1/* 2 * Copyright © 2015 Mozilla Foundation. 3 * Copyright © 2015 Google, Inc. 4 * 5 * This is part of HarfBuzz, a text shaping library. 6 * 7 * Permission is hereby granted, without written agreement and without 8 * license or royalty fees, to use, copy, modify, and distribute this 9 * software and its documentation for any purpose, provided that the 10 * above copyright notice and the following two paragraphs appear in 11 * all copies of this software. 12 * 13 * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR 14 * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES 15 * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN 16 * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH 17 * DAMAGE. 18 * 19 * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING, 20 * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND 21 * FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS 22 * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO 23 * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS. 24 * 25 * Mozilla Author(s): Jonathan Kew 26 * Google Author(s): Behdad Esfahbod 27 */ 28 29#ifndef HB_OT_SHAPE_COMPLEX_USE_MACHINE_HH 30#define HB_OT_SHAPE_COMPLEX_USE_MACHINE_HH 31 32#include "hb.hh" 33 34#include "hb-ot-shape-complex-syllabic.hh" 35 36/* buffer var allocations */ 37#define use_category() complex_var_u8_category() 38 39#define USE(Cat) use_syllable_machine_ex_##Cat 40 41enum use_syllable_type_t { 42 use_independent_cluster, 43 use_virama_terminated_cluster, 44 use_sakot_terminated_cluster, 45 use_standard_cluster, 46 use_number_joiner_terminated_cluster, 47 use_numeral_cluster, 48 use_symbol_cluster, 49 use_hieroglyph_cluster, 50 use_broken_cluster, 51 use_non_cluster, 52}; 53 54%%{ 55 machine use_syllable_machine; 56 alphtype unsigned char; 57 write exports; 58 write data; 59}%% 60 61%%{ 62 63# Categories used in the Universal Shaping Engine spec: 64# https://docs.microsoft.com/en-us/typography/script-development/use 65 66export O = 0; # OTHER 67 68export B = 1; # BASE 69export N = 4; # BASE_NUM 70export GB = 5; # BASE_OTHER 71export SUB = 11; # CONS_SUB 72export H = 12; # HALANT 73 74export HN = 13; # HALANT_NUM 75export ZWNJ = 14; # Zero width non-joiner 76export R = 18; # REPHA 77export S = 19; # SYM 78export CS = 43; # CONS_WITH_STACKER 79export HVM = 44; # HALANT_OR_VOWEL_MODIFIER 80export Sk = 48; # SAKOT 81export G = 49; # HIEROGLYPH 82export J = 50; # HIEROGLYPH_JOINER 83export SB = 51; # HIEROGLYPH_SEGMENT_BEGIN 84export SE = 52; # HIEROGLYPH_SEGMENT_END 85 86export FAbv = 24; # CONS_FINAL_ABOVE 87export FBlw = 25; # CONS_FINAL_BELOW 88export FPst = 26; # CONS_FINAL_POST 89export MAbv = 27; # CONS_MED_ABOVE 90export MBlw = 28; # CONS_MED_BELOW 91export MPst = 29; # CONS_MED_POST 92export MPre = 30; # CONS_MED_PRE 93export CMAbv = 31; # CONS_MOD_ABOVE 94export CMBlw = 32; # CONS_MOD_BELOW 95export VAbv = 33; # VOWEL_ABOVE / VOWEL_ABOVE_BELOW / VOWEL_ABOVE_BELOW_POST / VOWEL_ABOVE_POST 96export VBlw = 34; # VOWEL_BELOW / VOWEL_BELOW_POST 97export VPst = 35; # VOWEL_POST UIPC = Right 98export VPre = 22; # VOWEL_PRE / VOWEL_PRE_ABOVE / VOWEL_PRE_ABOVE_POST / VOWEL_PRE_POST 99export VMAbv = 37; # VOWEL_MOD_ABOVE 100export VMBlw = 38; # VOWEL_MOD_BELOW 101export VMPst = 39; # VOWEL_MOD_POST 102export VMPre = 23; # VOWEL_MOD_PRE 103export SMAbv = 41; # SYM_MOD_ABOVE 104export SMBlw = 42; # SYM_MOD_BELOW 105export FMAbv = 45; # CONS_FINAL_MOD UIPC = Top 106export FMBlw = 46; # CONS_FINAL_MOD UIPC = Bottom 107export FMPst = 47; # CONS_FINAL_MOD UIPC = Not_Applicable 108 109 110h = H | HVM | Sk; 111 112consonant_modifiers = CMAbv* CMBlw* ((h B | SUB) CMAbv? CMBlw*)*; 113medial_consonants = MPre? MAbv? MBlw? MPst?; 114dependent_vowels = VPre* VAbv* VBlw* VPst*; 115vowel_modifiers = HVM? VMPre* VMAbv* VMBlw* VMPst*; 116final_consonants = FAbv* FBlw* FPst*; 117final_modifiers = FMAbv* FMBlw* | FMPst?; 118 119complex_syllable_start = (R | CS)? (B | GB); 120complex_syllable_middle = 121 consonant_modifiers 122 medial_consonants 123 dependent_vowels 124 vowel_modifiers 125 (Sk B)* 126; 127complex_syllable_tail = 128 complex_syllable_middle 129 final_consonants 130 final_modifiers 131; 132number_joiner_terminated_cluster_tail = (HN N)* HN; 133numeral_cluster_tail = (HN N)+; 134symbol_cluster_tail = SMAbv+ SMBlw* | SMBlw+; 135 136virama_terminated_cluster = 137 complex_syllable_start 138 consonant_modifiers 139 h 140; 141sakot_terminated_cluster = 142 complex_syllable_start 143 complex_syllable_middle 144 Sk 145; 146standard_cluster = 147 complex_syllable_start 148 complex_syllable_tail 149; 150broken_cluster = 151 R? 152 (complex_syllable_tail | number_joiner_terminated_cluster_tail | numeral_cluster_tail | symbol_cluster_tail) 153; 154 155number_joiner_terminated_cluster = N number_joiner_terminated_cluster_tail; 156numeral_cluster = N numeral_cluster_tail?; 157symbol_cluster = (S | GB) symbol_cluster_tail?; 158hieroglyph_cluster = SB+ | SB* G SE* (J SE* (G SE*)?)*; 159independent_cluster = O; 160other = any; 161 162main := |* 163 independent_cluster => { found_syllable (use_independent_cluster); }; 164 virama_terminated_cluster => { found_syllable (use_virama_terminated_cluster); }; 165 sakot_terminated_cluster => { found_syllable (use_sakot_terminated_cluster); }; 166 standard_cluster => { found_syllable (use_standard_cluster); }; 167 number_joiner_terminated_cluster => { found_syllable (use_number_joiner_terminated_cluster); }; 168 numeral_cluster => { found_syllable (use_numeral_cluster); }; 169 symbol_cluster => { found_syllable (use_symbol_cluster); }; 170 hieroglyph_cluster => { found_syllable (use_hieroglyph_cluster); }; 171 broken_cluster => { found_syllable (use_broken_cluster); }; 172 other => { found_syllable (use_non_cluster); }; 173*|; 174 175 176}%% 177 178#define found_syllable(syllable_type) \ 179 HB_STMT_START { \ 180 if (0) fprintf (stderr, "syllable %d..%d %s\n", (*ts).second.first, (*te).second.first, #syllable_type); \ 181 for (unsigned i = (*ts).second.first; i < (*te).second.first; ++i) \ 182 info[i].syllable() = (syllable_serial << 4) | syllable_type; \ 183 syllable_serial++; \ 184 if (unlikely (syllable_serial == 16)) syllable_serial = 1; \ 185 } HB_STMT_END 186 187 188template <typename Iter> 189struct machine_index_t : 190 hb_iter_with_fallback_t<machine_index_t<Iter>, 191 typename Iter::item_t> 192{ 193 machine_index_t (const Iter& it) : it (it) {} 194 machine_index_t (const machine_index_t& o) : it (o.it) {} 195 196 static constexpr bool is_random_access_iterator = Iter::is_random_access_iterator; 197 static constexpr bool is_sorted_iterator = Iter::is_sorted_iterator; 198 199 typename Iter::item_t __item__ () const { return *it; } 200 typename Iter::item_t __item_at__ (unsigned i) const { return it[i]; } 201 unsigned __len__ () const { return it.len (); } 202 void __next__ () { ++it; } 203 void __forward__ (unsigned n) { it += n; } 204 void __prev__ () { --it; } 205 void __rewind__ (unsigned n) { it -= n; } 206 void operator = (unsigned n) 207 { unsigned index = (*it).first; if (index < n) it += n - index; else if (index > n) it -= index - n; } 208 void operator = (const machine_index_t& o) { *this = (*o.it).first; } 209 bool operator == (const machine_index_t& o) const { return (*it).first == (*o.it).first; } 210 bool operator != (const machine_index_t& o) const { return !(*this == o); } 211 212 private: 213 Iter it; 214}; 215struct 216{ 217 template <typename Iter, 218 hb_requires (hb_is_iterable (Iter))> 219 machine_index_t<hb_iter_type<Iter>> 220 operator () (Iter&& it) const 221 { return machine_index_t<hb_iter_type<Iter>> (hb_iter (it)); } 222} 223HB_FUNCOBJ (machine_index); 224 225 226 227static bool 228not_standard_default_ignorable (const hb_glyph_info_t &i) 229{ return !(i.use_category() == USE(O) && _hb_glyph_info_is_default_ignorable (&i)); } 230 231static inline void 232find_syllables_use (hb_buffer_t *buffer) 233{ 234 hb_glyph_info_t *info = buffer->info; 235 auto p = 236 + hb_iter (info, buffer->len) 237 | hb_enumerate 238 | hb_filter ([] (const hb_glyph_info_t &i) { return not_standard_default_ignorable (i); }, 239 hb_second) 240 | hb_filter ([&] (const hb_pair_t<unsigned, const hb_glyph_info_t &> p) 241 { 242 if (p.second.use_category() == USE(ZWNJ)) 243 for (unsigned i = p.first + 1; i < buffer->len; ++i) 244 if (not_standard_default_ignorable (info[i])) 245 return !_hb_glyph_info_is_unicode_mark (&info[i]); 246 return true; 247 }) 248 | hb_enumerate 249 | machine_index 250 ; 251 auto pe = p + p.len (); 252 auto eof = +pe; 253 auto ts = +p; 254 auto te = +p; 255 unsigned int act HB_UNUSED; 256 int cs; 257 %%{ 258 write init; 259 getkey (*p).second.second.use_category(); 260 }%% 261 262 unsigned int syllable_serial = 1; 263 %%{ 264 write exec; 265 }%% 266} 267 268#undef found_syllable 269 270#endif /* HB_OT_SHAPE_COMPLEX_USE_MACHINE_HH */ 271