In C,C++ if you write an integer in binary way(e.g. fwrite(&someInt,sizeof(int),1,fp)) say to file, you have tons of issues to consider if you want to read that integer from different machine - e.g., endianness issues, integer size may vary on other computer, float representation may vary on different computer, negative integers have different representation, etc. So directly reading and writing in binary mode may not work across different computers.
In this example below using Binary reader (from MSDN), do above issues also apply?
using System;
using System.IO;
class ConsoleApplication
{
const string fileName = "AppSettings.dat";
static void Main()
{
WriteDefaultValues();
DisplayValues();
}
public static void WriteDefaultValues()
{
using (BinaryWriter writer = new BinaryWriter(File.Open(fileName, FileMode.Create)))
{
writer.Write(1.250F);
writer.Write(@"c:\Temp");
writer.Write(10);
writer.Write(true);
}
}
public static void DisplayValues()
{
float aspectRatio;
string tempDirectory;
int autoSaveTime;
bool showStatusBar;
if (File.Exists(fileName))
{
using (BinaryReader reader = new BinaryReader(File.Open(fileName, FileMode.Open)))
{
aspectRatio = reader.ReadSingle();
tempDirectory = reader.ReadString();
autoSaveTime = reader.ReadInt32();
showStatusBar = reader.ReadBoolean();
}
Console.WriteLine("Aspect ratio set to: " + aspectRatio);
Console.WriteLine("Temp directory is: " + tempDirectory);
Console.WriteLine("Auto save time set to: " + autoSaveTime);
Console.WriteLine("Show status bar: " + showStatusBar);
}
}
}
What about BinaryFormatter class?
