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_UTF_PRIVATE_HH
28 #define HB_UTF_PRIVATE_HH
29
30 #include "hb-private.hh"
31
32
33 /* UTF-8 */
34
35 #define HB_UTF8_COMPUTE(Char, Mask, Len) \
36 if (Char < 128) { Len = 1; Mask = 0x7f; } \
37 else if ((Char & 0xe0) == 0xc0) { Len = 2; Mask = 0x1f; } \
38 else if ((Char & 0xf0) == 0xe0) { Len = 3; Mask = 0x0f; } \
39 else if ((Char & 0xf8) == 0xf0) { Len = 4; Mask = 0x07; } \
40 else Len = 0;
41
42 static inline const uint8_t *
hb_utf_next(const uint8_t * text,const uint8_t * end,hb_codepoint_t * unicode)43 hb_utf_next (const uint8_t *text,
44 const uint8_t *end,
45 hb_codepoint_t *unicode)
46 {
47 hb_codepoint_t c = *text, mask;
48 unsigned int len;
49
50 /* TODO check for overlong sequences? */
51
52 HB_UTF8_COMPUTE (c, mask, len);
53 if (unlikely (!len || (unsigned int) (end - text) < len)) {
54 *unicode = -1;
55 return text + 1;
56 } else {
57 hb_codepoint_t result;
58 unsigned int i;
59 result = c & mask;
60 for (i = 1; i < len; i++)
61 {
62 if (unlikely ((text[i] & 0xc0) != 0x80))
63 {
64 *unicode = -1;
65 return text + 1;
66 }
67 result <<= 6;
68 result |= (text[i] & 0x3f);
69 }
70 *unicode = result;
71 return text + len;
72 }
73 }
74
75 static inline const uint8_t *
hb_utf_prev(const uint8_t * text,const uint8_t * start,hb_codepoint_t * unicode)76 hb_utf_prev (const uint8_t *text,
77 const uint8_t *start,
78 hb_codepoint_t *unicode)
79 {
80 const uint8_t *end = text--;
81 while (start < text && (*text & 0xc0) == 0x80 && end - text < 4)
82 text--;
83
84 hb_codepoint_t c = *text, mask;
85 unsigned int len;
86
87 /* TODO check for overlong sequences? */
88
89 HB_UTF8_COMPUTE (c, mask, len);
90 if (unlikely (!len || (unsigned int) (end - text) != len)) {
91 *unicode = -1;
92 return end - 1;
93 } else {
94 hb_codepoint_t result;
95 unsigned int i;
96 result = c & mask;
97 for (i = 1; i < len; i++)
98 {
99 result <<= 6;
100 result |= (text[i] & 0x3f);
101 }
102 *unicode = result;
103 return text;
104 }
105 }
106
107
108 static inline unsigned int
hb_utf_strlen(const uint8_t * text)109 hb_utf_strlen (const uint8_t *text)
110 {
111 return strlen ((const char *) text);
112 }
113
114
115 /* UTF-16 */
116
117 static inline const uint16_t *
hb_utf_next(const uint16_t * text,const uint16_t * end,hb_codepoint_t * unicode)118 hb_utf_next (const uint16_t *text,
119 const uint16_t *end,
120 hb_codepoint_t *unicode)
121 {
122 hb_codepoint_t c = *text++;
123
124 if (unlikely (hb_in_range<hb_codepoint_t> (c, 0xd800, 0xdbff)))
125 {
126 /* high surrogate */
127 hb_codepoint_t l;
128 if (text < end && ((l = *text), likely (hb_in_range<hb_codepoint_t> (l, 0xdc00, 0xdfff))))
129 {
130 /* low surrogate */
131 *unicode = (c << 10) + l - ((0xd800 << 10) - 0x10000 + 0xdc00);
132 text++;
133 } else
134 *unicode = -1;
135 } else
136 *unicode = c;
137
138 return text;
139 }
140
141 static inline const uint16_t *
hb_utf_prev(const uint16_t * text,const uint16_t * start,hb_codepoint_t * unicode)142 hb_utf_prev (const uint16_t *text,
143 const uint16_t *start,
144 hb_codepoint_t *unicode)
145 {
146 hb_codepoint_t c = *--text;
147
148 if (unlikely (hb_in_range<hb_codepoint_t> (c, 0xdc00, 0xdfff)))
149 {
150 /* low surrogate */
151 hb_codepoint_t h;
152 if (start < text && ((h = *(text - 1)), likely (hb_in_range<hb_codepoint_t> (h, 0xd800, 0xdbff))))
153 {
154 /* high surrogate */
155 *unicode = (h << 10) + c - ((0xd800 << 10) - 0x10000 + 0xdc00);
156 text--;
157 } else
158 *unicode = -1;
159 } else
160 *unicode = c;
161
162 return text;
163 }
164
165
166 static inline unsigned int
hb_utf_strlen(const uint16_t * text)167 hb_utf_strlen (const uint16_t *text)
168 {
169 unsigned int l = 0;
170 while (*text++) l++;
171 return l;
172 }
173
174
175 /* UTF-32 */
176
177 static inline const uint32_t *
hb_utf_next(const uint32_t * text,const uint32_t * end HB_UNUSED,hb_codepoint_t * unicode)178 hb_utf_next (const uint32_t *text,
179 const uint32_t *end HB_UNUSED,
180 hb_codepoint_t *unicode)
181 {
182 *unicode = *text++;
183 return text;
184 }
185
186 static inline const uint32_t *
hb_utf_prev(const uint32_t * text,const uint32_t * start HB_UNUSED,hb_codepoint_t * unicode)187 hb_utf_prev (const uint32_t *text,
188 const uint32_t *start HB_UNUSED,
189 hb_codepoint_t *unicode)
190 {
191 *unicode = *--text;
192 return text;
193 }
194
195 static inline unsigned int
hb_utf_strlen(const uint32_t * text)196 hb_utf_strlen (const uint32_t *text)
197 {
198 unsigned int l = 0;
199 while (*text++) l++;
200 return l;
201 }
202
203
204 #endif /* HB_UTF_PRIVATE_HH */
205