如何避免异常进程无法访问该文件，因为它正由另一个进程使用。 在.net

Question

I am unzipping a.gz files by using the following code.Since i have huge number of files , I am using TPL tasks to run this code. But I used to get .NET exception:

The process cannot access the file because it is being used by another process.

How to correct this issue?
Code used is as follows:

private static void Decompress(FileInfo fileToDecompress)
{
    using (FileStream originalFileStream = fileToDecompress.OpenRead())
    {
        string currentFileName = fileToDecompress.FullName;
        string newFileName = currentFileName.Remove(currentFileName.Length - 
                                          fileToDecompress.Extension.Length);
        using (FileStream decompressedFileStream = File.Create(newFileName))
        {
            using (GZipStream decompressionStream = 
                     new GZipStream(originalFileStream, 
                                    CompressionMode.Decompress))
            {
                decompressionStream.CopyTo(decompressedFileStream);
            }
        }
    }
    //File.Move(myffile, Path.ChangeExtension(myffile, ".jpg"));
}

calling code is as shown below

var list = ftp.GetFileList(remotepath);

//-------------------
DateTime dt = DateTime.Now;
string st = String.Format("{0:yyyyMMdd}", dt);//20161120
Task[] myTasks = new Task[list.Count];
int i = 0;
foreach (string item in list)
{
     if (item.StartsWith("GExport_") && (!item.ToUpper().Contains("DUM")) && (item.Contains(st)) && (!item.ToUpper().Contains("BLK")))
     {
         4gpath = item;
        //Downloadfile()   
        ftp.Get(dtr["REMOTE_FILE_PATH"].ToString() + 4gpath , @localDestnDir + "\\" + dtr["SOURCE_PATH"].ToString());
        download_location_hw = dtr["LOCAL_FILE_PATH"].ToString();
        // Spin off a background task to process the file we just downloaded
        myTasks[i++] = Task.Factory.StartNew(() =>
        {
            //Extractfile()             
            ExtractZipfiles(download_location_hw + "//" + huwawei4gpath, dtr["REMOTE_FILE_PATH"].ToString(), 
             dtr["FTP_SERVER"].ToString(), dtr["FTP_USER_ID"].ToString(),
             dtr["TECH_CODE"].ToString(), dtr["VENDOR_CODE"].ToString());
        }
    }
}

private static void ExtractZipfiles(string download_location_hw,string remotepath,string server,string userid,string tech,string vendor)
{
    if (download_location_hw.Contains(".gz"))
    {
        DirectoryInfo directorySelected = new DirectoryInfo(Path.GetDirectoryName(download_location_hw));
        foreach (FileInfo fileToDecompress in directorySelected.GetFiles("*.gz"))
        {
            Decompress(fileToDecompress);
        }
    }

}

Answer 1

You're possibly extracting the zip files downloaded bu your tasks in the same folder:

download_location_hw = dtr["LOCAL_FILE_PATH"].ToString();
var directorySelected = new DirectoryInfo(Path.GetDirectoryName(download_location_hw));

If for somethat reason you have two files being targeted to the same local path, final foreach statement will have intersections with other tasks:

foreach (var fileToDecompress in directorySelected.GetFiles("*.gz"))
{
    Decompress(fileToDecompress);
}

Here you're searching for all *.gz files in folder, even those ones that can be still in download mode or being decompressed by other process. So you should be definitely sure each tasks is being run in the different folder.

If you want to simple correction of your code, you may try this (the main change here is not to enumerate all the archive files but simply get one from the task arguments, with FileInfo constructor ):

private static void ExtractZipfiles(string download_location_hw,string remotepath,string server,string userid,string tech,string vendor)
{
    if (download_location_hw.Contains(".gz") && File.Exists(download_location_hw))
    {
        Decompress(new FileInfo(download_location_hw));
    }
}

Answer 2

The various tasks are operating on the same download path and are causing a collission. For eg. Task 1 puts a.gz, b.gz while Task 2 puts c.gz under "MyDir".

But now both tasks are operating and trying to rename the files to a, b, c during decompression. So, it appears you need to create subdirectories for each of the tasks or prefix\\suffix the file names that are downloaded with unique identifiers.