2 * Copyright © 2009 Red Hat, Inc.
3 * Copyright © 2011 Codethink Limited
4 * Copyright © 2010,2011,2012 Google, Inc.
6 * This is part of HarfBuzz, a text shaping library.
8 * Permission is hereby granted, without written agreement and without
9 * license or royalty fees, to use, copy, modify, and distribute this
10 * software and its documentation for any purpose, provided that the
11 * above copyright notice and the following two paragraphs appear in
12 * all copies of this software.
14 * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
15 * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
16 * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
17 * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
20 * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
21 * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
22 * FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
23 * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
24 * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
26 * Red Hat Author(s): Behdad Esfahbod
27 * Codethink Author(s): Ryan Lortie
28 * Google Author(s): Behdad Esfahbod
31 #ifndef HB_UNICODE_PRIVATE_HH
32 #define HB_UNICODE_PRIVATE_HH
34 #include "hb-private.hh"
36 #include "hb-unicode.h"
37 #include "hb-object-private.hh"
40 extern HB_INTERNAL const uint8_t _hb_modified_combining_class[256];
46 #define HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS \
47 HB_UNICODE_FUNC_IMPLEMENT (combining_class) \
48 HB_UNICODE_FUNC_IMPLEMENT (eastasian_width) \
49 HB_UNICODE_FUNC_IMPLEMENT (general_category) \
50 HB_UNICODE_FUNC_IMPLEMENT (mirroring) \
51 HB_UNICODE_FUNC_IMPLEMENT (script) \
52 HB_UNICODE_FUNC_IMPLEMENT (compose) \
53 HB_UNICODE_FUNC_IMPLEMENT (decompose) \
54 HB_UNICODE_FUNC_IMPLEMENT (decompose_compatibility) \
55 /* ^--- Add new callbacks here */
57 /* Simple callbacks are those taking a hb_codepoint_t and returning a hb_codepoint_t */
58 #define HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS_SIMPLE \
59 HB_UNICODE_FUNC_IMPLEMENT (hb_unicode_combining_class_t, combining_class) \
60 HB_UNICODE_FUNC_IMPLEMENT (unsigned int, eastasian_width) \
61 HB_UNICODE_FUNC_IMPLEMENT (hb_unicode_general_category_t, general_category) \
62 HB_UNICODE_FUNC_IMPLEMENT (hb_codepoint_t, mirroring) \
63 HB_UNICODE_FUNC_IMPLEMENT (hb_script_t, script) \
64 /* ^--- Add new simple callbacks here */
66 struct hb_unicode_funcs_t {
67 hb_object_header_t header;
70 hb_unicode_funcs_t *parent;
74 #define HB_UNICODE_FUNC_IMPLEMENT(return_type, name) \
75 inline return_type name (hb_codepoint_t unicode) { return func.name (this, unicode, user_data.name); }
76 HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS_SIMPLE
77 #undef HB_UNICODE_FUNC_IMPLEMENT
79 inline hb_bool_t compose (hb_codepoint_t a, hb_codepoint_t b,
83 if (unlikely (!a || !b)) return false;
84 return func.compose (this, a, b, ab, user_data.compose);
87 inline hb_bool_t decompose (hb_codepoint_t ab,
88 hb_codepoint_t *a, hb_codepoint_t *b)
91 return func.decompose (this, ab, a, b, user_data.decompose);
94 inline unsigned int decompose_compatibility (hb_codepoint_t u,
95 hb_codepoint_t *decomposed)
97 unsigned int ret = func.decompose_compatibility (this, u, decomposed, user_data.decompose_compatibility);
98 if (ret == 1 && u == decomposed[0]) {
108 modified_combining_class (hb_codepoint_t unicode)
110 return _hb_modified_combining_class[combining_class (unicode)];
114 is_variation_selector (hb_codepoint_t unicode)
116 return unlikely (hb_in_ranges<hb_codepoint_t> (unicode,
117 0x180B, 0x180D, /* MONGOLIAN FREE VARIATION SELECTOR ONE..THREE */
118 0xFE00, 0xFE0F, /* VARIATION SELECTOR-1..16 */
119 0xE0100, 0xE01EF)); /* VARIATION SELECTOR-17..256 */
122 /* Default_Ignorable codepoints:
124 * Note that as of Oct 2012 (Unicode 6.2), U+180E MONGOLIAN VOWEL SEPARATOR
125 * is NOT Default_Ignorable, but it really behaves in a way that it should
126 * be. That has been reported to the Unicode Technical Committee for
127 * consideration. As such, we include it here, since Uniscribe removes it.
130 * http://unicode.org/cldr/utility/list-unicodeset.jsp?a=[:DI:]&abb=on&ucd=on&esc=on
132 * Last updated to the page with the following versions:
133 * Version 3.6; ICU version: 50.0.1.0; Unicode version: 6.1.0.0
137 * [\u00AD\u034F\u115F\u1160\u17B4\u17B5\u180B-\u180D\u200B-\u200F\u202A-\u202E\u2060-\u206F\u3164\uFE00-\uFE0F\uFEFF\uFFA0\uFFF0-\uFFF8\U0001D173-\U0001D17A\U000E0000-\U000E0FFF]
140 * 034F ;COMBINING GRAPHEME JOINER
141 * 115F ;HANGUL CHOSEONG FILLER
142 * 1160 ;HANGUL JUNGSEONG FILLER
143 * 17B4 ;KHMER VOWEL INHERENT AQ
144 * 17B5 ;KHMER VOWEL INHERENT AA
145 * 180B..180D ;MONGOLIAN FREE VARIATION SELECTOR THREE
146 * 200B..200F ;RIGHT-TO-LEFT MARK
147 * 202A..202E ;RIGHT-TO-LEFT OVERRIDE
148 * 2060..206F ;NOMINAL DIGIT SHAPES
149 * 3164 ;HANGUL FILLER
150 * FE00..FE0F ;VARIATION SELECTOR-16
151 * FEFF ;ZERO WIDTH NO-BREAK SPACE
152 * FFA0 ;HALFWIDTH HANGUL FILLER
153 * FFF0..FFF8 ;<unassigned-FFF8>
154 * 1D173..1D17A ;MUSICAL SYMBOL END PHRASE
155 * E0000..E0FFF ;<unassigned-E0FFF>
158 is_default_ignorable (hb_codepoint_t ch)
160 hb_codepoint_t plane = ch >> 16;
161 if (likely (plane == 0))
164 hb_codepoint_t page = ch >> 8;
166 case 0x00: return unlikely (ch == 0x00AD);
167 case 0x03: return unlikely (ch == 0x034F);
168 case 0x11: return hb_in_range<hb_codepoint_t> (ch, 0x115F, 0x1160);
169 case 0x17: return hb_in_range<hb_codepoint_t> (ch, 0x17B4, 0x17B5);
170 case 0x18: return hb_in_range<hb_codepoint_t> (ch, 0x180B, 0x180E);
171 case 0x20: return hb_in_ranges<hb_codepoint_t> (ch, 0x200B, 0x200F,
174 case 0x31: return unlikely (ch == 0x3164);
175 case 0xFE: return hb_in_range<hb_codepoint_t> (ch, 0xFE00, 0xFE0F) || ch == 0xFEFF;
176 case 0xFF: return hb_in_range<hb_codepoint_t> (ch, 0xFFF0, 0xFFF8) || ch == 0xFFA0;
177 default: return false;
184 case 0x01: return hb_in_range<hb_codepoint_t> (ch, 0x0001D173, 0x0001D17A);
185 case 0x0E: return hb_in_range<hb_codepoint_t> (ch, 0x000E0000, 0x000E0FFF);
186 default: return false;
193 #define HB_UNICODE_FUNC_IMPLEMENT(name) hb_unicode_##name##_func_t name;
194 HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
195 #undef HB_UNICODE_FUNC_IMPLEMENT
199 #define HB_UNICODE_FUNC_IMPLEMENT(name) void *name;
200 HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
201 #undef HB_UNICODE_FUNC_IMPLEMENT
205 #define HB_UNICODE_FUNC_IMPLEMENT(name) hb_destroy_func_t name;
206 HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
207 #undef HB_UNICODE_FUNC_IMPLEMENT
212 extern HB_INTERNAL const hb_unicode_funcs_t _hb_unicode_funcs_nil;
215 /* Modified combining marks */
219 * We permute the "fixed-position" classes 10-26 into the order
220 * described in the SBL Hebrew manual:
222 * http://www.sbl-site.org/Fonts/SBLHebrewUserManual1.5x.pdf
224 * (as recommended by:
225 * http://forum.fontlab.com/archive-old-microsoft-volt-group/vista-and-diacritic-ordering-t6751.0.html)
228 * https://bugzilla.mozilla.org/show_bug.cgi?id=662055
230 #define HB_MODIFIED_COMBINING_CLASS_CCC10 22 /* sheva */
231 #define HB_MODIFIED_COMBINING_CLASS_CCC11 15 /* hataf segol */
232 #define HB_MODIFIED_COMBINING_CLASS_CCC12 16 /* hataf patah */
233 #define HB_MODIFIED_COMBINING_CLASS_CCC13 17 /* hataf qamats */
234 #define HB_MODIFIED_COMBINING_CLASS_CCC14 23 /* hiriq */
235 #define HB_MODIFIED_COMBINING_CLASS_CCC15 18 /* tsere */
236 #define HB_MODIFIED_COMBINING_CLASS_CCC16 19 /* segol */
237 #define HB_MODIFIED_COMBINING_CLASS_CCC17 20 /* patah */
238 #define HB_MODIFIED_COMBINING_CLASS_CCC18 21 /* qamats */
239 #define HB_MODIFIED_COMBINING_CLASS_CCC19 14 /* holam */
240 #define HB_MODIFIED_COMBINING_CLASS_CCC20 24 /* qubuts */
241 #define HB_MODIFIED_COMBINING_CLASS_CCC21 12 /* dagesh */
242 #define HB_MODIFIED_COMBINING_CLASS_CCC22 25 /* meteg */
243 #define HB_MODIFIED_COMBINING_CLASS_CCC23 13 /* rafe */
244 #define HB_MODIFIED_COMBINING_CLASS_CCC24 10 /* shin dot */
245 #define HB_MODIFIED_COMBINING_CLASS_CCC25 11 /* sin dot */
246 #define HB_MODIFIED_COMBINING_CLASS_CCC26 26 /* point varika */
251 * Modify to move Shadda (ccc=33) before other marks. See:
252 * http://unicode.org/faq/normalization.html#8
253 * http://unicode.org/faq/normalization.html#9
255 #define HB_MODIFIED_COMBINING_CLASS_CCC27 28 /* fathatan */
256 #define HB_MODIFIED_COMBINING_CLASS_CCC28 29 /* dammatan */
257 #define HB_MODIFIED_COMBINING_CLASS_CCC29 30 /* kasratan */
258 #define HB_MODIFIED_COMBINING_CLASS_CCC30 31 /* fatha */
259 #define HB_MODIFIED_COMBINING_CLASS_CCC31 32 /* damma */
260 #define HB_MODIFIED_COMBINING_CLASS_CCC32 33 /* kasra */
261 #define HB_MODIFIED_COMBINING_CLASS_CCC33 27 /* shadda */
262 #define HB_MODIFIED_COMBINING_CLASS_CCC34 34 /* sukun */
263 #define HB_MODIFIED_COMBINING_CLASS_CCC35 35 /* superscript alef */
266 #define HB_MODIFIED_COMBINING_CLASS_CCC36 36 /* superscript alaph */
270 * Modify Telugu length marks (ccc=84, ccc=91).
271 * These are the only matras in the main Indic scripts range that have
272 * a non-zero ccc. That makes them reorder with the Halant that is
273 * ccc=9. Just zero them, we don't need them in our Indic shaper.
275 #define HB_MODIFIED_COMBINING_CLASS_CCC84 0 /* length mark */
276 #define HB_MODIFIED_COMBINING_CLASS_CCC91 0 /* ai length mark */
280 * Modify U+0E38 and U+0E39 (ccc=103) to be reordered before U+0E3A (ccc=9).
281 * Assign 3, which is unassigned otherwise.
282 * Uniscribe does this reordering too.
284 #define HB_MODIFIED_COMBINING_CLASS_CCC103 3 /* sara u / sara uu */
285 #define HB_MODIFIED_COMBINING_CLASS_CCC107 107 /* mai * */
288 #define HB_MODIFIED_COMBINING_CLASS_CCC118 118 /* sign u / sign uu */
289 #define HB_MODIFIED_COMBINING_CLASS_CCC122 122 /* mai * */
292 #define HB_MODIFIED_COMBINING_CLASS_CCC129 129 /* sign aa */
293 #define HB_MODIFIED_COMBINING_CLASS_CCC130 130 /* sign i */
294 #define HB_MODIFIED_COMBINING_CLASS_CCC132 132 /* sign u */
299 #define HB_UNICODE_GENERAL_CATEGORY_IS_MARK(gen_cat) \
301 (FLAG (HB_UNICODE_GENERAL_CATEGORY_SPACING_MARK) | \
302 FLAG (HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK) | \
303 FLAG (HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK)))
306 #endif /* HB_UNICODE_PRIVATE_HH */