The Exists method in the Directory class will return true only if a path represents a directory in the file system. The Exists method in the File class will return true only if the path represents a file in the file system. These behaviors can be combined to determine whether a path represents a file, represents a directory, or doesn’t exist in the file system. The following code fragment demonstrates this approach:

string x_path = "C:\mypath\myfile.txt";
if (Directory.Exists(x_path)) {
    Console.WriteLine("{0} is a directory", x_path);
} else if (File.Exists(x_path)) {
    Console.WriteLine("{0} is a file", x_path);
} else {
    Console.WriteLine("{0} does not exist", x_path);
}

Using bitwise operations is slightly more complicated than the foregoing approach. The FileInfo and DirectoryInfo classes provide an Attributes property, which is a bit flag enumeration containing file attributes, as listed in the System.IO.FileAttributes enumeration. This approach looks for the presence of the FileAttributes.Directory flag to differentiate between a file and a directory.

string x_path = "C:\mypath\myfile.txt";

FileInfo x_info = new FileInfo(x_path);

if ((ulong)x_info.Attributes == UInt64.MaxValue) {
    Console.WriteLine("{0} does not exist", x_path);
} else if ((x_info.Attributes & FileAttributes.Directory)
    == FileAttributes.Directory) {
        Console.WriteLine("{0} is a directory", x_path);
} else {
    Console.WriteLine("{0} is a file", x_path);
}

This fragment detects paths that relate to nonexistent files by checking to see whether the unsigned integer value of the attributes is the same as that of the UInt64.MaxValue field. The Microsoft documentation on the file attributes is a little hazy, and we’re not sure whether this is a bug or an expected behavior. For obvious reasons, we suggest using the Exists methods to tell files and directories apart.

The most direct way to create a stream is to create a new instance of FileStream, passing in a path string to the constructor. Because .NET streams can be used to read and write data, details must be supplied about how the file is opened. The basic FileStream constructor takes the following form:

FileStream(string path, FileMode mode, FileAccesss access);

The FileMode enumeration argument indicates how the file should be opened. Table 10-4 describes the different FileMode enumeration members.

The FileAccess argument specifies whether the file should be opened for reading, writing, or both, using values from the System.IO.FileAccess enumeration, which consists of Read, Write, and ReadWrite. The following statements demonstrate how these arguments can be combined:

// open an existing file for reading and writing
FileStream("myfile.txt", FileMode.Open, FileAccess.ReadWrite);

// truncate an existing file
FileStream("myfile.txt", FileMode.Truncate, FileAccess.Write);

// create a new file for reading
FileStream("mynewfile.txt", FileMode.CreateNew, FileAccess.Read);

The most comprehensive constructor takes the following form:

FileStream(string path, FileMode mode, FileAccesss access,
    FileShare share, int buffer, bool async);

The new elements are detailed next. In between the basic and comprehensive forms are several overloaded constructors that provide defaults for certain values. These are described in Table 10-5.

Instances of FileStream can also be created indirectly via the File and FileInfo classes. Both classes provide the same range of methods for this task, the difference being that the path name for the file must be passed into the static File methods. These classes can also be used to create instances of System.IO.StreamReader, which is covered in the Readers and Writers section later in this chapter. Table 10-6 lists the methods available in the File and FileInfo classes.