C Preprocessor - Cascaded Macros

[Michael Blackney]

I've been looking into cascaded macros and I can't seem to find if
what I'm doing is portable. I want to do something like:

#define THREE_ONES 1,1,1
#define ONE_TWO_THREE 1,2,3

#define JOIN_LISTS(a,b) JOIN_LISTS_PROCESSED(a,b)
#define JOIN_LISTS_PROCESSED(a1,b1,c1,a2,b2,c2) a1+a2,b1+b2,c1+c2

so that the statement:

{ JOIN_LISTS(THREE_ONES,ONE_TWO_THREE) },

will expand (one the first preprocessor pass) to:

{ JOIN_LISTS_PROCESSED(1,1,1,1,2,3) },

then expand on the second pass to:

{ 1+1, 1+2, 1+3 },

On my compiler (DJGPP) it works just fine - but is it portable? Do
all preprocessors use a similar replacement order?

--
michaelblackney at hotmail dot com
http://aburatan.sourceforge.net/
Latest version 0.95 2-5-4

[Chris Barts]

On Thu, 09 Sep 2004 16:08:15 -0700, Michael Blackney wrote:

I've been looking into cascaded macros and I can't seem to find if
what I'm doing is portable. I want to do something like:

#define THREE_ONES 1,1,1
#define ONE_TWO_THREE 1,2,3

#define JOIN_LISTS(a,b) JOIN_LISTS_PROCESSED(a,b)
#define JOIN_LISTS_PROCESSED(a1,b1,c1,a2,b2,c2) a1+a2,b1+b2,c1+c2
All of these definitions look portable. For maximum utility, you should
put in a lot more parentheses, like so:

(a1)+(a2),(b1)+(b2),(c1)+(c2)

Why? So you can have lists like this:
JOIN_LISTS_PROCESSED(x+5,2+3/2,5-6,4-23,5&6,7^2)

so that the statement:

{ JOIN_LISTS(THREE_ONES,ONE_TWO_THREE) },

will expand (one the first preprocessor pass) to:

{ JOIN_LISTS_PROCESSED(1,1,1,1,2,3) },

then expand on the second pass to:

{ 1+1, 1+2, 1+3 },

On my compiler (DJGPP) it works just fine - but is it portable? Do
all preprocessors use a similar replacement order?

I can't think of any reason a conforming compiler would misparse your
headers.