Second-Level Reference Label Generation Rules, 13 August 2026 Current

Issued 13 August 2026

Adds Javanese and Unified Canadian Aboriginal Syllabics; updates to Arabic, Japanese and Latin.

Common (integrated set) every writing system in this edition, merged, with the variants that cross between them 33946 code points 14894 variants
Overview and Summary 1.0mb
reference-lgr-2026-08-13.zip complete archive: all 67 files as XML, 1.2mb
SHA-256 96ca40db2b4bb8adda3b83241428135aab2e43ff5868699c18b2a541e01160b8

Script-Based

One ruleset per writing system.

Arabic (script) und-Arab 161 code points 254 variants
Armenian und-Armn 49 code points 2 variants
Balinese und-Bali 66 code points 38 variants
Bengali und-Beng 91 code points 26 variants
Cyrillic und-Cyrl 97 code points 4 variants
Devanagari (script) und-Deva 133 code points 76 variants
Ethiopic und-Ethi 322 code points 98 variants
Georgian und-Geor 44 code points
Greek und-Grek 47 code points 36 variants
Gujarati und-Gujr 86 code points 28 variants
Gurmukhi und-Guru 67 code points 2 variants
Hebrew (script) und-Hebr 38 code points 10 variants
Japanese (script) und-Jpan 6571 code points 2192 variants
Javanese und-Java 78 code points 40 variants
Kannada und-Knda 83 code points 20 variants
Khmer und-Khmr 94 code points 22 variants
Lao und-Laoo 73 code points 20 variants
Latin (script) und-Latn 232 code points 102 variants
Malayalam und-Mlym 91 code points 14 variants
Myanmar (script) und-Mymr 184 code points 50 variants
Oriya und-Orya 73 code points
Sinhala und-Sinh 303 code points 246 variants
Tamil (script) und-Taml 63 code points 6 variants
Telugu (script) und-Telu 74 code points
Thaana und-Thaa 61 code points 42 variants
Thai (script) und-Thai 95 code points 20 variants
Unified Canadian Aboriginal Syllabics und-Cans 316 code points 40 variants

Full Variant Sets

Rulesets spanning a collection of scripts that share variants.

Chinese (script, full variant set) und-Hani 19802 code points 12062 variants
Devanagari (script, full variant set) und-Deva 161 code points 146 variants
Korean (script, full variant set) und-Kore 15970 code points 658 variants
Latin (script, full variant set) und-Latn 302 code points 526 variants
Myanmar (script, full variant set) und-Mymr 192 code points 74 variants
Tamil (script, full variant set) und-Taml 69 code points 18 variants
Telugu (script, full variant set) und-Telu 108 code points 68 variants

Language-Based

One ruleset per language.

Arabic (language) ar 76 code points 140 variants
Belarusian be 47 code points
Bosnian bs 42 code points
Bosnian (Cyrillic) bs-Cyrl 41 code points
Bulgarian bg 43 code points
Chinese (language) zh 19722 code points 11728 variants
Danish da 44 code points
English en 58 code points 10 variants
Finnish fi 49 code points
French fr 54 code points 52 variants
German de 51 code points 2 variants
Hebrew (language) he 38 code points 10 variants
Hindi hi 91 code points 28 variants
Hungarian hu 47 code points
Icelandic is 47 code points
Inuktitut (Unified Canadian Aboriginal Syllabics) iu-Cans 173 code points 22 variants
Italian it 53 code points
Japanese (language) ja 6571 code points 26 variants
Korean (language) ko 11209 code points
Latvian lv 50 code points
Lithuanian lt 46 code points
Macedonian mk 44 code points
Montenegrin (Cyrillic) cnr-Cyrl 43 code points
Norwegian no 64 code points
Polish pl 46 code points
Portuguese pt 50 code points
Russian ru 44 code points
Serbian (Cyrillic) sr-Cyrl 41 code points
Spanish es 56 code points 2 variants
Swedish sv 62 code points
Thai (language) th 95 code points 20 variants
Ukrainian uk 50 code points

← About the Second-Level Reference Label Generation Rules