1 /* Recode Serbian text from Cyrillic to Latin script.
2 Copyright (C) 2006 Free Software Foundation, Inc.
3 Written by Danilo Šegan <danilo@gnome.org>, 2006,
4 and Bruno Haible <bruno@clisp.org>, 2006.
5
6 This program is free software; you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 2, or (at your option)
9 any later version.
10
11 This program is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
15
16 You should have received a copy of the GNU General Public License
17 along with this program; if not, write to the Free Software Foundation,
18 Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
19
20 #ifdef HAVE_CONFIG_H
21 # include <config.h>
22 #endif
23
24 /* Specification. */
25 #include "filters.h"
26
27 #include <stdlib.h>
28
29 #include "xalloc.h"
30
31
32 /* Table for Serbian Cyrillic to Latin transcription.
33 The table is indexed by the Unicode code point, in the range 0x0400..0x045f.
34 The longest table entry is three bytes long. */
35 static const char table[96][3 + 1] =
36 {
37 /* U+0400 */ "",
38 /* U+0401 */ "",
39 /* U+0402 */ "\xC4\x90", /* "Đ" */
40 /* U+0403 */ "",
41 /* U+0404 */ "",
42 /* U+0405 */ "",
43 /* U+0406 */ "",
44 /* U+0407 */ "",
45 /* U+0408 */ "J",
46 /* U+0409 */ "Lj",
47 /* U+040A */ "Nj",
48 /* U+040B */ "\xC4\x86", /* "Ć" */
49 /* U+040C */ "",
50 /* U+040D */ "",
51 /* U+040E */ "",
52 /* U+040F */ "D\xC5\xBE", /* "Dž" */
53 /* U+0410 */ "A",
54 /* U+0411 */ "B",
55 /* U+0412 */ "V",
56 /* U+0413 */ "G",
57 /* U+0414 */ "D",
58 /* U+0415 */ "E",
59 /* U+0416 */ "\xC5\xBD", /* "Ž" */
60 /* U+0417 */ "Z",
61 /* U+0418 */ "I",
62 /* U+0419 */ "",
63 /* U+041A */ "K",
64 /* U+041B */ "L",
65 /* U+041C */ "M",
66 /* U+041D */ "N",
67 /* U+041E */ "O",
68 /* U+041F */ "P",
69 /* U+0420 */ "R",
70 /* U+0421 */ "S",
71 /* U+0422 */ "T",
72 /* U+0423 */ "U",
73 /* U+0424 */ "F",
74 /* U+0425 */ "H",
75 /* U+0426 */ "C",
76 /* U+0427 */ "\xC4\x8C", /* "Č" */
77 /* U+0428 */ "\xC5\xA0", /* "Š" */
78 /* U+0429 */ "",
79 /* U+042A */ "",
80 /* U+042B */ "",
81 /* U+042C */ "",
82 /* U+042D */ "",
83 /* U+042E */ "",
84 /* U+042F */ "",
85 /* U+0430 */ "a",
86 /* U+0431 */ "b",
87 /* U+0432 */ "v",
88 /* U+0433 */ "g",
89 /* U+0434 */ "d",
90 /* U+0435 */ "e",
91 /* U+0436 */ "\xC5\xBE", /* "ž" */
92 /* U+0437 */ "z",
93 /* U+0438 */ "i",
94 /* U+0439 */ "",
95 /* U+043A */ "k",
96 /* U+043B */ "l",
97 /* U+043C */ "m",
98 /* U+043D */ "n",
99 /* U+043E */ "o",
100 /* U+043F */ "p",
101 /* U+0440 */ "r",
102 /* U+0441 */ "s",
103 /* U+0442 */ "t",
104 /* U+0443 */ "u",
105 /* U+0444 */ "f",
106 /* U+0445 */ "h",
107 /* U+0446 */ "c",
108 /* U+0447 */ "\xC4\x8D", /* "č" */
109 /* U+0448 */ "\xC5\xA1", /* "š" */
110 /* U+0449 */ "",
111 /* U+044A */ "",
112 /* U+044B */ "",
113 /* U+044C */ "",
114 /* U+044D */ "",
115 /* U+044E */ "",
116 /* U+044F */ "",
117 /* U+0450 */ "",
118 /* U+0451 */ "",
119 /* U+0452 */ "\xC4\x91", /* "đ" */
120 /* U+0453 */ "",
121 /* U+0454 */ "",
122 /* U+0455 */ "",
123 /* U+0456 */ "",
124 /* U+0457 */ "",
125 /* U+0458 */ "j",
126 /* U+0459 */ "lj",
127 /* U+045A */ "nj",
128 /* U+045B */ "\xC4\x87", /* "ć" */
129 /* U+045C */ "",
130 /* U+045D */ "",
131 /* U+045E */ "",
132 /* U+045F */ "d\xC5\xBE" /* "dž" */
133 };
134
135 /* Quick test for an uppercase character in the range U+0041..U+005A.
136 The argument must be a byte in the range 0..UCHAR_MAX. */
137 #define IS_UPPERCASE_LATIN(byte) \
138 ((unsigned char) ((byte) - 'A') <= 'Z' - 'A')
139
140 /* Quick test for an uppercase character in the range U+0400..U+042F.
141 The arguments must be bytes in the range 0..UCHAR_MAX. */
142 #define IS_UPPERCASE_CYRILLIC(byte1,byte2) \
143 ((byte1) == 0xd0 && (unsigned char) ((byte2) - 0x80) < 0x30)
144
145 void
serbian_to_latin(const char * input,size_t input_len,char ** output_p,size_t * output_len_p)146 serbian_to_latin (const char *input, size_t input_len,
147 char **output_p, size_t *output_len_p)
148 {
149 /* Loop through the input string, producing a replacement for each character.
150 Only characters in the range U+0400..U+045F (\xD0\x80..\xD1\x9F) need to
151 be handled, and more precisely only those for which a replacement exists
152 in the table. Other characters are copied without modification.
153 The characters U+0409, U+040A, U+040F are transliterated to uppercase or
154 mixed-case replacements ("LJ" / "Lj", "NJ" / "Nj", "DŽ" / "Dž"), depending
155 on the case of the surrounding characters.
156 Since we assume UTF-8 encoding, the bytes \xD0..\xD1 can only occur at the
157 beginning of a character; the second and further bytes of a character are
158 all in the range \x80..\xBF. */
159
160 /* Since sequences of 2 bytes are sequences of at most 3 bytes, the size
161 of the output will be at most 1.5 * input_len. */
162 size_t allocated = input_len + (input_len >> 1);
163 char *output = (char *) xmalloc (allocated);
164
165 const char *input_end = input + input_len;
166 const char *ip;
167 char *op;
168
169 for (ip = input, op = output; ip < input_end; )
170 {
171 unsigned char byte = (unsigned char) *ip;
172
173 /* Test for the first byte of a Cyrillic character. */
174 if ((byte >= 0xd0 && byte <= 0xd1) && (ip + 1 < input_end))
175 {
176 unsigned char second_byte = (unsigned char) ip[1];
177
178 /* Verify the second byte is valid. */
179 if (second_byte >= 0x80 && second_byte < 0xc0)
180 {
181 unsigned int uc = ((byte & 0x1f) << 6) | (second_byte & 0x3f);
182
183 if (uc >= 0x0400 && uc <= 0x045f)
184 {
185 /* Look up replacement from the table. */
186 const char *repl = table[uc - 0x0400];
187
188 if (repl[0] != '\0')
189 {
190 /* Found a replacement.
191 Now handle the special cases. */
192 if (uc == 0x0409 || uc == 0x040a || uc == 0x040f)
193 if ((ip + 2 < input_end
194 && IS_UPPERCASE_LATIN ((unsigned char) ip[2]))
195 || (ip + 3 < input_end
196 && IS_UPPERCASE_CYRILLIC ((unsigned char) ip[2],
197 (unsigned char) ip[3]))
198 || (ip >= input + 1
199 && IS_UPPERCASE_LATIN ((unsigned char) ip[-1]))
200 || (ip >= input + 2
201 && IS_UPPERCASE_CYRILLIC ((unsigned char) ip[-2],
202 (unsigned char) ip[-1])))
203 {
204 /* Use the upper-case replacement instead of
205 the mixed-case replacement. */
206 switch (uc)
207 {
208 case 0x0409:
209 repl = "LJ"; break;
210 case 0x040a:
211 repl = "NJ"; break;
212 case 0x040f:
213 repl = "D\xC5\xBD"/* "DŽ" */; break;
214 default:
215 abort ();
216 }
217 }
218
219 /* Use the replacement. */
220 *op++ = *repl++;
221 if (*repl != '\0')
222 {
223 *op++ = *repl++;
224 if (*repl != '\0')
225 {
226 *op++ = *repl++;
227 /* All replacements have at most 3 bytes. */
228 if (*repl != '\0')
229 abort ();
230 }
231 }
232 ip += 2;
233 continue;
234 }
235 }
236 }
237 }
238 *op++ = *ip++;
239 }
240
241 {
242 size_t output_len = op - output;
243
244 /* Verify that the allocated size was not exceeded. */
245 if (output_len > allocated)
246 abort ();
247 /* Shrink the result. */
248 if (output_len < allocated)
249 output = (char *) xrealloc (output, output_len);
250
251 /* Done. */
252 *output_p = output;
253 *output_len_p = output_len;
254 }
255 }
256