changing to US7ASCII

Grille12

unread,

Jun 29, 2011, 6:32:50 AM6/29/11

to

Hello,

I have a rather unusual request.
I need to test an application against a US7ASCCI oracle 11 server.
The server is a 32 bits Oracle 11g 11.2.0.1.0. I have no schema/databases
installed yet so I am not concerned by any data loss.
I do my test on a windows box. Unfortunately when installing oracle, I only
had these charset available (us7ascii is not in there).

WEMSWIN1252
AL32UTF8
AR8ISO8859P6
AR8MSWIN1256
BLT8ISO8859P13
BLT8MSWIN1257
CL8ISO8859P5
CL8MSWIN1251
EE8ISO8859P2
EE8MSWIN1250
EL8ISO8859P7
EL8MSWIN1253
IW8ISO8859P8
IW8MSWIN1255
JA16EUC
JA16EUCTILDE
JA16SJIS
JA16SJISTILDE
KO16MSWIN949
NE8ISO8859P10
NEEISO8859P4
TH8TISASCII
TR8MSWIN1254
VN8MSWIN1258
WE8ISO8859P15
WE8ISO8859P9
WE8MSWIN1252
ZHS16GBK
ZHT16HKSCS
ZHT16MSWIN950
ZHT32EUC

So I choosed WEMSWIN1252 (default one) in the hope to be able to change it
to US7ASCII later. Unfortunately, it does not seem to be possible.
Is there a way to change the charset of an oracle server (sitting on
windows) from WEMSWIN1252 to US7ASCCI?
Or, from which of the above available charset list can I change my server to
US7ASCII?

thanks for your help

Frank van Bortel

unread,

Jun 29, 2011, 9:09:39 AM6/29/11

to

You do realize databases store bits and bytes - not characters?

I fail to see the requirement to test against a US7ASCII database.
There's no such thing as a US7ASCII database. There's no
US7ASCII, for that matter - there's ASCII (as opposed to EBCDIC),
which is a 7 bits code. US7ASCII implies there are other
languages (ES7ASCII, anyone?) which there are none.
It also implies a different number of bits code, which there is
none.

There's something like environment setting which dictate the
use of a certain mapping of one code point to another in order
to mess up even more in the "standards" of computerland.
--

Regards,

Frank van Bortel
http://vanbortel.blogspot.com/2009/04/special-characters-part-i.html

Sybrand Bakker

unread,

Jul 1, 2011, 4:25:03 AM7/1/11

to

On Wed, 29 Jun 2011 12:32:50 +0200, "Grille12" <gril...@gmail.com>
wrote:

Oracle is moving away from single byte charactersets.
They are deprecating charactersets.
You might need to check in the documentation (Globalization Manual)
whether US7ASCII is still supported.

------
Sybrand Bakker
Senior Oracle DBA

Noons

unread,

Jul 1, 2011, 6:17:55 AM7/1/11

to

Grille12 wrote,on my timestamp of 29/06/2011 8:32 PM:

>
> So I choosed WEMSWIN1252 (default one) in the hope to be able to change it
> to US7ASCII later. Unfortunately, it does not seem to be possible.
> Is there a way to change the charset of an oracle server (sitting on
> windows) from WEMSWIN1252 to US7ASCCI?
> Or, from which of the above available charset list can I change my server to
> US7ASCII?

Normally, you do not change the server. You set the server -once!- to an
encoding that can cope with the character sets you need to work with, then you
set the client sessions via the NLS* parameters to the particular character sets
you need to work with.

UTF-8 encoding in the server is a superset of ASCII and will cope with most
other character sets other than the large multi-character/multi-alphabet
oriental ones.

For straight through ASCII, if you use AL32UTF8 in the server and whatever from
your above list at the client, you should be OK. If memory doesn't fail me.

(heck, it's been a while since the days of my dbs with google multi-encoded URLs!)

Here's an example of the different encodings and how to use them:
http://download.oracle.com/docs/cd/B19306_01/server.102/b14225/ch6unicode.htm#i1006826