what in the heck randomaccess

[adam2016]

so I'm still messing around with random access of a text file anyway I am trying to figure out the math in able to store the data correctly in a text file so I can load my person objects dynamically with out it them being stored in ram,

anyway I got some crazy output when I added + 4 to the end but when i type + 5 or any number greater or less than 4 I don't get the cray results

here is the results of the text file when I have 4 at the end
㨲ㄲ〺㈺㨱㨴㨱㔲㔺:㄀㔺㈺ㄺ㨰

and here is the results when I have 5 at the end

2:22:0:22:4:1:26:5: 1:5:2:10:

why the heck is some letters of the chinesse alphabet getting printed out ? lol

  #include <iostream>
#include <fstream>
#include <vector>
#include <map>

using namespace std;


class person{

  public:
      int data1;
      int data2;
      person(int d1,int d2){
        data1 = d1;
        data2 = d2;
      }
};

class index{

public:
    int locationID;
    int startByte;
    int size;

    index(){


    }

    index(int sb,int s){

       startByte = sb;
       size = s;
    }

     index(const index& other){

       startByte = other.startByte;
       locationID = other.locationID;
       size = other.size;
    }

    index& operator=(const index& other){

       startByte = other.startByte;
       locationID = other.locationID;
       size = other.size;

       return *this;
    }
};

int main()
{

    ofstream out;
    ifstream in;
    vector<person> people;
    map<int,index> indexs;

    person one(1,5);
    person two(2,10);

    people.push_back(one);
    people.push_back(two);

    out.open("ra2.txt");

    int size = people.size();
    out << size;
    out << ":";
    int indexSize = (int)(1 * 3 * size);
    int personStart = (int)(out.tellp() + indexSize + 1 + 2 + 3 * size + 4);
    out << personStart;
    out << ":";
    int indexStart = (int)out.tellp();
    int startPointer = personStart;

    out.seekp(personStart);

    for(int i = 0; i < people.size(); i++){

        out << people[i].data1;
        out << ":";
        out << people[i].data2;
        out << ":";


        index ind(startPointer,(int)(out.tellp() - startPointer));
        ind.locationID = i;
        indexs[i] = ind;

        startPointer = (int) (out.tellp());
    }

    out.seekp(indexStart);

    for(map<int,index>::iterator it = indexs.begin();it != indexs.end();it++){

        out << it->second.locationID;
        out << ":";
        out << it->second.startByte;
        out << ":";
        out << it->second.size;
        out << ":";
    }

   out.close();


}

[jonnin]

is the data in the text files fixed width such that each record is N bytes and you can skip from record to record via a multiple of N and internally skip fields by using the widths of each field in order totaled up?

random access files need to be fixed width data fields.

[adam2016]

Ho Jonnin it seems to work I tested the outputs and everything seems correct but for some reason and I can't figure out for the life of me why,the size of the first index is 4 and size of the second index is 5?

yet it still works great when skipping to that particular byte

 ifstream in;
    stringstream ss;
    map<int,index> inds;

    in.open("ra2.txt");

    int size;
    string sizeString;
    getline(in,sizeString,':');
    ss << sizeString;
    ss >> size;
    ss.clear();
    ss.str(string());
    int personStart;
    string psString;
    getline(in,psString,':');
    ss << psString;
    ss >> personStart;
    ss.clear();
    ss.str(string());

    while(in.tellg() < personStart){

        int locationId;
        string idString;

        getline(in,idString,':');
        ss << idString;
        ss >> locationId;
        ss.clear();
        ss.str(string());

        int startByte;
        string sbString;
        getline(in,sbString,':');
        ss << sbString;
        ss >> startByte;
        ss.clear();
        ss.str(string());

        int indexSize;
        string sizeString;
        getline(in,sizeString,':');
        ss << sizeString;
        ss >> indexSize;
        ss.clear();
        ss.str(string());

        index temp(startByte,indexSize);
        temp.locationID = locationId;

        inds[locationId] = temp;
    }

    for(map<int,index>::iterator it = inds.begin(); it != inds.end(); it++){


        in.seekg(it->second.startByte);
        int data1,data2;
        string d1String,d2String;

        ss.clear();
        ss.str(string());

        getline(in,d1String,':');
        ss << d1String;
        ss >> data1;
        ss.clear();
        ss.str(string());

        getline(in,d2String,':');
        ss << d2String;
        ss >> data2;
        ss.clear();
        ss.str(string());

        cout << "data 1: " << data1 << " data 2: " << data2 << endl;

    }

    int choice;
    cout << "enter a number you want 1 or 2 " << endl;
    cin >> choice;
    if(choice < 1 || choice > inds.size()){
        return 1;
    }else{

       in.seekg(inds[choice-1].startByte); // -1 because indexs start at 0
       int d1,d2;
       string s1,s2;

       getline(in,s1,':');
       ss << s1;
       ss >> d1;
       ss.clear();
       ss.str(string());

       getline(in,s2,':');
       ss << s2;
       ss >> d2;
       ss.clear();
       ss.str(string());

       cout << "data 1 : " << d1 << "data 2 : " << d2 << endl;
    }

[Thomas1965]

I can't figure out for the life of me why,the size of the first index is 4 and size of the second index is 5?

As jonnin pointed out the records need to be the same size, but they are not.

  person one(1,5);
  person two(2,10);

In text mode 10 obviously need 2 chars and 5 only one.
Normally you calculate the offset of a record by (rec number * sizeof(rec)) and set the file pointer to this position and read the record in one go. For random access normally a binary mode is used.