Cannot access the file because it is being used by another process

Question

My web method creates a pdf file in my %temp% folder and that works. I then want to add some custom fields (meta) to that file using the code below.

The class PdfStamper generates an IOException, whether I use its .Close() method or the using block just ends. The process that is still holding on to the file handle is the webdev web server itself (I'm debugging in VS2010 SP1).

private string AddCustomMetaData(string guid, int companyID, string filePath)
{
    try
    {
        PdfReader reader = new PdfReader(filePath);

        using (FileStream fs = new FileStream(filePath, FileMode.Open, FileAccess.ReadWrite, FileShare.ReadWrite))
        {
            PdfStamper st = new PdfStamper(reader, fs);
            Dictionary<string, string> info = reader.Info;
            info.Add("Guid", guid);
            info.Add("CompanyID", companyID.ToString());

            st.MoreInfo = info;
            st.Close();
        }

        reader.Close();

        return guid;
    }
    catch (Exception e)
    {
        return e.Message;
    }
}

No matter what I try, it keeps throwing the exception at st.Close();, to be more precise:

The process cannot access the file 'C:\Users[my username]\AppData\Local\Temp\53b96eaf-74a6-49d7-a715-6c2e866a63c3.pdf' because it is being used by another process.

Either I'm overlooking something obvious or there's a problem with the PdfStamper class I'm as of yet unaware of. Versions of itextsharp used are 5.3.3.0 and 5.4.0.0, the issue is the same.

Any insight would be greatly appreciated.

EDIT: I'm currently "coding around" the issue, but I haven't found any solution.

Answer 1

You're problem is that you are writing to a file while you are also reading from it. Unlike some file types (JPG, PNG, etc) that "load" all of the data into memory, iTextSharp reads the data as a stream. You either need to use two files and swap them at the end or you can force iTextSharp to "load" the first file by binding your PdfReader to a byte array of the file.

PdfReader reader = new PdfReader(System.IO.File.ReadAllBytes(filePath));

Answer 2

I suggest you to use the FileShare enumerator when you open the file, so Try to open a file with None sharing

   File.Open(fileName, FileMode.Open, FileAccess.Read, FileShare.None);

Answer 3

Try to .Dispose() your PDF reader (or whatever you use for creating it) when you save the file for the first time

Answer 4

Try this solution if you think its feasible for you - Once the webmethod creates file in Temp folder, you need to copy the file and paste it into other location or same location with different name and pass newly copied file path to your PDF reader.