lib/unigbrk/uc-is-grapheme-break.c

   1 /* Grapheme cluster break function.
   2    Copyright (C) 2010-2021 Free Software Foundation, Inc.
   3    Written by Ben Pfaff <blp@cs.stanford.edu>, 2010.
   4
   5    This program is free software: you can redistribute it and/or modify it
   6    under the terms of the GNU Lesser General Public License as published
   7    by the Free Software Foundation; either version 3 of the License, or
   8    (at your option) any later version.
   9
  10    This program 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 GNU
  13    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 "unigbrk.h"
  22
  23 /* Evaluates to true if there is an extended grapheme cluster break between
  24    code points with GBP_* values A and B, false if there is not.  The comments
  25    are the grapheme cluster boundary rules from in UAX #29. */
  26 #define UC_IS_GRAPHEME_BREAK(A, B)                                      \
  27   (/* GB1 and GB2 are covered--just use a GBP_CONTROL character, such   \
  28       as 0, for sot and eot. */                                         \
  29                                                                         \
  30    /* GB3 */                                                            \
  31    (A) == GBP_CR && (B) == GBP_LF ? false :                             \
  32                                                                         \
  33    /* GB4 */                                                            \
  34    (A) == GBP_CONTROL || (A) == GBP_CR || (A) == GBP_LF ? true :        \
  35                                                                         \
  36    /* GB5 */                                                            \
  37    (B) == GBP_CONTROL || (B) == GBP_CR || (B) == GBP_LF ? true :        \
  38                                                                         \
  39    /* GB6 */                                                            \
  40    (A) == GBP_L && ((B) == GBP_L || (B) == GBP_V                        \
  41                     || (B) == GBP_LV || (B) == GBP_LVT) ? false :       \
  42                                                                         \
  43    /* GB7 */                                                            \
  44    ((A) == GBP_LV || (A) == GBP_V)                                      \
  45    && ((B) == GBP_V || (B) == GBP_T) ? false :                          \
  46                                                                         \
  47    /* GB8 */                                                            \
  48    ((A) == GBP_LVT || (A) == GBP_T) && (B) == GBP_T ? false :           \
  49                                                                         \
  50    /* GB9 */                                                            \
  51    (B) == GBP_EXTEND || (B) == GBP_ZWJ ? false :                        \
  52                                                                         \
  53    /* GB9a */                                                           \
  54    (B) == GBP_SPACINGMARK ? false :                                     \
  55                                                                         \
  56    /* GB9b */                                                           \
  57    (A) == GBP_PREPEND ? false :                                         \
  58                                                                         \
  59    /* GB10 -- incomplete */                                             \
  60    ((A) == GBP_EB || (A) == GBP_EBG) && (B) == GBP_EM ? false :         \
  61                                                                         \
  62    /* GB11 */                                                           \
  63    (A) == GBP_ZWJ && ((B) == GBP_GAZ || (B) == GBP_EBG) ? false         \
  64                                                                         \
  65    /* GB999 */                                                          \
  66    : true)
  67
  68 #define UC_GRAPHEME_BREAKS_FOR(A)                                       \
  69   (  (UC_IS_GRAPHEME_BREAK(A, GBP_OTHER)       << GBP_OTHER)            \
  70    | (UC_IS_GRAPHEME_BREAK(A, GBP_CR)          << GBP_CR)               \
  71    | (UC_IS_GRAPHEME_BREAK(A, GBP_LF)          << GBP_LF)               \
  72    | (UC_IS_GRAPHEME_BREAK(A, GBP_CONTROL)     << GBP_CONTROL)          \
  73    | (UC_IS_GRAPHEME_BREAK(A, GBP_EXTEND)      << GBP_EXTEND)           \
  74    | (UC_IS_GRAPHEME_BREAK(A, GBP_PREPEND)     << GBP_PREPEND)          \
  75    | (UC_IS_GRAPHEME_BREAK(A, GBP_SPACINGMARK) << GBP_SPACINGMARK)      \
  76    | (UC_IS_GRAPHEME_BREAK(A, GBP_L)           << GBP_L)                \
  77    | (UC_IS_GRAPHEME_BREAK(A, GBP_V)           << GBP_V)                \
  78    | (UC_IS_GRAPHEME_BREAK(A, GBP_T)           << GBP_T)                \
  79    | (UC_IS_GRAPHEME_BREAK(A, GBP_LV)          << GBP_LV)               \
  80    | (UC_IS_GRAPHEME_BREAK(A, GBP_LVT)         << GBP_LVT)              \
  81    | (UC_IS_GRAPHEME_BREAK(A, GBP_RI)          << GBP_RI)               \
  82    | (UC_IS_GRAPHEME_BREAK(A, GBP_ZWJ)         << GBP_ZWJ)              \
  83    | (UC_IS_GRAPHEME_BREAK(A, GBP_EB)          << GBP_EB)               \
  84    | (UC_IS_GRAPHEME_BREAK(A, GBP_EM)          << GBP_EM)               \
  85    | (UC_IS_GRAPHEME_BREAK(A, GBP_GAZ)         << GBP_GAZ)              \
  86    | (UC_IS_GRAPHEME_BREAK(A, GBP_EBG)         << GBP_EBG))
  87
  88 static const unsigned long int gb_table[18] =
  89   {
  90     UC_GRAPHEME_BREAKS_FOR(0),  /* GBP_OTHER */
  91     UC_GRAPHEME_BREAKS_FOR(1),  /* GBP_CR */
  92     UC_GRAPHEME_BREAKS_FOR(2),  /* GBP_LF */
  93     UC_GRAPHEME_BREAKS_FOR(3),  /* GBP_CONTROL */
  94     UC_GRAPHEME_BREAKS_FOR(4),  /* GBP_EXTEND */
  95     UC_GRAPHEME_BREAKS_FOR(5),  /* GBP_PREPEND */
  96     UC_GRAPHEME_BREAKS_FOR(6),  /* GBP_SPACINGMARK */
  97     UC_GRAPHEME_BREAKS_FOR(7),  /* GBP_L */
  98     UC_GRAPHEME_BREAKS_FOR(8),  /* GBP_V */
  99     UC_GRAPHEME_BREAKS_FOR(9),  /* GBP_T */
 100     UC_GRAPHEME_BREAKS_FOR(10), /* GBP_LV */
 101     UC_GRAPHEME_BREAKS_FOR(11), /* GBP_LVT */
 102     UC_GRAPHEME_BREAKS_FOR(12), /* GBP_RI */
 103     UC_GRAPHEME_BREAKS_FOR(13), /* GBP_ZWJ */
 104     UC_GRAPHEME_BREAKS_FOR(14), /* GBP_EB */
 105     UC_GRAPHEME_BREAKS_FOR(15), /* GBP_EM */
 106     UC_GRAPHEME_BREAKS_FOR(16), /* GBP_GAZ */
 107     UC_GRAPHEME_BREAKS_FOR(17), /* GBP_EBG */
 108   };
 109
 110 bool
 111 uc_is_grapheme_break (ucs4_t a, ucs4_t b)
 112 {
 113   int a_gcp, b_gcp;
 114
 115   if ((a | b) < 0x300)
 116     {
 117       /* GB3 is the only relevant rule for this case. */
 118       return a != '\r' || b != '\n';
 119     }
 120
 121   a_gcp = uc_graphemeclusterbreak_property (a);
 122   b_gcp = uc_graphemeclusterbreak_property (b);
 123   return (gb_table[a_gcp] >> b_gcp) & 1;
 124 }