bearophile wrote:
This post is a partial copy of a post of mine from digitalmars.D.learn. Lot of
people seem to ignore that place.
This is the only way I have found to create a certain static array at compile
time:
int sumSqrt(int n) {
int result = 0;
while (n) {
int digit = n % 10;
n /= 10;
result += digit * digit;
}
return result;
}
template GenSquares(int n) {
static if (n < 0)
const int[] GenSquares = [];
else
const int[] GenSquares = GenSquares!(n-1) ~ [sumSqrt(n)];
}
void main() {
const int CHUNK = 1000;
static const auto squares = cast(int[CHUNK])GenSquares!(CHUNK-1);
}
That code works with D1, but gives a "recursive expansion" error with D2, this
is bad. Can D2 be fixed to have capabilities similar to D1?
It's a bug in D1, actually. The bug was fixed in D2 but not yet in D1.
As you increase the value, D1 will just silently segfault eventually. I
believe D1 will be fixed in the next release.
I have also tried to write a nicer version of the code that uses just one
compile-time function and no templates, and starts as:
struct S(int N) { int[N] a; }
S!(N) genSquares(int N)() { ...
But so far I haven't found a way. If such limit is real, then I think D2 has to
be improved to allow such things, because creating a fixed-size array at
compile time is one of the main purposes of compile-time functions.
It's bug 2569. Nothing fundamental.
Bye,
bearophile
