Conversion of '@' via IBM

[PelleNo]

Hello !

I load a flat file extracted from Unix.
I use a IBM mainframe as utility platform.

The file contains emailadresses and they look OK when I look at the file in the IBM. The '@' is represented by hex 'EC' (EBCDIC).

When I load the file into Teradata the '@' will show up as 'í' in my table.

I realize that this is a conversion problem between EBCDIC and ASCII, but how do I get the '@' to my table.

Best regards
Pelle

 

[dnoeth]

By default Teradata expects EBDIC-encoded data when running a session from a mainframe, so you have to switch to a different character set.
As i don't work with channel-attached systems i can't help you how to modify the character set.
Check the manuals for CHARSET...

Dieter

 

[PelleNo]

Thank you for your fast reply.

Normally I run with the character set 'SWEDISH' wich can handle our regional characters Å,Ä,Ö.(but not @)

I consulted the manuals and the character sets I find are :

EBCDIC
KATANAEBCDIC
KANJIEBCDIC5026_0I
KANJIEBCDIC5035_0I
KANJIEUC_OU
KANJISJIS_0S
ASCII

I set the character set with the JCL-parameter (for example) :
FDLOPT = 'CHARSET=EBCDIC'

If I run with the ASCII parameter I get an abend in the FASTLOAD :

0001 LOGON TDPP/FLOAD_AKTOR_P,

**** 11:03:48 CLI Error 282: .<..................&..+....<..<.
**** 11:03:48 Unable to log on Main SQL Session
**** 11:03:48 FastLoad cannot continue. Exiting.

If i run with the EBCDIC parameter i get even more characters wrong.

Are there any other charactersets that can handle both ÅÄÖ and @ ?

Any other suggestions ?

/Pelle

 

[dnoeth]

You can install your own character sets:
In the &quot;International Character Set Support&quot; manual there's an &quot;Example of Loading the Swedish Character Set Into DBC.CharTranslations&quot; in Chapter 3

Dieter

 

[PelleNo]

OK

I will have my system programmer have a look at the Swedish character set implemeted.

/Pelle