Archives of the TeradataForum
Message Posted: Mon, 06 Oct 2008 @ 18:04:29 GMT
| Subj: || || Re: Unicode: Load and Display International |
| From: || || Fred W Pluebell |
No. Though it may not seem to be so, the problem is likely on your mainframe side. The EBCDIC to internal (Latin or Unicode) translation you
are using is "reversible" (so the data appears correct when you export back as EBCDIC). But it's not "correct" as far as the internal code
assigned to those characters (so translation to a different character set does not work as intended).
The International Character Set Support manual includes several cautions about use of accented and non-Latin characters with the default ASCII
or EBCDIC client character sets. You may be able to use the Teradata-supplied EBCDIC037_0E installable character set instead, or you may need to
define one of your own that maps each accented character to the correct Latin/Unicode code point.
Note that the data already loaded is "wrong" internally so will need to be corrected. For example, you may be able to export the data using the
default EBCDIC character set, then re-load it (or apply updates) using the new character set.