![[first]](../icons/first.png){kind=icon}
![[previous]](../icons/n_left.png){kind=icon}
![[next]](../icons/n_right.png){kind=icon}
![[last]](../icons/last.png){kind=icon}
![[top]](../icons/n_top.png){kind=icon}
![[bottom]](../icons/bottom.png){kind=icon}
![[index]](../icons/index.png){kind=icon}
![[help]](../icons/help.png){kind=icon}
*/
1 /* Line breaking of strings.
2 Copyright (C) 2001-2003, 2006-2021 Free Software Foundation, Inc.
3 Written by Bruno Haible <bruno@clisp.org>, 2001.
4
5 This file is free software.
6 It is dual-licensed under "the GNU LGPLv3+ or the GNU GPLv2+".
7 You can redistribute it and/or modify it under either
8 - the terms of the GNU Lesser General Public License as published
9 by the Free Software Foundation; either version 3, or (at your
10 option) any later version, or
11 - the terms of the GNU General Public License as published by the
12 Free Software Foundation; either version 2, or (at your option)
13 any later version, or
14 - the same dual license "the GNU LGPLv3+ or the GNU GPLv2+".
15
16 This file is distributed in the hope that it will be useful,
17 but WITHOUT ANY WARRANTY; without even the implied warranty of
18 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
19 Lesser General Public License and the GNU General Public License
20 for more details.
21
22 You should have received a copy of the GNU Lesser General Public
23 License and of the GNU General Public License along with this
24 program. If not, see <https://www.gnu.org/licenses/>. */
25
26 #include <config.h>
27
28 /* Specification. */
29 #include "unilbrk.h"
30
31 #include <stdlib.h>
32 #include <string.h>
33
34 #include "c-ctype.h"
35 #include "uniconv.h"
36 #include "unilbrk/ulc-common.h"
37
38 /* Line breaking of a string in an arbitrary encoding.
39
40 We convert the input string to Unicode.
41
42 The standardized Unicode encodings are UTF-8, UCS-2, UCS-4, UTF-16,
43 UTF-16BE, UTF-16LE, UTF-7. UCS-2 supports only characters up to
44 \U0000FFFF. UTF-16 and variants support only characters up to
45 \U0010FFFF. UTF-7 is way too complex and not supported by glibc-2.1.
46 UCS-4 specification leaves doubts about endianness and byte order mark.
47 glibc currently interprets it as big endian without byte order mark,
48 but this is not backed by an RFC. So we use UTF-8. It supports
49 characters up to \U7FFFFFFF and is unambiguously defined. */
50
51 int
52 ulc_width_linebreaks (const char *s, size_t n,
/* ![[next]](../icons/right.png)
![[first]](../icons/n_first.png)
![[top]](../icons/top.png)
*/
53 int width, int start_column, int at_end_columns,
54 const char *o, const char *encoding,
55 char *p)
56 {
57 if (n > 0)
58 {
59 if (is_utf8_encoding (encoding))
60 return u8_width_linebreaks ((const uint8_t *) s, n, width, start_column, at_end_columns, o, encoding, p);
61 else
62 {
63 /* Convert the string to UTF-8 and build a translation table
64 from offsets into s to offsets into the translated string. */
65 size_t *offsets = (size_t *) malloc (n * sizeof (size_t));
66
67 if (offsets != NULL)
68 {
69 uint8_t *t;
70 size_t m;
71
72 t = u8_conv_from_encoding (encoding, iconveh_question_mark,
73 s, n, offsets, NULL, &m);
74 if (t != NULL)
75 {
76 char *memory =
77 (char *) (m > 0 ? malloc (m + (o != NULL ? m : 0)) : NULL);
78
79 if (m == 0 || memory != NULL)
80 {
81 char *q = (char *) memory;
82 char *o8 = (o != NULL ? (char *) (q + m) : NULL);
83 int res_column;
84 size_t i;
85
86 /* Translate the overrides to the UTF-8 string. */
87 if (o != NULL)
88 {
89 memset (o8, UC_BREAK_UNDEFINED, m);
90 for (i = 0; i < n; i++)
91 if (offsets[i] != (size_t)(-1))
92 o8[offsets[i]] = o[i];
93 }
94
95 /* Determine the line breaks of the UTF-8 string. */
96 res_column =
97 u8_width_linebreaks (t, m, width, start_column, at_end_columns, o8, encoding, q);
98
99 /* Translate the result back to the original string. */
100 memset (p, UC_BREAK_PROHIBITED, n);
101 for (i = 0; i < n; i++)
102 if (offsets[i] != (size_t)(-1))
103 p[i] = q[offsets[i]];
104
105 free (memory);
106 free (t);
107 free (offsets);
108 return res_column;
109 }
110 free (t);
111 }
112 free (offsets);
113 }
114 /* Impossible to convert. */
115 #if C_CTYPE_ASCII
116 if (is_all_ascii (s, n))
117 {
118 /* ASCII is a subset of UTF-8. */
119 return u8_width_linebreaks ((const uint8_t *) s, n, width, start_column, at_end_columns, o, encoding, p);
120 }
121 #endif
122 /* We have a non-ASCII string and cannot convert it.
123 Don't produce line breaks except those already present in the
124 input string. All we assume here is that the encoding is
125 minimally ASCII compatible. */
126 {
127 const char *s_end = s + n;
128 while (s < s_end)
129 {
130 *p = ((o != NULL && *o == UC_BREAK_MANDATORY) || *s == '\n'
131 ? UC_BREAK_MANDATORY
132 : UC_BREAK_PROHIBITED);
133 s++;
134 p++;
135 if (o != NULL)
136 o++;
137 }
138 /* We cannot compute widths in this case. */
139 }
140 }
141 }
142 return start_column;
143 }
144
145
146 #ifdef TEST
147
148 #include <stdio.h>
149 #include <locale.h>
150
151 /* Read the contents of an input stream, and return it, terminated with a NUL
152 byte. */
153 char *
154 read_file (FILE *stream)
/* ![[previous]](../icons/left.png)
*/
155 {
156 #define BUFSIZE 4096
157 char *buf = NULL;
158 int alloc = 0;
159 int size = 0;
160 int count;
161
162 while (! feof (stream))
163 {
164 if (size + BUFSIZE > alloc)
165 {
166 alloc = alloc + alloc / 2;
167 if (alloc < size + BUFSIZE)
168 alloc = size + BUFSIZE;
169 buf = realloc (buf, alloc);
170 if (buf == NULL)
171 {
172 fprintf (stderr, "out of memory\n");
173 exit (1);
174 }
175 }
176 count = fread (buf + size, 1, BUFSIZE, stream);
177 if (count == 0)
178 {
179 if (ferror (stream))
180 {
181 perror ("fread");
182 exit (1);
183 }
184 }
185 else
186 size += count;
187 }
188 buf = realloc (buf, size + 1);
189 if (buf == NULL)
190 {
191 fprintf (stderr, "out of memory\n");
192 exit (1);
193 }
194 buf[size] = '\0';
195 return buf;
196 #undef BUFSIZE
197 }
198
199 int
200 main (int argc, char * argv[])
/* ![[last]](../icons/n_last.png)
*/
201 {
202 setlocale (LC_CTYPE, "");
203 if (argc == 2)
204 {
205 /* Insert line breaks for a given width. */
206 int width = atoi (argv[1]);
207 char *input = read_file (stdin);
208 int length = strlen (input);
209 char *breaks = malloc (length);
210 int i;
211
212 ulc_width_linebreaks (input, length, width, 0, 0, NULL, locale_charset (), breaks);
213
214 for (i = 0; i < length; i++)
215 {
216 switch (breaks[i])
217 {
218 case UC_BREAK_POSSIBLE:
219 putc ('\n', stdout);
220 break;
221 case UC_BREAK_MANDATORY:
222 break;
223 case UC_BREAK_PROHIBITED:
224 break;
225 default:
226 abort ();
227 }
228 putc (input[i], stdout);
229 }
230
231 free (breaks);
232
233 return 0;
234 }
235 else
236 return 1;
237 }
238
239 #endif /* TEST */
![[bottom]](../icons/n_bottom.png){kind=icon}
*/