lib/unigbrk.in.h

   1 /* Grapheme cluster breaks in Unicode strings.
   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 #ifndef _UNIGBRK_H
  19 #define _UNIGBRK_H
  20
  21 /* Get bool.  */
  22 #include <stdbool.h>
  23
  24 /* Get size_t. */
  25 #include <stddef.h>
  26
  27 #include "unitypes.h"
  28
  29 #ifdef __cplusplus
  30 extern "C" {
  31 #endif
  32
  33 /* ========================================================================= */
  34
  35 /* Property defined in Unicode Standard Annex #29, section "Grapheme Cluster
  36    Boundaries"
  37    <https://unicode.org/reports/tr29/#Grapheme_Cluster_Boundaries> */
  38
  39 /* Possible values of the Grapheme_Cluster_Break property.
  40    This enumeration may be extended in the future.  */
  41 enum
  42 {
  43   GBP_OTHER        = 0,
  44   GBP_CR           = 1,
  45   GBP_LF           = 2,
  46   GBP_CONTROL      = 3,
  47   GBP_EXTEND       = 4,
  48   GBP_PREPEND      = 5,
  49   GBP_SPACINGMARK  = 6,
  50   GBP_L            = 7,
  51   GBP_V            = 8,
  52   GBP_T            = 9,
  53   GBP_LV           = 10,
  54   GBP_LVT          = 11,
  55   GBP_RI           = 12,
  56   GBP_ZWJ          = 13,
  57   GBP_EB           = 14,
  58   GBP_EM           = 15,
  59   GBP_GAZ          = 16,
  60   GBP_EBG          = 17
  61 };
  62
  63 /* Return the Grapheme_Cluster_Break property of a Unicode character. */
  64 extern int
  65        uc_graphemeclusterbreak_property (ucs4_t uc)
  66        _UC_ATTRIBUTE_CONST;
  67
  68 /* ========================================================================= */
  69
  70 /* Grapheme cluster breaks.  */
  71
  72 /* Returns true if there is a grapheme cluster boundary between Unicode code
  73    points A and B.  A "grapheme cluster" is an approximation to a
  74    user-perceived character, which sometimes corresponds to multiple code
  75    points.  For example, an English letter followed by an acute accent can be
  76    expressed as two consecutive Unicode code points, but it is perceived by the
  77    user as only a single character and therefore constitutes a single grapheme
  78    cluster.
  79
  80    Implements extended (not legacy) grapheme cluster rules, because UAX #29
  81    indicates that they are preferred.
  82
  83    Use A == 0 or B == 0 to indicate start of text or end of text,
  84    respectively. */
  85 extern bool
  86        uc_is_grapheme_break (ucs4_t a, ucs4_t b)
  87        _UC_ATTRIBUTE_CONST;
  88
  89 /* Returns the start of the next grapheme cluster following S, or NULL if the
  90    end of the string has been reached. */
  91 extern const uint8_t *
  92        u8_grapheme_next (const uint8_t *s, const uint8_t *end)
  93        _UC_ATTRIBUTE_PURE;
  94 extern const uint16_t *
  95        u16_grapheme_next (const uint16_t *s, const uint16_t *end)
  96        _UC_ATTRIBUTE_PURE;
  97 extern const uint32_t *
  98        u32_grapheme_next (const uint32_t *s, const uint32_t *end)
  99        _UC_ATTRIBUTE_PURE;
 100
 101 /* Returns the start of the previous grapheme cluster before S, or NULL if the
 102    start of the string has been reached. */
 103 extern const uint8_t *
 104        u8_grapheme_prev (const uint8_t *s, const uint8_t *start)
 105        _UC_ATTRIBUTE_PURE;
 106 extern const uint16_t *
 107        u16_grapheme_prev (const uint16_t *s, const uint16_t *start)
 108        _UC_ATTRIBUTE_PURE;
 109 extern const uint32_t *
 110        u32_grapheme_prev (const uint32_t *s, const uint32_t *start)
 111        _UC_ATTRIBUTE_PURE;
 112
 113 /* Determine the grapheme cluster boundaries in S, and store the result at
 114    p[0..n-1].  p[i] = 1 means that a new grapheme cluster begins at s[i].  p[i]
 115    = 0 means that s[i-1] and s[i] are part of the same grapheme cluster.  p[0]
 116    will always be 1.
 117  */
 118 extern void
 119        u8_grapheme_breaks (const uint8_t *s, size_t n, char *p);
 120 extern void
 121        u16_grapheme_breaks (const uint16_t *s, size_t n, char *p);
 122 extern void
 123        u32_grapheme_breaks (const uint32_t *s, size_t n, char *p);
 124 extern void
 125        ulc_grapheme_breaks (const char *s, size_t n, char *p);
 126 extern void
 127        uc_grapheme_breaks (const ucs4_t *s, size_t n, char *p);
 128
 129 /* ========================================================================= */
 130
 131 #ifdef __cplusplus
 132 }
 133 #endif
 134
 135
 136 #endif /* _UNIGBRK_H */