Implementing Character Encoding in a Pipeline Component

To support custom character encoding, you must implement a custom encoding class by deriving from the Microsoft .NET Framework Encoding class, then create a custom flat file pipeline component by inheriting from the standard Flat File Disassembler or Flat File Assembler component. You can supply a new encoding instance to the parsing engine by overriding the protected virtual method FFDasmComp.GetDataReader as shown in the following example.

/// <summary>  
/// Gets a data reader instance  
/// </summary>  
/// <param name="dataStream">Data stream</param>  
/// <param name="dataEncoding">Data encoding</param>  
/// <param name="detectEncodingFromByteOrderMarks">Detect encoding from a byte order mark</param>  
/// <returns>IDataReader instance</returns>  
      protected override IDataReader GetDataReader(Stream dataStream, Encoding dataEncoding, bool detectEncodingFromByteOrderMarks)  
      {  
         // Delegate call to the base implementation passing fixed UTF-7 encoding  
         return base.GetDataReader(dataStream, new CustomEncoding(), false);  
      }  

Using predefined encoding classes

The following encoding types are predefined by the Microsoft .NET Framework and can be used to construct the parser:

• ASCII

• UTF7

• UTF8

• Unicode (UTF16)

XmlReader xr = docspec.Parse(new DataReader(System.Text.Encoding.UTF8));  

Using supported code pages

Use the following code to support Shift-JIS (codepage 932).

XmlReader xr = docspec.Parse(new DataReader(System.Text.Encoding.GetEncoding(932)));  

Using a private encoding class

You can create your own encoding class that derives from the System.Text.Encoding abstract class and perform your own encoding and decoding.

class MyEncoding : System.Text.Encoding  
{  
   // overriding methods omitted  
}  
  
XmlReader xr = docspec.Parser(new DataReader(new MyEncoding()));  

Using a private DataReader class

You can create your own DataReader class that implements the IDataReader interface and performs reading without creating any encoding classes.

class MyDataReader : IDataReader  
{  
   // Implement data reader functions  
   // ...  
}  
  
XmlReader xr = docspec.Parse(new MyDataReader());  

See Also

Using the Parsing and Serializing Engines