Creating dictionary from text file and then accessing it

Question

I'm building a program. It uses data stored in text files in following format {Index:Value}. Example:

1:10,86;
2:11,65;
3:13,32;
4:13,53;
5:13,93;
...
1500:1565,99;

There is rougly 1490-1500 lines in each file.

I need those files read in a dictionary and access its Index as integer and Value as double both by different functions. Something like:

Data.ByIndex((integer) 3) - returns (double) 13,32

Now getting Index by Value would be little more tricky:

Data.ByClosestValue((double) 12,0) - returns (integer) 2
Data.ByClosestValue((double) 13) - returns (integer) 3
Data.ByClosestLower((double) 13,5) - returns (integer) 3
Data.ByClosestHigher((double) 13,5) - returns (integer) 5

There are few key moments:

1. It has to be really fast. There is usually 10-15 data files read in the same time and the dictionary accessed several times for each file.
2. No LINQ if possible.

For now I went with following:

1. File.ReadAllText() method seems the fastest to me.
2. Getting entries by using Split(';') and the Split(':').
3. Dictionary <string, string>. The reason I went with string types is that if will be faster to read from text data by using Split() function.

What will be the optimal solution?

Update: most people would suggest storing data in a database and I agree that would be the best solution, but unfortunately I have no control over those data files. These data files are formed in another program and there is a strict requirement it must be editable by human via notepad or whatever.

Answer 1

I have a strong believe that reading files to memory, parsing values to int and double (so you can store them to IDictionary<int, double> before usage will be the most effective solution in this case. You can use SortedDictionary to have it more effective. Your ByIndex() function will be trivial:

double ByIndex(int index)
{
  double value = 0.0;
  Data.TryGet(index, out value);
  return value;
}

Other functions:

int ByClosestValue(double val)
{
  int closest = -1;
  foreach(var v in Data)
  { 
    if (Math.Round(v.Value, 0) == Math.Round(val, 0) 
    {
      closest =v.Key; 
      break;
    }
  }
  return closest;
}

ByClosestLower and ByClosesHigher are almost the same - you only need to call Floor() and Ceiling() instead of Round(). You can use also this: Efficiently find nearest dictionary key to your advantage