copy file + progress bar
- Thread starter rlangi05
- Start date
To do it you have to read the byte count as the copy is in progress. No basic IO file copy method provide this info, but with one IO.FileStream for reading and one for writing you read-write byte-by-byte and can update a ProgressBar during the copy process.
This example should help, there is a ProgressBar with maxvalue 100 on form for progress percent:
This example should help, there is a ProgressBar with maxvalue 100 on form for progress percent:
VB.NET:
Sub copyfilewithprogress()
Dim path As String = Application.StartupPath & "\"
Dim mediafile As String = "Amy Diamond - What's in it for me.mp3"
Dim fi As New IO.FileInfo(mediafile)
Dim sr As New IO.FileStream(path & mediafile, IO.FileMode.Open) 'source file
Dim sw As New IO.FileStream(path & "copy of " & mediafile, IO.FileMode.Create) 'target file, defaults overwrite
Dim len As Long = sr.Length - 1
For i As Long = 0 To len
sw.WriteByte(sr.ReadByte)
If i Mod 1000 = 0 Then 'only update UI every 1 Kb copied
ProgressBar1.Value = i * 100 / len
Application.DoEvents()
End If
Next
ProgressBar1.Value = 0
End Sub
BGM
Member
Thank you for that! I want to note, however, that if you add a function for cancelling the operation, you have to call also, in the cancellation function:
If you don't do this, then your files will be locked by your application and you won't be able to restart the copy process.
@JohnH - Aside from the progress bar, how does this function differ from file.copy? I know file.copy will return an error if the files are missing.
VB.NET:
sr.close
sw.Close@JohnH - Aside from the progress bar, how does this function differ from file.copy? I know file.copy will return an error if the files are missing.
FileStream with FileMode.Open requires an existing file, as documented FileNotFoundException is thrown when not.
Also see that this thread is very old and for .Net 1.1. For .Net 2.0 and onwards you can use My.Computer.FileSystem object, its CopyFile method has options for progress UI and also cancellation.
BGM
Member
Actually, the filestream works perfectly (.net 4). I attached it to a background worker. It updates my progressbar and allows me to cancel the copy progress. I filestream the original file to a temp file in the new directory, then, if the progress = 100, and the filesize of the old file matches the filesize of the new file, then I give that temp file the correct name. If the progress was cancelled or there was an error, then I delete the tempfile.
I am thinking, however, of using checksums to verify that the new file matches the old file, however. I saw in debug.print that the filesize can differ by one or two bytes, so there wasn't much use in comparing file sizes based on the length of the streamed file.
Yes, I realized that, so I wrote my own Exists() function to account for that.
I am thinking, however, of using checksums to verify that the new file matches the old file, however. I saw in debug.print that the filesize can differ by one or two bytes, so there wasn't much use in comparing file sizes based on the length of the streamed file.
Of course is does, all other file IO tools use it also, including FileSystem.CopyFile. What I'm saying is that many additions has been made to .Net framework since the first version, to make common programming tasks much simpler for the developer.
A plain byte copy of a file can't differ by even a single byte, if it does something is wrong.
BGM
Member
That's what I thought too. I think I had the debug.print in the wrong spot... It is copying perfectly now.
That snippet you placed there has helped me out alot! I had tried so many other options. Thank you very much. I shall say a prayer for you.
BGM
Member
Mr. John H - How do I modify this function to allow for smaller files? I have a file that I use for testing that is ony 14kb, and I can't copy it. I think it has to do with the mod function. How can I keep the ui updating every 1kb, but still qualify the code to copy the file too? I admit, I don't completely understand what is going on.
Now, I have the io function wrapped into a worker.DoWork function... so you may notice the FileCopyWorker.ReportProgress(progress) - that is so I can run the io process on a worker thread.
Now, I have the io function wrapped into a worker.DoWork function... so you may notice the FileCopyWorker.ReportProgress(progress) - that is so I can run the io process on a worker thread.
VB.NET:
Private Sub FileCopyWorker_DoWork(ByVal sender As Object, ByVal e As System.ComponentModel.DoWorkEventArgs) Handles FileCopyWorker.DoWork
Try Dim temppath As String = GetDirFromPath(whatdestinationfile) & "temp"
Dim fi As New IO.FileInfo(temppath)
Dim sr As New IO.FileStream(whatoriginationfile, IO.FileMode.Open) 'source file
Dim sw As New IO.FileStream(temppath, IO.FileMode.Create) 'target file, defaults overwrite
Dim len As Long = sr.Length - 1
MessengerLogger.Info("Please wait whilst I upload " & whatoriginationfile)
Dim i As long
For i = 0 To len
If FileCopyWorker.CancellationPending Then
MessengerLogger.Error("The transfer was cancelled at: " & progress & "%.")
Exit For
End If
sw.WriteByte(sr.ReadByte)
If i Mod 1000 = 0 Then 'only update UI every 1 Kb copied
progress = i * 100 / len
FileCopyWorker.ReportProgress(progress)
Debug.Print(progress)
Application.DoEvents()
End If
Next
sr.close
sw.close
If progress = 100 then
'Debug.Print("finished:" & progress)
File.Move(temppath, whatdestinationfile)
Else
'Debug.Print("cancelled:" & progress)
File.Delete(temppath)
end if
Catch IOerror As Exception
MessengerLogger.Error("There was an error in the procedure: " & IOerror.Message & " :CopyFileWithProgress().")
End Try
End SubI see no reason the code would't work on any size file. It copies one-by-one byte from source to target file, while reporting every 1kb progress - that's it and that's all.
Maybe your problem is that you look for "progress=100", instead use a Boolean to indicate when operation has completed.
I would also like to add, the code example was just that, a basic example to show how to read bytes from a source and write them to a target, while calculating progress. Normally for file copy/move operations you would read/write larger chunks of bytes each loop iteration, resulting in wastly faster IO operations.
Maybe your problem is that you look for "progress=100", instead use a Boolean to indicate when operation has completed.
I would also like to add, the code example was just that, a basic example to show how to read bytes from a source and write them to a target, while calculating progress. Normally for file copy/move operations you would read/write larger chunks of bytes each loop iteration, resulting in wastly faster IO operations.
BGM
Member
Thanks for helping out. I figured the copy progress was a little slow. I don't understand why progress=100 wouldn't work. I test for that, because if progress is < 100, then either there was an error or the user cancelled the progress, which allows me to do some cleanup or notifications to the user.
I understand it was a basic example, which is what I appreciated. I think the reason it doesn't copy smaller files (and it doesn't) is because it only does the work if imod 1000 = 0. If i doesn't evaluate correctly in the mod, then it doesn't work right.
As you can see from my tests, here below, it has nothing to do with progress = 100. It has to do with the mod function.
Now, I understand that the mod function is supposed to provide an interval for the progress to update, but i mod 1000 either isn't in the right place, or isn't the right expression, I think.
I plugged this code into FastSharp (try it if you haven't - it lets you quickly evaluate expressions without using the VS IDE, and its free, too!)
Look at the output for t=14:
Now, we make t = 1400
Finally, with t = 14000
However, if t=1000
but t=100
I understand it was a basic example, which is what I appreciated. I think the reason it doesn't copy smaller files (and it doesn't) is because it only does the work if imod 1000 = 0. If i doesn't evaluate correctly in the mod, then it doesn't work right.
As you can see from my tests, here below, it has nothing to do with progress = 100. It has to do with the mod function.
Now, I understand that the mod function is supposed to provide an interval for the progress to update, but i mod 1000 either isn't in the right place, or isn't the right expression, I think.
I plugged this code into FastSharp (try it if you haven't - it lets you quickly evaluate expressions without using the VS IDE, and its free, too!)
VB.NET:
dim i as long
dim progress as long
dim m as long
dim t as long = 14
for i = 0 to t
m = i mod 1000
' console.write(vbcr & "mod=" & m)
if m = 0 then
progress = i * 100 / t
console.write(vbcr & "i =" & i & vbcr)
console.write("progress=" & progress & "%")
end if
nextLook at the output for t=14:
VB.NET:
i =0
progress=0%Now, we make t = 1400
VB.NET:
i =0
progress=0%
i =1000
progress=71%Finally, with t = 14000
VB.NET:
i =0
progress=0%
i =1000
progress=7%
i =2000
progress=14%
i =3000
progress=21%
i =4000
progress=29%
i =5000
progress=36%
i =6000
progress=43%
i =7000
progress=50%
i =8000
progress=57%
i =9000
progress=64%
i =10000
progress=71%
i =11000
progress=79%
i =12000
progress=86%
i =13000
progress=93%
i =14000
progress=100%However, if t=1000
VB.NET:
i =0
progress=0%
i =1000
progress=100%but t=100
VB.NET:
i =0
progress=0%BGM
Member
Hey, I think I might have gotten it - and you may have been right. I moved things around a bit...
VB.NET:
[COLOR=#0000ff][B]Private[/B][/COLOR] [COLOR=#0000ff][B]Sub[/B][/COLOR] FileCopyWorker_DoWork([COLOR=#0000ff][B]ByVal[/B][/COLOR] sender [COLOR=#0000ff][B]As[/B][/COLOR] [COLOR=#0c8910][B]Object[/B][/COLOR], [COLOR=#0000ff][B]ByVal[/B][/COLOR] e [COLOR=#0000ff][B]As[/B][/COLOR] System.ComponentModel.DoWorkEventArgs) [COLOR=#0000ff][B]Handles[/B][/COLOR] FileCopyWorker.DoWork
whattempdestinationfile = GetDirFromPath(whatdestinationfile) & [COLOR=#f99304]"temp"[/COLOR]
[COLOR=#0000ff][B]Try[/B][/COLOR]
[COLOR=#0000ff][B]Dim[/B][/COLOR] fi [COLOR=#0000ff][B]As[/B][/COLOR] [COLOR=#fc0a02]New[/COLOR] IO.FileInfo(whattempdestinationfile)
[COLOR=#0000ff][B]Dim[/B][/COLOR] sr [COLOR=#0000ff][B]As[/B][/COLOR] [COLOR=#fc0a02]New[/COLOR] IO.FileStream(whatoriginationfile, IO.FileMode.Open) [COLOR=#a9aaa9]'source file[/COLOR]
[COLOR=#0000ff][B]Dim[/B][/COLOR] sw [COLOR=#0000ff][B]As[/B][/COLOR] [COLOR=#fc0a02]New[/COLOR] IO.FileStream(whattempdestinationfile, IO.FileMode.Create) [COLOR=#a9aaa9]'target file, defaults overwrite[/COLOR]
[COLOR=#0000ff][B]Dim[/B][/COLOR] len [COLOR=#0000ff][B]As[/B][/COLOR] [COLOR=#0c8910][B]Long[/B][/COLOR] = sr.Length - 1
MessengerLogger.Info([COLOR=#f99304]"Please wait whilst I upload "[/COLOR] & whatoriginationfile)
[COLOR=#0000ff][B]Dim[/B][/COLOR] i [COLOR=#0000ff][B]As[/B][/COLOR] [COLOR=#0c8910][B]long[/B][/COLOR]
[COLOR=#0000ff][B]For[/B][/COLOR] i = 0 [COLOR=#0000ff][B]To[/B][/COLOR] len
[COLOR=#0000ff][B]If[/B][/COLOR] FileCopyWorker.CancellationPending [COLOR=#0000ff][B]Then[/B][/COLOR]
MessengerLogger.[COLOR=#0000ff][B]Error[/B][/COLOR]([COLOR=#f99304]"The transfer was cancelled at: "[/COLOR] & progress & [COLOR=#f99304]"%."[/COLOR])
[COLOR=#0000ff][B]Exit[/B][/COLOR] [COLOR=#0000ff][B]For[/B][/COLOR]
[COLOR=#0000ff][B]End[/B][/COLOR] [COLOR=#0000ff][B]If[/B][/COLOR]
sw.WriteByte(sr.ReadByte)
progress = i * 100 / len
[COLOR=#0000ff][B]If[/B][/COLOR] i [COLOR=#fc0a02]Mod[/COLOR] 1000 = 0 [COLOR=#fc0a02]Or[/COLOR] progress = 100 [COLOR=#0000ff][B]Then[/B][/COLOR] [COLOR=#a9aaa9]'only update UI every 1 Kb copied[/COLOR]
FileCopyWorker.ReportProgress(progress)
Application.DoEvents()
[COLOR=#0000ff][B]End[/B][/COLOR] [COLOR=#0000ff][B]If[/B][/COLOR]
[COLOR=#0000ff][B]Next[/B][/COLOR]
sr.close
sw.close
[COLOR=#0000ff][B]Catch[/B][/COLOR] IOerror [COLOR=#0000ff][B]As[/B][/COLOR] Exception
MessengerLogger.[COLOR=#0000ff][B]Error[/B][/COLOR]([COLOR=#f99304]"There was an error in the procedure: "[/COLOR] & IOerror.Message & [COLOR=#f99304]" :CopyFileWithProgress()."[/COLOR])
[COLOR=#0000ff][B]End[/B][/COLOR] [COLOR=#0000ff][B]Try[/B][/COLOR]
[COLOR=#0000ff][B]End[/B][/COLOR] [COLOR=#0000ff][B]Sub[/B][/COLOR]
[\code]If copy for example is stopped after 995 out of 1000 bytes then progress would be at 100, but copy operation was not completed.
There are only two possible outcomes if a line is executed after the copy loop, either the operation was cancelled (CancellationPending) or it was completed.
How could it not? The progress reporting is unrelated to the copy operation.
If an exception is thrown those are not called. Either utilize the Using statement, or put them in Finally block of the Try statement.
Now you are calculating percentage for each byte, which is unnecessary. While not a big waste of CPU cycles, it still is, always something to think about.
It if was better than VS IDE I would use it
Faster? I doubt it produces code faster than VS intellisense.
BGM
Member
Okay, I see that now. The IO functions run regardless unless the user cancels the process, in which case it gets interrupted, or there is an error, but my progressbar is only affected by the i mod part. It is important, therefore, to make sure sw.writebyte(sr.readbyte) is NOT inside the mod block.
I don't see what to do to make it faster or more efficient. To be guaranteed of 100 for progress, despite the mod function, I could do this:
VB.NET:
If i mod 1000 = 0 or i = len then...Good point about the exception. I should have realized that.
If I use Finally, then I can't dim the filestream declarations locally. I don't understand how to use Using, unless it is a resource.
I added, per your advice a boolean that is defined as false if the user cancels the operation, or if there was an IO error, but true if i = len.
VB.NET:
For i = 0 To len
If FileCopyWorker.CancellationPending Then
FileStatus = false
MessengerLogger.Warn("You cancelled the transfer at: " & progress & "%.")
Exit For
End If
sw.WriteByte(sr.ReadByte)
progress = i * 100 / len
If i Mod 1000 = 0 or i = len Then 'only update UI every 1 Kb copied
If i = len then FileStatus = true
FileCopyWorker.ReportProgress(progress)
Application.DoEvents()
End If
NextOkay, I agree, but what do I do? How can I make sure the progressbar follows the IO stream more efficiently. Maybe mod 10000 instead of mod 1000?
Yeah, I suppose VS is as good, but Fastsharp gives me a separate window, and I can leave my immediates window docked. I can't tell that it is slower or faster, actually.
Why? 100% progress is something you can report after the copy loop if not cancelled. You can even do that from RunWorkerCompleted event handler.
I guess that depends on the file size copied. Look at Windows copy, you rarely see progress indication for a single file copy under 1mb, the operation is simply too fast and completes almost before it is possible to display a progress GUI. So time is a factor here, maybe you could disregard the particular byte progress and just look at the time since last progress update, for example only report progress once per second?
Any disposable class instance is a resource. The code sample is using a TextWriter instance for example. Though Using statement is not such a good idea here, the Using block swallows all exceptions, and here you need to know if an exception is thrown.
BGM
Member
Mr. JohnH, I thank you again for your time!
I declared the filestreams outside of the sub and defined them new within the try loop, which allowed me to use Finally to close the filestreams.
I think my poor brain is finally getting it.
Then, in the RunWorkerCompleted function, I compare size of the old and new files and made sure the progressbar received a value of 100 upon success. The catch statement sets FileStatus to false if there was an exception.
Private Sub FileCopyWorker_RunWorkerCompleted(ByVal sender As Object, ByVal e As System.ComponentModel.RunWorkerCompletedEventArgs) Handles
I declared the filestreams outside of the sub and defined them new within the try loop, which allowed me to use Finally to close the filestreams.
I think my poor brain is finally getting it.
VB.NET:
If FileCopyWorker.CancellationPending Then
FileStatus = false
MessengerLogger.Warn("You cancelled the transfer at: " & progress & "%.")
Exit For
End If
sw.WriteByte(sr.ReadByte)
If i Mod 1000 = 0 Then 'only update UI every 1 Kb copied
progress = i * 100 / len
FileCopyWorker.ReportProgress(progress)
Application.DoEvents()
End If
If i = len then FileStatus = true
NextThen, in the RunWorkerCompleted function, I compare size of the old and new files and made sure the progressbar received a value of 100 upon success. The catch statement sets FileStatus to false if there was an exception.
Private Sub FileCopyWorker_RunWorkerCompleted(ByVal sender As Object, ByVal e As System.ComponentModel.RunWorkerCompletedEventArgs) Handles
VB.NET:
FileCopyWorker.RunWorkerCompleted
If filestatus then
File.Move(whattempdestinationfile, whatdestinationfile)
If GetFileSize(whatdestinationfile) = GetFileSize(whatoriginationfile) then
MessengerLogger.Info(msg_finished)
MakeThreadedNotifications() 'that is, send an email to notify someone
End If
whatprogressbar.Value = 100
Else
MessengerLogger.Warn(msg_toobad)
File.Delete(whattempdestinationfile)
End If
whatstartbutton.Enabled = True
whatcancelbutton.Enabled = False
End SubDo you think it is the monitoring of the byte progress that makes the action slower? I understand about the progress being too fast to monitor for small files.
