so i use BOOST.EXTENTION to load modules. I have a special file that describes each module. I read variables from that file.
so such example:
shared_library m("my_module_name");
// Call a function that returns an int and takes a float parameter.
int result = m.get<int, float>("function_name")(5.0f);
m.close();
for me would turn into:
shared_library m("my_module_name");
// Call a function that returns an int and takes a float parameter.
int result = m.get<myMap["TYPE_1_IN_STRING_FORM"], myMap["TYPE_2_IN_STRING_FORM"]>("function_name")(5.0f);
m.close();
How to create such map that would map standard and costume types?
Update:
may be with variant:
shared_library m("my_module_name");
int result = m.get<boost::variant< int, float, ... other types we want to support >, boost::variant< int, float, ... other types we want to support > >("function_name")(5.0f);
m.close();
can halp? so we would not care as long as all types we want are declared in it?
For that, you would need a heterogeneous map - that is, its elements can be of different types. Furthermore you would need the ability to return types from functions, not just variables.
Now, a heterogeneous map would be possible with Boost.Variant or a simple union, but that binds it to compile time: we need to know every type that is possible to create that variant/union.
Of course a Boost.Any would be possible to store everything and its dog, but the problem strikes again: you need to extract the real type out of that Boost.Any again. The problem repeats itself. And if you know the real type, you can aswell just make a variant/union and save yourself the any_cast trouble.
Now, for another troublesome thing:
m.get<myMap["TYPE_1_IN_STRING_FORM"], myMap["TYPE_2_IN_STRING_FORM"]>
To make the above line work, you'd need two features that C++ doesn't have: the ability to return types and runtime templates. Lets ignore the first point for a moment.
Templates are compile-time, and the get function is such a template. Now, to use that template, your myMap would need to be able to return types at compile-time, while getting populated at runtime. See the contradiction? That's why runtime templates would be needed.
Sadly, exactly those three things are not possible (or extremely hard and very very limited) in C++ at runtime: heterogeneous data types (without constant size), returning types and templates.
Everything that involves types needs to be done at compile-time. This blogpost by #Gman somewhat correlates with that problem. It's definitly worth a read if you want to know what C++ just can't do.
So, to conclude: You'll need to rethink and refactor your problem and solution. :|
Related
As i started experimenting more in depth with C++1x features i ran into some thinking. For example when there is this construct
template<unsigned int N> unsigned int functionForTest(const char (&a)[N]);
and the usage of it like
functionForTest("Hello"); ---> const char [6]
functionForTest("Hello World") ---> const char [12];
then c++ ends up instantiating 2 functions with 2 different parameter types and that means increase in binary size if this function is used with different sizes. How efficient is that? Is it compiler specific? Isn't the traditional C-like array and size passing to function much more efficient here?
This is how i build g++ -std=c++17 -Xlinker -Map=output.map compilerDiffs.cpp -o test.exe and thats a sample of the map file inspected to come to this conclusion
samples of Map file
Generics ("templates" in C++) are a HUGE win for any type-safe language:
https://www.geeksforgeeks.org/generics-in-c/
The simple idea is to pass data type as a parameter so that we don’t
need to write the same code for different data types. For example, a
software company may need sort() for different data types. Rather than
writing and maintaining the multiple codes, we can write one sort()
and pass data type as a parameter.
The advantages of Generic Programming are
Code Reusability
Avoid Function Overloading
Once written it can be used for multiple times and cases.
And yes - that means if you instantiate a template for two different types ... then your compiler will generate two different functions. That's not an "inefficiency" - that's the whole POINT. You've written one, "generic" function; the compiler takes care of the rest.
You no longer have to "re-write" the same function over and over again for each specific type.
That's a "win".
The problem above is that "templates" are simply the wrong choice for your particular example. You'd probably want a "std::string" instead (in which case there's no need for "N". Alternatively, maybe you'd want to pass "N" as a function parameter.
Templates are Good. But you need to use the right tool for the right job :)
It's easy to do something like that in Python, but implementing it in C++ seems to be more challenging.
I actually have some solution to this, but I'd like to see if you can see any better solution.
Here's what I want to do.
I have a list of values of different types (string, integer, can be also instance of some class etc.). Now here's the first problem - in C++ (unlike in Python) all values in vector/array have to be of the same type.
The solution I can see is that I can use std::any like this: vector<std::any> list.
I also have an array/vector of functions (or pointers to functions) with different parameter types and returned values - one function can accept string and integer and return a char and other can accept a char and return an int. Here's another problem: in C++ you can have an array/vector of functions only if they have the same parameters and returned values (as far as I know) because in your declaration of the vector you need to define the parameter types and the returned value.
The other problem is that I need to retrieve the information about the parameters and the returned value for each function. In other words, having those functions, I need to know that this function accepts 2 strings and 1 integer and returns a char for example. In Python I can use inspect.signature function to retrieve information about type annotations of a function. In C++, I don't know if there is a way to do this.
The solution I can see here is to use std::any again (although I will use another solution, I will explain why later).
The solution I can see to this problem is that I won't retrieve that information but instead the user of the class which accepts this vector of functions will simply have to specify what are the parameter types and returned value for each function. In other words, the solution I can see is that I won't be retrieving the information about parameter types programmatically.
The other problem I have is that later I need to call one of those functions with some parameters. In Python I do this like this:
arguments = [1, 'str', some_object] // here I prepare a list of arguments (they are of different types)
func(**arguments)
In C++ I can do unpacking as well, but not if the parameters are of different types.
The solution I can see here is as follows. Those functions in the vector will all accepts only argument which is vector<std::any> args which will simply contain all of the arguments. Later when I want to call the function, I will simply construct a vector with std::any values and pass it as an argument. This would also solve the previous problem of not being able to store vector of functions with different parameters.
Can you see better solutions?
You might wonder what I need all of this is for. I do some program synthesis stuff and I need to programmatically construct programs from existing functions. I'm writing a library and I want the user of my library to be able to specify those base functions out of which I construct programs. In order to do what I want, I need to know what are the parameters and returned values of those functions and I need to call them later.
I believe what you are looking for is std::apply. You can use std::tuple instead of std::vector to store a list of values of different types -- as long as the types are known at compile-time. Then std::apply(f, t) in C++ is basically the same as f(*t) in Python.
I have a list of values of different types (string, integer, can be also instance of some class etc.).
A type which is a union of subtypes is called a sum type or tagged union. C++ has the template std::variant for that.
Now here's the first problem - in C++ (unlike in Python) all values in vector/array have to be of the same type.
Of course, so use cleverly C++ containers. You might want some std::map or std::vector of your particular instance of std::variant.
I also have an array/vector of functions
You probably want some std::vector of std::function-s and code with C++ lambda expressions
You should read a good C++ programming book
I'm writing a library and I want the user of my library to be able to specify those base functions out of which I construct programs.
You could get inspiration from SWIG and consider generating some C++ code in your library. So write (in Python or C++) your C++ metaprogram (generating some C++ code, like ANTLR does) which generates the user code, and your user would adapt his build automation tool for such a need (like users of GNU bison do).
You might also consider embedding Guile (or Lua) in your application.
PS. You might be interested by other programming languages like Ocaml, Go, Scheme (with Guile, and read SICP), Common Lisp (with SBCL), or Rust.
(i'm using Visual C+++ 2010)
suppose i have defined a tuple like this:
typedef std::tr1::tuple<
int //i want to set its default value to 9
, double //i want to set its default value to 3.3
, int //i want to set its default value to 2
, double //i want to set its default value to -7.2
> Mytuple;
i can do that in a struct. but i wonder if it is possible to do that in std::tr1::tuple.
Besides, i want to know when shoud i use std::tr1:tuple or struct?
anyone can help me?
A tuple is not a magical replacement for a struct. Their purposes are very different. A struct is, first and foremost, a language construct. A tuple is a library construct.
Structs get to have default values because the language says that you can write constructors to give them default values. The language then takes responsibility to call that constructor to initialize the struct. tuple, as a library construct, has no such capabilities, anymore than you can give std::vector<T> a default T that it will always automatically use. You can provide initial values for each member, but you can't give it defaults.
And if you think about it, you wouldn't want to. Imagine if someone could say that every tuple<int, float> was always created with 3 and 54.221. Even if some other code created it that knew nothing about this rule, it would have to be followed, (just as constructors for the type are used everywhere that type is used). Remember: every tuple<int, float> is the same type.
Really, a tuple is a substitute for the inability to perform reflection on a struct and do compile-time iteration over its members. The other main reason they exist is to be able to have compile-time dynamic structures (that is, the ability to create aggregates of types based on compile-time arguments, rather than a static list directly written into a file).
So unless you need to use std::tie (for effectively returning multiple values), iteration over members (ie: call some template function for each member of an object), or some similar specialized code, you should be using a struct.
I am creating an array template class that needs to be able to handle as broad an input as possible. It will work no problem with int, float, char but I would like to be able to test for strings and other types of arrays so I can sort the main array. I would like the functions to be as generic as possible so it deals with everything the same way.
Essentially, I want to know how to test an unknown variable for its data type in templates.
The technique you search for is called "Traits". With a trait you determine the type of a variable by using template specialization.
See http://accu.org/index.php/journals/442 for a good explanation with sample code.
You might be interested in using Boost.TypeTraits
If you want to sort your array, just use std::sort. Why do you think you need to treat std::string different from an int? They mostly work the same.
I need an array to hold member-function-pointers of different classes. How can I define the array?
The code should look like this:
arr[0] = &CMyClass::FuncX;
arr[1] = &CYourClass::FuncY;
arr[2] = &CHerClass::FuncZ;
I tried void*, but it doesn't work.
You can't; they are all different types and arrays are homogeneous.
Regardless what the arguments are or what the return value is, there is an implicit this which is unique to the class type. The type of a class member pointer is:
return_value (class_type::*)(parameters);
As you can see, because they belong to different classes they will always be a different type. Even if it were the same class, the return_value and parameters would have to be consistent to create an array, otherwise you'd still have different types.
What's the bigger picture? Boost.Bind with Boost.Function comes to mind. Also, virtual functions may solve your problem.
As others have pointed out, you can't store pointers to different kinds of functions directly. You might want to look at the Command template, e.g., from Modern C++ Design, which at least lets you put different invokable "things" (pointers or smart pointers to functions, functors, member functions) into a single thing.
On its own, that probably won't be sufficient -- you'll (apparently) end up with the template instantiated over different types, which produces different types. Those types will all use the same syntax, but won't all go into an array (which demands a single type).
Depending on your constraints, (compile-time vs. run-time indexing, in particular) you may be able to use a Boost::tuple to store a collection of command objects. You can treat that a bit like an array, using numeric indexing to get to an individual item. Unlike a normal array, however:
the syntax is a bit ugly, and
The indexing has to be done at compile-time (using compile-time constants).
Without know the parameters or return types of the function its hard to define them for you look at this page to get the gist of it or post the declaration of the functions.
Others have noted why you can't do this. But even if you could, what would you be able to do with it. In order to call a member function pointer, you need to an object of the appropriate type to call it on. So you would need to know the type of each of the member function pointers. You need to take a step back and figure out what it is that you are trying to accomplish.