Multibyte characters must not be used in database names or user names, because certain conditions may apply or it may not be possible to connect to some clients.
Related notes and constraints are described below.
1) Configuring the client encoding system
The client encoding system must be configured when the names are created.
See
Refer to "Character Set Support" in "Server Administration" in the PostgreSQL Documentation for information on how to configure the client encoding system.
2) Encoding system of names used for connection
Ensure that the encoding system of names used for connection is the same as that of the database that was connected when these names were created.
The reasons for this are as follows:
Storage system for names in Fujitsu Enterprise Postgres
The system catalog saves encoded names by using the encoding system of the database at the time the names were created.
Encoding conversion policy when connected
When connected, names sent from the client are matched with names in the system catalog without performing encoding conversion.
Accordingly, if the database that was connected when the names were defined uses the EUC_JP encoding system, but the database name is specified using UTF-8 encoding, then the database will be considered to be non-existent.
3) Connection constraints
The table below shows the connection constraints for each client type, based on the following assumptions:
The conditions described in 1) and 2) above are satisfied.
The database name and user names use the same encoding system.
Client type | Client operating system |
|---|---|
JDBC driver | Cannot be connected |
ODBC driver | No connection constraints |
SQLEmbedded SQL in C | No connection constraints |
psql command | No connection constraints |
