Re: Japanese character set conversion

On 22/11/2007, at 2:39 AM, Åke Olsson wrote:

We have just started a project for porting an iseries application for use in Japan.

The database, back end and some functions will reside on an iseries.

A lot of the UI will be in Java (Unicode). (Web interface)

One of the users mentioned a problem with conversion from iseries DBCS to Unicode where "some" characters were "lost in translation".

Highly unlikely. This sort of thing usually happens only when CCSIDs that do not support round-trip conversion are used.

The only real "gotcha" with Japanese is that in CCSID 5026 the "normal" lower-case Latin characters are occupied by Katakana characters.

It is possible for a character in one CCSID to not have a counterpart in another CCSID in which case it will be translated into the substitution character (x'1a' I think).

Do you have specific examples of the problem?

I wonder if anyone has knowledge of such a problem. In that case - is anyone aware of a solution or work-around.

One solution would be to store the host data in Unicode thus avoiding any translation.


Regards,
Simon Coulter.
--------------------------------------------------------------------
FlyByNight Software OS/400, i5/OS Technical Specialists

http://www.flybynight.com.au/
Phone: +61 2 6657 8251 Mobile: +61 0411 091 400 /"\
Fax: +61 2 6657 8251 \ /
X
ASCII Ribbon campaign against HTML E-Mail / \
--------------------------------------------------------------------