Can someone help me understand this code?
long long n,u,m,b;main(e,r){for(;n++||(e=getchar()|32)>=0;b=
"ynwtsflrabg"[n%=11]-e?b:b*8+n)for(r=b%64-25;e<47&&b;b/=8)for
(n=19;n;"1+DIY/.K430x9G(kC["[n]-42&255^b||(m+=n>15?n:n>9?m%
u*~-u:~r?n+!r*16:n*16,b=0))u=1ll<<6177%n--*4;printf("%llx",m);}
I found some explanations, but they are very brief. What I know so far: This code outputs a decimal number for a given input of spelled numbers. It searches for some combinations of characters, that form the numbers and stores those characters in octal representation instead of hexadecimal. It assigns some values to the found combinations. Then it adds those values to the result (for words like "five", "thirteen" etc.), or somehow shifts the digits (for words like "thousand" etc.). That's all I understood from the explanations I found: https://www.ioccc.org/2012/kang/hint.html, http://j.mearie.org/post/7462182919/spelt-number-to-decimal.
I tried to understand this code expression by expression, but I only came to an endless loop, where I don't understand what one variable is for, since it is only modified later on, but then it depends on a next variable, that is also modified later in the loops. I just don't know where to start.
I tried to comment every single piece of the code + remove some obfuscations and ambiguities:
long long n, u, m, b; // = 0
int main(int e, int r) {
/* loop1:*/ for(;
/* --condition: */
/* if not:*/ n++ || // n++ == 0 <=> n == 0 // Set n += 1
/* then:*/
/* if not:*/ (e = getchar() | 32) >= 0; // Set e <- input char
// `| 32` transforms upper case to lower case
// getchar < 0 <=> EOF
/* then break*/
/* --increment: */
b =
/* if:*/ "ynwtsflrabg"[n %= 11] - e ? // Set n %= 11
/* then b=*/ b :
/* else b=*/ b * 8 + n)
/* loop2:*/ for(r = (b % 64) - 25; // Set r
/* --condition: */
/* if not:*/ (e < 47) && b;
/* break*/
/* --increment: */
b /= 8) // Set b
/* loop3:*/ for(n = 19; // Set n
/* --condition: */
/* if not:*/ n; // <=> n != 0
/* then break */
/* --increment: */
/* if not:*/ ("1+DIY/.K430x9G(kC["[n] - 42) & 255 ^ b ||
/* <=> {7, 1, 26, 31, 47, 5, 4, 33, 10, 9, 6, 78,
15, 29, 254, 65, 25, 49, 214}[n] == b */
/* then:*/ (m += // Set m
/* if:*/ n > 15 ?
/* then m+=*/ n :
/* else if:*/ n > 9 ?
/* then m+=*/ m % u * ~-u :
// <=> m % u * (u - 1)
/* else if:*/ ~(int)r ?
/* then m+=*/ n + !(int)r * 16 :
/* else m+=*/ n * 16,
b = 0)) // Set b
/* --body: */ u = 1ll << (6177 % n-- * 4); // Set u, n
// <=> u = pow(2, 6177 % n-- * 4)
printf("%llx\n", m);
}
Most of this rewritten code may look even scarier than the original, but I just needed some step by step walk-through, so maybe it will help someone else too.
It seems that the output is generated in the last loop, at m += .... But I can't understand the condition above it: "1+DIY/.K430x9G(kC["[n] - 42) & 255 ^ b. It translates to: {7, 1, 26, 31, 47, 5, 4, 33, 10, 9, 6, 78, 15, 29, 254, 65, 25, 49, 214}[n] == b, where my {...} notation behaves like a string literal containing characters corresponding to those numbers - which seem meaningless to me at this point.
Edit
Someone has voted to close this question for being unclear. To make clear what I need:
I would appreciate any hints or insights on any of the techniques used in this code. Like for example:
- Author of the code used deliberate obfuscation:
i["foo"]instead of"foo"[i]. - Author shortened the
(u - 1)expression as~-u
My main problems are with parts:
"1+DIY/.K430x9G(kC["[n]-42&255^b- Why those characters?m+=n>15?n:n>9?m%u*~-u:~r?n+!r*16:n*16,b=0)- Why those conditions onn?e<47- Why47- the/character?
