1 /* Line breaking of UTF-8 strings.
2 Copyright (C) 2001-2003, 2006-2010 Free Software Foundation, Inc.
3 Written by Bruno Haible <bruno@clisp.org>, 2001.
5 This program is free software: you can redistribute it and/or modify it
6 under the terms of the GNU Lesser General Public License as published
7 by the Free Software Foundation; either version 3 of the License, or
8 (at your option) any later version.
10 This program is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 Lesser General Public License for more details.
15 You should have received a copy of the GNU Lesser General Public License
16 along with this program. If not, see <http://www.gnu.org/licenses/>. */
26 #include "unilbrk/lbrktables.h"
27 #include "uniwidth/cjk.h"
31 u8_possible_linebreaks (const uint8_t *s, size_t n, const char *encoding, char *p)
33 int LBP_AI_REPLACEMENT = (is_cjk_encoding (encoding) ? LBP_ID : LBP_AL);
34 const uint8_t *s_end = s + n;
35 int last_prop = LBP_BK; /* line break property of last non-space character */
36 char *seen_space = NULL; /* Was a space seen after the last non-space character? */
37 char *seen_space2 = NULL; /* At least two spaces after the last non-space? */
39 /* Don't break inside multibyte characters. */
40 memset (p, UC_BREAK_PROHIBITED, n);
45 int count = u8_mbtouc_unsafe (&uc, s, s_end - s);
46 int prop = unilbrkprop_lookup (uc);
50 /* Mandatory break. */
51 *p = UC_BREAK_MANDATORY;
60 /* Resolve property values whose behaviour is not fixed. */
64 /* Resolve ambiguous. */
65 prop = LBP_AI_REPLACEMENT;
68 /* This is arbitrary. */
72 /* We don't handle complex scripts yet.
73 Treat LBP_SA like LBP_XX. */
75 /* This is arbitrary. */
80 /* Deal with spaces and combining characters. */
84 /* Don't break just before a space. */
85 *p = UC_BREAK_PROHIBITED;
86 seen_space2 = seen_space;
89 else if (prop == LBP_ZW)
91 /* Don't break just before a zero-width space. */
92 *p = UC_BREAK_PROHIBITED;
97 else if (prop == LBP_CM)
99 /* Don't break just before a combining character, except immediately after a
101 if (last_prop == LBP_ZW)
103 /* Break after zero-width space. */
104 *p = UC_BREAK_POSSIBLE;
105 /* A combining character turns a preceding space into LBP_ID. */
110 *p = UC_BREAK_PROHIBITED;
111 /* A combining character turns a preceding space into LBP_ID. */
112 if (seen_space != NULL)
115 seen_space = seen_space2;
117 goto lookup_via_table;
124 /* prop must be usable as an index for table 7.3 of UTR #14. */
125 if (!(prop >= 0 && prop < sizeof (unilbrk_table) / sizeof (unilbrk_table[0])))
128 if (last_prop == LBP_BK)
130 /* Don't break at the beginning of a line. */
131 *q = UC_BREAK_PROHIBITED;
133 else if (last_prop == LBP_ZW)
135 /* Break after zero-width space. */
136 *q = UC_BREAK_POSSIBLE;
140 switch (unilbrk_table [last_prop] [prop])
143 *q = UC_BREAK_POSSIBLE;
146 *q = (seen_space != NULL ? UC_BREAK_POSSIBLE : UC_BREAK_PROHIBITED);
149 *q = UC_BREAK_PROHIBITED;
172 /* Read the contents of an input stream, and return it, terminated with a NUL
175 read_file (FILE *stream)
183 while (! feof (stream))
185 if (size + BUFSIZE > alloc)
187 alloc = alloc + alloc / 2;
188 if (alloc < size + BUFSIZE)
189 alloc = size + BUFSIZE;
190 buf = realloc (buf, alloc);
193 fprintf (stderr, "out of memory\n");
197 count = fread (buf + size, 1, BUFSIZE, stream);
209 buf = realloc (buf, size + 1);
212 fprintf (stderr, "out of memory\n");
221 main (int argc, char * argv[])
225 /* Display all the break opportunities in the input string. */
226 char *input = read_file (stdin);
227 int length = strlen (input);
228 char *breaks = malloc (length);
231 u8_possible_linebreaks ((uint8_t *) input, length, "UTF-8", breaks);
233 for (i = 0; i < length; i++)
237 case UC_BREAK_POSSIBLE:
238 /* U+2027 in UTF-8 encoding */
239 putc (0xe2, stdout); putc (0x80, stdout); putc (0xa7, stdout);
241 case UC_BREAK_MANDATORY:
242 /* U+21B2 (or U+21B5) in UTF-8 encoding */
243 putc (0xe2, stdout); putc (0x86, stdout); putc (0xb2, stdout);
245 case UC_BREAK_PROHIBITED:
250 putc (input[i], stdout);