Charsets and Collations

Explains how EDB2MySQL maps EDB collations to MySQL/MariaDB collations during schema conversion.

EDB Collation Notation

EDB collations are expressed as a base plus optional modifiers:

Component Values Meaning
Base UNI, ANSI Binary (byte-by-byte) ordering
Base PTG, ENU, DEU, FRA, … Locale-specific linguistic ordering
Modifier _CI Case-insensitive
Modifier _AI Accent-insensitive
Modifier _KI Kana-insensitive (no MySQL equivalent — ignored)
Modifier _WI Width-insensitive (no MySQL equivalent — ignored)

Examples: UNI (binary), UNI_CI (case-insensitive), PTG_CI (Portuguese, case-insensitive), ENU_CI_AI (English, case- and accent-insensitive).

EDB silently ignores _AI on UNI and ANSI bases — UNI_AI behaves identically to UNI. EDB2MySQL accounts for this during mapping.

EDB Character Set

The EDB session is either ANSI or Unicode, which determines the collation families used:

Shown in the Session panel on the Source Database tab after opening a session.

How Collation Mapping Works

For each text column, EDB2MySQL looks up the collation map with four keys and picks the first eligible candidate collation. A COLLATE clause is emitted on a column only when its resolved collation differs from the table default.

The default charset and collation (set on the Target Database tab) is used as the fallback when no map entry matches.

How Collation Mapping Works — full explanation of candidate conditions and fallback behaviour.

The Collation Map

Stored in collation_map.json in %ProgramData%\Easygate Software\EDB2MySQL\. Loaded at startup; built-in defaults are used if the file is absent or invalid.

Customise the map via the Edit Collations Map button on the Source Database or Target Database tabs.

See Also

(C) 2026 Easygate, Lda