Changes in Oracle AI Database Globalization Support Guide for Oracle AI Database 26ai

The following are changes in Oracle AI Database Globalization Support Guide for Oracle AI Database 26ai.

New Features

• New Features in Oracle AI Database 26ai, Release Update 23.26.3

• New Features in Oracle AI Database 26ai, Release Update 23.26.2

• New Features in Oracle AI Database 26ai

New Features in Oracle AI Database 26ai, Release Update 23.26.3

• Support for Unicode 17.0, keeping Oracle Database current with the latest Unicode Standard and expanding multilingual coverage for newly encoded scripts, characters, and symbols.

  Unicode 17.0 includes a new currency symbol, the SAUDI RIYAL SIGN (Unicode code point U+20C1), that the Saudi Central Bank introduced in 2025 to represent the currency of Saudi Arabia. The new symbol can be used by configuring the NLS_CURRENCY parameter through the client environment, database session, application-specific logon configuration, or individual SQL formatting functions. To display the symbol correctly, the client character set and application font must support Unicode code point U+20C1.

  Saudi Riyal Symbol

  
  Description of the illustration saudi_riyal_symbol.png

  See "Unicode Support".

New Features in Oracle AI Database 26ai, Release Update 23.26.2

• A set of new EBCDIC client-only character sets whose conversion to Unicode strictly follows mappings defined by the IBM Character Data Representation Architecture (CDRA).

  See "Choosing a Source Character Set for Data Migration".

• A set of collations that emulate the binary ordering of the new EBCDIC client-only character sets.

  See "Emulating EBCDIC Binary Ordering".

New Features in Oracle AI Database 26ai

• Support for Unicode 15.0, a major version of the Unicode Standard that supersedes all its previous versions.

  See "Unicode Support".

• Enhanced Ideographic Variation Sequence (IVS) support:

  • A new UCA collation UCA1210_JAPANESE_IVS provides IVS support when processing Japanese text.

    See "The UCA Japanese IVS Collation".

  • String operations that use Unicode composed characters now count an IVS as one complete character. These operations include the SQL functions INSTRC, LENGTHC, and SUBSTRC, and the SQL condition LIKEC.

    See "SQL Functions for Different Length Semantics" and LIKE Conditions for Different Length Semantics.

• A new territory TURKIYE, which supports the official name of the country of Türkiye.

  See "Territories".

• A new linguistic sort ICELANDIC_IST, which implements the Icelandic Standard IST 130:2004.

  See "Linguistic Collations".

• The implementation of the Oracle client character set ZHS32GB18030 is updated to support the latest GB18030-2022 standard.

  See "Client-Only Character Sets".

• The UTL_I18N package provides three new procedures that detect the most likely Oracle character sets and Oracle languages for a given data sample.

  See DETECT_CHARSET, DETECT_LANGUAGE, and DETECT_LANGUAGE_CHARSET in Oracle AI Database PL/SQL Packages and Types Reference.

• The process of upgrading time zone data using the DBMS_DST package has been enhanced as follows:

  • The upgrade is optimized to take the actual data contents of a table into account. Only tables impacted by a daylight saving time rule change will undergo a data change. See "Upgrade Window".

  • You can monitor the actions performed on tables during the upgrade process by viewing the upgrade history table. See "About the Upgrade History Table".

• The new TIME_AT_DBTIMEZONE initialization parameter allows you to specify whether time-dependent database operations use the time zone of the database host system or the time zone of the PDB.

  See Oracle AI Database Reference.