1 /*
2 * Copyright © 2010,2012 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 #include "hb-ot-shape-complex.hh"
28
29
30 static bool
compose_hebrew(const hb_ot_shape_normalize_context_t * c,hb_codepoint_t a,hb_codepoint_t b,hb_codepoint_t * ab)31 compose_hebrew (const hb_ot_shape_normalize_context_t *c,
32 hb_codepoint_t a,
33 hb_codepoint_t b,
34 hb_codepoint_t *ab)
35 {
36 /* Hebrew presentation-form shaping.
37 * https://bugzilla.mozilla.org/show_bug.cgi?id=728866
38 * Hebrew presentation forms with dagesh, for characters U+05D0..05EA;
39 * Note that some letters do not have a dagesh presForm encoded.
40 */
41 static const hb_codepoint_t sDageshForms[0x05EAu - 0x05D0u + 1] = {
42 0xFB30u, /* ALEF */
43 0xFB31u, /* BET */
44 0xFB32u, /* GIMEL */
45 0xFB33u, /* DALET */
46 0xFB34u, /* HE */
47 0xFB35u, /* VAV */
48 0xFB36u, /* ZAYIN */
49 0x0000u, /* HET */
50 0xFB38u, /* TET */
51 0xFB39u, /* YOD */
52 0xFB3Au, /* FINAL KAF */
53 0xFB3Bu, /* KAF */
54 0xFB3Cu, /* LAMED */
55 0x0000u, /* FINAL MEM */
56 0xFB3Eu, /* MEM */
57 0x0000u, /* FINAL NUN */
58 0xFB40u, /* NUN */
59 0xFB41u, /* SAMEKH */
60 0x0000u, /* AYIN */
61 0xFB43u, /* FINAL PE */
62 0xFB44u, /* PE */
63 0x0000u, /* FINAL TSADI */
64 0xFB46u, /* TSADI */
65 0xFB47u, /* QOF */
66 0xFB48u, /* RESH */
67 0xFB49u, /* SHIN */
68 0xFB4Au /* TAV */
69 };
70
71 bool found = (bool) c->unicode->compose (a, b, ab);
72
73 #ifdef HB_NO_OT_SHAPE_COMPLEX_HEBREW_FALLBACK
74 return found;
75 #endif
76
77 if (!found && !c->plan->has_gpos_mark)
78 {
79 /* Special-case Hebrew presentation forms that are excluded from
80 * standard normalization, but wanted for old fonts. */
81 switch (b) {
82 case 0x05B4u: /* HIRIQ */
83 if (a == 0x05D9u) { /* YOD */
84 *ab = 0xFB1Du;
85 found = true;
86 }
87 break;
88 case 0x05B7u: /* patah */
89 if (a == 0x05F2u) { /* YIDDISH YOD YOD */
90 *ab = 0xFB1Fu;
91 found = true;
92 } else if (a == 0x05D0u) { /* ALEF */
93 *ab = 0xFB2Eu;
94 found = true;
95 }
96 break;
97 case 0x05B8u: /* QAMATS */
98 if (a == 0x05D0u) { /* ALEF */
99 *ab = 0xFB2Fu;
100 found = true;
101 }
102 break;
103 case 0x05B9u: /* HOLAM */
104 if (a == 0x05D5u) { /* VAV */
105 *ab = 0xFB4Bu;
106 found = true;
107 }
108 break;
109 case 0x05BCu: /* DAGESH */
110 if (a >= 0x05D0u && a <= 0x05EAu) {
111 *ab = sDageshForms[a - 0x05D0u];
112 found = (*ab != 0);
113 } else if (a == 0xFB2Au) { /* SHIN WITH SHIN DOT */
114 *ab = 0xFB2Cu;
115 found = true;
116 } else if (a == 0xFB2Bu) { /* SHIN WITH SIN DOT */
117 *ab = 0xFB2Du;
118 found = true;
119 }
120 break;
121 case 0x05BFu: /* RAFE */
122 switch (a) {
123 case 0x05D1u: /* BET */
124 *ab = 0xFB4Cu;
125 found = true;
126 break;
127 case 0x05DBu: /* KAF */
128 *ab = 0xFB4Du;
129 found = true;
130 break;
131 case 0x05E4u: /* PE */
132 *ab = 0xFB4Eu;
133 found = true;
134 break;
135 }
136 break;
137 case 0x05C1u: /* SHIN DOT */
138 if (a == 0x05E9u) { /* SHIN */
139 *ab = 0xFB2Au;
140 found = true;
141 } else if (a == 0xFB49u) { /* SHIN WITH DAGESH */
142 *ab = 0xFB2Cu;
143 found = true;
144 }
145 break;
146 case 0x05C2u: /* SIN DOT */
147 if (a == 0x05E9u) { /* SHIN */
148 *ab = 0xFB2Bu;
149 found = true;
150 } else if (a == 0xFB49u) { /* SHIN WITH DAGESH */
151 *ab = 0xFB2Du;
152 found = true;
153 }
154 break;
155 }
156 }
157
158 return found;
159 }
160
161
162 const hb_ot_complex_shaper_t _hb_ot_complex_shaper_hebrew =
163 {
164 nullptr, /* collect_features */
165 nullptr, /* override_features */
166 nullptr, /* data_create */
167 nullptr, /* data_destroy */
168 nullptr, /* preprocess_text */
169 nullptr, /* postprocess_glyphs */
170 HB_OT_SHAPE_NORMALIZATION_MODE_DEFAULT,
171 nullptr, /* decompose */
172 compose_hebrew,
173 nullptr, /* setup_masks */
174 HB_TAG ('h','e','b','r'), /* gpos_tag. https://github.com/harfbuzz/harfbuzz/issues/347#issuecomment-267838368 */
175 nullptr, /* reorder_marks */
176 HB_OT_SHAPE_ZERO_WIDTH_MARKS_BY_GDEF_LATE,
177 true, /* fallback_position */
178 };
179