The character set support in Greenplum Database allows you to store text in a variety of character sets, including single-byte character sets such as the ISO 8859 series and multiple-byte character sets such as EUC (Extended Unix Code), UTF-8, and Mule internal code. All supported character sets can be used transparently by clients, but a few are not supported for use within the server (that is, as a server-side encoding)1. The default character set is selected while initializing your Greenplum Database array using gpinitsystem. It can be overridden when you create a database, so you can have multiple databases each with a different character set.

Parent topic: Greenplum Database Reference Guide

Setting the Character Set

gpinitsystem defines the default character set for a Greenplum Database system by reading the setting of the ENCODING parameter in the gp_init_config file at initialization time. The default character set is UNICODE or UTF8.

You can create a database with a different character set besides what is used as the system-wide default. For example:

=> CREATE DATABASE korean WITH ENCODING 'EUC_KR';

Important

Although you can specify any encoding you want for a database, it is unwise to choose an encoding that is not what is expected by the locale you have selected. The LC_COLLATE and LC_CTYPE settings imply a particular encoding, and locale-dependent operations (such as sorting) are likely to misinterpret data that is in an incompatible encoding.

Since these locale settings are frozen by gpinitsystem, the apparent flexibility to use different encodings in different databases is more theoretical than real.

One way to use multiple encodings safely is to set the locale to C or POSIX during initialization time, thus deactivating any real locale awareness.

Character Set Conversion Between Server and Client

Greenplum Database supports automatic character set conversion between server and client for certain character set combinations. The conversion information is stored in the master pg_conversion system catalog table. Greenplum Database comes with some predefined conversions or you can create a new conversion using the SQL command CREATE CONVERSION.

To enable automatic character set conversion, you have to tell Greenplum Database the character set (encoding) you would like to use in the client. There are several ways to accomplish this:

• Using the \encoding command in psql, which allows you to change client encoding on the fly.

• Using SET client_encoding TO.

  To set the client encoding, use the following SQL command:

  => SET CLIENT_ENCODING TO '<value>';

  To query the current client encoding:

  => SHOW client_encoding;

  To return to the default encoding:

  => RESET client_encoding;

• Using the PGCLIENTENCODING environment variable. When PGCLIENTENCODING is defined in the client’s environment, that client encoding is automatically selected when a connection to the server is made. (This can subsequently be overridden using any of the other methods mentioned above.)

• Setting the configuration parameter client_encoding. If client_encoding is set in the master postgresql.conffile, that client encoding is automatically selected when a connection to Greenplum Database is made. (This can subsequently be overridden using any of the other methods mentioned above.)

If the conversion of a particular character is not possible (suppose you chose EUC_JP for the server and LATIN1 for the client, then some Japanese characters do not have a representation in LATIN1) then an error is reported.

If the client character set is defined as SQL_ASCII, encoding conversion is deactivated, regardless of the server’s character set. The use of SQL_ASCII is unwise unless you are working with all-ASCII data. SQL_ASCII is not supported as a server encoding.

1Not all APIs support all the listed character sets. For example, the JDBC driver does not support MULE_INTERNAL, LATIN6, LATIN8, and LATIN10.

2The SQL_ASCII setting behaves considerably differently from the other settings. Byte values 0-127 are interpreted according to the ASCII standard, while byte values 128-255 are taken as uninterpreted characters. If you are working with any non-ASCII data, it is unwise to use the SQL_ASCII setting as a client encoding. SQL_ASCII is not supported as a server encoding.