Code-Set-Conversion Files

The code-set-conversion file describes how to map each character in a particular source code set to the characters of a particular target code set. IBM Informix products can perform a given code-set conversion if code-set-conversion files exist to describe the mapping between the two code sets.

When an IBM Informix product needs to obtain code-set-conversion information, it accesses one of the GLS code-set-conversion files in the following table.

In these paths, INFORMIXDIR is the environment variable that specifies the directory in which you install the IBM Informix product, gls is the subdirectory that contains the GLS files, and Y represents the version number for the code-set-conversion object-file format.

This rest of this section describes the remaining elements in the pathname of GLS code-set-conversion files.

Code-Set-Conversion Source and Object Files

Each code-set-conversion file has the following two forms:

• The code-set-conversion source file is an ASCII file that describes the mapping to use for one direction of the code-set conversion.

  This file has a .cv extension and serves as documentation for the corresponding object file.

• The code-set-conversion object file is a compiled form of the code-set-conversion information.

  IBM Informix products use the object file to obtain code-set-conversion information quickly. Object code-set-conversion files have a .cvo file extension.

The header of the code-set-conversion source file (.cv) lists the two code sets that it converts and the direction of the conversion.

Code-Set-Conversion Filenames

To conform to DOS 8.3 naming conventions, GLS code-set-conversion files use a condensed form of the code-set names, code1code2, in their filenames. The 8-character name of each code-set-conversion file is derived from the hexadecimal representation of the code-set numbers of the source code set (code1) and the target code set (code2).

For example, the ISO8859-1 code set has an IBM CCSID number of 819 in decimal and 0333 in hexadecimal. The IBM CCSID 437 code set, a common IBM UNIX code set, has a hexadecimal value of 01b5. Therefore, the 033301b5.cv code-set-conversion file describes the conversion from the CCSID 819 code set to the CCSID 437 code set.

Required for Code-Set Conversion

IBM Informix products use the Code-Set Name-Mapping file to translate between code-set names and the more compact code-set numbers. You can use the registry file to find the hexadecimal values that correspond to code-set names or code-set numbers.

Most code-set conversion requires two code-set-conversion files. One file supports conversion of characters in code set A to their counterparts in code set B. Another supports the conversion in the return direction (from B to A). Such conversions are called two-way code-set conversions. For example, the code-set conversion between the CCSID 437 code set (hexadecimal 01b5 code number) and the CCSID 819 code set (or ISO8859-1 with a hexademical 0333 code number) requires the following two code-set-conversion files:

• The 01b50333.cv file describes the mappings to use when IBM Informix products convert characters in the CCSID 437 code set to those in the ISO8859-1 code set.
• The 033301b5.cv file describes the mappings to use when IBM Informix products convert characters in the ISO8859-1 code set to those in the CCSID 437 code set.

To be able to convert between these two code sets, an IBM Informix product must be able to locate both these code-set-conversion object files. Performing the conversion on only one direction would result in mismatched characters. For more information on mismatched characters, see Performing Code-Set Conversion.

The following table shows some of the code-set conversions that IBM Informix products can support, along with their associated code-set-conversion source filenames.