fails to
find a matching class member function. In this case, other namespaces
not considered during normal lookup may be searched where the set of
namespaces to be searched depends on the types of the function
arguments.
So, I was expecting below program would compile fine, but it doesn't:
namespace N1 {
class A {};
void foo (A *p) {}
}
namespace N2 {
void foo (N1::A &p) {}
}
int main () {
N1::A xa;
foo(&xa); // ok
foo(xa); // error: cannot convert ‘N1::A’ to ‘N1::A*’ for argument ‘1’ to ‘void N1::foo(N1::A*)’
}
I searched several questions in SO, but couldn't find which lists the requirements or situations in simple word which suggests: When the ADL kicks in ?
A little more detailed answer would be really helpful to me and future visitors.

It shouldn't compiles. A is in namespace N1. How compiler should knows, that you want to call N2::foo?
n3376 3.4.2/2
For each argument type T in the function call, there is a set of zero or more associated namespaces and a
set of zero or more associated classes to be considered. The sets of namespaces and classes is determined
entirely by the types of the function arguments (and the namespace of any template template argument).
Typedef names and using-declarations used to specify the types do not contribute to this set. The sets of
namespaces and classes are determined in the following way:
If T is a class type (including unions), its associated classes are: the class itself; the class of which it is a
member, if any; and its direct and indirect base classes. Its associated namespaces are the namespaces
of which its associated classes are members. Furthermore, if T is a class template specialization,
its associated namespaces and classes also include: the namespaces and classes associated with the
types of the template arguments provided for template type parameters (excluding template template
parameters); the namespaces of which any template template arguments are members; and the classes
of which any member templates used as template template arguments are members. [ Note: Non-type
template arguments do not contribute to the set of associated namespaces. — end note ]

ADL kicks in pretty much always, as soon as your function takes user a user defined type. It's kicking in for foo here: xa is defined in N1, so foo is searched in N1 as well as in the global namespace. (Without ADL, foo would only be searched in the global namespace.)
And I don't see why you would expect the second call to foo to compile. The type of xa is defined in N1, so ADL adds N1 to the search path, but there's nothing in the expression to imply N2.

It says "other namespaces" are searched. It doesn't say "all namespaces" are searched.
The rules for what extra namespaces are included in ADL are a little complicated, but the most important one is the namespace in which A is defined. That's why your first foo is found. Your second foo can't be found because namespace N2 is nothing to do with anything, and it is not searched by ADL.

If unqualified Name Look-up fails, then look-up proceeds using the argument of the function's call.
Example
func(A x);
Then compiler will look at namespace up starting at namespace including class A. One example is this
// argument_dependent_name_koenig_lookup_on_functions.cpp
namespace A
{
struct X
{
};
void f(const X&)
{
}
}
int main()
{
// The compiler finds A::f() in namespace A, which is where
// the type of argument x is defined. The type of x is A::X.
A::X x;
f(x);
}
More here
http://msdn.microsoft.com/en-us/library/60bx1ys7.aspx

The compiler stops lookup once he has found a function with a matching name. It does not continue searching if the argument types or accessibility (public/protected/private) actually prevent using the function in the current context. Hence in your example, the compiler has no change to "see" N2::foo, since N1::foo is found first.
Note that in your example, N2::foo wouldn't be found even if N1::foo did not exist, as you have no reference to N2 anywhere inside main, so N2 will not be searched at all.

Related

Question about resolution of free function names in C++ templates

This program works as expected:
#include <iostream>
template <typename T>
void output(T t) {
prt(t);
}
struct It {
It(int* p) : p(p) {}
int* p;
};
void prt(It it) {
std::cout << *(it.p) << std::endl;
}
int main() {
int val = 12;
It it(&val);
output(it);
return 0;
}
When you compile and execute this, it prints "12" as it should. Even though the function prt, required by the output template function, is defined after output, prt is visible at the point of instantiation, and therefore everything works.
The program below is very similar to the program above, but it fails to compile:
#include <iostream>
template <typename T>
void output(T t) {
prt(t);
}
void prt(int* p) {
std::cout << (*p) << std::endl;
}
int main() {
int val = 12;
output(&val);
return 0;
}
This code is trying to do the same thing as the previous example, but this fails in gcc 8.2 with the error message:
'prt' was not declared in this scope, and no declarations were found by
argument-dependent lookup at the point of instantiation [-fpermissive]
The only thing that changed is that the argument passed to output is a built-in type, rather than a user-defined type. But I didn't think that should matter for name resolution. So my question is: 1) why does the second example fail?; and 2) why does one example fail and the other succeeds?
The Standard rule that applies here is found in [temp.dep.candidate]:
For a function call where the postfix-expression is a dependent name, the candidate functions are found using the usual lookup rules ([basic.lookup.unqual], [basic.lookup.argdep]) except that:
For the part of the lookup using unqualified name lookup, only function declarations from the template definition context are found.
For the part of the lookup using associated namespaces ([basic.lookup.argdep]), only function declarations found in either the template definition context or the template instantiation context are found.
In both examples, unqualified name lookup finds no declarations of prt, since there were no such declarations before the point where the template was defined. So we move on to argument-dependent lookup, which looks only in the associated namespaces of the argument types.
Class It is a member of the global namespace, so the global namespace is the one associated namespace, and the one declaration is visible within that namespace in the template instantiation context.
A pointer type U* has the same associated namespaces as type U, and a fundamental type has no associated namespaces at all. So since the only argument type int* is a pointer to fundamental type, there are no associated namespaces, and argument-dependent lookup can't possibly find any declarations in the second program.
I can't exactly say why the rules were designed this way, but I would guess the intent is that a template should either use the specific declared functions it meant to use, or else use a function as an extensible customization point, but those user customizations need to be closely related to a user-defined type they will work with. Otherwise, it becomes possible to change the behavior of a template that really meant to use one specific function or function template declaration by providing a better overload for some particular case. Admittedly, this is more from the viewpoint of when there is at least one declaration in the template definition context, not when that lookup finds nothing at all, but then we get into cases where SFINAE was counting on not finding something, etc.

Declare function after template defined

Let's say I have a template function:
template <class T>
void tfoo( T t )
{
foo( t );
}
later I want to use it with a type, so I declare/define a function and try to call it:
void foo( int );
int main()
{
tfoo(1);
}
and I am getting error from g++:
‘foo’ was not declared in this scope, and no declarations were found by argument-dependent lookup at the point of instantiation [-fpermissive]
foo( t );
why it cannot find void foo(int) at the point of instantiation? It is declared at that point. Is there a way to make it work (without moving declaration of foo before template)?
foo in your case is a dependent name, since function choice depends on the type if the argument and the argument type depends on the template parameter. This means that foo is looked up in accordance with the rules of dependent lookup.
The difference between dependent and non-dependent lookup is that in case of dependent lookup ADL-nominated namespaces are seen as extended: they are extended with extra names visible from the point of template instantiation (tfoo call in your case). That includes the names, which appeared after the template declaration. The key point here is that only ADL-nominated namespaces are extended in this way.
(By ADL-nominated namespace I refer to namespace associated with function argument type and therefore brought into consideration by the rules of dependent name lookup. See "3.4.2 Argument-dependent name lookup")
In your case the argument has type int. int is a fundamental type. Fundamental types do not have associated namespaces (see "3.4.2 Argument-dependent name lookup"), which means that it does not nominate any namespace through ADL. In your example ADL is not involved at all. Dependent name lookup for foo in this case is no different from non-dependent lookup. It will not be able to see your foo, since it is declared below the template.
Note the difference with the following example
template <class T> void tfoo( T t )
{
foo( t );
}
struct S {};
void foo(S s) {}
int main()
{
S s;
tfoo(s);
}
This code will compile since argument type S is a class type. It has an associated namespace - the global one - and it adds (nominates) that global namespace for dependent name lookup. Such ADL-nominated namespaces are seen by dependent lookup in their updated form (as seen from the point of the call). This is why the lookup can see foo and completes successfully.
It is a rather widespread misconception when people believe that the second phase of so called "two-phase lookup" should be able to see everything that was additionally declared below template definition all the way to the point of instantiation (point of the call in this case).
No, the second phase does not see everything. It can see the extra stuff only in namespaces associated with function arguments. All other namespaces do not get updated. They are seen as if observed from the point of template definition.

Namespace causes sub-optimal template overload resolution

This is very similar to this question, but I'm not sure the answer there is entirely applicable to the minimal code I've put together that demonstrates the issue. (My code does not use trailing-return types, and there are some other differences as well.) Additionally, the issue of whether MSVC's behavior is legal doesn't seem to be addressed.
In short, I'm seeing the compiler select a generic function template instantiation rather than a more-specific overload when the function template is inside a namespace.
Consider the following set of namespace and class definitions:
namespace DoStuffUtilNamespace
{
template<typename UNKNOWN>
void doStuff(UNKNOWN& foo)
{
static_assert(sizeof(UNKNOWN) == -1, "CANNOT USE DEFAULT INSTANTIATION!");
}
}
class UtilForDoingStuff
{
public:
template <typename UNKNOWN>
void doStuffWithObjectRef(UNKNOWN& ref)
{
DoStuffUtilNamespace::doStuff(ref);
}
};
class MyClassThatCanDoStuff { };
namespace DoStuffUtilNamespace
{
using ::MyClassThatCanDoStuff; // No effect.
void doStuff(MyClassThatCanDoStuff& foo) { /* No assertion! */ }
}
... and the following use-cases:
int main()
{
MyClassThatCanDoStuff foo;
DoStuffUtilNamespace::MyClassThatCanDoStuff scoped_foo;
UtilForDoingStuff util;
DoStuffUtilNamespace::doStuff(foo); // Compiles
DoStuffUtilNamespace::doStuff(scoped_foo); // Compiles
util.doStuffWithObjectRef(foo); // Triggers static assert
util.doStuffWithObjectRef(scoped_foo); // Triggers static assert
}
If the entire DoStuffUtilNamespace is eliminated and all its members are moved to global scope, this compiles fine with G++ and Clang++.
With the namespace, doStuff is of course a dependent name. According to the top-voted answer on the similar question, the standard says:
In resolving dependent names, names from the following sources are considered:
Declarations that are visible at the point of definition of the template.
Declarations from namespaces associated with the types of the function arguments both from the instantiation context and from the definition context.
This seems a little odd to me; I don't understand why the first bullet point would specify that the declarations must be visible at the point of definition of the template rather than at the point of instantiation, since the second bullet point explicitly specifies that some declarations visible only at the point of instantiation are allowed. (If someone would like to offer a rationale, I'd appreciate it, but that's not my question because it's my understanding that questions of the form "why did the standards committee decide X" are off topic.)
So I think that explains why util.doStuffWithObjectRef(foo); triggers the static assertion: doStuff(MyClassThatCanDoStuff&) hasn't been declared at the point of definition of UtilForDoingStuff::doStuffWithObjectRef<UNKNOWN>(UNKNOWN&). And indeed moving the class UtilForDoingStuff definition after the doStuff overload has been defined seems to fix the issue.
But what exactly does the standard mean by "namespaces associated with the types of the function arguments"? Shouldn't the using ::MyClassThatCanDoStuff declaration, together with the explicit scoping of the scoped_foo instance type within the namespace, trigger argument-dependent lookup, and shouldn't this look-up find the non-asserting definition of doStuff()?
Also, the entire code is compiled without error using clang++ -ftemplate-delayed-parsing, which emulates MSVC's template-parsing behavior. This seems preferable, at least in this particular case, because the ability to add new declarations to a namespace at any time is one of the primary appeals of namespaces. But, as noted above, it doesn't quite seem to follow the letter of the law, according to the standard. Is it permissible, or is it an instance of non-conformance?
EDIT:: As pointed out by KIIV, there is a workaround; the code compiles if template specialization is used instead of overloading. I would still like to know the answers to my questions about the standard.
With the namespace, doStuff is of course a dependent name.
You are starting from the wrong premise. There is no ADL for a qualified call like DoStuffUtilNamespace::doStuff(ref). [basic.lookup.argdep]/p1, emphasis mine:
When the postfix-expression in a function call (5.2.2) is an
unqualified-id, other namespaces not considered during the usual
unqualified lookup (3.4.1) may be searched, and in those namespaces,
namespace-scope friend function or function template declarations
(11.3) not otherwise visible may be found.
DoStuffUtilNamespace::doStuff is a qualified-id, not an unqualified-id. ADL doesn't apply.
For this reason, DoStuffUtilNamespace::doStuff is also not a dependent name. [temp.dep]/p1:
In an expression of the form:
postfix-expression ( expression-listopt)
where the postfix-expression is an unqualified-id, the
unqualified-id denotes a dependent name if [...]. If an operand of an operator is a type-dependent expression, the operator also denotes
a dependent name. Such names are unbound and are looked up at the
point of the template instantiation (14.6.4.1) in both the context of
the template definition and the context of the point of instantiation
(The italicization of dependent name indicate that this paragraph is defining the term.)
Instead, per [temp.nondep]/p1:
Non-dependent names used in a template definition are found using the
usual name lookup and bound at the point they are used.
which doesn't find your later overload declaration.
Specialization works because it's still the same function template declaration that's used; you just supplied a different implementation than the default one.
But what exactly does the standard mean by "namespaces associated with
the types of the function arguments"? Shouldn't the using ::MyClassThatCanDoStuff declaration, together
with the explicit scoping of the scoped_foo instance type within the
namespace, trigger argument-dependent lookup
No. using-declarations do not affect ADL. [basic.lookup.argdep]/p2, emphasis mine:
For each argument type T in the function call, there is a set of
zero or more associated namespaces and a set of zero or more
associated classes to be considered. The sets of namespaces and
classes is determined entirely by the types of the function arguments
(and the namespace of any template template argument).
Typedef names and using-declarations used to specify the types do not contribute to this set. The sets of namespaces and classes are
determined in the following way:
If T is a fundamental type, [...]
If T is a class type (including unions), its associated classes are: the class itself; the class of which it is a member, if any; and its
direct and indirect base classes. Its associated namespaces are the
innermost enclosing namespaces of its associated classes. Furthermore,
if T is a class template specialization, its associated namespaces and
classes also include: the namespaces and classes associated with the
types of the template arguments provided for template type parameters
(excluding template template parameters); the namespaces of which any
template template arguments are members; and the classes of which any
member templates used as template template arguments are members. [
Note: Non-type template arguments do not contribute to the set of associated namespaces. —end note ]
[...]
With template specialization I can get it work:
namespace DoStuffUtilNamespace
{
template<typename UNKNOWN>
void doStuff(UNKNOWN& foo)
{
static_assert(sizeof(UNKNOWN) == -1, "CANNOT USE DEFAULT INSTANTIATION!");
}
}
class UtilForDoingStuff
{
public:
template <typename UNKNOWN>
void doStuffWithObjectRef(UNKNOWN& ref)
{
DoStuffUtilNamespace::doStuff(ref);
}
};
class MyClassThatCanDoStuff { };
namespace DoStuffUtilNamespace
{
using ::MyClassThatCanDoStuff;
template <> void doStuff<MyClassThatCanDoStuff>(MyClassThatCanDoStuff& foo) { /* No assertion! */ }
}
int main()
{
MyClassThatCanDoStuff foo;
DoStuffUtilNamespace::MyClassThatCanDoStuff scoped_foo;
UtilForDoingStuff util;
DoStuffUtilNamespace::doStuff(foo); // Compiles
DoStuffUtilNamespace::doStuff(scoped_foo); // Compiles
util.doStuffWithObjectRef(foo); // Compiles
util.doStuffWithObjectRef(scoped_foo); // Compiles
}
Declarations from namespaces associated with the types of the function arguments both from the instantiation context and from the definition context.
Example with the following code which prints B::foo Demo
namespace A
{
template <typename T>
void foo(const T&) {std::cout << "A::foo" << std::endl;}
template <typename T>
void bar(const T& t) {
foo(t); // thank to ADL, it will also look at B::foo for B::S.
}
}
namespace B
{
struct S {};
void foo(const S&) {std::cout << "B::foo" << std::endl;}
}
int main()
{
B::S s;
A::bar(s);
}
So when calling ?::foo(const B::S&), the second bullet point adds B::foo to the list of overloads.
why template-specialization works in this case
There is only one function:
template<>
void DoStuffUtilNamespace::doStuff<MyClassThatCanDoStuff>(MyClassThatCanDoStuff& foo);
even if it is defined later.
Note that the fact that there is a specialization should be known in the translation unit, else the program is ill formed (doesn't respect ODR).
while overloading doesn't.
You think:
So I think that explains why util.doStuffWithObjectRef(foo); triggers the static assertion: doStuff(MyClassThatCanDoStuff&) hasn't been declared at the point of definition of UtilForDoingStuff::doStuffWithObjectRef<UNKNOWN>(UNKNOWN&). And indeed moving the class UtilForDoingStuff definition after the doStuff overload has been defined seems to fix the issue.
Exactly.

Confusion over argument dependent lookup and friend function definition

I have the following program:
struct Foo {
friend void foo (int) {}
operator int () const { return 0; }
};
int main() {
foo(Foo()); // This compiles fine
foo(0); // This fails to find `foo()`
return 0;
}
I could not figure out what part of the standard defines the ADL rule that says the call to foo(0) should fail, while the call to foo(Foo()) should succeed. Can someone shed some light?
Reopened over my own deletion, eh? Well, after some prompting from juanchopanza, I read over the relevant part of the standard, C++11 §3.4.2, more carefully. I browsed it way too quickly before. To wit, paragraph 2 spells out:
For each argument type T in the function call, there is a set of zero or more associated namespaces and a set of zero or more associated classes to be considered. The sets of namespaces and classes is determined entirely by the types of the function arguments (and the namespace of any template template argument). Typedef names and using-declarations used to specify the types do not contribute to this set. The sets of namespaces and classes are determined in the following way:
If T is a fundamental type, its associated sets of namespaces and classes are both empty.
If T is a class type (including unions), its associated classes are: the class itself; the class of which it is a member, if any; and its direct and indirect base classes. Its associated namespaces are the namespaces of which its associated classes are members. Furthermore, if T is a class template specialization, its associated namespaces and classes also include: the namespaces and classes associated with the types of the template arguments provided for template type parameters (excluding template template parameters); the namespaces of which any template template arguments are members; and the classes of which any member templates used as template template arguments are members.
...
The first bullet explains why passing in an int argument fails. The second bullet explains why passing in a Foo instance works.
Looking at your code:
int main() {
foo(Foo()); // This declares a global function "foo" taking a "Foo" as an argument
foo(0); // "foo" takes a "Foo" above, and theres no way to convert an int to a Foo...

Why will two-phase lookup fail to choose overloaded version of 'swap'?

I am studying this fascinating answer to a subtle question regarding the best practice to implement the swap function for user-defined types. (My question was initially motivated by a discussion of the illegality of adding types to namespace std.)
I will not re-print the code snippet from the above-linked answer here.
Instead, I would like to understand the answer.
The answer I've linked above states, beneath the first code snippet, in regards to overloading swap in namespace std (rather than specializing it in that namespace):
If your compiler prints out something different then it is not
correctly implementing "two-phase lookup" for templates.
The answer then goes on to point out that specializing swap in namespace std (as opposed to overloading it) produces a different result (the desired result in the case of specialization).
However, the answer proceeds with an additional case: specializing swap for a user-defined template class - in which case, again, the desired result is not achieved.
Unfortunately, the answer simply states the facts; it does not explain why.
Can someone please elaborate on that answer, and describe the process of lookup in the two specific code snippets provided in that answer:
overloading swap in namespace std for a user-defined non-template class (as in the first code snippet of the linked answer)
specializing swap in namespace std for a user-defined template class (as in the final code snippet of the linked answer)
In both cases, the generic std::swap is called, rather than the user-defined swap. Why?
(This will shed light on the nature of two-phase lookup, and the reason for the best practice for implementing user-defined swap; thanks.)
Preamble with plenty of Standardese
The call to swap() in the example entails a dependent name because its arguments begin[0] and begin[1] depend on the template parameter T of the surrounding algorithm() function template. Two-phase name lookup for such dependent names is defined in the Standard as follows:
14.6.4.2 Candidate functions [temp.dep.candidate]
1 For a function call where the postfix-expression is a dependent name,
the candidate functions are found using the usual lookup rules (3.4.1,
3.4.2) except that:
— For the part of the lookup using unqualified name lookup (3.4.1), only function declarations from the template definition
context are found.
— For the part of the lookup using associated
namespaces (3.4.2), only function declarations found in either the
template definition context or the template instantiation context are
found.
Unqualified lookup is defined by
3.4.1 Unqualified name lookup [basic.lookup.unqual]
1 In all the cases listed in 3.4.1, the scopes are searched for a
declaration in the order listed in each of the respective categories;
name lookup ends as soon as a declaration is found for the name. If no
declaration is found, the program is ill-formed.
and argument-dependent lookup (ADL) as
3.4.2 Argument-dependent name lookup [basic.lookup.argdep]
1 When the postfix-expression in a function call (5.2.2) is an
unqualified-id, other namespaces not considered during the usual
unqualified lookup (3.4.1) may be searched, and in those namespaces,
namespace-scope friend function or function template declarations
(11.3) not otherwise visible may be found. These modifications to the
search depend on the types of the arguments (and for template template
arguments, the namespace of the template argument).
Applying the Standard to the example
The first example calls exp::swap(). This is not a dependent name and does not require two-phase name lookup. Because the call to swap is qualified, ordinary lookup takes place which finds only the generic swap(T&, T&) function template.
The second example (what #HowardHinnant calls "the modern solution") calls swap() and also has an overload swap(A&, A&) in the same namespace as where class A lives (the global namespace in this case). Because the call to swap is unqualified, both ordinary lookup and ADL take place at the point of definition (again only finding the generic swap(T&, T&)) but another ADL takes place at the point of instantiation (i.e where exp::algorithm() is being called in main()) and this picks up swap(A&, A&) which is a better match during overload resolution.
So far so good. Now for the encore: the third example calls swap() and has a specialization template<> swap(A&, A&) inside namespace exp. The lookup is the same as in the second example, but now ADL does not pick up the template specialization because it is not in an associated namespace of class A. However, even though the specialization template<> swap(A&, A&) does not play a role during overload resolution, it is still instantiated at the point of use.
Finally, the fourth example calls swap() and has an overload template<class T> swap(A<T>&, A<T>&) inside namespace exp for template<class T> class A living in the global namespace. The lookup is the same as in the third example, and again ADL does not pick up the overload swap(A<T>&, A<T>&) because it is not in an associated namespace of the class template A<T>. And in this case, there is also no specialization that has to be instantiated at the point of use, so the generic swap(T&, T&) is being callled here.
Conclusion
Even though you are not allowed to add new overloads to namespace std, and only explicit specializations, it would not even work because of the various intricacies of two-phase name lookup.
It is impossible to overload swap in namespace std for a user defined type. Introduction an overload (as opposed to a specialization) in namespace std is undefined behavior (illegal under the standard, no diagnosis required).
It is impossible to specialize a function in general for a template class (as opposed to a template class instance -- ie, std::vector<int> is an instance, while std::vector<T> is the entire template class). What appears to be a specialization is actually an overload. So the first paragraph applies.
The best practice for implementing user-defined swap is to introduce a swap function or overload in the same namespace as your template or class lives in.
Then, if swap is called in the right context (using std::swap; swap(a,b);), which is how it is called in std library, ADL will kick in, and your overload will be found.
The other option is to do a full specialization of swap in std for your particular type. This is impossible (or impractical) for template classes, as you need to specialize for each and every instance of your template class that exists. For other classes, it is fragile, as specialization applies to only that particular type: subclasses will have to be respecialized in std as well.
In general, specialization of functions is extremely fragile, and you are better off introducing overrides. As you cannot introduce overrides into std, the only place they will be reliably found from is in your own namespace. Such overrides in your own namespace are preferred over overrides in std as well.
There are two ways to inject a swap into your namespace. Both work for this purpose:
namespace test {
struct A {};
struct B {};
void swap(A&, A&) { std::cout << "swap(A&,A&)\n"; }
struct C {
friend void swap(C&, C&) { std::cout << "swap(C&, C&)\n"; }
};
void bob() {
using std::swap;
test::A a, b;
swap(a,b);
test::B x, y;
swap(x, y);
C u, v;
swap(u, v);
}
}
void foo() {
using std::swap;
test::A a, b;
swap(a,b);
test::B x, y;
swap(x, y);
test::C u, v;
swap(u, v);
test::bob();
}
int main() {
foo();
return 0;
}
the first is to inject it into the namespace directly, the second is to include it as an inline friend. The inline friend for "external operators" is a common pattern that basically means you can only find swap via ADL, but in this particular context does not add much.