enc/koi8_u.c

   1 /**********************************************************************
   2   koi8_u.c -  Oniguruma (regular expression library)
   3 **********************************************************************/
   4 /*-
   5  * Copyright (c) 2002-2008  K.Kosako  <sndgk393 AT ybb DOT ne DOT jp>
   6  * All rights reserved.
   7  *
   8  * Redistribution and use in source and binary forms, with or without
   9  * modification, are permitted provided that the following conditions
  10  * are met:
  11  * 1. Redistributions of source code must retain the above copyright
  12  *    notice, this list of conditions and the following disclaimer.
  13  * 2. Redistributions in binary form must reproduce the above copyright
  14  *    notice, this list of conditions and the following disclaimer in the
  15  *    documentation and/or other materials provided with the distribution.
  16  *
  17  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
  18  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  19  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  20  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
  21  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  22  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  23  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  25  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  26  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  27  * SUCH DAMAGE.
  28  */
  29
  30 #include "regenc.h"
  31
  32 #define ENC_KOI8_U_TO_LOWER_CASE(c) EncKOI8_U_ToLowerCaseTable[c]
  33 #define ENC_IS_KOI8_U_CTYPE(code,ctype) \
  34   ((EncKOI8_U_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
  35
  36 static const UChar EncKOI8_U_ToLowerCaseTable[256] = {
  37   '\000', '\001', '\002', '\003', '\004', '\005', '\006', '\007',
  38   '\010', '\011', '\012', '\013', '\014', '\015', '\016', '\017',
  39   '\020', '\021', '\022', '\023', '\024', '\025', '\026', '\027',
  40   '\030', '\031', '\032', '\033', '\034', '\035', '\036', '\037',
  41   '\040', '\041', '\042', '\043', '\044', '\045', '\046', '\047',
  42   '\050', '\051', '\052', '\053', '\054', '\055', '\056', '\057',
  43   '\060', '\061', '\062', '\063', '\064', '\065', '\066', '\067',
  44   '\070', '\071', '\072', '\073', '\074', '\075', '\076', '\077',
  45   '\100', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
  46   '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
  47   '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
  48   '\170', '\171', '\172', '\133', '\134', '\135', '\136', '\137',
  49   '\140', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
  50   '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
  51   '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
  52   '\170', '\171', '\172', '\173', '\174', '\175', '\176', '\177',
  53   '\200', '\201', '\202', '\203', '\204', '\205', '\206', '\207',
  54   '\210', '\211', '\212', '\213', '\214', '\215', '\216', '\217',
  55   '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227',
  56   '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\237',
  57   '\240', '\241', '\242', '\243', '\244', '\245', '\246', '\247',
  58   '\250', '\251', '\252', '\253', '\254', '\255', '\256', '\257',
  59   '\260', '\261', '\262', '\243', '\244', '\265', '\246', '\247',
  60   '\270', '\271', '\272', '\273', '\274', '\255', '\276', '\277',
  61   '\300', '\301', '\302', '\303', '\304', '\305', '\306', '\307',
  62   '\310', '\311', '\312', '\313', '\314', '\315', '\316', '\317',
  63   '\320', '\321', '\322', '\323', '\324', '\325', '\326', '\327',
  64   '\330', '\331', '\332', '\333', '\334', '\335', '\336', '\337',
  65   '\300', '\301', '\302', '\303', '\304', '\305', '\306', '\307',
  66   '\310', '\311', '\312', '\313', '\314', '\315', '\316', '\317',
  67   '\320', '\321', '\322', '\323', '\324', '\325', '\326', '\327',
  68   '\330', '\331', '\332', '\333', '\334', '\335', '\336', '\337'
  69 };
  70
  71 static const unsigned short EncKOI8_U_CtypeTable[256] = {
  72   0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
  73   0x4008, 0x420c, 0x4209, 0x4208, 0x4208, 0x4208, 0x4008, 0x4008,
  74   0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
  75   0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
  76   0x4284, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
  77   0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
  78   0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0,
  79   0x78b0, 0x78b0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
  80   0x41a0, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x74a2,
  81   0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
  82   0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
  83   0x74a2, 0x74a2, 0x74a2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x51a0,
  84   0x41a0, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x70e2,
  85   0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
  86   0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
  87   0x70e2, 0x70e2, 0x70e2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x4008,
  88   0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
  89   0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
  90   0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
  91   0x00a0, 0x00a0, 0x0284, 0x00a0, 0x00a0, 0x10a0, 0x01a0, 0x00a0,
  92   0x00a0, 0x00a0, 0x00a0, 0x30e2, 0x30e2, 0x00a0, 0x30e2, 0x30e2,
  93   0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x30e2, 0x00a0, 0x00a0,
  94   0x00a0, 0x00a0, 0x00a0, 0x34a2, 0x34a2, 0x00a0, 0x34a2, 0x34a2,
  95   0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x34a2, 0x00a0, 0x00a0,
  96   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
  97   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
  98   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
  99   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
 100   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
 101   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
 102   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
 103   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2
 104 };
 105
 106 static int
 107 koi8_u_mbc_case_fold(OnigCaseFoldType flag ARG_UNUSED,
 108                      const UChar** pp, const UChar* end ARG_UNUSED,
 109                      UChar* lower, OnigEncoding enc ARG_UNUSED)
 110 {
 111   const UChar* p = *pp;
 112
 113   *lower = ENC_KOI8_U_TO_LOWER_CASE(*p);
 114   (*pp)++;
 115   return 1;
 116 }
 117
 118 #if 0
 119 static int
 120 koi8_u_is_mbc_ambiguous(OnigCaseFoldType flag, const UChar** pp,
 121                         const UChar* end, OnigEncoding enc ARG_UNUSED)
 122 {
 123   int v;
 124   const UChar* p = *pp;
 125
 126   (*pp)++;
 127   v = (EncKOI8_U_CtypeTable[*p] & (BIT_CTYPE_UPPER | BIT_CTYPE_LOWER));
 128   return (v != 0 ? TRUE : FALSE);
 129 }
 130 #endif
 131
 132 static int
 133 koi8_u_is_code_ctype(OnigCodePoint code, unsigned int ctype,
 134                      OnigEncoding enc ARG_UNUSED)
 135 {
 136   if (code < 256)
 137     return ENC_IS_KOI8_U_CTYPE(code, ctype);
 138   else
 139     return FALSE;
 140 }
 141
 142 static const OnigPairCaseFoldCodes CaseFoldMap[] = {
 143   { 0xa3, 0xb3 },
 144   { 0xa4, 0xb4 },
 145   { 0xa6, 0xb6 },
 146   { 0xa7, 0xb7 },
 147   { 0xad, 0xbd },
 148
 149   { 0xc0, 0xe0 },
 150   { 0xc1, 0xe1 },
 151   { 0xc2, 0xe2 },
 152   { 0xc3, 0xe3 },
 153   { 0xc4, 0xe4 },
 154   { 0xc5, 0xe5 },
 155   { 0xc6, 0xe6 },
 156   { 0xc7, 0xe7 },
 157   { 0xc8, 0xe8 },
 158   { 0xc9, 0xe9 },
 159   { 0xca, 0xea },
 160   { 0xcb, 0xeb },
 161   { 0xcc, 0xec },
 162   { 0xcd, 0xed },
 163   { 0xce, 0xee },
 164   { 0xcf, 0xef },
 165
 166   { 0xd0, 0xf0 },
 167   { 0xd1, 0xf1 },
 168   { 0xd2, 0xf2 },
 169   { 0xd3, 0xf3 },
 170   { 0xd4, 0xf4 },
 171   { 0xd5, 0xf5 },
 172   { 0xd6, 0xf6 },
 173   { 0xd7, 0xf7 },
 174   { 0xd8, 0xf8 },
 175   { 0xd9, 0xf9 },
 176   { 0xda, 0xfa },
 177   { 0xdb, 0xfb },
 178   { 0xdc, 0xfc },
 179   { 0xdd, 0xfd },
 180   { 0xde, 0xfe },
 181   { 0xdf, 0xff }
 182 };
 183
 184 static int
 185 koi8_u_apply_all_case_fold(OnigCaseFoldType flag,
 186                            OnigApplyAllCaseFoldFunc f,
 187                            void* arg, OnigEncoding enc ARG_UNUSED)
 188 {
 189   return onigenc_apply_all_case_fold_with_map(
 190              numberof(CaseFoldMap), CaseFoldMap, 0,
 191              flag, f, arg);
 192 }
 193
 194 static int
 195 koi8_u_get_case_fold_codes_by_str(OnigCaseFoldType flag,
 196                   const OnigUChar* p, const OnigUChar* end,
 197                   OnigCaseFoldCodeItem items[], OnigEncoding enc ARG_UNUSED)
 198 {
 199   return onigenc_get_case_fold_codes_by_str_with_map(
 200              numberof(CaseFoldMap), CaseFoldMap, 0,
 201              flag, p, end, items);
 202 }
 203
 204 OnigEncodingDefine(koi8_u, KOI8_U) = {
 205   onigenc_single_byte_mbc_enc_len,
 206   "KOI8-U",       /* name */
 207   1,             /* max enc length */
 208   1,             /* min enc length */
 209   onigenc_is_mbc_newline_0x0a,
 210   onigenc_single_byte_mbc_to_code,
 211   onigenc_single_byte_code_to_mbclen,
 212   onigenc_single_byte_code_to_mbc,
 213   koi8_u_mbc_case_fold,
 214   koi8_u_apply_all_case_fold,
 215   koi8_u_get_case_fold_codes_by_str,
 216   onigenc_minimum_property_name_to_ctype,
 217   koi8_u_is_code_ctype,
 218   onigenc_not_support_get_ctype_code_range,
 219   onigenc_single_byte_left_adjust_char_head,
 220   onigenc_always_true_is_allowed_reverse_match,
 221   onigenc_single_byte_ascii_only_case_map,
 222   0,
 223   ONIGENC_FLAG_NONE,
 224 };