Creating and using legacy code tables

You use the method described here if you do not want to extend the JVM for user-defined character sets as described in Custom Charset Provider but instead the extension is only to apply to BeanConnect.

If the predefined code tables do not satisfy your requirements, you can create code tables of your own. A Java sample source for a user-defined code table is supplied with the product. You will find this example in the JavaDoc for the package net.fsc.beanta.encoding in the Encoding.CustomEncoder class.

To avoid problems with the class loaders when using the classes of your own code tables, you should add the classes to the file BeanConnect.rar.

You select the required code table at deployment time as follows:

• In the configuration property encoding, specify the name of the required code table using custom:<my-code-table>.

• Set the configuration property encodingActive to true to activate it.

You select the required code table at runtime as follows:

• In the Java program activate the required code table with the setEncoding() method of the EISConnection or the OltpMessageContext interface as follows:

  OwnTable myTable = new OwnTable();
  connection.setEncoding(new Encoding.CustomEncoder(myTable));
  connection.setEncodingActive(true);
  

For compatibility reasons, not only the procedure described for using customer-defined code tables is supported, but also the following procedure for utilizing user-defined code tables in which conversion is implemented with two Java programs:

OwnTable_ByteToChar myTableByteToChar = new 
OwnTable_ByteToChar();
OwnTable_CharToByte myTableCharToByte = new 
OwnTable_CharToByte();
setEncoding(new Encoding.Custom(myTableByteToChar,
                                  myTableCharToByte));

These code tables only support 1-byte encoding. This means that the encoding source and destination for one char can only be one byte. If you need 2-byte code tables, we recommend the use of a custom charset which provides all options of code conversion.