Question

I have the following problem: I have some Excel-sheets and must export them into DOS-CSV format (for some reasons). The consequence is, that the german umlaute (Ä,Ü,Ö,ä,ü,ö) are not exported correctly. In a next step these CSV-files must be imported into a Winform application. Is it now possible to get back the correkt characters Ä,Ü,Ö,ä,ü,ö during the import?

Answer 1

If you choose the DOS-CSV format, Excel is going to encode the document using the 437 codepage (found that here). You can convert it back to UTF-8 using a little bit of code:

Encoding dosEncoding = Encoding.GetEncoding(437);
string original = String.Empty;

using (StreamReader sr = new StreamReader(@"D:\Path\To\output.csv", dosEncoding))
{
    original = sr.ReadToEnd();
    sr.Close();
}

byte[] encBytes = dosEncoding.GetBytes(original);
byte[] utf8Bytes = Encoding.Convert(dosEncoding, Encoding.UTF8, encBytes);

string converted = Encoding.UTF8.GetString(utf8Bytes);

I tested this by putting Ä,Ü,Ö,ä,ü,ö into a cell and then saving it as a DOS formatted CSV file in Excel. Looking at the document, Excel turned it into Ž,š,™,„,,”.

Running it through the above code turned it back into Ä,Ü,Ö,ä,ü,ö.