sl@0: /*
sl@0: *
sl@0: * (C) Copyright IBM Corp. 1998-2005 - All Rights Reserved
sl@0: *
sl@0: */
sl@0:
sl@0: #include "LETypes.h"
sl@0: #include "LEScripts.h"
sl@0: #include "OpenTypeTables.h"
sl@0: #include "OpenTypeUtilities.h"
sl@0: #include "IndicReordering.h"
sl@0:
sl@0: U_NAMESPACE_BEGIN
sl@0:
sl@0: #define ARRAY_SIZE(array) (sizeof array / sizeof array[0])
sl@0:
sl@0: // Split matra table indices
sl@0: #define _x1 (1 << CF_INDEX_SHIFT)
sl@0: #define _x2 (2 << CF_INDEX_SHIFT)
sl@0: #define _x3 (3 << CF_INDEX_SHIFT)
sl@0: #define _x4 (4 << CF_INDEX_SHIFT)
sl@0: #define _x5 (5 << CF_INDEX_SHIFT)
sl@0: #define _x6 (6 << CF_INDEX_SHIFT)
sl@0: #define _x7 (7 << CF_INDEX_SHIFT)
sl@0: #define _x8 (8 << CF_INDEX_SHIFT)
sl@0: #define _x9 (9 << CF_INDEX_SHIFT)
sl@0:
sl@0: // simple classes
sl@0: #define _xx (CC_RESERVED)
sl@0: #define _ma (CC_VOWEL_MODIFIER | CF_POS_ABOVE)
sl@0: #define _mp (CC_VOWEL_MODIFIER | CF_POS_AFTER)
sl@0: #define _sa (CC_STRESS_MARK | CF_POS_ABOVE)
sl@0: #define _sb (CC_STRESS_MARK | CF_POS_BELOW)
sl@0: #define _iv (CC_INDEPENDENT_VOWEL)
sl@0: // Added by Nokia -- Devanagai independent vowel A -> special case
sl@0: #define _ia (CC_INDEPENDENT_VOWEL_A)
sl@0:
sl@0: #define _i2 (CC_INDEPENDENT_VOWEL_2)
sl@0: #define _ct (CC_CONSONANT | CF_CONSONANT)
sl@0: #define _cn (CC_CONSONANT_WITH_NUKTA | CF_CONSONANT)
sl@0: // Added by Nokia -- Special case Kannada and Bengali Ra to satisfy usecase (Initial)Ra + ZWJ + Virama + Consonant
sl@0: #define _cr (CC_CONSONANT_KANNADA_BENGALI_RA | CF_CONSONANT)
sl@0:
sl@0: #define _nu (CC_NUKTA)
sl@0: #define _dv (CC_DEPENDENT_VOWEL)
sl@0: // Added by Nokia -- Devanagari vowel CANDRA E -> special case with independent vowel A
sl@0: #define _de (CC_DEPENDENT_VOWEL_CANDRA_E | CF_POS_ABOVE)
sl@0:
sl@0: // Added by Nokia -- Gurmukhi bearer ARA
sl@0: #define _ba (CC_GUR_BEARER_A)
sl@0: // Added by Nokia -- Gurmukhi bearer IRI
sl@0: #define _bi (CC_GUR_BEARER_I)
sl@0: // Added by Nokia -- Gurmukhi bearer URA
sl@0: #define _bu (CC_GUR_BEARER_U)
sl@0:
sl@0: // Added by Nokia -- Gurmukhi dependent vowel which combined with bearer ARA to form an independent vowel
sl@0: #define _ga (CC_GUR_DEPENDENT_VOWEL_A)
sl@0: // Added by Nokia -- Gurmukhi dependent vowel which combined with bearer IRI to form an independent vowel
sl@0: #define _gi (CC_GUR_DEPENDENT_VOWEL_I)
sl@0: // Added by Nokia -- Gurmukhi dependent vowel which combined with bearer URA to form an independent vowel
sl@0: #define _gu (CC_GUR_DEPENDENT_VOWEL_U)
sl@0:
sl@0: #define _gar (_ga | CF_POS_AFTER)
sl@0: #define _gaa (_ga | CF_POS_ABOVE)
sl@0: #define _gil (_gi | CF_POS_BEFORE)
sl@0: #define _gir (_gi | CF_POS_AFTER)
sl@0: #define _gia (_gi | CF_POS_ABOVE)
sl@0: #define _gub (_gu | CF_POS_BELOW)
sl@0: #define _gua (_gu | CF_POS_ABOVE)
sl@0:
sl@0: #define _dl (_dv | CF_POS_BEFORE)
sl@0: #define _db (_dv | CF_POS_BELOW)
sl@0: #define _da (_dv | CF_POS_ABOVE)
sl@0: #define _dr (_dv | CF_POS_AFTER)
sl@0: #define _lm (_dv | CF_LENGTH_MARK)
sl@0: #define _l1 (CC_SPLIT_VOWEL_PIECE_1 | CF_POS_BEFORE)
sl@0: #define _a1 (CC_SPLIT_VOWEL_PIECE_1 | CF_POS_ABOVE)
sl@0: #define _r2 (CC_SPLIT_VOWEL_PIECE_2 | CF_POS_AFTER)
sl@0: #define _m2 (CC_SPLIT_VOWEL_PIECE_2 | CF_LENGTH_MARK)
sl@0: #define _m3 (CC_SPLIT_VOWEL_PIECE_3 | CF_LENGTH_MARK)
sl@0: #define _vr (CC_VIRAMA)
sl@0:
sl@0: // split matras
sl@0: #define _s1 (_dv | _x1)
sl@0: #define _s2 (_dv | _x2)
sl@0: #define _s3 (_dv | _x3)
sl@0: #define _s4 (_dv | _x4)
sl@0: #define _s5 (_dv | _x5)
sl@0: #define _s6 (_dv | _x6)
sl@0: #define _s7 (_dv | _x7)
sl@0: #define _s8 (_dv | _x8)
sl@0: #define _s9 (_dv | _x9)
sl@0:
sl@0: // consonants with special forms
sl@0: // NOTE: this assumes that no consonants with nukta have
sl@0: // special forms... (Bengali RA?)
sl@0: #define _bb (_ct | CF_BELOW_BASE)
sl@0: #define _pb (_ct | CF_POST_BASE)
sl@0: #define _vt (_bb | CF_VATTU)
sl@0: #define _rv (_vt | CF_REPH)
sl@0: #define _rp (_pb | CF_REPH)
sl@0: #define _rb (_bb | CF_REPH)
sl@0:
sl@0: // Added by Nokia -- for Kannada Ra & Bengali Ra
sl@0: #define _b2 (_cr | CF_BELOW_BASE)
sl@0: #define _kr (_b2 | CF_REPH)
sl@0:
sl@0: // Added by Nokia -- for Bengali Ya
sl@0: #define _by (_pb | CF_CONSONANT_BENGALI_YA)
sl@0:
sl@0: // Added by Nokia -- for Tamil independent vowel O
sl@0: #define _to (CC_INDEPENDENT_VOWEL_TAMIL_O | CF_POS_AFTER)
sl@0:
sl@0: // Added by Nokia -- for Malayalam CHILLU
sl@0: #define _pbc (_pb | CF_CONSONANT_MLYL_CHILLU)
sl@0: #define _bbc (_bb | CF_CONSONANT_MLYL_CHILLU)
sl@0: #define _ctc (_ct | CF_CONSONANT_MLYL_CHILLU)
sl@0:
sl@0: //
sl@0: // Character class tables
sl@0: //
sl@0: static const IndicClassTable::CharClass devaCharClasses[] =
sl@0: {
sl@0: _xx, _ma, _ma, _mp, _iv, _ia, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, // 0900 - 090F
sl@0: _iv, _iv, _iv, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, // 0910 - 091F
sl@0: _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _cn, _ct, _ct, _ct, _ct, _ct, _ct, // 0920 - 092F
sl@0: _rv, _cn, _ct, _ct, _cn, _ct, _ct, _ct, _ct, _ct, _xx, _xx, _nu, _xx, _dr, _dl, // 0930 - 093F
sl@0: _dr, _db, _db, _db, _db, _de, _da, _da, _da, _dr, _dr, _dr, _dr, _vr, _xx, _xx, // 0940 - 094F
sl@0: _xx, _sa, _sb, _sa, _sa, _xx, _xx, _xx, _cn, _cn, _cn, _cn, _cn, _cn, _cn, _cn, // 0950 - 095F
sl@0: _iv, _iv, _db, _db, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0960 - 096F
sl@0: _xx // 0970
sl@0: };
sl@0:
sl@0: // Nokia: Changed 09AF character from _pb to _by which represents a Bengali Ya which may or may not take
sl@0: // post base form, depending on its position within a consonant cluster.
sl@0: // Also changed 09B0 from _rb to special case _kr to satiisfy the usecase Ra + ZWJ + Virama + Consonant.
sl@0: // Changed 09E2 and 09E3 to _db here, not _dv since the Bengali rendering rules want them below
sl@0: // the base consonant.
sl@0: static const IndicClassTable::CharClass bengCharClasses[] =
sl@0: {
sl@0: _xx, _ma, _mp, _mp, _xx, _i2, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _xx, _i2, // 0980 - 098F
sl@0: _iv, _xx, _xx, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, // 0990 - 099F
sl@0: _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _xx, _ct, _ct, _bb, _ct, _ct, _by, // 09A0 - 09AF
sl@0: _kr, _xx, _ct, _xx, _xx, _xx, _ct, _ct, _ct, _ct, _xx, _xx, _nu, _xx, _r2, _dl, // 09B0 - 09BF
sl@0: _dr, _db, _db, _db, _db, _xx, _xx, _l1, _dl, _xx, _xx, _s1, _s2, _vr, _xx, _xx, // 09C0 - 09CF
sl@0: _xx, _xx, _xx, _xx, _xx, _xx, _xx, _m2, _xx, _xx, _xx, _xx, _cn, _cn, _xx, _cn, // 09D0 - 09DF
sl@0: _iv, _iv, _db, _db, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 09E0 - 09EF
sl@0: _ct, _ct, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx // 09F0 - 09FA
sl@0: };
sl@0:
sl@0: // Nokia: Changed 0A05 from _iv to _ba and 0A72&0A73 from _ct to _bi&_bu, changed 0A3E,0A48&0A4C
sl@0: // from _d* to _ga*, 0A3F,0A40&0A47 from _d* to _gi*, 0A41,0A42&0A4B from _d* to _gu* to support
sl@0: // decompositions of the independent vowels into a sequence of a vowel bearer and a depending vowel sign.
sl@0: // Also changed 0A2F from _bb to _pb since YA is expected to post-base when combined in constant cluster.
sl@0: static const IndicClassTable::CharClass punjCharClasses[] =
sl@0: {
sl@0: _xx, _ma, _ma, _mp, _xx, _ba, _iv, _iv, _iv, _iv, _iv, _xx, _xx, _xx, _xx, _iv, // 0A00 - 0A0F
sl@0: _iv, _xx, _xx, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, // 0A10 - 0A1F
sl@0: _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _xx, _ct, _ct, _ct, _ct, _ct, _pb, // 0A20 - 0A2F
sl@0: _vt, _xx, _ct, _cn, _xx, _bb, _cn, _xx, _ct, _bb, _xx, _xx, _nu, _xx, _gar, _gil, // 0A30 - 0A3F
sl@0: _gir, _gub, _gub, _xx, _xx, _xx, _xx, _gia, _gaa, _xx, _xx, _gua, _gaa, _vr, _xx, _xx, // 0A40 - 0A4F
sl@0: _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _cn, _cn, _cn, _ct, _xx, _cn, _xx, // 0A50 - 0A5F
sl@0: _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0A60 - 0A6F
sl@0: _ma, _ma, _bi, _bu, _xx // 0A70 - 0A74
sl@0: };
sl@0:
sl@0: static const IndicClassTable::CharClass gujrCharClasses[] =
sl@0: {
sl@0: _xx, _ma, _ma, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _iv, _xx, _iv, // 0A80 - 0A8F
sl@0: _iv, _iv, _xx, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, // 0A90 - 0A9F
sl@0: _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _xx, _ct, _ct, _ct, _ct, _ct, _ct, // 0AA0 - 0AAF
sl@0: _rv, _xx, _ct, _ct, _xx, _ct, _ct, _ct, _ct, _ct, _xx, _xx, _nu, _xx, _dr, _dl, // 0AB0 - 0ABF
sl@0: _dr, _db, _db, _db, _db, _da, _xx, _da, _da, _dr, _xx, _dr, _dr, _vr, _xx, _xx, // 0AC0 - 0ACF
sl@0: _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0AD0 - 0ADF
sl@0: _iv, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx // 0AE0 - 0AEF
sl@0: };
sl@0:
sl@0: static const IndicClassTable::CharClass oryaCharClasses[] =
sl@0: {
sl@0: _xx, _ma, _mp, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _xx, _iv, // 0B00 - 0B0F
sl@0: _iv, _xx, _xx, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, // 0B10 - 0B1F
sl@0: _ct, _ct, _ct, _ct, _bb, _ct, _ct, _ct, _bb, _xx, _ct, _ct, _bb, _bb, _bb, _pb, // 0B20 - 0B2F
sl@0: _rb, _xx, _bb, _bb, _xx, _ct, _ct, _ct, _ct, _ct, _xx, _xx, _nu, _xx, _r2, _da, // 0B30 - 0B3F
sl@0: _dr, _db, _db, _db, _xx, _xx, _xx, _l1, _s1, _xx, _xx, _s2, _s3, _vr, _xx, _xx, // 0B40 - 0B4F
sl@0: _xx, _xx, _xx, _xx, _xx, _xx, _m2, _m2, _xx, _xx, _xx, _xx, _cn, _cn, _xx, _cn, // 0B50 - 0B5F
sl@0: _iv, _iv, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0B60 - 0B6F
sl@0: _xx, _ct // 0B70 - 0B71
sl@0: };
sl@0:
sl@0: // Nokia: Changed 0B92 to _to for the exceptional case of Tamil independent vowel O
sl@0: static const IndicClassTable::CharClass tamlCharClasses[] =
sl@0: {
sl@0: _xx, _xx, _ma, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _xx, _xx, _iv, _iv, // 0B80 - 0B8F
sl@0: _iv, _xx, _to, _iv, _iv, _ct, _xx, _xx, _xx, _ct, _ct, _xx, _ct, _xx, _ct, _ct, // 0B90 - 0B9F
sl@0: _xx, _xx, _xx, _ct, _ct, _xx, _xx, _xx, _ct, _ct, _ct, _xx, _xx, _xx, _ct, _ct, // 0BA0 - 0BAF
sl@0: _ct, _ct, _ct, _ct, _ct, _ct, _xx, _ct, _ct, _ct, _xx, _xx, _xx, _xx, _r2, _dr, // 0BB0 - 0BBF
sl@0: _da, _dr, _dr, _xx, _xx, _xx, _l1, _l1, _dl, _xx, _s1, _s2, _s3, _vr, _xx, _xx, // 0BC0 - 0BCF
sl@0: _xx, _xx, _xx, _xx, _xx, _xx, _xx, _m2, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0BD0 - 0BDF
sl@0: _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0BE0 - 0BEF
sl@0: _xx, _xx, _xx // 0BF0 - 0BF2
sl@0: };
sl@0:
sl@0: // FIXME: Should some of the bb's be pb's? (KA, NA, MA, YA, VA, etc. (approx 13))
sl@0: // Nokia: Changed 0C31 character to _bb as with _ct the consonant cluster is not rendered correctly.
sl@0: static const IndicClassTable::CharClass teluCharClasses[] =
sl@0: {
sl@0: _xx, _mp, _mp, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _iv, _iv, // 0C00 - 0C0F
sl@0: _iv, _xx, _iv, _iv, _iv, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, // 0C10 - 0C1F
sl@0: _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _xx, _bb, _bb, _bb, _bb, _bb, _bb, // 0C20 - 0C2F
sl@0: _bb, _bb, _bb, _bb, _xx, _bb, _bb, _bb, _bb, _bb, _xx, _xx, _xx, _xx, _da, _da, // 0C30 - 0C3F
sl@0: _da, _dr, _dr, _dr, _dr, _xx, _a1, _da, _s1, _xx, _da, _da, _da, _vr, _xx, _xx, // 0C40 - 0C4F
sl@0: _xx, _xx, _xx, _xx, _xx, _da, _m2, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0C50 - 0C5F
sl@0: _iv, _iv, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx // 0C60 - 0C6F
sl@0: };
sl@0:
sl@0: // U+CC3 and U+CC4 are _lm here not _dr since the Kannada rendering
sl@0: // rules want them below and to the right of the entire cluster
sl@0: //
sl@0: // There's some information about this in:
sl@0: //
sl@0: // http://brahmi.sourceforge.net/docs/KannadaComputing.html
sl@0:
sl@0: // Nokia: Changed 0CBC character class to _nu (NUKTA), as post Unicode 4.0, this
sl@0: // code point is classified as the Kannada NUKTA character.
sl@0: // Also changed 0CB0 from _rb to special case _cr to satiisfy the usecase Ra + ZWJ + Virama + Consonant
sl@0: // see _cr and _rb above
sl@0: static const IndicClassTable::CharClass kndaCharClasses[] =
sl@0: {
sl@0: _xx, _xx, _mp, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _iv, _iv, // 0C80 - 0C8F
sl@0: _iv, _xx, _iv, _iv, _iv, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, // 0C90 - 0C9F
sl@0: _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _bb, _xx, _bb, _bb, _bb, _bb, _bb, _bb, // 0CA0 - 0CAF
sl@0: _kr, _ct, _bb, _bb, _xx, _bb, _bb, _bb, _bb, _bb, _xx, _xx, _nu, _xx, _dr, _da, // 0CB0 - 0CBF
sl@0: _s1, _dr, _r2, _lm, _lm, _xx, _a1, _s2, _s3, _xx, _s4, _s5, _da, _vr, _xx, _xx, // 0CC0 - 0CCF
sl@0: _xx, _xx, _xx, _xx, _xx, _m3, _m2, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _bb, _xx, // 0CD0 - 0CDF
sl@0: _iv, _iv, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx // 0CE0 - 0CEF
sl@0: };
sl@0:
sl@0: // FIXME: this is correct for old-style Malayalam (MAL) but not for reformed Malayalam (MLR)
sl@0: // FIXME: should there be a REPH for old-style Malayalam?
sl@0: static const IndicClassTable::CharClass mlymCharClasses[] =
sl@0: {
sl@0: _xx, _xx, _mp, _mp, _xx, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _iv, _xx, _iv, _iv, // 0D00 - 0D0F
sl@0: // _iv, _xx, _iv, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _bb, // 0D10 - 0D1F //old
sl@0: _iv, _xx, _iv, _iv, _iv, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, _ct, // 0D10 - 0D1F
sl@0: // _ct, _ct, _ct, _bb, _ct, _bb, _bb, _ct, _ct, _xx, _ct, _ct, _ct, _ct, _ct, _pb, // 0D20 - 0D2F //old
sl@0: _ct, _ct, _ct, _ctc, _ct, _ct, _ct, _ct, _ctc, _xx, _ct, _ct, _ct, _ct, _ct, _pb, // 0D20 - 0D2F
sl@0: // _pb, _cn, _bb, _ct, _ct, _pb, _ct, _ct, _ct, _ct, _xx, _xx, _xx, _xx, _r2, _dr, // 0D30 - 0D3F //old
sl@0: _pbc, _cn, _bbc, _ctc, _ct, _pb, _ct, _ct, _ct, _ct, _xx, _xx, _xx, _xx, _r2, _dr, // 0D30 - 0D3F
sl@0: _dr, _dr, _dr, _dr, _xx, _xx, _l1, _l1, _dl, _xx, _s1, _s2, _s3, _vr, _xx, _xx, // 0D40 - 0D4F
sl@0: _xx, _xx, _xx, _xx, _xx, _xx, _xx, _m2, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, // 0D50 - 0D5F
sl@0: _iv, _iv, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx, _xx // 0D60 - 0D6F
sl@0: };
sl@0:
sl@0: //
sl@0: // Split matra tables
sl@0: //
sl@0: static const SplitMatra bengSplitTable[] = {{0x09C7, 0x09BE}, {0x09C7, 0x09D7}};
sl@0:
sl@0: static const SplitMatra oryaSplitTable[] = {{0x0B47, 0x0B56}, {0x0B47, 0x0B3E}, {0x0B47, 0x0B57}};
sl@0:
sl@0: static const SplitMatra tamlSplitTable[] = {{0x0BC6, 0x0BBE}, {0x0BC7, 0x0BBE}, {0x0BC6, 0x0BD7}};
sl@0:
sl@0: static const SplitMatra teluSplitTable[] = {{0x0C46, 0x0C56}};
sl@0:
sl@0: static const SplitMatra kndaSplitTable[] = {{0x0CBF, 0x0CD5}, {0x0CC6, 0x0CD5}, {0x0CC6, 0x0CD6}, {0x0CC6, 0x0CC2},
sl@0: {0x0CC6, 0x0CC2, 0x0CD5}};
sl@0:
sl@0: static const SplitMatra mlymSplitTable[] = {{0x0D46, 0x0D3E}, {0x0D47, 0x0D3E}, {0x0D46, 0x0D57}};
sl@0:
sl@0: //
sl@0: // Script Flags
sl@0: //
sl@0:
sl@0: // FIXME: post 'GSUB' reordering of MATRA_PRE's for Malayalam and Tamil
sl@0: // FIXME: reformed Malayalam needs to reorder VATTU to before base glyph...
sl@0: // FIXME: eyelash RA only for Devanagari??
sl@0: #define DEVA_SCRIPT_FLAGS (SF_EYELASH_RA | SF_NO_POST_BASE_LIMIT)
sl@0: #define BENG_SCRIPT_FLAGS (SF_REPH_AFTER_BELOW | SF_NO_POST_BASE_LIMIT)
sl@0: #define PUNJ_SCRIPT_FLAGS (SF_NO_POST_BASE_LIMIT | SF_MATRAS_AFTER_POSTBASE)
sl@0: #define GUJR_SCRIPT_FLAGS (SF_NO_POST_BASE_LIMIT)
sl@0: #define ORYA_SCRIPT_FLAGS (SF_REPH_AFTER_BELOW | SF_NO_POST_BASE_LIMIT)
sl@0: #define TAML_SCRIPT_FLAGS (SF_MPRE_FIXUP | SF_NO_POST_BASE_LIMIT)
sl@0: #define TELU_SCRIPT_FLAGS (SF_MATRAS_AFTER_BASE | 3)
sl@0: #define KNDA_SCRIPT_FLAGS (SF_MATRAS_AFTER_BASE | 3)
sl@0: #define MLYM_SCRIPT_FLAGS (SF_MPRE_FIXUP | SF_NO_POST_BASE_LIMIT)
sl@0:
sl@0: //
sl@0: // Indic Class Tables
sl@0: //
sl@0: static const IndicClassTable devaClassTable = {0x0900, 0x0970, 2, DEVA_SCRIPT_FLAGS, devaCharClasses, NULL};
sl@0:
sl@0: static const IndicClassTable bengClassTable = {0x0980, 0x09FA, 3, BENG_SCRIPT_FLAGS, bengCharClasses, bengSplitTable};
sl@0:
sl@0: static const IndicClassTable punjClassTable = {0x0A00, 0x0A74, 2, PUNJ_SCRIPT_FLAGS, punjCharClasses, NULL};
sl@0:
sl@0: static const IndicClassTable gujrClassTable = {0x0A80, 0x0AEF, 2, GUJR_SCRIPT_FLAGS, gujrCharClasses, NULL};
sl@0:
sl@0: static const IndicClassTable oryaClassTable = {0x0B00, 0x0B71, 3, ORYA_SCRIPT_FLAGS, oryaCharClasses, oryaSplitTable};
sl@0:
sl@0: static const IndicClassTable tamlClassTable = {0x0B80, 0x0BF2, 3, TAML_SCRIPT_FLAGS, tamlCharClasses, tamlSplitTable};
sl@0:
sl@0: static const IndicClassTable teluClassTable = {0x0C00, 0x0C6F, 3, TELU_SCRIPT_FLAGS, teluCharClasses, teluSplitTable};
sl@0:
sl@0: static const IndicClassTable kndaClassTable = {0x0C80, 0x0CEF, 4, KNDA_SCRIPT_FLAGS, kndaCharClasses, kndaSplitTable};
sl@0:
sl@0: static const IndicClassTable mlymClassTable = {0x0D00, 0x0D6F, 3, MLYM_SCRIPT_FLAGS, mlymCharClasses, mlymSplitTable};
sl@0:
sl@0: //
sl@0: // IndicClassTable addresses
sl@0: //
sl@0: static const IndicClassTable * const indicClassTables[] = {
sl@0: NULL, /* 'zyyy' (COMMON) */
sl@0: NULL, /* 'qaai' (INHERITED) */
sl@0: NULL, /* 'arab' (ARABIC) */
sl@0: NULL, /* 'armn' (ARMENIAN) */
sl@0: &bengClassTable, /* 'beng' (BENGALI) */
sl@0: NULL, /* 'bopo' (BOPOMOFO) */
sl@0: NULL, /* 'cher' (CHEROKEE) */
sl@0: NULL, /* 'qaac' (COPTIC) */
sl@0: NULL, /* 'cyrl' (CYRILLIC) */
sl@0: NULL, /* 'dsrt' (DESERET) */
sl@0: &devaClassTable, /* 'deva' (DEVANAGARI) */
sl@0: NULL, /* 'ethi' (ETHIOPIC) */
sl@0: NULL, /* 'geor' (GEORGIAN) */
sl@0: NULL, /* 'goth' (GOTHIC) */
sl@0: NULL, /* 'grek' (GREEK) */
sl@0: &gujrClassTable, /* 'gujr' (GUJARATI) */
sl@0: &punjClassTable, /* 'guru' (GURMUKHI) */
sl@0: NULL, /* 'hani' (HAN) */
sl@0: NULL, /* 'hang' (HANGUL) */
sl@0: NULL, /* 'hebr' (HEBREW) */
sl@0: NULL, /* 'hira' (HIRAGANA) */
sl@0: &kndaClassTable, /* 'knda' (KANNADA) */
sl@0: NULL, /* 'kata' (KATAKANA) */
sl@0: NULL, /* 'khmr' (KHMER) */
sl@0: NULL, /* 'laoo' (LAO) */
sl@0: NULL, /* 'latn' (LATIN) */
sl@0: &mlymClassTable, /* 'mlym' (MALAYALAM) */
sl@0: NULL, /* 'mong' (MONGOLIAN) */
sl@0: NULL, /* 'mymr' (MYANMAR) */
sl@0: NULL, /* 'ogam' (OGHAM) */
sl@0: NULL, /* 'ital' (OLD-ITALIC) */
sl@0: &oryaClassTable, /* 'orya' (ORIYA) */
sl@0: NULL, /* 'runr' (RUNIC) */
sl@0: NULL, /* 'sinh' (SINHALA) */
sl@0: NULL, /* 'syrc' (SYRIAC) */
sl@0: &tamlClassTable, /* 'taml' (TAMIL) */
sl@0: &teluClassTable, /* 'telu' (TELUGU) */
sl@0: NULL, /* 'thaa' (THAANA) */
sl@0: NULL, /* 'thai' (THAI) */
sl@0: NULL, /* 'tibt' (TIBETAN) */
sl@0: NULL, /* 'cans' (CANADIAN-ABORIGINAL) */
sl@0: NULL, /* 'yiii' (YI) */
sl@0: NULL, /* 'tglg' (TAGALOG) */
sl@0: NULL, /* 'hano' (HANUNOO) */
sl@0: NULL, /* 'buhd' (BUHID) */
sl@0: NULL, /* 'tagb' (TAGBANWA) */
sl@0: NULL, /* 'brai' (BRAILLE) */
sl@0: NULL, /* 'cprt' (CYPRIOT) */
sl@0: NULL, /* 'limb' (LIMBU) */
sl@0: NULL, /* 'linb' (LINEAR_B) */
sl@0: NULL, /* 'osma' (OSMANYA) */
sl@0: NULL, /* 'shaw' (SHAVIAN) */
sl@0: NULL, /* 'tale' (TAI_LE) */
sl@0: NULL, /* 'ugar' (UGARITIC) */
sl@0: NULL /* 'hrkt' (KATAKANA_OR_HIRAGANA) */
sl@0: };
sl@0:
sl@0: IndicClassTable::CharClass IndicClassTable::getCharClass(LEUnicode ch) const
sl@0: {
sl@0: if (ch == C_SIGN_ZWJ) {
sl@0: return CF_CONSONANT | CC_ZERO_WIDTH_MARK;
sl@0: }
sl@0:
sl@0: if (ch == C_SIGN_ZWNJ) {
sl@0: return CC_ZERO_WIDTH_MARK;
sl@0: }
sl@0:
sl@0: if (ch < firstChar || ch > lastChar) {
sl@0: return CC_RESERVED;
sl@0: }
sl@0:
sl@0: return classTable[ch - firstChar];
sl@0: }
sl@0:
sl@0: const IndicClassTable *IndicClassTable::getScriptClassTable(le_int32 scriptCode)
sl@0: {
sl@0: if (scriptCode < 0 || scriptCode >= scriptCodeCount) {
sl@0: return NULL;
sl@0: }
sl@0:
sl@0: return indicClassTables[scriptCode];
sl@0: }
sl@0:
sl@0: le_int32 IndicReordering::getWorstCaseExpansion(le_int32 scriptCode)
sl@0: {
sl@0: const IndicClassTable *classTable = IndicClassTable::getScriptClassTable(scriptCode);
sl@0:
sl@0: if (classTable == NULL) {
sl@0: return 1;
sl@0: }
sl@0:
sl@0: return classTable->getWorstCaseExpansion();
sl@0: }
sl@0:
sl@0: U_NAMESPACE_END