X-Git-Url: https://git.saurik.com/apple/icu.git/blobdiff_plain/b75a7d8f3b4adbae880cab104ce2c6a50eee4db2..6de34a005df0e4a02d88750703b950237b77d9c8:/icuSources/layout/IndicClassTables.cpp diff --git a/icuSources/layout/IndicClassTables.cpp b/icuSources/layout/IndicClassTables.cpp index 0800ffd2..46e1b782 100644 --- a/icuSources/layout/IndicClassTables.cpp +++ b/icuSources/layout/IndicClassTables.cpp @@ -1,7 +1,6 @@ /* - * %W% %E% * - * (C) Copyright IBM Corp. 1998-2003 - All Rights Reserved + * (C) Copyright IBM Corp. 1998-2004 - All Rights Reserved * */ @@ -30,18 +29,26 @@ enum // simple classes _xx = IndicClassTable::CC_RESERVED, - _ma = IndicClassTable::CC_MODIFYING_MARK_ABOVE, - _mp = IndicClassTable::CC_MODIFYING_MARK_POST, + _ma = IndicClassTable::CC_VOWEL_MODIFIER | IndicClassTable::CF_POS_ABOVE, + _mp = IndicClassTable::CC_VOWEL_MODIFIER | IndicClassTable::CF_POS_AFTER, + _sa = IndicClassTable::CC_STRESS_MARK | IndicClassTable::CF_POS_ABOVE, + _sb = IndicClassTable::CC_STRESS_MARK | IndicClassTable::CF_POS_BELOW, _iv = IndicClassTable::CC_INDEPENDENT_VOWEL, + _i2 = IndicClassTable::CC_INDEPENDENT_VOWEL_2, _ct = IndicClassTable::CC_CONSONANT | IndicClassTable::CF_CONSONANT, _cn = IndicClassTable::CC_CONSONANT_WITH_NUKTA | IndicClassTable::CF_CONSONANT, _nu = IndicClassTable::CC_NUKTA, _dv = IndicClassTable::CC_DEPENDENT_VOWEL, - _dl = _dv | IndicClassTable::CF_MATRA_PRE, - _db = _dv | IndicClassTable::CF_MATRA_BELOW, - _da = _dv | IndicClassTable::CF_MATRA_ABOVE, - _dr = _dv | IndicClassTable::CF_MATRA_POST, + _dl = _dv | IndicClassTable::CF_POS_BEFORE, + _db = _dv | IndicClassTable::CF_POS_BELOW, + _da = _dv | IndicClassTable::CF_POS_ABOVE, + _dr = _dv | IndicClassTable::CF_POS_AFTER, _lm = _dv | IndicClassTable::CF_LENGTH_MARK, + _l1 = IndicClassTable::CC_SPLIT_VOWEL_PIECE_1 | IndicClassTable::CF_POS_BEFORE, + _a1 = IndicClassTable::CC_SPLIT_VOWEL_PIECE_1 | IndicClassTable::CF_POS_ABOVE, + _r2 = IndicClassTable::CC_SPLIT_VOWEL_PIECE_2 | IndicClassTable::CF_POS_AFTER, + _m2 = IndicClassTable::CC_SPLIT_VOWEL_PIECE_2 | IndicClassTable::CF_LENGTH_MARK, + _m3 = IndicClassTable::CC_SPLIT_VOWEL_PIECE_3 | IndicClassTable::CF_LENGTH_MARK, _vr = IndicClassTable::CC_VIRAMA, // split matras @@ -71,31 +78,31 @@ enum // static const IndicClassTable::CharClass devaCharClasses[] = { - _xx, _ma, _ma, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, // 0900 - 090F + _xx, _ma, _ma, _mp, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, // 0900 - 090F _iv, _iv, _iv, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, // 0910 - 091F _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _cn, _ct, _ct, _ct, _ct, _ct, _ct, // 0920 - 092F - _rv, _cn, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _xx, _xx, _nu, _xx, _dr, _dl, // 0930 - 093F + _rv, _cn, _ct, _ct, _cn, _ct, _ct, _ct, _ct, _ct, _xx, _xx, _nu, _xx, _dr, _dl, // 0930 - 093F _dr, _db, _db, _db, _db, _da, _da, _da, _da, _dr, _dr, _dr, _dr, _vr, _xx, _xx, // 0940 - 094F - _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _cn, _cn, _cn, _cn, _cn, _cn, _cn, _cn, // 0950 - 095F + _xx, _sa, _sb, _sa, _sa, _xx, _xx, _xx, _cn, _cn, _cn, _cn, _cn, _cn, _cn, _cn, // 0950 - 095F _iv, _iv, _db, _db, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0960 - 096F _xx // 0970 }; static const IndicClassTable::CharClass bengCharClasses[] = { - _xx, _ma, _mp, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _xx, _iv, // 0980 - 098F + _xx, _ma, _mp, _mp, _xx, _i2, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _xx, _i2, // 0980 - 098F _iv, _xx, _xx, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, // 0990 - 099F _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _xx, _ct, _ct, _bb, _ct, _ct, _pb, // 09A0 - 09AF - _rv, _xx, _ct, _xx, _xx, _xx, _ct, _ct, _ct, _ct, _xx, _xx, _nu, _xx, _dr, _dl, // 09B0 - 09BF - _dr, _db, _db, _db, _db, _xx, _xx, _dl, _dl, _xx, _xx, _s1, _s2, _vr, _xx, _xx, // 09C0 - 09CF - _xx, _xx, _xx, _xx, _xx, _xx, _xx, _dr, _xx, _xx, _xx, _xx, _cn, _cn, _xx, _cn, // 09D0 - 09DF + _rv, _xx, _ct, _xx, _xx, _xx, _ct, _ct, _ct, _ct, _xx, _xx, _nu, _xx, _r2, _dl, // 09B0 - 09BF + _dr, _db, _db, _db, _db, _xx, _xx, _l1, _dl, _xx, _xx, _s1, _s2, _vr, _xx, _xx, // 09C0 - 09CF + _xx, _xx, _xx, _xx, _xx, _xx, _xx, _m2, _xx, _xx, _xx, _xx, _cn, _cn, _xx, _cn, // 09D0 - 09DF _iv, _iv, _dv, _dv, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 09E0 - 09EF _ct, _ct, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx // 09F0 - 09FA }; static const IndicClassTable::CharClass punjCharClasses[] = { - _xx, _xx, _ma, _xx, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _xx, _xx, _xx, _iv, // 0A00 - 0A0F + _xx, _ma, _ma, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _xx, _xx, _xx, _iv, // 0A00 - 0A0F _iv, _xx, _xx, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, // 0A10 - 0A1F _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _xx, _ct, _ct, _ct, _ct, _ct, _bb, // 0A20 - 0A2F _vt, _xx, _ct, _ct, _xx, _bb, _ct, _xx, _ct, _bb, _xx, _xx, _nu, _xx, _dr, _dl, // 0A30 - 0A3F @@ -121,9 +128,9 @@ static const IndicClassTable::CharClass oryaCharClasses[] = _xx, _ma, _ma, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _xx, _iv, // 0B00 - 0B0F _iv, _xx, _xx, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, // 0B10 - 0B1F _ct, _ct, _ct, _ct, _bb, _ct, _ct, _ct, _bb, _xx, _ct, _ct, _bb, _bb, _bb, _pb, // 0B20 - 0B2F - _rv, _xx, _bb, _bb, _xx, _xx, _ct, _ct, _ct, _ct, _xx, _xx, _nu, _xx, _dr, _da, // 0B30 - 0B3F - _dr, _db, _db, _db, _xx, _xx, _xx, _dl, _s1, _xx, _xx, _s2, _s3, _vr, _xx, _xx, // 0B40 - 0B4F - _xx, _xx, _xx, _xx, _xx, _xx, _da, _dr, _xx, _xx, _xx, _xx, _cn, _cn, _xx, _cn, // 0B50 - 0B5F + _rv, _xx, _bb, _bb, _xx, _xx, _ct, _ct, _ct, _ct, _xx, _xx, _nu, _xx, _r2, _da, // 0B30 - 0B3F + _dr, _db, _db, _db, _xx, _xx, _xx, _l1, _s1, _xx, _xx, _s2, _s3, _vr, _xx, _xx, // 0B40 - 0B4F + _xx, _xx, _xx, _xx, _xx, _xx, _m2, _m2, _xx, _xx, _xx, _xx, _cn, _cn, _xx, _cn, // 0B50 - 0B5F _iv, _iv, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0B60 - 0B6F _xx // 0B70 }; @@ -133,9 +140,9 @@ static const IndicClassTable::CharClass tamlCharClasses[] = _xx, _xx, _ma, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _xx, _xx, _iv, _iv, // 0B80 - 0B8F _iv, _xx, _iv, _iv, _iv, _ct, _xx, _xx, _xx, _ct, _ct, _xx, _ct, _xx, _ct, _ct, // 0B90 - 0B9F _xx, _xx, _xx, _ct, _ct, _xx, _xx, _xx, _ct, _ct, _ct, _xx, _xx, _xx, _ct, _ct, // 0BA0 - 0BAF - _ct, _ct, _ct, _ct, _ct, _ct, _xx, _ct, _ct, _ct, _xx, _xx, _xx, _xx, _dr, _dr, // 0BB0 - 0BBF - _da, _dr, _dr, _xx, _xx, _xx, _dl, _dl, _dl, _xx, _s1, _s2, _s3, _vr, _xx, _xx, // 0BC0 - 0BCF - _xx, _xx, _xx, _xx, _xx, _xx, _xx, _dr, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0BD0 - 0BDF + _ct, _ct, _ct, _ct, _ct, _ct, _xx, _ct, _ct, _ct, _xx, _xx, _xx, _xx, _r2, _dr, // 0BB0 - 0BBF + _da, _dr, _dr, _xx, _xx, _xx, _l1, _l1, _dl, _xx, _s1, _s2, _s3, _vr, _xx, _xx, // 0BC0 - 0BCF + _xx, _xx, _xx, _xx, _xx, _xx, _xx, _m2, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0BD0 - 0BDF _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0BE0 - 0BEF _xx, _xx, _xx // 0BF0 - 0BF2 }; @@ -147,20 +154,25 @@ static const IndicClassTable::CharClass teluCharClasses[] = _iv, _xx, _iv, _iv, _iv, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, // 0C10 - 0C1F _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _xx, _bb, _bb, _bb, _bb, _bb, _bb, // 0C20 - 0C2F _bb, _ct, _bb, _bb, _xx, _bb, _bb, _bb, _bb, _bb, _xx, _xx, _xx, _xx, _da, _da, // 0C30 - 0C3F - _da, _dr, _dr, _dr, _dr, _xx, _da, _da, _s1, _xx, _da, _da, _da, _vr, _xx, _xx, // 0C40 - 0C4F - _xx, _xx, _xx, _xx, _xx, _da, _db, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0C50 - 0C5F + _da, _dr, _dr, _dr, _dr, _xx, _a1, _da, _s1, _xx, _da, _da, _da, _vr, _xx, _xx, // 0C40 - 0C4F + _xx, _xx, _xx, _xx, _xx, _da, _m2, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0C50 - 0C5F _iv, _iv, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx // 0C60 - 0C6F }; -// FIXME: is 0CD5 a dr or an lm?? +// U+CC3 and U+CC4 are _lm here not _dr since the Kannada rendering +// rules want them below and to the right of the entire cluster +// +// There's some information about this in: +// +// http://brahmi.sourceforge.net/docs/KannadaComputing.html static const IndicClassTable::CharClass kndaCharClasses[] = { _xx, _xx, _mp, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _iv, // 0C80 - 0C8F _iv, _xx, _iv, _iv, _iv, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, // 0C90 - 0C9F _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _xx, _bb, _bb, _bb, _bb, _bb, _bb, // 0CA0 - 0CAF _rb, _ct, _bb, _bb, _xx, _bb, _bb, _bb, _bb, _bb, _xx, _xx, _xx, _xx, _dr, _da, // 0CB0 - 0CBF - _s1, _dr, _dr, _dr, _dr, _xx, _da, _s2, _s3, _xx, _s4, _s5, _da, _vr, _xx, _xx, // 0CC0 - 0CCF - _xx, _xx, _xx, _xx, _xx, _lm, _dr, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _ct, _xx, // 0CD0 - 0CDF + _s1, _dr, _r2, _lm, _lm, _xx, _a1, _s2, _s3, _xx, _s4, _s5, _da, _vr, _xx, _xx, // 0CC0 - 0CCF + _xx, _xx, _xx, _xx, _xx, _m3, _m2, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _ct, _xx, // 0CD0 - 0CDF _iv, _iv, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx // 0CE0 - 0CEF }; @@ -170,10 +182,10 @@ static const IndicClassTable::CharClass mlymCharClasses[] = { _xx, _xx, _mp, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _iv, _iv, // 0D00 - 0D0F _iv, _xx, _iv, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _bb, // 0D10 - 0D1F - _ct, _ct, _ct, _bb, _ct, _bb, _bb, _ct, _ct, _xx, _ct, _ct, _ct, _ct, _ct, _ct, // 0D20 - 0D2F - _pb, _cn, _bb, _ct, _ct, _pb, _ct, _ct, _ct, _ct, _xx, _xx, _xx, _xx, _dr, _dr, // 0D30 - 0D3F - _dr, _db, _db, _db, _xx, _xx, _dl, _dl, _dl, _xx, _s1, _s2, _s3, _vr, _xx, _xx, // 0D40 - 0D4F - _xx, _xx, _xx, _xx, _xx, _xx, _xx, _dr, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0D50 - 0D5F + _ct, _ct, _ct, _bb, _ct, _bb, _bb, _ct, _ct, _xx, _ct, _ct, _ct, _ct, _ct, _pb, // 0D20 - 0D2F + _pb, _cn, _bb, _ct, _ct, _pb, _ct, _ct, _ct, _ct, _xx, _xx, _xx, _xx, _r2, _dr, // 0D30 - 0D3F + _dr, _dr, _dr, _dr, _xx, _xx, _l1, _l1, _dl, _xx, _s1, _s2, _s3, _vr, _xx, _xx, // 0D40 - 0D4F + _xx, _xx, _xx, _xx, _xx, _xx, _xx, _m2, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0D50 - 0D5F _iv, _iv, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx // 0D60 - 0D6F }; @@ -234,7 +246,7 @@ static const IndicClassTable mlymClassTable = {0x0D00, 0x0D6F, 3, MLYM_SCRIPT_FL // // IndicClassTable addresses // -static const IndicClassTable *indicClassTables[] = { +static const IndicClassTable * const indicClassTables[] = { NULL, /* 'zyyy' (COMMON) */ NULL, /* 'qaai' (INHERITED) */ NULL, /* 'arab' (ARABIC) */ @@ -288,7 +300,8 @@ static const IndicClassTable *indicClassTables[] = { NULL, /* 'osma' (OSMANYA) */ NULL, /* 'shaw' (SHAVIAN) */ NULL, /* 'tale' (TAI_LE) */ - NULL /* 'ugar' (UGARITIC) */ + NULL, /* 'ugar' (UGARITIC) */ + NULL /* 'hrkt' (KATAKANA_OR_HIRAGANA) */ }; IndicClassTable::CharClass IndicClassTable::getCharClass(LEUnicode ch) const