An addition to the Microsoft .NET framework 2.0 to 1.x is the support for FTP. All these days we need to rely on third-party libraries that pretty well satisfies most of our needs, but for sure, there is an extra pleasure with the use of the .Net framework library classes. The code included is not designed to be a full-fledged reusable library, but rather an easy to use and reusable piece of code that is easily comprehensible and can be reused and tweaked to fit your specific needs. Therefore the code for each functionality (upload, download, delete and so on) can be easy picked up separately and reused. The main motive behind this article was the unavailability of .Net 2.0 FTP sample codes and their usage in C#, maybe because it's a new entrant to the .Net scenario, or the third-party implementations available were working pretty well, that this area of the .Net 2.0 library hasn't gotten enough focus.
Background
I started working on this FTP module as part of my official work, but the requirement soon changed and I need to do it for .Net 1.1. So, I haven't travelled deeper into the Rabbit hole. But I believe this gives a good, instant start for using the FTP support in .Net 2.0.
Using the code
Don't forget to add the following directive:
using System.Net;
using System.IO;
The following steps can be considered as a generic procedure of getting an FTP request executed using a "FtpWebRequest" object:
-
Create an "FtpWebRequest" object over an FTP server URI
-
Set the FTP method to execute (upload, download, and so on)
-
Set options (SSL support, transfer as binary or not and so on) for the FTP webrequest
-
Set the login credentials (username and password)
-
Execute the request
-
Recieve the response stream (if required)
-
Close the FTP Request, in addition to any open streams
One point to watch out while coding for any FTP application is to have the settings for the FTP request proper to suit the FTP server and its specific configurations. A "FtpWebRequest" object exposes many poperties to have these settings in place.
The sample for the upload functionality is as follows.
First a URI is created that represents the FTP address along with the filename (directory structure included). This URI is used to create the "FtpWebRequest" instance.
Then the properties of the "FtpWebRequest" object are set, that determines the settings for the FTP request.
Some of its important properties are:
-
Credentials: specifies the username and password to login to the FTP server.
-
KeepAlive: specifies if the control connection should be closed or not after the request is completed. By default it is set to true.
-
UseBinary: denotes the datatype for file transfers. The 2 modes of file transfer in this case are Binary and ASCII. At the bit level both vary in the 8th bit of a byte. ASCII uses the 8th bit as an insignificant bit for error control, whereas for binary all the 8 bits are significant. So take care when you go for the ASCII transmission. To be simple, all those files that open and read well in Notepad are safe as ASCII. Executables, formatted documents and so on should be sent using binary mode. BTW sending ASCII files as binary works fine most of the time.
-
UsePassive: specifies whether to use active or passive mode. Earlier an active FTP worked fine with all clients, but nowm since most of the random ports will blocked by firewalls, the active mode may fail. The passive FTP is helpful in this case. But still it causes issues at the server. The higher ports requested by the client on the server may also be blocked by a firewall. But since FTP servers will need to make their servers accessible to the greatest number of clients, they will almost certainly need to support passive FTP. The reason why passive mode is considered safe is that it ensures all data flow initiation comes from inside (the client) the network rather than from the outside (the server).
-
Contentlength: setting this property is useful for the server we make requests to but is not of much use for us (the client) because FtpWebRequest usually ignores this property value, so it will not be available for our use in most of the cases. But if we set this property then the FTP server will get an idea in advance about the size of the file it should expect (in the case of an upload).
-
Method: Denotes what action (command) to take in the current request (upload, download, filelist and so on). It is set a value defined in the WebRequestMethods.FTP structure.
private void Upload(string filename)
{
FileInfo fileInf = new FileInfo(filename);
string uri = "ftp://" + ftpServerIP + "/" + fileInf.Name;
FtpWebRequest reqFTP;
// Create FtpWebRequest object from the Uri provided
reqFTP = (FtpWebRequest)FtpWebRequest.Create(new Uri("ftp://" + ftpServerIP + "/" + fileInf.Name));
// Provide the WebPermission Credintials
reqFTP.Credentials = new NetworkCredential(ftpUserID, ftpPassword);
// By default KeepAlive is true, where the control connection is not closed
// after a command is executed.
reqFTP.KeepAlive = false;
// Specify the command to be executed.
reqFTP.Method = WebRequestMethods.Ftp.UploadFile;
// Specify the data transfer type.
reqFTP.UseBinary = true;
// Notify the server about the size of the uploaded file
reqFTP.ContentLength = fileInf.Length;
// The buffer size is set to 2kb
int buffLength = 2048;
byte[] buff = new byte[buffLength];
int contentLen;
// Opens a file stream (System.IO.FileStream) to read the file to be uploaded
FileStream fs = fileInf.OpenRead();
try
{
// Stream to which the file to be upload is written
Stream strm = reqFTP.GetRequestStream();
// Read from the file stream 2kb at a time
contentLen = fs.Read(buff, 0, buffLength);
// Until Stream content ends
while (contentLen != 0)
{
// Write Content from the file stream to the FTP Upload Stream
strm.Write(buff, 0, contentLen);
contentLen = fs.Read(buff, 0, buffLength);
}
// Close the file stream and the Request Stream
strm.Close();
fs.Close();
}
catch (Exception ex)
{
MessageBox.Show(ex.Message, "Upload Error");
}
}
The preceding is a sample code for a FTP Upload (PUT). The underlying sub-command used is STOR. Here an FtpWebRequest object is made for the specified file on the FTP server. Various properties are set for the request, namely Credentials, KeepAlive, Method, UseBinary and ContentLength.
The file in our local machine is opened and the contents are written to the FTP request stream. Here a buffer of size 2KB is used as an appropriate size suited for upload of larger or smaler files.
private void Download(string filePath, string fileName)
{
FtpWebRequest reqFTP;
try
{
//filePath = <<The full path where the file is to be created. the>>,
//fileName = <<Name of the file to be createdNeed not name on FTP server. name name()>>
FileStream outputStream = new FileStream(filePath + "\\" + fileName, FileMode.Create);
reqFTP = (FtpWebRequest)FtpWebRequest.Create(new Uri("ftp://" + ftpServerIP + "/" + fileName));
reqFTP.Method = WebRequestMethods.Ftp.DownloadFile;
reqFTP.UseBinary = true;
reqFTP.Credentials = new NetworkCredential(ftpUserID, ftpPassword);
FtpWebResponse response = (FtpWebResponse)reqFTP.GetResponse();
Stream ftpStream = response.GetResponseStream();
long cl = response.ContentLength;
int bufferSize = 2048;
int readCount;
byte[] buffer = new byte[bufferSize];
readCount = ftpStream.Read(buffer, 0, bufferSize);
while (readCount > 0)
{
outputStream.Write(buffer, 0, readCount);
readCount = ftpStream.Read(buffer, 0, bufferSize);
}
ftpStream.Close();
outputStream.Close();
response.Close();
}
catch (Exception ex)
{
MessageBox.Show(ex.Message);
}
}
The preceding is sample code to download a file from the FTP server. Unlike the Upload functionality described above, Download would require the response stream that will contain the content of the file requested. Here the file to download is specified as part of the URI that in turn is used for the creation of the "FtpWebRequest" object. To "GET" the file requested, get the response of the "FtpWebRequest" object using the "GetResponse" method. This new response object built provides the response stream that contains the file content as a stream that you can easily convert to a file stream to get the file in place.
Note: We have the flexibility to set the location and name of the file under which it is to be saved on our local machine.
public string[] GetFileList()
{
string[] downloadFiles;
StringBuilder result = new StringBuilder();
FtpWebRequest reqFTP;
try
{
reqFTP = (FtpWebRequest)FtpWebRequest.Create(new Uri("ftp://" + ftpServerIP + "/"));
reqFTP.UseBinary = true;
reqFTP.Credentials = new NetworkCredential(ftpUserID, ftpPassword);
reqFTP.Method = WebRequestMethods.Ftp.ListDirectory;
WebResponse response = reqFTP.GetResponse();
StreamReader reader = new StreamReader(response.GetResponseStream());
string line = reader.ReadLine();
while (line != null)
{
result.Append(line);
result.Append("\n");
line = reader.ReadLine();
}
// to remove the trailing '\n'
result.Remove(result.ToString().LastIndexOf('\n'), 1);
reader.Close();
response.Close();
return result.ToString().Split('\n');
}
catch (Exception ex)
{
System.Windows.Forms.MessageBox.Show(ex.Message);
downloadFiles = null;
return downloadFiles;
}
}
The preceding is a sample block of code for getting the file list on the FTP server. The URI is built specifying the FTP server address/name and the required path if any. In the preceding example the root folder is specified for the creation of the "FtpWebRequest" object. Here the response stream is used for the creation of a "StreamReader" object, that has the entire list of file names on the server separated by "\r\n" which is a newline and carriage-return together. You can get the entire file list ("\r\n" separated) using the "ReadToEnd" method of the StreamReader object. The preceding implementation, reads each file name and creates a StringBuilder object by appending each file name. The resultant StringBuilder object is split into a stirng array and returned. I am sure there are better ways to do it. A better way is to remove the entire "\r" instances from the entire list (returned by <<StreamReader>>.ReadToEnd()) and split the resultant string using a "\n" delimiter. Anyway I did not want to spend more of my energy and time pondering over it ;-).
The implementations for Rename, Delete, GetFileSize, FileListDetails and MakeDir are very similar to the preceding pieces of code and the attached code is easily comprehensible.
Note: For renaming, the new name can be assigned to the "RenameTo" property of the "FtpWebRequest" object. For MakeDirectory, the name of the new directory can be specified as part of the URI used to create the "FtpWebRequest" object.
Points of Interest
Please take note of the following points while coding in this area:
-
Unless the "EnableSsl" property is true, all data and commands, including your user name and password information, are sent to the server in clear text. Anyone monitoring network traffic can view your credentials and use them to connect to the server. If you are connecting to an FTP server that requires credentials and supports Secure Sockets Layer (SSL), you should set "EnableSsl" to true.
-
If you do not have the proper WebPermission to access the FTP resource, a "SecurityException" exception is thrown.
-
Requests are sent to the server by calling the "GetResponse" method. When the requested operation completes, an FtpWebResponse object is returned. The "FtpWebResponse" object provides the status of the operation and any data downloaded from the server. That is, the "StatusCode" property of the FtpWebResponse object provides the latest status code returned by the FTP server. The "StatusDescription" property of the FtpWebResponse object provides the description of the status code returned.