标记约1500页时，iTextSharp内存不足异常

Question

I am attempting to create a single pdf document from a list of approximately 5000 forms that I am stamping from a list of objects in my ASP.NET MVC website written in C#. I am getting an Out of Memory Exception once I perform the stamping process of approximately 1500 forms. If I restrict my list of objects to under 1500 objects I do not get the error.

Here is the data object which mimics the form data:

public class MyPageData
{
    public string Name { get; set; }
    public string Description { get; set; }
    public decimal DueAmount { get; set; }
    public decimal PaidAmount { get; set; }
}

The code to create the document. NOTE The template is a pdf form and the Stream is a FileStream object created with write access and no file sharing.

public class DocumentBuilder
{
    public static void FillDocument(IList<MyPageData> pages, string template, Stream outputFile)
    {
        using (Document document = new Document())
        {
            using (PdfSmartCopy copy = new PdfSmartCopy(document, outputFile))
            {
                document.Open();
                foreach (var page in pages)
                {
                    PdfReader reader = new PdfReader(StampPage(template, page));
                    copy.AddPage(copy.GetImportedPage(reader, 1));
                    copy.FreeReader(reader);
                }
            }
        }
    }

    private static byte[] StampPage(string template, MyPageData page)
    {
        PdfReader reader = new PdfReader(template);
        using (MemoryStream stream = new MemoryStream())
        {
            using (PdfStamper stamper = new PdfStamper(reader, stream))
            {
                AcroFields fields = stamper.AcroFields;
                FillFields(fields, page);
                stamper.FormFlattening = true;
                stamper.Writer.Flush();
            }
            return stream.ToArray();
        }
    }

    private static void FillFields(AcroFields fields, MyPageData page)
    {
        fields.SetField("Name", page.Name);
        fields.SetField("Description", page.Description);
        fields.SetField("DueAmount", page.DueAmount.ToString());
        fields.SetField("PaidAmout", page.PaidAmount.ToString());
    }
}

I have been all over the iText and iTextSharp document and I believe my code is correct. I did attempt to zip the document as demonstrated Here but the error is still thrown when stamping. Any thoughts or suggestion?

Answer 1

(not an answer - yet)

I just tried your code with the following which runs 5,000 without issue for me. The first block creates a template to match your's, the second block populates some sample data base on your class and the third actually implements your logic. Is there anything drastically different between your code and this? Does this give you an OOM exception?

//Creates our sample template
var templateFile = Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.Desktop), "template.pdf");
using (var fs = new FileStream(templateFile, FileMode.Create, FileAccess.Write, FileShare.None)) {
    using (var doc = new Document()) {
        using (var writer = PdfWriter.GetInstance(doc, fs)) {
            doc.Open();

            var fields = new string[] { "Name", "Description", "DueAmount", "PaidAmout" };
            var y = 700;
            foreach (var f in fields) {
                var tf = new TextField(writer, new iTextSharp.text.Rectangle(100, y, 400, y + 50), f);
                writer.AddAnnotation(tf.GetTextField());
                y -= 200;
            }

            doc.Close();
        }
    }
}

//Create our final PDF
var pages = new List<MyPageData>();
for (decimal i = 0; i < 5000; i++) {
    pages.Add(new MyPageData() { Name = "Chris", Description = "Cheese", DueAmount = i, PaidAmount = i * 0.5m });
}

//Create our file
var outputFile = Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.Desktop), "output.pdf");
using (var fs = new FileStream(outputFile, FileMode.Create, FileAccess.Write, FileShare.None)) {
    DocumentBuilder.FillDocument(pages, templateFile, fs);
}

EDIT

I upped it to 75 fields and although it takes a lot longer I still don't get an OOM. For testing, I changed your MyPageData class to:

public class MyPageData {
    public string[] StringData { get; set; }
}

And your DocumentBuilder 's FillFields to:

private static void FillFields(AcroFields fields, MyPageData page) {
    var fieldCount = 75;
    for (var j = 0; j < fieldCount; j++) {
        fields.SetField("Field_" + j.ToString(), page.StringData[j]);
    }
}

And my template and sample data creation to:

var fieldCount = 75;

//Creates our sample template
var templateFile = Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.Desktop), "template.pdf");
using (var fs = new FileStream(templateFile, FileMode.Create, FileAccess.Write, FileShare.None)) {
    using (var doc = new Document()) {
        using (var writer = PdfWriter.GetInstance(doc, fs)) {
            doc.Open();

            var y = 800;
            //Create 75 fields
            for (var j = 0; j < fieldCount; j++) {
                var tf = new TextField(writer, new iTextSharp.text.Rectangle(100, y, 400, y + 10), "Field_" + j.ToString());
                tf.Options = tf.Options | TextField.MULTILINE;
                writer.AddAnnotation(tf.GetTextField());
                y -= 10;
            }

            doc.Close();
        }
    }
}

//Create our sample data
var pages = new List<MyPageData>();
for (decimal i = 0; i < 5000; i++) {
    var stringData = new List<string>();
    for (var j = 0; j < fieldCount; j++) {
        stringData.Add("Data_" + j.ToString());
    }
    pages.Add(new MyPageData() { StringData = stringData.ToArray() });
}