C# 中的 OfficeOpenXML EPPlus - Excel 文件在写入一定大小的数据后会损坏

Question

I am using the OfficeOpenXML library to export excel data in my .NET MVC Application. However I noticed something funny going on. I am able to download the file but when opening it I am prompted with the following message. If I hit "Yes" all the data is shown without any issues.

I noticed this prompt only shows up after a certain amount of data is written to the excel file. And not a huge amount. If I write 1000 rows of data as shown below the prompt shows, but 10 rows it doesn't show.

Code below:

public static ExcelPackage GenerateExcelFile<T>(ExcelPackage pck)
{
    ExcelWorksheet ws = pck.Workbook.Worksheets.Add("Data");
    DataTable dt = new DataTable();

    dt.Columns.Add(new DataColumn { ColumnName = "A" });
    dt.Columns.Add(new DataColumn { ColumnName = "B" });
    dt.Columns.Add(new DataColumn { ColumnName = "C" });
    dt.Columns.Add(new DataColumn { ColumnName = "D" });
    dt.Columns.Add(new DataColumn { ColumnName = "E" });

    for (int i = 0; i < 1000; i++)
    {
        DataRow row = dt.NewRow();
        row["A"] = "Value";
        row["B"] = "Value";
        row["C"] = "Value";
        row["D"] = "Value";
        row["E"] = "Value";
        dt.Rows.Add(row);
    }
    ws.Cells["A1"].LoadFromDataTable(dt, true);
    return pck;
}

This is taken from the following controller:

public HttpResponseMessage ExportGridToExcelWithFilters ()
{
    using (ExcelPackage pck = new ExcelPackage())
    {
        var excelData = ExcelExportVM.GenerateExcelFile(pck);
        using (MemoryStream ms = new MemoryStream())
        {
            excelData.SaveAs(ms);
            HttpResponseMessage responsePackage = new HttpResponseMessage();
            responsePackage.Content = new ByteArrayContent(ms.GetBuffer());
            responsePackage.Content.Headers.ContentType = new MediaTypeHeaderValue("application/vnd.openxmlformats-officedocument.spreadsheetml.sheet");
            return responsePackage;
        }
    }
}

Changing the above code to produce less data:

for (int i = 0; i < 10; i++) //FROM 1000
{
    DataRow row = dt.NewRow();
    row["A"] = "Value";
    row["B"] = "Value";
    row["C"] = "Value";
    row["D"] = "Value";
    row["E"] = "Value";
    dt.Rows.Add(row);
}  

It now works and I am able to open the file without entering recovery mode.

Answer 1

I've had the same issue, this was caused by memorystream and the bad code :)

When you handle the stream after the file save, I can assume you are trying to send it somewhere, for example HttpResponse, and if you do MemoryStream to byte[] conversion there, that should not be done with GetBuffer() method, call ToArray() instead.