I have two macros FOO2 and FOO3:
#define FOO2(x,y) ...
#define FOO3(x,y,z) ...
I want to define a new macro FOO as follows:
#define FOO(x,y) FOO2(x,y)
#define FOO(x,y,z) FOO3(x,y,z)
But this doesn't work because macros do not overload on number of arguments.
Without modifying FOO2 and FOO3, is there some way to define a macro FOO (using __VA_ARGS__ or otherwise) to get the same effect of dispatching FOO(x,y) to FOO2, and FOO(x,y,z) to FOO3?
Simple as:
#define GET_MACRO(_1,_2,_3,NAME,...) NAME
#define FOO(...) GET_MACRO(__VA_ARGS__, FOO3, FOO2)(__VA_ARGS__)
So if you have these macros:
FOO(World, !) # expands to FOO2(World, !)
FOO(foo,bar,baz) # expands to FOO3(foo,bar,baz)
If you want a fourth one:
#define GET_MACRO(_1,_2,_3,_4,NAME,...) NAME
#define FOO(...) GET_MACRO(__VA_ARGS__, FOO4, FOO3, FOO2)(__VA_ARGS__)
FOO(a,b,c,d) # expeands to FOO4(a,b,c,d)
Naturally, if you define FOO2, FOO3 and FOO4, the output will be replaced by those of the defined macros.
To add on to netcoder's answer, you CAN in fact do this with a 0-argument macro, with the help of the GCC ##__VA_ARGS__ extension:
#define GET_MACRO(_0, _1, _2, NAME, ...) NAME
#define FOO(...) GET_MACRO(_0, ##__VA_ARGS__, FOO2, FOO1, FOO0)(__VA_ARGS__)
Here is a more general solution:
// get number of arguments with __NARG__
#define __NARG__(...) __NARG_I_(__VA_ARGS__,__RSEQ_N())
#define __NARG_I_(...) __ARG_N(__VA_ARGS__)
#define __ARG_N( \
_1, _2, _3, _4, _5, _6, _7, _8, _9,_10, \
_11,_12,_13,_14,_15,_16,_17,_18,_19,_20, \
_21,_22,_23,_24,_25,_26,_27,_28,_29,_30, \
_31,_32,_33,_34,_35,_36,_37,_38,_39,_40, \
_41,_42,_43,_44,_45,_46,_47,_48,_49,_50, \
_51,_52,_53,_54,_55,_56,_57,_58,_59,_60, \
_61,_62,_63,N,...) N
#define __RSEQ_N() \
63,62,61,60, \
59,58,57,56,55,54,53,52,51,50, \
49,48,47,46,45,44,43,42,41,40, \
39,38,37,36,35,34,33,32,31,30, \
29,28,27,26,25,24,23,22,21,20, \
19,18,17,16,15,14,13,12,11,10, \
9,8,7,6,5,4,3,2,1,0
// general definition for any function name
#define _VFUNC_(name, n) name##n
#define _VFUNC(name, n) _VFUNC_(name, n)
#define VFUNC(func, ...) _VFUNC(func, __NARG__(__VA_ARGS__)) (__VA_ARGS__)
// definition for FOO
#define FOO(...) VFUNC(FOO, __VA_ARGS__)
Define your functions:
#define FOO2(x, y) ((x) + (y))
#define FOO3(x, y, z) ((x) + (y) + (z))
// it also works with C functions:
int FOO4(int a, int b, int c, int d) { return a + b + c + d; }
Now you can use FOO with 2, 3 and 4 arguments:
FOO(42, 42) // will use makro function FOO2
FOO(42, 42, 42) // will use makro function FOO3
FOO(42, 42, 42, 42) // will call FOO4 function
Use it for default arguments:
#define func(...) VFUNC(func, __VA_ARGS__)
#define func2(a, b) func4(a, b, NULL, NULL)
#define func3(a, b, c) func4(a, b, c, NULL)
// real function:
int func4(int a, int b, void* c, void* d) { /* ... */ }
Use it for functions with possible infinite number of arguments:
#define SUM(...) VFUNC(SUM, __VA_ARGS__)
#define SUM2(a, b) ((a) + (b))
#define SUM3(a, b, c) ((a) + (b) + (c))
#define SUM4(a, b, c) ((a) + (b) + (c) + (d))
// ...
PS: __NARG__ is copied from Laurent Deniau & Roland Illig here: https://groups.google.com/group/comp.std.c/browse_thread/thread/77ee8c8f92e4a3fb/346fc464319b1ee5?pli=1
I was just researching this myself, and I came across this here. The author added default argument support for C functions via macros.
I'll try to briefly summarize the article. Basically, you need to define a macro that can count arguments. This macro will return 2, 1, 0, or whatever range of arguments it can support. Eg:
#define _ARG2(_0, _1, _2, ...) _2
#define NARG2(...) _ARG2(__VA_ARGS__, 2, 1, 0)
With this, you need to create another macro that takes a variable number of arguments, counts the arguments, and calls the appropriate macro. I've taken your example macro and combined it with the article's example. I have FOO1 call function a() and FOO2 call function a with argument b (obviously, I'm assuming C++ here, but you can change the macro to whatever).
#define FOO1(a) a();
#define FOO2(a,b) a(b);
#define _ARG2(_0, _1, _2, ...) _2
#define NARG2(...) _ARG2(__VA_ARGS__, 2, 1, 0)
#define _ONE_OR_TWO_ARGS_1(a) FOO1(a)
#define _ONE_OR_TWO_ARGS_2(a, b) FOO2(a,b)
#define __ONE_OR_TWO_ARGS(N, ...) _ONE_OR_TWO_ARGS_ ## N (__VA_ARGS__)
#define _ONE_OR_TWO_ARGS(N, ...) __ONE_OR_TWO_ARGS(N, __VA_ARGS__)
#define FOO(...) _ONE_OR_TWO_ARGS(NARG2(__VA_ARGS__), __VA_ARGS__)
So if you have
FOO(a)
FOO(a,b)
The preprocessor expands that to
a();
a(b);
I would definitely read the article that I linked. It's very informative and he mentions that NARG2 won't work on empty arguments. He follows this up here.
Here is a more compact version of the answer above. With example.
#include <iostream>
using namespace std;
#define OVERLOADED_MACRO(M, ...) _OVR(M, _COUNT_ARGS(__VA_ARGS__)) (__VA_ARGS__)
#define _OVR(macroName, number_of_args) _OVR_EXPAND(macroName, number_of_args)
#define _OVR_EXPAND(macroName, number_of_args) macroName##number_of_args
#define _COUNT_ARGS(...) _ARG_PATTERN_MATCH(__VA_ARGS__, 9,8,7,6,5,4,3,2,1)
#define _ARG_PATTERN_MATCH(_1,_2,_3,_4,_5,_6,_7,_8,_9, N, ...) N
//Example:
#define ff(...) OVERLOADED_MACRO(ff, __VA_ARGS__)
#define ii(...) OVERLOADED_MACRO(ii, __VA_ARGS__)
#define ff3(c, a, b) for (int c = int(a); c < int(b); ++c)
#define ff2(c, b) ff3(c, 0, b)
#define ii2(a, b) ff3(i, a, b)
#define ii1(n) ii2(0, n)
int main() {
ff (counter, 3, 5)
cout << "counter = " << counter << endl;
ff (abc, 4)
cout << "abc = " << abc << endl;
ii (3)
cout << "i = " << i << endl;
ii (100, 103)
cout << "i = " << i << endl;
return 0;
}
Run:
User@Table 13:06:16 /c/T
$ g++ test_overloaded_macros.cpp
User@Table 13:16:26 /c/T
$ ./a.exe
counter = 3
counter = 4
abc = 0
abc = 1
abc = 2
abc = 3
i = 0
i = 1
i = 2
i = 100
i = 101
i = 102
Note that having both _OVR and _OVR_EXPAND may look redundant, but it's necessary for the preprocessor to expand the _COUNT_ARGS(__VA_ARGS__) part, which otherwise is treated as a string.
Here's a spin off from Evgeni Sergeev's answer. This one supports zero argument overloads as well!
I tested this with GCC and MinGW. It ought to work with old and new versions of C++. Note that I wouldn't guarantee it for MSVC... But with some tweaking, I'm confident it could be made to work with that too.
I also formatted this to be pasted into a header file (which I called macroutil.h). If you do that, you can just include this header whatever you need the feature, and not look at the nastiness involved in the implementation.
#ifndef MACROUTIL_H
#define MACROUTIL_H
//-----------------------------------------------------------------------------
// OVERLOADED_MACRO
//
// used to create other macros with overloaded argument lists
//
// Example Use:
// #define myMacro(...) OVERLOADED_MACRO( myMacro, __VA_ARGS__ )
// #define myMacro0() someFunc()
// #define myMacro1( arg1 ) someFunc( arg1 )
// #define myMacro2( arg1, arg2 ) someFunc( arg1, arg2 )
//
// myMacro();
// myMacro(1);
// myMacro(1,2);
//
// Note the numerical suffix on the macro names,
// which indicates the number of arguments.
// That is the REQUIRED naming convention for your macros.
//
//-----------------------------------------------------------------------------
// OVERLOADED_MACRO
// derived from: https://stackguides.com/questions/11761703/overloading-macro-on-number-of-arguments
// replaced use of _COUNT_ARGS macro with VA_NUM_ARGS defined below
// to support of zero argument overloads
#define OVERLOADED_MACRO(M, ...) _OVR(M, VA_NUM_ARGS(__VA_ARGS__)) (__VA_ARGS__)
#define _OVR(macroName, number_of_args) _OVR_EXPAND(macroName, number_of_args)
#define _OVR_EXPAND(macroName, number_of_args) macroName##number_of_args
//#define _COUNT_ARGS(...) _ARG_PATTERN_MATCH(__VA_ARGS__, 15,14,13,12,11,10,9,8,7,6,5,4,3,2,1)
#define _ARG_PATTERN_MATCH(_1,_2,_3,_4,_5,_6,_7,_8,_9,_10,_11,_12,_13,_14,_15, N, ...) N
// VA_NUM_ARGS
// copied from comments section of:
// http://efesx.com/2010/07/17/variadic-macro-to-count-number-of-arguments/
// which itself was derived from:
// https://gustedt.wordpress.com/2010/06/08/detect-empty-macro-arguments/
#define _ARG16(_0, _1, _2, _3, _4, _5, _6, _7, _8, _9, _10, _11, _12, _13, _14, _15, ...) _15
#define HAS_COMMA(...) _ARG16(__VA_ARGS__, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0)
#define HAS_NO_COMMA(...) _ARG16(__VA_ARGS__, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1)
#define _TRIGGER_PARENTHESIS_(...) ,
#define HAS_ZERO_OR_ONE_ARGS(...) \
_HAS_ZERO_OR_ONE_ARGS( \
/* test if there is just one argument, eventually an empty one */ \
HAS_COMMA(__VA_ARGS__), \
/* test if _TRIGGER_PARENTHESIS_ together with the argument adds a comma */ \
HAS_COMMA(_TRIGGER_PARENTHESIS_ __VA_ARGS__), \
/* test if the argument together with a parenthesis adds a comma */ \
HAS_COMMA(__VA_ARGS__ (~)), \
/* test if placing it between _TRIGGER_PARENTHESIS_ and the parenthesis adds a comma */ \
HAS_COMMA(_TRIGGER_PARENTHESIS_ __VA_ARGS__ (~)) \
)
#define PASTE5(_0, _1, _2, _3, _4) _0 ## _1 ## _2 ## _3 ## _4
#define _HAS_ZERO_OR_ONE_ARGS(_0, _1, _2, _3) HAS_NO_COMMA(PASTE5(_IS_EMPTY_CASE_, _0, _1, _2, _3))
#define _IS_EMPTY_CASE_0001 ,
#define _VA0(...) HAS_ZERO_OR_ONE_ARGS(__VA_ARGS__)
#define _VA1(...) HAS_ZERO_OR_ONE_ARGS(__VA_ARGS__)
#define _VA2(...) 2
#define _VA3(...) 3
#define _VA4(...) 4
#define _VA5(...) 5
#define _VA6(...) 6
#define _VA7(...) 7
#define _VA8(...) 8
#define _VA9(...) 9
#define _VA10(...) 10
#define _VA11(...) 11
#define _VA12(...) 12
#define _VA13(...) 13
#define _VA14(...) 14
#define _VA15(...) 15
#define _VA16(...) 16
#define VA_NUM_ARGS(...) VA_NUM_ARGS_IMPL(__VA_ARGS__, PP_RSEQ_N(__VA_ARGS__) )
#define VA_NUM_ARGS_IMPL(...) VA_NUM_ARGS_N(__VA_ARGS__)
#define VA_NUM_ARGS_N( \
_1, _2, _3, _4, _5, _6, _7, _8, _9,_10, \
_11,_12,_13,_14,_15,_16,N,...) N
#define PP_RSEQ_N(...) \
_VA16(__VA_ARGS__),_VA15(__VA_ARGS__),_VA14(__VA_ARGS__),_VA13(__VA_ARGS__), \
_VA12(__VA_ARGS__),_VA11(__VA_ARGS__),_VA10(__VA_ARGS__), _VA9(__VA_ARGS__), \
_VA8(__VA_ARGS__),_VA7(__VA_ARGS__),_VA6(__VA_ARGS__),_VA5(__VA_ARGS__), \
_VA4(__VA_ARGS__),_VA3(__VA_ARGS__),_VA2(__VA_ARGS__),_VA1(__VA_ARGS__), \
_VA0(__VA_ARGS__)
//-----------------------------------------------------------------------------
#endif // MACROUTIL_H
Maybe you can use this macro to count the number of arguments.
#define VA_NUM_ARGS(...) VA_NUM_ARGS_IMPL(__VA_ARGS__, 5,4,3,2,1)
#define VA_NUM_ARGS_IMPL(_1,_2,_3,_4,_5,N,...) N
This seems to work fine on GCC, Clang and MSVC. It's a cleaned up version of some of the answers here
#define _my_BUGFX(x) x
#define _my_NARG2(...) _my_BUGFX(_my_NARG1(__VA_ARGS__,_my_RSEQN()))
#define _my_NARG1(...) _my_BUGFX(_my_ARGSN(__VA_ARGS__))
#define _my_ARGSN(_1,_2,_3,_4,_5,_6,_7,_8,_9,_10,N,...) N
#define _my_RSEQN() 10,9,8,7,6,5,4,3,2,1,0
#define _my_FUNC2(name,n) name ## n
#define _my_FUNC1(name,n) _my_FUNC2(name,n)
#define GET_MACRO(func,...) _my_FUNC1(func,_my_BUGFX(_my_NARG2(__VA_ARGS__))) (__VA_ARGS__)
#define FOO(...) GET_MACRO(FOO,__VA_ARGS__)
