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.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. 41C = 1; 42V = 2; 43N = 3; 44H = 4; 45ZWNJ = 5; 46ZWJ = 6; 47M = 7; 48SM = 8; 49A = 10; 50PLACEHOLDER = 11; 51DOTTEDCIRCLE = 12; 52RS = 13; 53Repha = 15; 54Ra = 16; 55CM = 17; 56Symbol= 18; 57CS = 19; 58 59c = (C | Ra); # is_consonant 60n = ((ZWNJ?.RS)? (N.N?)?); # is_consonant_modifier 61z = ZWJ|ZWNJ; # is_joiner 62reph = (Ra H | Repha); # possible reph 63 64cn = c.ZWJ?.n?; 65forced_rakar = ZWJ H ZWJ Ra; 66symbol = Symbol.N?; 67matra_group = z*.M.N?.(H | forced_rakar)?; 68syllable_tail = (z?.SM.SM?.ZWNJ?)? A*; 69halant_group = (z?.H.(ZWJ.N?)?); 70final_halant_group = halant_group | H.ZWNJ; 71medial_group = CM?; 72halant_or_matra_group = (final_halant_group | matra_group*); 73 74complex_syllable_tail = (halant_group.cn)* medial_group halant_or_matra_group syllable_tail; 75 76consonant_syllable = (Repha|CS)? cn complex_syllable_tail; 77vowel_syllable = reph? V.n? (ZWJ | complex_syllable_tail); 78standalone_cluster = ((Repha|CS)? PLACEHOLDER | reph? DOTTEDCIRCLE).n? complex_syllable_tail; 79symbol_cluster = symbol syllable_tail; 80broken_cluster = reph? n? complex_syllable_tail; 81other = any; 82 83main := |* 84 consonant_syllable => { found_syllable (consonant_syllable); }; 85 vowel_syllable => { found_syllable (vowel_syllable); }; 86 standalone_cluster => { found_syllable (standalone_cluster); }; 87 symbol_cluster => { found_syllable (symbol_cluster); }; 88 broken_cluster => { found_syllable (broken_cluster); }; 89 other => { found_syllable (non_indic_cluster); }; 90*|; 91 92 93}%% 94 95#define found_syllable(syllable_type) \ 96 HB_STMT_START { \ 97 if (0) fprintf (stderr, "syllable %d..%d %s\n", ts, te, #syllable_type); \ 98 for (unsigned int i = ts; i < te; i++) \ 99 info[i].syllable() = (syllable_serial << 4) | indic_##syllable_type; \ 100 syllable_serial++; \ 101 if (unlikely (syllable_serial == 16)) syllable_serial = 1; \ 102 } HB_STMT_END 103 104static void 105find_syllables_indic (hb_buffer_t *buffer) 106{ 107 unsigned int p, pe, eof, ts, te, act; 108 int cs; 109 hb_glyph_info_t *info = buffer->info; 110 %%{ 111 write init; 112 getkey info[p].indic_category(); 113 }%% 114 115 p = 0; 116 pe = eof = buffer->len; 117 118 unsigned int syllable_serial = 1; 119 %%{ 120 write exec; 121 }%% 122} 123 124#undef found_syllable 125 126#endif /* HB_OT_SHAPE_COMPLEX_INDIC_MACHINE_HH */ 127