I've been struggling with trying to read the configuration files on a device via FTP. The files are generated dynamically when the FTP request is sent, and the device does not support the FTPGetFileSize command.
I've lost count of the number of examples from the Web that I've tried in an attempt to solve this. Part of the problem is likely due to my inexperience with file streams. So, I'm looking for a kind soul that might be able to provide some insight.
In the code below, I'm trying two different methods. The first is a straight read of the entire file, the second attempts to control the read based whether there is anything coming in over the stream.
Both methods work on other devices that support the FTPGetFileSize command. But neither method works for this device.
Project form
Imports are:
Imports System.Net
Imports System.IO
Imports System.Text
The cmdGet_Click event code looks like this:
Dim sLocal_Path As String = "C:\Temp\FTP_Test\" + txtIPAddress.Text + "\"
System.IO.Directory.CreateDirectory(sLocal_Path)
' Complete the local path
sLocal_Path += txtFileName.Text
If System.IO.File.Exists(sLocal_Path) Then
System.IO.File.Delete(sLocal_Path) ' Make sure we don't have a collision.
End If
' Set up the Remote Path.
Dim sRemote_Path As String = "ftp://" + txtIPAddress.Text + "/gen:" + txtFileName.Text
Dim request As FtpWebRequest = CType(WebRequest.Create(sRemote_Path), FtpWebRequest)
request.Method = WebRequestMethods.Ftp.DownloadFile
request.Credentials = New NetworkCredential(FTP_User, FTP_Password)
request.UsePassive = True
request.Timeout = 15000
request.UseBinary = False
Dim response As FtpWebResponse = CType(request.GetResponse(), FtpWebResponse)
Dim responseStream As Stream = response.GetResponseStream()
Dim reader As StreamReader = New StreamReader(responseStream)
Dim writeFile As StreamWriter = New StreamWriter(sLocal_Path)
If RadioReadToEnd.Checked Then
' Write the FTP stream to the target file
writeFile.Write(reader.ReadToEnd()) ' works with static files.
Else
' Write the FTP stream as long as we have data coming in.
Dim ByteCount As Integer = 0
Threading.Thread.Sleep(500)
While Not (reader.Peek() = -1)
ByteCount += 1
If ByteCount > 4095 Then ' Go up for air after every block.
ByteCount = 0 ' Reset the byte counter for the next block.
Application.DoEvents()
End If
writeFile.Write(Chr(reader.Read))
End While
End If
reader.Close()
reader.Dispose()
writeFile.Close()
writeFile.Dispose()
responseStream.Close()
responseStream.Dispose()
Note that I put a delay in just before the While Loop in the hopes that maybe something would show up in the stream to keep the loop from terminating. Apparently I need some way to prime the pump.
Edit: The DOS FTP Get works fine. My assumption is that it has some way to get around the FTPGetFileSize not being supported. Is there a way to get the WebClient FTP Methods to do the same thing with a dynamically generated file?
I need a nudge (okay, a kick) in the right direction. Any suggestions are greatly appreciated!
It is now working! The problem turned out to be a missing '/' at the root of the path.
This particular device has a drive reference in it's path that other devices my routine talks to did not have. So, between that and the error being thrown regarding the file size, I was put off of the scent.
Thanks to Martin Prikryl's comment about showing a log, I went off to create a Wireshark recording, and that's when I noticed the missing slash when I compared it to the DOS FTP Get recording. So, while I'm a little red-faced, I hope my experience helps someone else who runs into a similar problem. But mostly, I'm very relieved that I did not have to go the route of writing an FTP Client that dealt with raw commands such as POST and RETR! The WebClient tools are now working very well for this application.
Related
Title says it all, How do I save a file from the application resources to the desktop for example?
In the future it will create a directory and store it in there, essentially I'm making a game installer, but none of this matters right now.
I have tried about 6 different code methods and all of them failed, kept throwing errors about not being able to convert an icon/image to a 1-dimensional array. Or that "Icon" or "Image" is not a member of System.Drawing.Icon
Any help would be appreciated
Thank you!
EDIT: Posting some code I've tried.
#1 ('Length' is not a member of System.Drawing.Icon)
Dim File01 As System.IO.FileStream = New System.IO.FileStream("C:\Users\" + Environment.UserName + "\Desktop\" + "EXAMPLE_Icon.ico", IO.FileMode.Create) File01.Write(My.Resources.EXAMPLE_Icon, 0, My.Resources.EXAMPLE_Icon.length) File01.Close()
#2 ('System.Drawing.Icon' cannot be converted to '1-dimensional array of Byte)
Dim filePath = Path.Combine("C:\Users\" + Environment.UserName + "\Desktop\", "EXAMPLE_Icon.ico") File.WriteAllBytes(filePath, My.Resources.EXAMPLE_Icon)
#3 ('RawFormat' is not a member of 'System.Drawing.Icon')
Dim filePath = Path.Combine("C:\Users\" + Environment.UserName + "\Desktop\", "EXAMPLE_Icon.ico") Using icon = My.Resources.EXAMPLE_Icon icon.Save("C:\Users\" + Environment.UserName + "\Desktop\", icon.RawFormat) End Using
Build the destination path using Path.Combine() when the path is composited.
Many known System locations have a predefined value in Environment.SpecialFolder: in your case, Environment.SpecialFolder.Desktop. Environment.GetFolderPath() accepts one of these values and returns the correct file system path.
This returns the Path of the Desktop folder for the current User:
Environment.GetFolderPath(Environment.SpecialFolder.Desktop)
The Icon object has a Save() method that accepts a Stream as argument.
You can pass a FileStream to save the Icon to a File, preserving its format, or a MemoryStream, then use File.WriteAllBytes() to save the MemoryStream's buffer calling its ToArray() method.
Using a Resource name, as a string, using ResourceManager.GetObject():
Dim iconPath = Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.Desktop), "myIcon.ico")
Using stream As New FileStream(iconPath, FileMode.Create, FileAccess.Write, FileShare.None),
ico = TryCast(My.Resources.ResourceManager.GetObject("myIcon"), Icon)
ico?.Save(stream)
End Using
Or, depending on what fits better in the context of your operations:
Using stream As New FileStream(iconPath, FileMode.Create, FileAccess.Write, FileShare.None)
My.Resources.myIcon.Save(stream)
End Using
GC.Collect()
Note: here, I'm calling GC.Collect() after calling Save() directly on the Icon object generated by My.Resources. The Docs don't exactly explain it: My.Resources is a factory, it doesn't return the object stored in the Project's Resources, it generates a new object (a different one each time) from the data stored as a Resource. With My.Resources.myIcon.Save(stream), we create a new object, save its data to disc, but never actually dispose of it. Calling GC.Collect() right after, the memory used is reclaimed immediately. If we don't, that memory is never reclaimed (we're leaking).
It's probably better to assign the generated object a to temporary variable declared with a Using statement. The two methods that follow can make use of GC.Collect(), calling it right after, to reclaim the allocated memory immediately, but it's not strictly required, we're not really leaking resources.
Open up Visual Studio's Diagnostic Tools to test the behavior with and without GC.Collect(), calling these methods multiple times.
Using stream As New FileStream(iconPath, FileMode.Create, FileAccess.Write, FileShare.None),
ico = My.Resources.myIcon
ico.Save(stream)
End Using
' GC.Collect() <= Not strictly required, but test the difference
Or using a MemoryStream:
Using ms As New MemoryStream(),
ico As Icon = My.Resources.myIcon
ico.Save(ms)
ms.Position = 0
File.WriteAllBytes(iconPath, ms.ToArray())
End Using
' GC.Collect() <= Not strictly required, but test the difference
I'm creating a file on the fly on a WebAPI call, and sending that file back to the client.
I think I'm misunderstanding flush/close on a FileStream:
Dim path As String = tempFolder & "\" & fileName
Dim result As New HttpResponseMessage(HttpStatusCode.OK)
Dim stream As New FileStream(path, FileMode.Open)
With result
.Content = New StreamContent(stream)
.Content.Headers.ContentDisposition = New Headers.ContentDispositionHeaderValue("attachment")
.Content.Headers.ContentDisposition.FileName = fileName
.Content.Headers.ContentType = New Headers.MediaTypeHeaderValue("application/octet-stream")
.Content.Headers.ContentLength = stream.Length
End With
'stream.Flush()
'stream.Close()
'Directory.Delete(tempFolder, True)
Return result
You can see where I've commented things out above.
Questions:
Does the stream flush/close itself?
How can I delete the tempFolder after returning the result?
On top of all this, it would be great to know how to generate the file and send it to the user without writing it to the file system first. I'm confident this is possible, but I'm not sure how. I'd love to be able to understand how to do this, and solve my current problem.
Update:
I went ahead with accepted answer, and found it to be quite simple:
Dim ReturnStream As MemoryStream = New MemoryStream()
Dim WriteStream As StreamWriter = New StreamWriter(ReturnStream)
With WriteStream
.WriteLine("...")
End With
WriteStream.Flush()
WriteStream.Close()
Dim byteArray As Byte() = ReturnStream.ToArray()
ReturnStream.Flush()
ReturnStream.Close()
Then I was able to stream the content as bytearraycontent:
With result
.Content = New ByteArrayContent(byteArray)
...
End With
On top of all this, it would be great to know how to generate the file and send it to the user without writing it to the file system first. I'm confident this is possible, but I'm not sure how. I'd love to be able to understand how to do this, and solve my current problem.
To do the same thing without writing a file to disk, you might look into the MemoryStream class. As you'd guess, it streams data from memory like the FileStream does from a file. The two main steps would be:
Take your object in memory and instead of writing it to a file, you'd serialize it into a MemoryStream using a BinaryFormatter or other method (see that topic on another StackOverflow Q here: How to convert an object to a byte array in C#).
Pass the MemoryStream to the StreamContent method, exactly the same way you're passing the FileStream now.
For a long time I have been trying to debug why my parsing counts were off when downloading files to be parsed and been made to look really dumb about this. I did some debugging and found that the file I download when trying to decompress using GZipStream shows that it misses data from the original file. Here is my code for decompressing:
Using originalFileStream As FileStream = fileItem.OpenRead()
Dim currentFileName As String = fileItem.FullName
Dim newFileName = currentFileName.Remove(currentFileName.Length - fileItem.Extension.Length)
newFile = newFileName
Using decompressedFileStream As FileStream = File.Create(newFileName)
Using decompressionStream As GZipStream = New GZipStream(originalFileStream, CompressionMode.Decompress)
decompressionStream.CopyTo(decompressedFileStream)
Console.WriteLine("Decompressed: {0}", fileItem.Name)
decompressionStream.Close()
originalFileStream.Close()
End Using
End Using
End Using
Now what I do is return the newfile to the calling function and read the contents from there:
Dim responseData As String = inputFile.ReadToEnd
Now pasting the url in the browser and downloading from there and then opening using winrar I can see the data is not the same. Now this does not happen all the time as some files parse and decompress correctly. Each downloaded file has check counter to compare how many posts I am supposed to be parsing from it and that triggered me to see the mismatch in counts.
EDIT
Here is what I found in addition. If I read the problem file (as I said that only some files happen this way) by individual lines I will get all the data:
Dim rData As String = inputFile.ReadLine
If Not rData = "" Then
While Not inputFile.EndOfStream
rData = inputFile.ReadLine + vbNewLine + rData
End While
getIndividualPosts(rData)
End If
Now if I try to read an individual line from a file that is not problematic it will return nothing and so I will have to readtoEnd. Can anyone explain this odd behavior and is it related to the GZIPSTREAM or some error in my code.
We have a servlet that will deliver PDF reports to a browser. We also have a IIS server running .net apps and we want to return the PDF from the servlet as a stream to the .Net app and then the .Net app would render the PDF to the browser (we are using this technique for reason I don't need to go into here). I am not much of a VB/ Visual Studio devloper by this code works by using a web request:
Dim BUFFER_SIZE As Integer = 1024
' Create a request for the URL.
Dim serveraction As String = "https://OurSeverName/ServletContext/Dispatch?action=ajaxRunReport&reportName="
Dim request As WebRequest = _
WebRequest.Create(serveraction + ReportName.Text)
' Get the response.
Dim res As WebResponse = request.GetResponse()
' Get the stream containing content returned by the server.
Dim dataStream As Stream = res.GetResponseStream()
' Open the stream using a BinaryReader for easy access.
Dim reader As New BinaryReader(dataStream)
' Read the content.
Response.ContentType = "application/pdf"
Response.AddHeader("content-disposition", "inline; filename=reportfile.pdf")
Dim bytes = New Byte(BUFFER_SIZE - 1) {}
While reader.Read(bytes, 0, BUFFER_SIZE) > 0
Response.BinaryWrite(bytes)
End While
reader.Close()
' Clean up the streams and the response.
Response.Flush()
Response.Close()
The only issue is, even though the code runs quickly, it takes 20-30 seconds to render the PDF in Chrome and IE but only a few seconds in FireFox. Any idea why there is a delay in rendering the PDF? Is there a better way to stream a PDF from one server to another?
There were just a couple of very subtle tweaks that were needed (and they seem pretty insignificant and non-intuitive to me).
I added the following before setting the content type:
Response.Clear()
Response.ClearHeaders()
And I added the following after reader.Close()
Response.End()
That was it. Now the PDF files stream nicely from the Java servlet to the IIS server and to the end user's browser.
I have wrote a web service, in a nutshell it uses openpop to get email messages does stuff with the content to insert into databases and saves attachments which are images. That works fine when i save images locally, it does exactley what it is suppose to. Now an added requirment was to save images to an FTP directory, so i can create my folders dynamically (they are created based upon timestamp) and that works well. My problem comes from when i try to save them to the ftp. Yes my user name and password are correct, otherwise i wouldn't be creating the directory.
Private Sub UploadFile(ByVal fileToSave As FileInfo, ByVal path As String)
Dim UploadRequest As FtpWebRequest = DirectCast(WebRequest.Create("ftp://UserName:Passowrd#999.99.999.9" & path), FtpWebRequest)
UploadRequest.Credentials = New NetworkCredential("PicService", "grean.matching18")
UploadRequest.Method = System.Net.WebRequestMethods.Ftp.UploadFile
UploadRequest.UseBinary = True
UploadRequest.UsePassive = True
' Const BufferSize As Integer = 2048
' Dim content(BufferSize - 1) As Byte, dataRead As Integer
Dim bFile() As Byte = System.IO.File.ReadAllBytes(fileToSave.ToString)
'UploadRequest.ContentLength = content.Length
Using FileStream1 As FileStream = fileToSave.OpenRead()
Try
'open request to send
Using RequestStream As Stream = UploadRequest.GetRequestStream
End Using
Catch ex As Exception
Finally
'ensure file closed
FileStream1.Close()
End Try
End Using
End Sub
I have tried using Passive False and Binary False as well, i did more research on my stack trace.
And found this article but no solution as of yet. Any input would be appreciated, i am also posting another question on windows services for different issue. If you would like to take a shot at it, the other question isnt about ftp but permissions for a service on windows server 2003
This may not be the solution but I've found that the URI string has to be 'just right' and that what is 'just right' varies by the ftp server.
So ftp://server/directory/file works on some servers but needs to be ftp://server//directory/file to work on others (note the double slash after the server name)
Aso, your URI has 'password' spelled incorrectly: ftp://UserName:Passowrd#999.99.999.9 and you are supplying the credentials in a separate code line as well.