How to read an unknown contents of a text file into a 2d array

Question

I am trying to read the unknown contents of a text file into a 2D array and have it come out looking like:

M [0] [0]=2 M [0] [1]=1 M [0] [2]=0
M [1] [0]=0 M [1] [1]=1 M [1] [2]=3
M [2] [0]=8 M [2] [1]=9 M [2] [2]=1
M [3] [0]=3 M [3] [1]=5 M [3] [2]=2

when the text file looks like this:

2
1 0
0 1
3
8 9 1
3 5 2
-2 3 -1
0

The zero at the end shows the end of the file.

My problem is the array can be a max size of 10X10 so there is no way of knowing what the size of the 2D array is and how to get it to look like i have shown above.

Any ideas?

Answer 1

For some dimensions N x M

char c;
int x;
for(int i = 0; i < N; i++)
{
for(int j = 0; j < M; j++)
{
  c = fgetc(file);
  if(c == ' ') continue;
  if(c == EOF) return;
  if(c == '-')
  {
      c = fgetc(file);
      x = -1 * ((int)c);
  } else {
      x = c;
  }
  if(x == 0)
  {
      array[i][j] = x;
  } else {
      return;
  }
}
}

But if you're talking about "what's the size of the matrix required to store these" then you're going to need a way to figure out what dimensions you want.

Answer 2

Just use 'fstream'. It ignores new lines and works just like 'iostream'. You just need to keep track of your matrix row and column.

//open "myFileName.txt" with an input file stream
std::ifstream inputFile("myFileName.txt");

while(!inputFile.eof()) //Check for end-of-file character
{
    //do this while keeping track of your matrix position
    inputFile >> M [curRow] [curColumn]
}

inputFile.close();

And don't forget to include the library:

#include <fstream>

Edit: The >> operator will also attempt to auto-cast the input as whatever type you are using:

double dTemp;
int iTemp;
std::string sTemp;

std::ifstream in("myFile.txt");

in >> dTemp; //input cast as double
in >> iTemp; //input cast as integer
in >> sTemp; //input cast as string

in.close();

Edit: Get the number of elements of the file

int temp, numberOfInputs;
while(!inputFile.eof())
{
    inputFile >> temp;
    ++numberOfInputs;
}

inputFile.seekg(0, std::ios::beg); //Go to beginning of file

Once you have the number of inputs you can use that to figure out the number of rows and colums.

Answer 3

Try:

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


int main()
{
    std::ifstream   file("plop.dat");
    if(!file)
    {
        std::cerr << "Failed to open File\n";
        return 1;
    }

    // Keep looping until we are done.
    while(true)
    {
        int size;
        file >> size;

        // You said a size zero indicates termination.
        if (size == 0)
        {    break;
        }

        // Create a square 2D vector (Vector inside a Vector)
        std::vector<std::vector<int> >      matrix(size, std::vector<int>(size, 0));

        // Loop over each axis
        for(int x = 0;x < size; ++x)
        {
            for(int y = 0;y < size; ++y)
            {
                // Read one number for each location.
                // The operator >> automatically skips white space
                // White Space includes newline and tab. So it should work
                // perfectly if the input is OK. But it is hard to detect
                // an error in the format of the file.
                file >> matrix[x][y];
            }
        }
    }
}

~

Answer 4

So all row of the array have 3 values?
Simply read the values into the array keeping a count of which column you are in and ignore the newlines?

Look at getline

Answer 5

getline and stringstream to read, vector< vector<int> > to store.

Edit: Oh, so the size is always N*N? Then just read using while(cin>>x) { if (cin.good()) ...; } into a vector<int>, check the total size, and split into vector< vector<int> >

Answer 6

Single and multi-dimensional C arrays are just contiguous pieces of memory. The difference is more syntactical in terms of what indexing into the array does: for a 2-dimensional array it just multiplies one dimension by the size of the other dimension before adding the second dimension to find the offset.

So to solve:

1. read the values into a std::vector (other answers have sugegsted ways to do this)
2. work out the size (integer square root of myArray.size())

3. calculate the index as, e.g.:

   int idx(int size, int d1, int d2)
   {
       return (d1*size)+d2;
   }
   

4. Return the vector element at the index e.g.:

   for (int d1 = 0; d1 < size; d1++)
   {
       for (int d2 = 0; d2 < size; d2++)
           std::cout << "M [" << d1 << "] [" << d2 << "]=" << myArray[idx(size, d1, d2)] << " ";
       std::cout << std::endl;
   }
   

gives me:

$ g++ arr.cpp && ./a.out
M[0][0]=2 M[0][1]=1 M[0][2]=0
M[1][0]=0 M[1][1]=1 M[1][2]=3
M[2][0]=8 M[2][1]=9 M[2][2]=1