要说明这一点，一个名为Document1.pdf文档保存到使用Access UI中的附件栏。该PDF文件的开头是这样的：

如果我们用以下code，试图提取PDF文件保存到磁盘

然后得到的文件将包括在文件的开头的元数据（在此情况下，20字节）

Adob​​e Reader的能够打开这个文件，因为它是强大到足以忽略任何垃圾的％PDF-1.4签署之前，可能会出现在文件中。不幸的是不是所有的文件格式和应用程序是如此宽容外来字节在文件的开头。

在仅官方及贸易;从附件解压文件的方式在Access领域是使用ACE DAO的 .SaveToFile 方法字段2 的对象，像这样：

请注意，如果您尝试使用的Field2对象的。价值你仍然会得到元数据的字节序列的开始;在 .SaveToFile 的过程就是剥离出来。

We are working on a project where we need to migrate data stored in an Access database to a cache database. The Access database contains columns with a data type of Attachment; some of the tuples contain multiple attachments. I am able to obtain the filenames of these files by using .FileName, but I'm unsure how to determine when one file ends and another starts in .FileData.

I am using the following to obtain this data:

System.Data.OleDb.OleDbCommand command= new System.Data.OleDb.OleDbCommand();
command.CommandText = "select [Sheet1].[pdf].FileData,* from [Sheet1]";
command.Connection = conn;
System.Data.OleDb.OleDbDataReader rdr = command.ExecuteReader();

(My original answer to this question was misleading. It worked okay for PDF files that were subsequently opened with Adobe Reader, but it did not always work properly for other types of files. The following is the corrected version.)

Unfortunately we cannot directly retrieve the contents of a file in an Access Attachment field using OleDb. The Access Database Engine prepends some metadata to the binary contents of the file, and that metadata is included if we retrieve the .FileData via OleDb.

To illustrate, a document named "Document1.pdf" is saved to an Attachment field using the Access UI. The beginning of that PDF file looks like this:

If we use the following code to try and extract the PDF file to disk

using (OleDbCommand cmd = new OleDbCommand())
{
    cmd.Connection = con;
    cmd.CommandText = 
            "SELECT Attachments.FileData " +
            "FROM AttachTest " +
            "WHERE Attachments.FileName='Document1.pdf'";
    using (OleDbDataReader rdr = cmd.ExecuteReader())
    {
        rdr.Read();
        byte[] fileData = (byte[])rdr[0];
        using (var fs = new FileStream(
                @"C:UsersGordDesktopFromFileData.pdf", 
                FileMode.Create, FileAccess.Write))
        {
            fs.Write(fileData, 0, fileData.Length);
            fs.Close();
        }
    }
}

then the resulting file will include the metadata at the beginning of the file (20 bytes in this case)

Adobe Reader is able to open this file because it is robust enough to ignore any "junk" that may appear in the file before the '%PDF-1.4' signature. Unfortunately not all file formats and applications are so forgiving of extraneous bytes at the beginning of the file.

The only Official™ way of extracting files from an Attachment field in Access is to use the .SaveToFile method of an ACE DAO Field2 object, like so:

// required COM reference: Microsoft Office 14.0 Access Database Engine Object Library
//
// using Microsoft.Office.Interop.Access.Dao; ...
var dbe = new DBEngine();
Database db = dbe.OpenDatabase(@"C:UsersPublicDatabase1.accdb");
Recordset rstMain = db.OpenRecordset(
        "SELECT Attachments FROM AttachTest WHERE ID=1",
        RecordsetTypeEnum.dbOpenSnapshot);
Recordset2 rstAttach = rstMain.Fields["Attachments"].Value;
while ((!"Document1.pdf".Equals(rstAttach.Fields["FileName"].Value)) && (!rstAttach.EOF))
{
    rstAttach.MoveNext();
}
if (rstAttach.EOF)
{
    Console.WriteLine("Not found.");
}
else
{
    Field2 fld = (Field2)rstAttach.Fields["FileData"];
    fld.SaveToFile(@"C:UsersGordDesktopFromSaveToFile.pdf");
}
db.Close();

Note that if you try to use the .Value of the Field2 object you will still get the metadata at the beginning of the byte sequence; the .SaveToFile process is what strips it out.