Romulan Spelling

The Romulans use a language that can be approximated with the English alphabet and normal punctuation marks. Their spelling rules are even stranger than English, however. In particular, they have a rule that states:

G before P except after E or when pronounced X as in npgukbor or wpguk.

Operationally, you can detect the X pronounciation by the string PGUK appearing in the word. Also, the Romulan rules for capitalization are different from ours, so capital letters can appear anywhere in a word.

Input and Output

Given a file containing lines of Romulan text, you are to output the text with spelling corrected according to the G before P except after E rule given above. No input line will contain more than 70 characters including spaces.

For example, the input file corresponding to the romulan translation of the quote:

"I received the wierd piece of pie from my neighbor sam who in turn recieved the weird peice of pei from his nieghbor harry," might well be ...

Sample Input

I rpEpgvpd tKp wgprd tgpEp of tgp from my npguKbor sam wKo gn turn
rpEgpvpd tKp wpgrd tpgEp of tpg from Kgs ngpuKbor Karry,

Sample Output

I rpEpgvpd tKp wgprd tgpEp of tgp from my npguKbor sam wKo gn turn
rpEpgvpd tKp wgprd tgpEp of tgp from Kgs npguKbor Karry,

---

題目描述：

「G before P except after E or when pronounced X as in npgukbor or wpguk. 」
現在我們要從一般的英文單字中，轉換成 羅幕倫人 用的發音語法。
這是個艱難的任務，正如你所看到的那令人敬畏的題目說明。
出題者大概是程式寫多了，說話就跟 code 一樣。
具體翻譯如下：
「
'G' 必須在 'P' 前面，因此先把所有可能轉換對於相鄰的 pg -> gp。
除非 'E' 在前面，則保留原本的順序，即 epg -X-> egp。
除非字根 X 出現，也保留原本順序，其中 X = pguk。
」
照個順序做才會正確，否則都是徒然。
有很多模稜兩可的情況，例如 eggpuk 應該被念成 epgguk。

#include <stdio.h>
#include <string.h>
#include <algorithm>
using namespace std;
char tolower(char c) {
    if(c >= 'A' && c <= 'Z')
        c |= 32;
    return c;
}
void romulan(char s[]) {
    int update = 0, i;
    static char ts[1024];
    strcpy(ts, s);
    do {
        do {
            update = 0;
            for(i = 0; s[i]; i++) {
                if(tolower(s[i]) == 'p' && tolower(s[i+1]) == 'g') {
                    swap(s[i], s[i+1]);
                    update = 1;
                }
            }
        } while(update);
        do {
            update = 0;
            for(i = 0; s[i]; i++) {
                if(tolower(s[i]) == 'g' && tolower(s[i+1]) == 'p') {
                    if(i-1 >= 0 && tolower(s[i-1]) == 'e')
                        swap(s[i], s[i+1]), update = 1;
                    else if(tolower(s[i+2]) == 'u' && tolower(s[i+3]) == 'k')
                        swap(s[i], s[i+1]), update = 1;
                }
            }
        } while(update);
        update = strcmp(ts, s);
        strcpy(ts, s);
    } while(update);
}
int main() {
    char s[1024], word[1024];
    while(gets(s)) {
        int i, j;
        for(i = 0; s[i]; i++) {
            if((s[i]|32) < 'a' || (s[i]|32) > 'z') {
                putchar(s[i]);
            } else {
                for(j = 0; (s[i]|32) >= 'a' && (s[i]|32) <= 'z'; i++, j++)
                    word[j] = s[i];
                word[j] = '\0';
                romulan(word);
                printf("%s", word);
                i--;
            }
        }
        puts("");
    }
    return 0;
}