CPACKAGE, CSERIESTOTEXT, and CTEXT function usage

Resolving The Problem

In a situation where an input file and type tree use the Latin5 character set, and the output tree is set to UTF-8 with the objective to convert the input from Latin5 to UTF-8, the following syntax has been incorrectly used:

=CPACKAGE(input,"UTF-8")

This syntax means whatever the input used (latin5 , native or other), the CPACKAGE function will evaluate considering 'input' is UTF-8.

The result will be incorrect.

Solution

If you need to CPACKAGE input from Latin 5 you must use the correct character set:

CPACKAGE(input,"ibm-920_P100-1995")
Note : "ibm" character set are only available sicne version 8.1.x

As your output tree is set to UTF-8, TX will CPACKAGE your input considering it as Latin-5 ( ibm-920_P100-1995 ) and will convert it to UTF-8 when building the output.


Why "ibm-920_P100-1995"

Syntax:

CPACKAGE (single-object-expression , character-set-of-object-content )

In Websphere transformation extender 8.1.x If you search in the help Character setand CPACKAGEyou will find the Character set codes for CPACKAGE, CSERIESTOTEXT, and CTEXT.

This will show you that the Latin5 Data language has "ibm-920_P100-1995" as the Character set code