SNA Server Support for ASCII-EBCDIC Character Conversion

MORE INFORMATION

1. Determine conversion requirements
   
   The CONVERT API call supports ASCII-to-EBCDIC and EBCDIC-to-ASCII conversion of characters. The following character set (char_set) table options are supported by the CONVERT call:

         char_set   Description
         --------   -----------------------------------------------------
         SV_A      Only the following characters are converted:
   
                      $ # @ A-Z 0-9
   
         SV_AE     Only the following characters are converted:
   
                      . $ # @ a-z A-Z 0-9
   
         SV_G      SNA Server uses a user-defined type G conversion
                   table for character conversion. See below for more
                   information about type G conversion table.

   See Appendix A of the IBM SNA Formats Guide (IBM doc #GA27-3136) for a description of the A and AE character sets. For SV_A and SV_AE conversions, trailing blanks (that is, blanks at the end of the source string) are converted to blanks in either direction. In contrast, embedded blanks are converted to 0x00 (NULL). If any source character is converted to 0x00, the CONVERSION_ERROR secondary return_code is returned, though the conversion will still be completed.
   
   Type G conversion table
   -----------------------
   The Type G conversion table is user-defined. SNA Server ships with a default type G conversion table called COMTBLG.DAT. This conversion table is automatically installed on SNA Server Windows NT and SNA Server Windows 95 client machines, to <snaroot>\SYSTEM\COMTBLG.DAT. Here is how to enable the this type G conversion table on the other client platforms:

   On Windows 3.x, copy the COMTBLG.DAT file to the client machine and add a COMTBLG entry to the [wnap] section of WIN.INI to point to this file. For example:

   COMTBLG=C:\SNA.WIN\COMTBLG.DAT

   The COMTBLG.DAT conversion table is composed of 32 lines of 32 characters each. Each line represents 16 printable hexadecimal characters followed by a carriage return and line feed. The first 16 lines provide ASCII-to-EBCDIC conversion information while the second 16 lines provide EBCDIC-to-ASCII conversion information.

   When SNA Server performs the conversion, it uses the numeric equivalent of each incoming character as a 0-origin index into the conversion table. This index specifies the table location containing the hexadecimal value of the converted character. For example, assume that the 32nd position in the table contained 0x40. SNA Server would convert an incoming character with a value of 32 to a value of 64 (0x40).
2. If the above conversion tables are insufficient, a custom conversion table must be generated.
   
   The GET_CP_CONVERT_TABLE function supports the generation of a custom conversion table which can then be used to create a custom conversion table to replace the COMTBLG.DAT file shipped with SNA Server. The application programmer passes the desired source code page (source_cp) and target code page (target_cp), and receives the desired 256-byte conversion table. These conversion tables can be concatenated together to form a custom <snaroot>\system\COMTBLG.DAT file to replace the version included with SNA Server. This conversion table can then be used when calling the CONVERT function when the SV_G conversion table option is specified. The GET_CP_CONVERT_TABLE function supports the following source and target code pages:

            ASCII code pages (also referred to as PC Character Sets)
            --------------------------------------------------------
   
      437 - English.  US IBM PC. Includes characters for English and most
            other European languages. Countries using this code page include
            those from Latin America, the United States and Australia.
   
      850 - Multilingual (Latin I).  Includes characters for most of the
            languages that use the Latin alphabet.  This includes Belgium,
            Canada(French), Denmark, Finland, France, Germany, Italy,
            Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, and
            United Kingdom.  Latin America, the United States and Australia
            may also use this code page.
   
      860 - Portuguese.  Includes specific character support for the
            Portuguese of Portugal.
   
      863 - Canadian-French.  Includes characters for English and
            Canadian-French
   
      865 - Slavic/Nordic (Latin II). Includes characters for the Slavic
            languages that use the Latin alphabet.</WWITEM></WWSTEPS>

           EBCDIC code pages (also referred to as host code pages)
           -------------------------------------------------------
   
      037 - United States, Australia, Belgium, Canada(English and,
            French), Netherlands, Portugal
   
      273 - Germany
   
      277 - Denmark, Norway
   
      278 - Sweden, Finland
   
      280 - Italy
   
      284 - Latin America, Spain
   
      285 - United Kingdom
   
      297 - France
   
      500 - International, Switzerland (German and French)

   NOTE: User defined code pages range from 65280 - 65534
3. If the GET_CP_CONVERT_TABLE is not sufficient: