1 /* -*- buffer-read-only: t -*- vi: set ro: */
2 /* DO NOT EDIT! GENERATED AUTOMATICALLY! */
3 /* Iterating through multibyte strings: macros for multi-byte encodings.
4 Copyright (C) 2001, 2005, 2007, 2009-2014 Free Software Foundation, Inc.
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 3 of the License, or
9 (at your option) any later version.
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.
16 You should have received a copy of the GNU General Public License
17 along with this program. If not, see <http://www.gnu.org/licenses/>. */
19 /* Written by Bruno Haible <bruno@clisp.org>. */
21 /* The macros in this file implement forward iteration through a
22 multi-byte string, without knowing its length a-priori.
24 With these macros, an iteration loop that looks like
27 for (iter = buf; *iter != '\0'; iter++)
35 for (mbui_init (iter, buf); mbui_avail (iter); mbui_advance (iter))
37 do_something (mbui_cur_ptr (iter), mb_len (mbui_cur (iter)));
40 The benefit of these macros over plain use of mbrtowc is:
41 - Handling of invalid multibyte sequences is possible without
42 making the code more complicated, while still preserving the
43 invalid multibyte sequences.
45 Compared to mbiter.h, the macros here don't need to know the string's
46 length a-priori. The downside is that at each step, the look-ahead
47 that guards against overrunning the terminating '\0' is more expensive.
48 The mbui_* macros are therefore suitable when there is a high probability
49 that only the first few multibyte characters need to be inspected.
50 Whereas the mbi_* macros are better if usually the iteration runs
51 through the entire string.
54 is a type usable for variable declarations.
56 mbui_init (iter, startptr)
57 initializes the iterator, starting at startptr.
60 returns true if there are more multibyte characters available before
61 the end of string is reached. In this case, mbui_cur (iter) is
62 initialized to the next multibyte character.
65 advances the iterator by one multibyte character.
68 returns the current multibyte character, of type mbchar_t. All the
69 macros defined in mbchar.h can be used on it.
72 return a pointer to the beginning of the current multibyte character.
74 mbui_reloc (iter, ptrdiff)
75 relocates iterator when the string is moved by ptrdiff bytes.
77 mbui_copy (&destiter, &srciter)
78 copies srciter to destiter.
80 Here are the function prototypes of the macros.
82 extern void mbui_init (mbui_iterator_t iter, const char *startptr);
83 extern bool mbui_avail (mbui_iterator_t iter);
84 extern void mbui_advance (mbui_iterator_t iter);
85 extern mbchar_t mbui_cur (mbui_iterator_t iter);
86 extern const char * mbui_cur_ptr (mbui_iterator_t iter);
87 extern void mbui_reloc (mbui_iterator_t iter, ptrdiff_t ptrdiff);
88 extern void mbui_copy (mbui_iterator_t *new, const mbui_iterator_t *old);
100 /* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before
102 BSD/OS 4.1 has a bug: <stdio.h> and <time.h> must be included before
109 #include "strnlen1.h"
111 #ifndef _GL_INLINE_HEADER_BEGIN
112 #error "Please include config.h first."
114 _GL_INLINE_HEADER_BEGIN
115 #ifndef MBUITER_INLINE
116 # define MBUITER_INLINE _GL_INLINE
121 bool in_shift; /* true if next byte may not be interpreted as ASCII */
122 mbstate_t state; /* if in_shift: current shift state */
123 bool next_done; /* true if mbui_avail has already filled the following */
124 struct mbchar cur; /* the current character:
125 const char *cur.ptr pointer to current character
126 The following are only valid after mbui_avail.
127 size_t cur.bytes number of bytes of current character
128 bool cur.wc_valid true if wc is a valid wide character
129 wchar_t cur.wc if wc_valid: the current character
134 mbuiter_multi_next (struct mbuiter_multi *iter)
140 /* Handle most ASCII characters quickly, without calling mbrtowc(). */
141 if (is_basic (*iter->cur.ptr))
143 /* These characters are part of the basic character set. ISO C 99
144 guarantees that their wide character code is identical to their
147 iter->cur.wc = *iter->cur.ptr;
148 iter->cur.wc_valid = true;
152 assert (mbsinit (&iter->state));
153 iter->in_shift = true;
155 iter->cur.bytes = mbrtowc (&iter->cur.wc, iter->cur.ptr,
156 strnlen1 (iter->cur.ptr, MB_CUR_MAX),
158 if (iter->cur.bytes == (size_t) -1)
160 /* An invalid multibyte sequence was encountered. */
162 iter->cur.wc_valid = false;
163 /* Whether to set iter->in_shift = false and reset iter->state
164 or not is not very important; the string is bogus anyway. */
166 else if (iter->cur.bytes == (size_t) -2)
168 /* An incomplete multibyte character at the end. */
169 iter->cur.bytes = strlen (iter->cur.ptr);
170 iter->cur.wc_valid = false;
171 /* Whether to set iter->in_shift = false and reset iter->state
172 or not is not important; the string end is reached anyway. */
176 if (iter->cur.bytes == 0)
178 /* A null wide character was encountered. */
180 assert (*iter->cur.ptr == '\0');
181 assert (iter->cur.wc == 0);
183 iter->cur.wc_valid = true;
185 /* When in the initial state, we can go back treating ASCII
186 characters more quickly. */
187 if (mbsinit (&iter->state))
188 iter->in_shift = false;
191 iter->next_done = true;
195 mbuiter_multi_reloc (struct mbuiter_multi *iter, ptrdiff_t ptrdiff)
197 iter->cur.ptr += ptrdiff;
201 mbuiter_multi_copy (struct mbuiter_multi *new_iter, const struct mbuiter_multi *old_iter)
203 if ((new_iter->in_shift = old_iter->in_shift))
204 memcpy (&new_iter->state, &old_iter->state, sizeof (mbstate_t));
206 memset (&new_iter->state, 0, sizeof (mbstate_t));
207 new_iter->next_done = old_iter->next_done;
208 mb_copy (&new_iter->cur, &old_iter->cur);
211 /* Iteration macros. */
212 typedef struct mbuiter_multi mbui_iterator_t;
213 #define mbui_init(iter, startptr) \
214 ((iter).cur.ptr = (startptr), \
215 (iter).in_shift = false, memset (&(iter).state, '\0', sizeof (mbstate_t)), \
216 (iter).next_done = false)
217 #define mbui_avail(iter) \
218 (mbuiter_multi_next (&(iter)), !mb_isnul ((iter).cur))
219 #define mbui_advance(iter) \
220 ((iter).cur.ptr += (iter).cur.bytes, (iter).next_done = false)
222 /* Access to the current character. */
223 #define mbui_cur(iter) (iter).cur
224 #define mbui_cur_ptr(iter) (iter).cur.ptr
227 #define mbui_reloc(iter, ptrdiff) mbuiter_multi_reloc (&iter, ptrdiff)
229 /* Copying an iterator. */
230 #define mbui_copy mbuiter_multi_copy
232 _GL_INLINE_HEADER_END
234 #endif /* _MBUITER_H */