store strings in different languages (diff. code pages) in one table field

[tepperis]

hi,

the DB2 v. 7 resides on a Z/OS mainframe.

we want to support different languages with special local characters
for one kind of information.

if someone is using hungarian, the string may contain special hungarian
characters. if polish is used polish characters should be used...

data comes from a web interface.
the output is given back to the intranet, but also to an oracle based
system and to other applications inside the mainframe.

as UNICODE/UTF-8 is not supported by version 7. and different
code pages may probably be in use on the web part
automatic conversions between encoding schemes will not work.
one single code page of the database or table is not sufficient.

the table contains a country code which will satisfy our needs.
we don't need a special language identifier.

what kind of data type is recommended for the field? (currently CHAR)
may be the option 'FOR BIT DATA' helpfull?

tia
michael tepperis

[Artur]

CHAR / VARCHAR FOR BIT DATA might be helpful in your case.

the only difference to char/varchar is that the characters will not be
converted between different code pages when sent them from server to
client, and vice versa . So, you can think that you are operating on
BLOB type, with character "style" interface (in terms on API: inserts,
updates). So, you get what you have inserted.

However you should be aware that it won't work if you are using clients
that are accessing >>the same data<< from different code pages (eg.
unix, windows).
-- Artur Wronski