Neoview Character Sets Administrator's Guide (R2.3)

Table 1-3 Features, Behaviors, and Limitations of the Neoview Character Set Configurations
LimitationsFeatures and BehaviorsConfiguration
For this release, the ISO88591 configuration has
these limitations and restrictions:
• To ensure the correct representation of all
character data sent to UCS2 columns:
— All character data sent to UCS2 columns in
SQL statement character string literals must
be encoded and stored as ISO8859-1
characters.
— All character data for UCS2 columns that is
bound as parameters is encoded and stored
as UCS2 characters.
• If they want to ensure compatible client locales,
users are required to use only 7-bit ASCII or
ISO8859-1 characters for all SQL identifiers and
user character data. For more information about
compatible client locales, see the troubleshooting
information in Table 5-1 (page 35).
• SQL EMS messages are displayed correctly only
if they are encoded in ISO8859-1. NDCS EMS
messages are displayed correctly only if they are
encoded in 7-bit ASCII.
• SQL string functions operate only on single-byte
characters.
• Neoview DB Admin displays only table and
column names that are encoded in ISO8859-1.
• Neoview Management Dashboard displays only
ASCII-formatted characters.
The ISO88591 configuration replicates the
Neoview character set environment for Release
2.2. It allows users to store data encoded in any
character set—including ISO8859-1 through
ISO8859-15 and East Asian multibyte character
sets—in ISO88591 columns.
• The Neoview database stores and retrieves
all client locale character-encoded table
names, column names, and character literals
as if they were encoded in 8-bit ISO8859-1
characters.
• The default column character set definition
is ISO88591.
• Uses binary collation.
• All error messages are sent in the client locale
character encoding.
• The Neoview database assumes that all EMS
event messages are in UTF8 format.
• The ISO88591 configuration supports the use
of Release 2.2 or Release 2.3 Neoview ODBC
and Neoview JDBC drivers. For more
information, see “Compatibility Between
Neoview ODBC and JDBC Drivers and the
Neoview Database” (page 18).
ISO88591
For this release, the SJIS configuration has these
limitations and restrictions:
• To ensure the correct representation of all
character data sent to UCS2 columns:
— All character data sent to UCS2 columns in
SQL statement character string literals must
be encoded and stored as SJIS characters.
— All character data for UCS2 columns that is
bound as parameters is encoded and stored
as UCS2 characters.
• SQL string functions do not support SJIS
characters in ISO88591 columns. They assume
each byte is a separate character.
• When you use TMFCOM and FUP to reference
SQL tables, use Guardian names instead of ANSI
names in SQL identifiers, particularly when
using multibyte characters.
• Table names, column names, and character
literals are stored in ISO88591 columns as
SJIS characters using the Microsoft codepage
932 (MS932).
• Requires Release 2.3 Neoview ODBC and
Neoview JDBC drivers. If you connect a
Release 2.2 driver to a Release 2.3 Neoview
platform using the SJIS configuration, the
connection fails and a connection error is
generated. For more information, see
“Compatibility Between Neoview ODBC and
JDBC Drivers and the Neoview Database”
(page 18).
• Compatible data from EUC-JP or UTF8
character sets is translated to SJIS.
• The default column character set definition
is ISO88591.
• The size of an ISO88591 column indicates the
number of bytes in the column, regardless of
whether or not the column contains SJIS
characters.
• Uses binary collation, which for SJIS
characters is also JIS collation.
• All EMS messages are sent in UTF8 format.
SJIS
Unicode
Neoview Character Set Configurations 17