Question

I have a several DBF files generated by a third party that I need to be able to query. I am having trouble because all of the column types have been defined as characters, but the data within some of these fields actually contain binary data. If I try to read these fields using an OleDbDataReader as anything other than a string or character array, I get an InvalidCastException thrown, but I need to be able to read them as a binary value or at least cast/convert them after they are read. The columns that actually DO contain text are being returned as expected.

For example, the very first column is defined as a character field with a length of 2 bytes, but the field contains a 16-bit integer.

I have written the following test code to read the first column and convert it to the appropriate data type, but the value is not coming out right.

The first row of the database has a value of 17365 (0x43D5) in the first column. Running the following code, what I end up getting is 17215 (0x433F). I’m pretty sure it has to do with using the ASCII encoding to get the bytes from the string returned by the data reader, but I’m not sure of another way to get the value into the format that I need, other that to write my own DBF reader and bypass ADO.NET altogether which I don’t want to do unless I absolutely have to. Any help would be greatly appreciated.

        byte[] c0;
        int i0; 

        string con = @"Provider=Microsoft.Jet.OLEDB.4.0;Data Source=C:\ASTM;Extended Properties=dBASE III;User ID=Admin;Password=;";

        using (OleDbConnection c = new OleDbConnection(con))
        {
            c.Open();
            OleDbCommand cmd = c.CreateCommand();
            cmd.CommandText = "SELECT * FROM astm2007";
            OleDbDataReader dr = cmd.ExecuteReader();
            while (dr.Read())
            {
                c0 = Encoding.ASCII.GetBytes(dr.GetValue(0).ToString());

                i0 = BitConverter.ToInt16(c0, 0);
            }
            dr.Dispose();
        }

Answer 1

I am pretty sure that you are correct about the ASCII character conversions. I looked a bit for the supported scalar functions for the Jet engine but was not able to find them … or rather I found scalar functions listed but no syntax. The CONVERT function is probably what you want. Something like:

SELECT CONVERT(twobytefield, SQL_BINARY) from astm2007

Then you could call dr.GetBytes() to read the raw data. However, I was not able to construct a statement using that function that the Jet engine liked.

If you are not able to get the conversion working, another possibility is to use the Advantage .NET Data Provider. Or the OLE DB provider (but the .NET data provider might be a better fit since you are using C#). That provider reads DBF files and supports the CONVERT scalar function. It has a free local engine.

Since you mention you are going to try it and since I tested it to make sure I wasn’t lying, here is the code snippet I used:

AdsConnection conn = new AdsConnection( 
   @"data source=c:\path;chartype=ansi;ServerType=local;TableType=cdx;" );
conn.Open();
AdsCommand cmd = conn.CreateCommand();
cmd.CommandText = "select cast(somefield as sql_binary) from sometable";
cmd.CommandType = CommandType.Text;
AdsExtendedReader rdr = cmd.ExecuteExtendedReader();
rdr.Read();
byte[] c0 = rdr.GetBytes( 0 );
int i0 = BitConverter.ToInt16( c0, 0 );
Console.WriteLine( "val = {0}", i0 );