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_KHMER_MACHINE_HH 28#define HB_OT_SHAPE_COMPLEX_KHMER_MACHINE_HH 29 30#include "hb.hh" 31 32%%{ 33 machine khmer_syllable_machine; 34 alphtype unsigned char; 35 write data; 36}%% 37 38%%{ 39 40# Same order as enum khmer_category_t. Not sure how to avoid duplication. 41C = 1; 42V = 2; 43ZWNJ = 5; 44ZWJ = 6; 45PLACEHOLDER = 11; 46DOTTEDCIRCLE = 12; 47Coeng= 14; 48Ra = 16; 49Robatic = 20; 50Xgroup = 21; 51Ygroup = 22; 52VAbv = 26; 53VBlw = 27; 54VPre = 28; 55VPst = 29; 56 57c = (C | Ra | V); 58cn = c.((ZWJ|ZWNJ)?.Robatic)?; 59joiner = (ZWJ | ZWNJ); 60xgroup = (joiner*.Xgroup)*; 61ygroup = Ygroup*; 62 63# This grammar was experimentally extracted from what Uniscribe allows. 64 65matra_group = VPre? xgroup VBlw? xgroup (joiner?.VAbv)? xgroup VPst?; 66syllable_tail = xgroup matra_group xgroup (Coeng.c)? ygroup; 67 68 69broken_cluster = (Coeng.cn)* (Coeng | syllable_tail); 70consonant_syllable = (cn|PLACEHOLDER|DOTTEDCIRCLE) broken_cluster; 71other = any; 72 73main := |* 74 consonant_syllable => { found_syllable (consonant_syllable); }; 75 broken_cluster => { found_syllable (broken_cluster); }; 76 other => { found_syllable (non_khmer_cluster); }; 77*|; 78 79 80}%% 81 82#define found_syllable(syllable_type) \ 83 HB_STMT_START { \ 84 if (0) fprintf (stderr, "syllable %d..%d %s\n", ts, te, #syllable_type); \ 85 for (unsigned int i = ts; i < te; i++) \ 86 info[i].syllable() = (syllable_serial << 4) | khmer_##syllable_type; \ 87 syllable_serial++; \ 88 if (unlikely (syllable_serial == 16)) syllable_serial = 1; \ 89 } HB_STMT_END 90 91static void 92find_syllables_khmer (hb_buffer_t *buffer) 93{ 94 unsigned int p, pe, eof, ts, te, act HB_UNUSED; 95 int cs; 96 hb_glyph_info_t *info = buffer->info; 97 %%{ 98 write init; 99 getkey info[p].khmer_category(); 100 }%% 101 102 p = 0; 103 pe = eof = buffer->len; 104 105 unsigned int syllable_serial = 1; 106 %%{ 107 write exec; 108 }%% 109} 110 111#undef found_syllable 112 113#endif /* HB_OT_SHAPE_COMPLEX_KHMER_MACHINE_HH */ 114