1The graphemeCluster directory contains files used to modify the default Grapheme Cluster Break (GCB) 2(https://unicode.org/reports/tr29/) algorithm to add support for not splitting Indic aksaras. 3 4The modifications are: 5 61. Adding 3 new character categories to https://unicode.org/reports/tr29/#Grapheme_Cluster_Break_Property_Values 7 8 Virama=[\p{Gujr}\p{sc=Telu}\p{sc=Mlym}\p{sc=Orya}\p{sc=Beng}\p{sc=Deva}&\p{Indic_Syllabic_Category=Virama}] 9 10 LinkingConsonant=[\p{Gujr}\p{sc=Telu}\p{sc=Mlym}\p{sc=Orya}\p{sc=Beng}\p{sc=Deva}&\p{Indic_Syllabic_Category=Consonant}] 11 12 ExtCccZwj=[\p{gcb=Extend}-\p{ccc=0}] \p{gcb=ZWJ}] 13 14Note that these categories are not GCB property values: 15In fact, they overlap the GCB property values. 16It is not necessary for the rules to have disjoint categories. 17The list of scripts can be added to over time, as test files for them become available. 18 192. Adding a rule to https://unicode.org/reports/tr29/#Grapheme_Cluster_Boundary_Rules 20 21 9.3) LinkingConsonant ExtCccZwj* Virama ExtCccZwj* × LinkingConsonant 22 233. Adding test files supplied by India to org.unicode.cldr.unittest.data.graphemeCluster/* 24 25 TestSegmenter-Bengali.txt 26 TestSegmenter-Devanagari.txt 27 TestSegmenter-Gujarati.txt 28 TestSegmenter-Malayalam.txt 29 TestSegmenter-Odia.txt 30 TestSegmenter-Telugu.txt 31 324. Adding modified files in this directory, which can be used in place of the default files from 33 https://unicode.org/Public/12.0.0/ucd/auxiliary/ 34 35 GraphemeBreakTest.html 36 GraphemeBreakTest.txt 37 38Note: The GraphemeBreakProperty.txt file is unmodified, as those properties don't change. 39