lib/unilbrk/u32-width-linebreaks.c

   1 /* Line breaking of UTF-32 strings.
   2    Copyright (C) 2001-2003, 2006-2025 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 "unilbrk/internal.h"
  32 #include "unilbrk/lbrktables.h"
  33 #include "uniwidth.h"
  34
  35 static int
  36 u32_width_linebreaks_internal (const uint32_t *s, size_t n,
  37                               int width, int start_column, int at_end_columns,
  38                               const char *o, const char *encoding, int cr,
  39                               char *p)
  40 {
  41   const uint32_t *s_end;
  42   char *last_p;
  43   int last_column;
  44   int piece_width;
  45
  46   u32_possible_linebreaks_loop (s, n, encoding, cr, p);
  47
  48   s_end = s + n;
  49   last_p = NULL;
  50   last_column = start_column;
  51   piece_width = 0;
  52   while (s < s_end)
  53     {
  54       ucs4_t uc = *s;
  55
  56       /* Respect the override.  */
  57       if (o != NULL && *o != UC_BREAK_UNDEFINED)
  58         *p = *o;
  59
  60       if (*p == UC_BREAK_POSSIBLE
  61           || *p == UC_BREAK_MANDATORY || *p == UC_BREAK_CR_BEFORE_LF)
  62         {
  63           /* An atomic piece of text ends here.  */
  64           if (last_p != NULL && last_column + piece_width > width)
  65             {
  66               /* Insert a line break.  */
  67               *last_p = UC_BREAK_POSSIBLE;
  68               last_column = 0;
  69             }
  70         }
  71
  72       if (*p == UC_BREAK_MANDATORY || *p == UC_BREAK_CR_BEFORE_LF)
  73         {
  74           /* uc is a line break character.  */
  75           /* Start a new piece at column 0.  */
  76           last_p = NULL;
  77           last_column = 0;
  78           piece_width = 0;
  79         }
  80       else
  81         {
  82           /* uc is not a line break character.  */
  83           int w;
  84
  85           if (*p == UC_BREAK_POSSIBLE)
  86             {
  87               /* Start a new piece.  */
  88               last_p = p;
  89               last_column += piece_width;
  90               piece_width = 0;
  91               /* No line break for the moment, may be turned into
  92                  UC_BREAK_POSSIBLE later, via last_p. */
  93             }
  94
  95           *p = UC_BREAK_PROHIBITED;
  96
  97           w = uc_width (uc, encoding);
  98           if (w >= 0) /* ignore control characters in the string */
  99             piece_width += w;
 100         }
 101
 102       s++;
 103       p++;
 104       if (o != NULL)
 105         o++;
 106     }
 107
 108   /* The last atomic piece of text ends here.  */
 109   if (last_p != NULL && last_column + piece_width + at_end_columns > width)
 110     {
 111       /* Insert a line break.  */
 112       *last_p = UC_BREAK_POSSIBLE;
 113       last_column = 0;
 114     }
 115
 116   return last_column + piece_width;
 117 }
 118
 119 #if defined IN_LIBUNISTRING
 120 /* For backward compatibility with older versions of libunistring.  */
 121
 122 # undef u32_width_linebreaks
 123
 124 int
 125 u32_width_linebreaks (const uint32_t *s, size_t n,
 126                       int width, int start_column, int at_end_columns,
 127                       const char *o, const char *encoding,
 128                       char *p)
 129 {
 130   return u32_width_linebreaks_internal (s, n,
 131                                         width, start_column, at_end_columns,
 132                                         o, encoding, -1, p);
 133 }
 134
 135 #endif
 136
 137 int
 138 u32_width_linebreaks_v2 (const uint32_t *s, size_t n,
 139                          int width, int start_column, int at_end_columns,
 140                          const char *o, const char *encoding,
 141                          char *p)
 142 {
 143   return u32_width_linebreaks_internal (s, n,
 144                                         width, start_column, at_end_columns,
 145                                         o, encoding, LBP_CR, p);
 146 }