contrib/diffutils/lib/mbiter.h

   1 /* Iterating through multibyte strings: macros for multi-byte encodings.
   2    Copyright (C) 2001, 2005, 2007, 2009-2013 Free Software Foundation, Inc.
   3
   4    This program is free software: you can redistribute it and/or modify
   5    it under the terms of the GNU General Public License as published by
   6    the Free Software Foundation; either version 3 of the License, or
   7    (at your option) any later version.
   8
   9    This program is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12    GNU General Public License for more details.
  13
  14    You should have received a copy of the GNU General Public License
  15    along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
  16
  17 /* Written by Bruno Haible <bruno@clisp.org>.  */
  18
  19 /* The macros in this file implement forward iteration through a
  20    multi-byte string.
  21
  22    With these macros, an iteration loop that looks like
  23
  24       char *iter;
  25       for (iter = buf; iter < buf + buflen; iter++)
  26         {
  27           do_something (*iter);
  28         }
  29
  30    becomes
  31
  32       mbi_iterator_t iter;
  33       for (mbi_init (iter, buf, buflen); mbi_avail (iter); mbi_advance (iter))
  34         {
  35           do_something (mbi_cur_ptr (iter), mb_len (mbi_cur (iter)));
  36         }
  37
  38    The benefit of these macros over plain use of mbrtowc is:
  39    - Handling of invalid multibyte sequences is possible without
  40      making the code more complicated, while still preserving the
  41      invalid multibyte sequences.
  42
  43    mbi_iterator_t
  44      is a type usable for variable declarations.
  45
  46    mbi_init (iter, startptr, length)
  47      initializes the iterator, starting at startptr and crossing length bytes.
  48
  49    mbi_avail (iter)
  50      returns true if there are more multibyte characters available before
  51      the end of string is reached. In this case, mbi_cur (iter) is
  52      initialized to the next multibyte character.
  53
  54    mbi_advance (iter)
  55      advances the iterator by one multibyte character.
  56
  57    mbi_cur (iter)
  58      returns the current multibyte character, of type mbchar_t.  All the
  59      macros defined in mbchar.h can be used on it.
  60
  61    mbi_cur_ptr (iter)
  62      return a pointer to the beginning of the current multibyte character.
  63
  64    mbi_reloc (iter, ptrdiff)
  65      relocates iterator when the string is moved by ptrdiff bytes.
  66
  67    mbi_copy (&destiter, &srciter)
  68      copies srciter to destiter.
  69
  70    Here are the function prototypes of the macros.
  71
  72    extern void          mbi_init (mbi_iterator_t iter,
  73                                   const char *startptr, size_t length);
  74    extern bool          mbi_avail (mbi_iterator_t iter);
  75    extern void          mbi_advance (mbi_iterator_t iter);
  76    extern mbchar_t      mbi_cur (mbi_iterator_t iter);
  77    extern const char *  mbi_cur_ptr (mbi_iterator_t iter);
  78    extern void          mbi_reloc (mbi_iterator_t iter, ptrdiff_t ptrdiff);
  79    extern void          mbi_copy (mbi_iterator_t *new, const mbi_iterator_t *old);
  80  */
  81
  82 #ifndef _MBITER_H
  83 #define _MBITER_H 1
  84
  85 #include <assert.h>
  86 #include <stdbool.h>
  87 #include <stddef.h>
  88 #include <string.h>
  89
  90 /* Tru64 with Desktop Toolkit C has a bug: <stdio.h> must be included before
  91    <wchar.h>.
  92    BSD/OS 4.1 has a bug: <stdio.h> and <time.h> must be included before
  93    <wchar.h>.  */
  94 #include <stdio.h>
  95 #include <time.h>
  96 #include <wchar.h>
  97
  98 #include "mbchar.h"
  99
 100 _GL_INLINE_HEADER_BEGIN
 101 #ifndef MBITER_INLINE
 102 # define MBITER_INLINE _GL_INLINE
 103 #endif
 104
 105 struct mbiter_multi
 106 {
 107   const char *limit;    /* pointer to end of string */
 108   bool in_shift;        /* true if next byte may not be interpreted as ASCII */
 109   mbstate_t state;      /* if in_shift: current shift state */
 110   bool next_done;       /* true if mbi_avail has already filled the following */
 111   struct mbchar cur;    /* the current character:
 112         const char *cur.ptr             pointer to current character
 113         The following are only valid after mbi_avail.
 114         size_t cur.bytes                number of bytes of current character
 115         bool cur.wc_valid               true if wc is a valid wide character
 116         wchar_t cur.wc                  if wc_valid: the current character
 117         */
 118 };
 119
 120 MBITER_INLINE void
 121 mbiter_multi_next (struct mbiter_multi *iter)
 122 {
 123   if (iter->next_done)
 124     return;
 125   if (iter->in_shift)
 126     goto with_shift;
 127   /* Handle most ASCII characters quickly, without calling mbrtowc().  */
 128   if (is_basic (*iter->cur.ptr))
 129     {
 130       /* These characters are part of the basic character set.  ISO C 99
 131          guarantees that their wide character code is identical to their
 132          char code.  */
 133       iter->cur.bytes = 1;
 134       iter->cur.wc = *iter->cur.ptr;
 135       iter->cur.wc_valid = true;
 136     }
 137   else
 138     {
 139       assert (mbsinit (&iter->state));
 140       iter->in_shift = true;
 141     with_shift:
 142       iter->cur.bytes = mbrtowc (&iter->cur.wc, iter->cur.ptr,
 143                                  iter->limit - iter->cur.ptr, &iter->state);
 144       if (iter->cur.bytes == (size_t) -1)
 145         {
 146           /* An invalid multibyte sequence was encountered.  */
 147           iter->cur.bytes = 1;
 148           iter->cur.wc_valid = false;
 149           /* Whether to set iter->in_shift = false and reset iter->state
 150              or not is not very important; the string is bogus anyway.  */
 151         }
 152       else if (iter->cur.bytes == (size_t) -2)
 153         {
 154           /* An incomplete multibyte character at the end.  */
 155           iter->cur.bytes = iter->limit - iter->cur.ptr;
 156           iter->cur.wc_valid = false;
 157           /* Whether to set iter->in_shift = false and reset iter->state
 158              or not is not important; the string end is reached anyway.  */
 159         }
 160       else
 161         {
 162           if (iter->cur.bytes == 0)
 163             {
 164               /* A null wide character was encountered.  */
 165               iter->cur.bytes = 1;
 166               assert (*iter->cur.ptr == '\0');
 167               assert (iter->cur.wc == 0);
 168             }
 169           iter->cur.wc_valid = true;
 170
 171           /* When in the initial state, we can go back treating ASCII
 172              characters more quickly.  */
 173           if (mbsinit (&iter->state))
 174             iter->in_shift = false;
 175         }
 176     }
 177   iter->next_done = true;
 178 }
 179
 180 MBITER_INLINE void
 181 mbiter_multi_reloc (struct mbiter_multi *iter, ptrdiff_t ptrdiff)
 182 {
 183   iter->cur.ptr += ptrdiff;
 184   iter->limit += ptrdiff;
 185 }
 186
 187 MBITER_INLINE void
 188 mbiter_multi_copy (struct mbiter_multi *new_iter, const struct mbiter_multi *old_iter)
 189 {
 190   new_iter->limit = old_iter->limit;
 191   if ((new_iter->in_shift = old_iter->in_shift))
 192     memcpy (&new_iter->state, &old_iter->state, sizeof (mbstate_t));
 193   else
 194     memset (&new_iter->state, 0, sizeof (mbstate_t));
 195   new_iter->next_done = old_iter->next_done;
 196   mb_copy (&new_iter->cur, &old_iter->cur);
 197 }
 198
 199 /* Iteration macros.  */
 200 typedef struct mbiter_multi mbi_iterator_t;
 201 #define mbi_init(iter, startptr, length) \
 202   ((iter).cur.ptr = (startptr), (iter).limit = (iter).cur.ptr + (length), \
 203    (iter).in_shift = false, memset (&(iter).state, '\0', sizeof (mbstate_t)), \
 204    (iter).next_done = false)
 205 #define mbi_avail(iter) \
 206   ((iter).cur.ptr < (iter).limit && (mbiter_multi_next (&(iter)), true))
 207 #define mbi_advance(iter) \
 208   ((iter).cur.ptr += (iter).cur.bytes, (iter).next_done = false)
 209
 210 /* Access to the current character.  */
 211 #define mbi_cur(iter) (iter).cur
 212 #define mbi_cur_ptr(iter) (iter).cur.ptr
 213
 214 /* Relocation.  */
 215 #define mbi_reloc(iter, ptrdiff) mbiter_multi_reloc (&iter, ptrdiff)
 216
 217 /* Copying an iterator.  */
 218 #define mbi_copy mbiter_multi_copy
 219
 220 _GL_INLINE_HEADER_END
 221
 222 #endif /* _MBITER_H */