lib/mbsspn.c

   1 /* Searching a string for a character outside a given set of characters.
   2    Copyright (C) 1999, 2002, 2006-2025 Free Software Foundation, Inc.
   3    Written by Bruno Haible <bruno@clisp.org>, 2007.
   4
   5    This file is free software: you can redistribute it and/or modify
   6    it under the terms of the GNU Lesser General Public License as
   7    published by the Free Software Foundation, either version 3 of the
   8    License, or (at your option) any later version.
   9
  10    This file 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
  13    GNU Lesser General Public License for more details.
  14
  15    You should have received a copy of the GNU Lesser General Public License
  16    along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
  17
  18 #include <config.h>
  19
  20 /* Specification.  */
  21 #include <string.h>
  22
  23 #include <stdlib.h>
  24
  25 #if GNULIB_MCEL_PREFER
  26 # include "mcel.h"
  27 #else
  28 # include "mbuiterf.h"
  29 #endif
  30
  31 /* Find the first occurrence in the character string STRING of any character
  32    not in the character string REJECT.  Return the number of bytes from the
  33    beginning of the string to this occurrence, or to the end of the string
  34    if none exists.  */
  35 size_t
  36 mbsspn (const char *string, const char *reject)
  37 {
  38   /* Optimize two cases.  */
  39   if (reject[0] == '\0')
  40     return 0;
  41   if (reject[1] == '\0')
  42     {
  43       unsigned char uc = (unsigned char) reject[0];
  44       const char *iter = string;
  45
  46       if (MB_CUR_MAX > 1)
  47         {
  48 #if GNULIB_MCEL_PREFER
  49           for (mcel_t g; *iter; iter += g.len)
  50             {
  51               g = mcel_scanz (iter);
  52               if (! (g.len == 1 && (unsigned char) *iter == uc))
  53                 break;
  54             }
  55 #else
  56           mbuif_state_t state;
  57           for (mbuif_init (state); mbuif_avail (state, iter); )
  58             {
  59               mbchar_t cur = mbuif_next (state, iter);
  60               if (!(mb_len (cur) == 1 && (unsigned char) *iter == uc))
  61                 break;
  62               iter += mb_len (cur);
  63             }
  64 #endif
  65         }
  66       else
  67         {
  68           for (; *iter != '\0'; iter++)
  69             if ((unsigned char) *iter != uc)
  70               break;
  71         }
  72       return iter - string;
  73     }
  74   /* General case.  */
  75   if (MB_CUR_MAX > 1)
  76     {
  77 #if GNULIB_MCEL_PREFER
  78       for (size_t i = 0; ; )
  79         {
  80           char c = string[i];
  81           if (!c)
  82             return i;
  83           mcel_t g = mcel_scanz (string + i);
  84           if (g.len == 1)
  85             {
  86               if (!mbschr (reject, c))
  87                 return i;
  88             }
  89           else
  90             {
  91               for (char const *aiter = reject; ; )
  92                 {
  93                   if (!*aiter)
  94                     return i;
  95                   mcel_t a = mcel_scanz (aiter);
  96                   if (mcel_cmp (a, g) == 0)
  97                     break;
  98                   aiter += a.len;
  99                 }
 100             }
 101           i += g.len;
 102         }
 103 #else
 104       mbuif_state_t state;
 105       const char *iter;
 106       for (mbuif_init (state), iter = string; mbuif_avail (state, iter); )
 107         {
 108           mbchar_t cur = mbuif_next (state, iter);
 109           if (mb_len (cur) == 1)
 110             {
 111               if (mbschr (reject, *iter) == NULL)
 112                 goto found;
 113             }
 114           else
 115             {
 116               mbuif_state_t astate;
 117               const char *aiter;
 118               for (mbuif_init (astate), aiter = reject; ; )
 119                 {
 120                   if (!mbuif_avail (astate, aiter))
 121                     goto found;
 122                   mbchar_t acur = mbuif_next (astate, aiter);
 123                   if (mb_equal (acur, cur))
 124                     break;
 125                   aiter += mb_len (acur);
 126                 }
 127             }
 128           iter += mb_len (cur);
 129         }
 130      found:
 131       return iter - string;
 132 #endif
 133     }
 134   else
 135     return strspn (string, reject);
 136 }