I have not been able to find an answer on this, but my case is:
// vulkan_glfw_backend.hpp
struct alignas(8) VulkanGlfwWindowContext;
class MY_API VulkanGlfwBackend
{
// [...]
private:
VulkanGlfwWindowContext* mpContext;
};
And the source file, where I want to have the implementation:
// vulkan_glfw_backend.cpp
#include "vulkan_glfw_backend.hpp"
struct VulkanGlfwWindowContext
{
int numWindows;
GLFWwindow* windows[MAX_WINDOWS];
};
Initially, my compiler complained because it couldn't determine the alignment requirements for the class, which I suppose makes sense. Then I added the alignas attribute. But now I get an error message, which I cannot understand the logical reason for:
vulkan_glfw_backend.cpp:113:51: error: invalid use of incomplete type ‘struct VulkanGlfwWindowContext’
113 | for (int i = 0; i < mpContext->numWindows; i++)
Since I declare it explicitly as a pointer inside a class, the storage requirements should be clear. Additionally, it's an implementation detail how the memory layout of the struct looks, which the compiler knows at compile-time, because it is defined at the top of the source file.
So, why is this going wrong, and can I fix it somehow? For several reasons I desire to have an opaque type. Thanks!
Answering my own question :)
So this is one of those days, where I'm not completely posting the entire code as is (my apologies!). So, the entire thing is inside a namespace:
// header
namespace mynamespace
{
// vulkan_glfw_backend.hpp
struct alignas(8) VulkanGlfwWindowContext;
}
//source
#include "vulkan_glfw_backend.hpp"
using namespace mynamespace;
struct VulkanGlfwWindowContext
{
int numWindows;
GLFWwindow* windows[MAX_WINDOWS];
};
The reason for this going wrong - again, I don't understand the logics behind it. Anyways, this is the fix:
//source
struct mynamespace::VulkanGlfwWindowContext
{
int numWindows;
GLFWwindow* windows;
};
Apparently, C++ doesn't understand the definition without fully-qualified namespace prepended to the name. I have seen the same issue with functions - it is not enough to open the namespace, it must be included in the definition.
This confuses me, as it is not required when implementing class member functions, where the name of the class is sufficient, e.g. someclass::somemember() {}.
But it works now!
Related
I have a class called scratch and have used scratch.h to declare it.
Now I have another class called scratch2 under scratch2.h and want to create an object of scratch as a shared pointer.
This is the syntax I used inside scratch2 class declartion:
std::shared_ptr<scratch> newObject(new scratch());
But I am getting this error: Error: Expected type specifier
So I tried this instead:
std::shared_ptr<scratch> newObject2 = std::make_shared<scratch>();
which works fine. Can anyone please tell me why the first one isn't working?
My scratch.h code:
#ifndef _SCRATCH_
#define _SCRATCH_
#include <iostream>
class scratch {
private:
int _a;
float _b;
std::string _s;
public:
scratch();
scratch(int a, float b, std::string n);
~scratch();
};
#endif
and my scratch2.h:
#ifndef _SCRATCH_2_
#define _SCRATCH_2_
#include "scratch.h"
#include <memory>
class scratch2 {
std::shared_ptr<scratch> newObject(new scratch()); // Expected a type specifier error occurs here
std::shared_ptr<scratch> newObject2 = std::make_shared<scratch>(); // works fine here
};
#endif
Because in the context of declaring class members:
std::shared_ptr<scratch> newObject(new scratch());
This initially looks to the compiler as a class method declaration. C++'s syntax is very complicated. You can look at the entire declaration and understand what it's trying to do, but the compiler is parsing keywords one keyword at a time, and sees this:
type name( ...
inside a class declaration, and this starts to look like a class method declaration, and that's what the compiler tried to parse, and failed.
The formal specification of the C++ language spills a lot of ink on the subject of how things should be declared, mindful of the current state of compiler technology.
You need to work with the compiler, and use an alternate syntax that's unambiguous:
std::shared_ptr<scratch> newObject = std::shared_ptr<scratch>(new scratch());
Verified with gcc 5.3
Inside of a class definition, there are only two ways you're allowed to initialize your members. You can use = and you can use {}. You are not allowed to use ():
struct foo {
int x = 4; // OK
int y{7}; // OK
int z(12); // error
};
Admittedly, the compiler error in this case is extremely unhelpful.
Please take a look at the following code. First, a namespace holding a class is created. Then, the namespace is extended and the class we defined before is used. Here is a live demo.
// File one
namespace system {
class module {
// ...
};
}
// File two including file one
namespace system {
struct entry {
entry(module *module);
module *module;
};
}
Compiling this under gcc 4.8.1 without the permissive flag produces two compiler errors. Explicitly mentioning the namespace like system::module solves the problem, even though it is the same namespace we're already in.
error: declaration of ‘system::module* system::entry::module’ [-fpermissive]
module *module;
^
error: changes meaning of ‘module’ from ‘class system::module’ [-fpermissive]
class module {
^
Why is this code not standard conform? What other options do I have except explicitly mentioning the namespace every time.
As already pointed out in another answer, the compiler is complaining that you have changed the semantics of an identifier from a type to a variable name (module).
You should give your types a name that will make them easily distinguishable from variables. You should give member variables names that can be easily distinguished from regular variables. This increases code readability and maintainability. And, it has the side-effect of making the problem you face a non-issue:
// File one
namespace system {
class module_type {
// ...
};
}
// File two including file one
namespace system {
struct entry {
entry(module_type *module);
module_type *module_;
};
}
In this statement
module *module;
you redefined name module.
So in the member function declaration you have to use elaborated type name
entry( class module *module);
In any case it is a bad idea to redefine names such a way.
// File one
namespace system {
class module {
// ...
};
}
// File two including file one
namespace system {
struct entry {
entry(module *module);
module *module;
};
}
where I cannot really come up with more descriptive names
Naming is one of the most difficult challenges programmer's face, and yet, when shown a way (or 2 or n), it will forever be easier.
I offer the following - from a Coding standard I have long used.
class and namespace names are Capitalized, at least the 1st char
variable names are lower case (or camel case with lower case 1st char)
// File one
namespace System {
class Module {
// ...
};
}
// File two including file one
namespace System {
struct Entry {
Entry(Module* module);
Module* module;
};
}
So?
Search for and read some of the coding standards. They exist for a reason. Once you learn one good idea, it will serve the rest of your career.
I am programming on linux using g++ and I often encounter the problem that I need to use a class or data type in a header file which I define later, either at a later point in the header or in another header file.
For instance look at this header file:
class example
{
mydatatype blabla;
};
struct mydatatype
{
int blablainteger;
char blablachar;
};
This will give error because mydatatype is used before its defined
so usually I change it like this:
struct mydatatype; // <-- class prototype
class example
{
mydatatype *blabla; // <-- now a pointer to the data type
// I will allocate the data during runtime with the new operator
};
struct mydatatype
{
int blablainteger;
char blablachar;
};
Now it works. I could often just put the definition above, or include the header which is needed, but I don't want to include headers in a header or juggle with the definition order, it always gets messy.
The solution I showed usually works, but now I have encountered a new phenomenon. This time the datatype is not a class but a typedef, I cant use prototypes for a typedef and I don't want to use the actual datatype which the typedef incorporates.. it's messy too.
Is there any solution to this?
Firstly, the solution you've thought of (prototype and pointer), is unneeded, and slower than just implementing it without the pointer.
The "proper" solution for this, would be creating seperate headers for each type, and then include them in your other header. That way it will always be defined! You can even make them so that they include eachother.
However, if you've ever opened a .h file provided by g++, you've most likely seen this at the start of the header:
#ifndef SOMETHING_H
#define SOMETHING_H
// Code
#endif /* SOMETHING_H */
This is to solve the issue of types redefining themselves.
If they weren't there, and you included the header file multiple times, the types would be redefined, and an error would be thrown. This makes it so that the types are always present, but never included twice.
I hope that helps!
Place each class/type in it's own header file, and then include the relevant header file in other headers where you need it. Use an inclusion guard in each header e.g.:
// SomeHeaderFile.h
#ifndef SOME_HEADER_FILE_H
#define SOME_HEADER_FILE_H
// code
#endif
I disagree that this is messy - it allows you have an organised structure to you project, it allows each class to operate independently of others and without worrying about order, and it's a good idea to place each class in it's own file anyway.
You could just define the class inside the other class like
template<class T>
class vertex {
private:
class edge {
public:
vertex<T> *to;
double weight;
edge() {
weight = INFINITY;
to = NULL;
};
} *paths;
T data;
unsigned nof_paths;
public:
vertex(T val) {
data = val;
paths = NULL;
nof_paths = 0;
}
void addPathTo(vertex<T>*&);
edge* getAllPaths() {
return paths;
};
};
Obviously this works for small classes... if your class is ENORMOUS you'll be better using separate header files like the other guys said.
Let's say the Acme company releases a useful library with an extremely ugly C API. I'd like to wrap the structs and related functions in C++ classes. It seems like I can't use the same names for the wrapper classes, because the original library is not inside a namespace.
Something like this is not possible, right?
namespace AcmesUglyStuff {
#include <acme_stuff.h> // declares a struct Thing
}
class Thing {
public:
...
private:
AcmesUglyStuff::Thing thing;
};
Linking will be a problem.
The only way I can think to wrap the library, and not pollute my namespace with the C library names, is a hack like this, reserving space in the class:
// In mything.h
namespace wrapper {
class Thing {
public:
...
private:
char impl[SIZE_OF_THING_IN_C_LIB];
};
}
// In thing.cc
#include <acme_stuff.h>
wrapper::Thing::Thing() {
c_lib_function((::Thing*)impl); // Thing here referring to the one in the C lib
}
Is that the only way? I'd like to avoid putting prefixes on all my class names, like XYThing, etc.
Seems like you're making this harder than it needs to be.
#include "acme_stuff.h" // puts all of its names in global namespace
namespace acme {
class Thing {
public:
// whatever
private:
::Thing thing;
};
}
Now just use acme::Thing rather than Thing.
If it's really important to you to not have the C names in the global namespace, then you need a level of indirection:
namespace acme {
class Thing {
public:
Thing();
~Thing();
// whatever
private:
void *acme_thing;
};
}
In your implementation file, #include "acme_stuff.h", in your constructor create a new ::Thing object and store its address in acme_thing, in your destructor delete it, and in your member functions cast acme_thing to type ::Thing*.
It's not a good idea to try to name something the exact same thing as something else. (I mean equal fully-qualified names, including all namespaces.) If some library has already grabbed the obvious best name in the global namespace, you'll need to pick a different name.
You could put your class Thing in a namespace as Pete Becker suggests, and then use ::Thing to access Acme's Thing. That would be fine if you're prepared to always access your class through it's fully namespace-qualified name (e.g. My::Thing). It's tempting to try using My::Thing; or using namespace My;, but that won't work, because any translation unit that includes the definition of your class (e.g. via a header file you create) must necessarily pull Acme's Thing into the global namespace first (otherwise an "Undefined symbol" compilation error would occur when parsing the definition of My::Thing).
Is it really a C API? Try to extern "C" {} to whole included header to solve the linking problem.
namespace AcmesUglyStuff {
extern "C" {
#include <acme_stuff.h>
}
}
I got a global object of type "unnamed-struct" and i'm trying to define it. I don't want to pollute my global namespace with such useless type (it will be used only once).
Global.h
extern struct {
int x;
} A;
Is there any correct way to define such object?
I was trying this:
Global.cpp
struct {
int x;
} A = { 0 };
But VS2012 throws "error C2371: 'A' : redefinition; different basic types". Thanks.
One possible solution: create another file Global_A.cpp that does not include Global.h, and define A there. By the equivalent-definition rule this will be valid, as long as the anonymous struct definitions are equivalent.
This is still a bad idea, and most compilers will warn about it e.g. (gcc): warning: non-local variable `<anonymous struct> A' uses anonymous type.
There is no way to do just this simply because it would be error prone: some time in the future someone (probably even you) may want to modify this structure's definition and he might forget to do this in the header and source files accordingly. So you will have to invent a name for this structure and use its name the source file and leave its definition up to the header.
I don't know if this helps you, and this my first post... So take it easy.
I just ran into this issue and what I did was I added a function that manipulates the struct to the file that contained the anonymous struct. That way I can call that function from any file in my project, and it manipulates the values in the struct for me.
Here is an example:
header.c has this anonymous struct:
struct
{
char line1[80];
char line2[80];
char line3[80];
} header;
I want to manipulate those values in "interface.c" because I am making a command line interface in another file. My first instinct was to use an extern, but it seems like adding the following function to header.c is just as good or better (Some people discourage the use of externs when avoidable).
void changeHeaders(char *one, char *two, char *three);
void changeHeaders(char *one, char *two, char *three)
{
strcpy(header.line1, one);
printf("\nHeader 1: %s", header.line1);
strcpy(header.line2, two);
printf("\nHeader 2: %s", header.line2);
strcpy(header.line3, three);
printf("\nHeader 3: %s", header.line3);
}
Now as long as I include the prototype for that function I can manipulate those struct variables from any file by using that function. Hope that helps someone.