lib/uninorm/decomposition.c

   1 /* Decomposition of Unicode characters.
   2    Copyright (C) 2009-2024 Free Software Foundation, Inc.
   3    Written by Bruno Haible <bruno@clisp.org>, 2009.
   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 "uninorm.h"
  30
  31 #include "uninorm/decomposition-table.h"
  32
  33 int
  34 uc_decomposition (ucs4_t uc, int *decomp_tag, ucs4_t *decomposition)
  35 {
  36   if (uc >= 0xAC00 && uc < 0xD7A4)
  37     {
  38       /* Hangul syllable.  See Unicode standard, chapter 3, section
  39          "Hangul Syllable Decomposition",  See also the clarification at
  40          <https://www.unicode.org/versions/Unicode5.1.0/>, section
  41          "Clarification of Hangul Jamo Handling".  */
  42       unsigned int t;
  43
  44       uc -= 0xAC00;
  45       t = uc % 28;
  46
  47       *decomp_tag = UC_DECOMP_CANONICAL;
  48       if (t == 0)
  49         {
  50           unsigned int v, l;
  51
  52           uc = uc / 28;
  53           v = uc % 21;
  54           l = uc / 21;
  55
  56           decomposition[0] = 0x1100 + l;
  57           decomposition[1] = 0x1161 + v;
  58           return 2;
  59         }
  60       else
  61         {
  62 #if 1 /* Return the pairwise decomposition, not the full decomposition.  */
  63           decomposition[0] = 0xAC00 + uc - t; /* = 0xAC00 + (l * 21 + v) * 28; */
  64           decomposition[1] = 0x11A7 + t;
  65           return 2;
  66 #else
  67           unsigned int v, l;
  68
  69           uc = uc / 28;
  70           v = uc % 21;
  71           l = uc / 21;
  72
  73           decomposition[0] = 0x1100 + l;
  74           decomposition[1] = 0x1161 + v;
  75           decomposition[2] = 0x11A7 + t;
  76           return 3;
  77 #endif
  78         }
  79     }
  80   else if (uc < 0x110000)
  81     {
  82       unsigned short entry = decomp_index (uc);
  83       if (entry != (unsigned short)(-1))
  84         {
  85           const unsigned char *p;
  86           unsigned int element;
  87           unsigned int length;
  88
  89           p = &gl_uninorm_decomp_chars_table[3 * (entry & 0x7FFF)];
  90           element = (p[0] << 16) | (p[1] << 8) | p[2];
  91           /* The first element has 5 bits for the decomposition type.  */
  92           *decomp_tag = (element >> 18) & 0x1f;
  93           length = 1;
  94           for (;;)
  95             {
  96               /* Every element has an 18 bits wide Unicode code point.  */
  97               *decomposition = element & 0x3ffff;
  98               /* Bit 23 tells whether there are more elements,  */
  99               if ((element & (1 << 23)) == 0)
 100                 break;
 101               p += 3;
 102               element = (p[0] << 16) | (p[1] << 8) | p[2];
 103               decomposition++;
 104               length++;
 105             }
 106           return length;
 107         }
 108     }
 109   return -1;
 110 }