iTextSharp + FileStream = Corrupt PDF file

问题:

I am trying to create a pdf file with iTextSharp. My attempt writes the content of the pdf to a MemoryStream so I can write the result both into file and a database BLOB. The file gets created, has a size of about 21kB and it looks like a pdf when opend with Notepad++. But my PDF viewer says it's currupted. Here is a little code snippet (only tries to write to a file, not to a database):

Document myDocument = new Document();
MemoryStream myMemoryStream = new MemoryStream();
PdfWriter myPDFWriter = PdfWriter.GetInstance(myDocument, myMemoryStream);
myDocument.Open();
// Content of the pdf gets inserted here
using (FileStream fs = File.Create("D:\\...\\aTestFile.pdf"))
{
    myMemoryStream.WriteTo(fs);
}
myMemoryStream.Close();

Where is the mistake I make?

Thank you, Norbert

回答1:

I think your problem was that you weren't properly adding content to your PDF. This is done through the Document.Add() method and you finish up by calling Document.Close().

When you call Document.Close() however, your MemoryStream also closes so you won't be able to write it to your FileStream as you have. You can get around this by storing the content of your MemoryStream to a byte array.

The following code snippet works for me:

using (MemoryStream myMemoryStream = new MemoryStream()) {
    Document myDocument = new Document();
    PdfWriter myPDFWriter = PdfWriter.GetInstance(myDocument, myMemoryStream);

    myDocument.Open();

    // Add to content to your PDF here...
    myDocument.Add(new Paragraph("I hope this works for you."));

    // We're done adding stuff to our PDF.
    myDocument.Close();

    byte[] content = myMemoryStream.ToArray();

    // Write out PDF from memory stream.
    using (FileStream fs = File.Create("aTestFile.pdf")) {
        fs.Write(content, 0, (int)content.Length);
    }
}

回答2:

I had similar issue. My file gets downloaded but the file size will be 13Bytes. I resolved the issue when I used binary writer to write my file

byte[] bytes = new byte[0];
//pass in your API response into the bytes initialized

using (StreamWriter streamWriter = new StreamWriter(FilePath, true))
{
   BinaryWriter binaryWriter = new BinaryWriter(streamWriter.BaseStream);
   binaryWriter.Write(bytes);
}

回答3:

Just some thoughts - what happens if you replace the memory stream with a file stream? Does this give you the result you need? This will at least tell you where the problem could be.

If this does work, how do the files differ (in size and binary representation)?

Just a guess, but have you tried seeking to the beginning of the memory stream before writing?

myMemoryStream.Seek(0, SeekOrigin.Begin);

回答4:

Try double checking your code that manipulates the PDF with iText. Make sure you're calling the appropriate EndText method of any PdfContentByte objects, and make sure you call myDocument.Close() before writing the file to disk. Those are things I've had problems with in the past when generating PDFs with iTextSharp.

回答5:

documentobject.Close();
using (FileStream fs = System.IO.File.Create(path)){                        
    Memorystreamobject.WriteTo(fs);
}