CrtInvertedIdx

// ./CrtInvertedIdx moon.fidx.sort > sun.iidx
#include <iostream>
#include <fstream>

using namespace std;

int main(int argc, char* argv[])
{
    ifstream ifsImgInfo(argv[1]);//打开输入流
    if (!ifsImgInfo) {
        cerr << "Cannot open " << argv[1] << " for input\n";
        return -1;
    }

    string strLine,strDocNum,tmp1="";
    int cnt = 0;
    while (getline(ifsImgInfo, strLine)) {
        string::size_type idx;
        string tmp;


        idx = strLine.find("\t");
        tmp = strLine.substr(0,idx);
        //得到索引词
        if (tmp.size()<2 || tmp.size() > 8) continue;//不可能构成中文或者大于最大匹配单元

        if (tmp1.empty()) tmp1=tmp;

        if (tmp == tmp1) {//如果是相同的索引词,则将文档编号追加到strDocNum中
            strDocNum = strDocNum + " " + strLine.substr(idx+1);//索引词相同,则加起来
        }
        else {//索引词不同
            if ( strDocNum.empty() )
                strDocNum = strDocNum + " " + strLine.substr(idx+1);

            cout << tmp1 << "\t" << strDocNum << endl;//输出
            tmp1 = tmp;
            strDocNum.clear();
            strDocNum = strDocNum + " " + strLine.substr(idx+1);
        }

        cnt++;
        //if (cnt==100) break;
    }
    cout << tmp1 << "\t" << strDocNum << endl;

    return 0;
}