1 /* 2 * Copyright 2011,2012 Google, Inc. 3 * 4 * This is part of HarfBuzz, a text shaping library. 5 * 6 * Permission is hereby granted, without written agreement and without 7 * license or royalty fees, to use, copy, modify, and distribute this 8 * software and its documentation for any purpose, provided that the 9 * above copyright notice and the following two paragraphs appear in 10 * all copies of this software. 11 * 12 * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR 13 * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES 14 * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN 15 * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH 16 * DAMAGE. 17 * 18 * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING, 19 * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND 20 * FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS 21 * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO 22 * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS. 23 * 24 * Google Author(s): Behdad Esfahbod 25 */ 26 27 #ifndef HB_OT_SHAPE_COMPLEX_INDIC_MACHINE_HH 28 #define HB_OT_SHAPE_COMPLEX_INDIC_MACHINE_HH 29 30 #include "hb-private.hh" 31 32 %%{ 33 machine indic_syllable_machine; 34 alphtype unsigned char; 35 write data; 36 }%% 37 38 %%{ 39 40 # Same order as enum indic_category_t. Not sure how to avoid duplication. 41 X = 0; 42 C = 1; 43 V = 2; 44 N = 3; 45 H = 4; 46 ZWNJ = 5; 47 ZWJ = 6; 48 M = 7; 49 SM = 8; 50 VD = 9; 51 A = 10; 52 NBSP = 11; 53 DOTTEDCIRCLE = 12; 54 RS = 13; 55 Coeng = 14; 56 Repha = 15; 57 Ra = 16; 58 CM = 17; 59 60 c = (C | Ra)CM*; # is_consonant 61 n = ((ZWNJ?.RS)? (N.N?)?); # is_consonant_modifier 62 z = ZWJ|ZWNJ; # is_joiner 63 h = H | Coeng; # is_halant_or_coeng 64 reph = (Ra H | Repha); # possible reph 65 66 cn = c.ZWJ?.n?; 67 forced_rakar = ZWJ H ZWJ Ra; 68 matra_group = z{0,3}.M.N?.(H | forced_rakar)?; 69 syllable_tail = (Coeng (cn|V))? (SM.ZWNJ?)? (VD VD?)?; 70 place_holder = NBSP | DOTTEDCIRCLE; 71 halant_group = (z?.h.(ZWJ.N?)?); 72 final_halant_group = halant_group | h.ZWNJ; 73 halant_or_matra_group = (final_halant_group | (h.ZWJ)? matra_group{0,4}); 74 75 76 consonant_syllable = Repha? (cn.halant_group){0,4} cn A? halant_or_matra_group? syllable_tail; 77 vowel_syllable = reph? V.n? (ZWJ | (halant_group.cn){0,4} halant_or_matra_group? syllable_tail); 78 standalone_cluster = reph? place_holder.n? (halant_group.cn){0,4} halant_or_matra_group? syllable_tail; 79 broken_cluster = reph? n? (halant_group.cn){0,4} halant_or_matra_group syllable_tail; 80 other = any; 81 82 main := |* 83 consonant_syllable => { found_syllable (consonant_syllable); }; 84 vowel_syllable => { found_syllable (vowel_syllable); }; 85 standalone_cluster => { found_syllable (standalone_cluster); }; 86 broken_cluster => { found_syllable (broken_cluster); }; 87 other => { found_syllable (non_indic_cluster); }; 88 *|; 89 90 91 }%% 92 93 #define found_syllable(syllable_type) \ 94 HB_STMT_START { \ 95 if (0) fprintf (stderr, "syllable %d..%d %s\n", last, p+1, #syllable_type); \ 96 for (unsigned int i = last; i < p+1; i++) \ 97 info[i].syllable() = (syllable_serial << 4) | syllable_type; \ 98 last = p+1; \ 99 syllable_serial++; \ 100 if (unlikely (syllable_serial == 16)) syllable_serial = 1; \ 101 } HB_STMT_END 102 103 static void 104 find_syllables (hb_buffer_t *buffer) 105 { 106 unsigned int p, pe, eof, ts HB_UNUSED, te HB_UNUSED, act HB_UNUSED; 107 int cs; 108 hb_glyph_info_t *info = buffer->info; 109 %%{ 110 write init; 111 getkey info[p].indic_category(); 112 }%% 113 114 p = 0; 115 pe = eof = buffer->len; 116 117 unsigned int last = 0; 118 unsigned int syllable_serial = 1; 119 %%{ 120 write exec; 121 }%% 122 } 123 124 #endif /* HB_OT_SHAPE_COMPLEX_INDIC_MACHINE_HH */ 125