enc/iso_8859_4.c

   1 /**********************************************************************
   2   iso8859_4.c -  Oniguruma (regular expression library)
   3 **********************************************************************/
   4 /*-
   5  * Copyright (c) 2002-2007  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_ISO_8859_4_TO_LOWER_CASE(c) EncISO_8859_4_ToLowerCaseTable[c]
  33 #define ENC_IS_ISO_8859_4_CTYPE(code,ctype) \
  34   ((EncISO_8859_4_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
  35
  36 static const UChar EncISO_8859_4_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', '\261', '\242', '\263', '\244', '\265', '\266', '\247',
  58   '\250', '\271', '\272', '\273', '\274', '\255', '\276', '\257',
  59   '\260', '\261', '\262', '\263', '\264', '\265', '\266', '\267',
  60   '\270', '\271', '\272', '\273', '\274', '\277', '\276', '\277',
  61   '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
  62   '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
  63   '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\327',
  64   '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\337',
  65   '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
  66   '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
  67   '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
  68   '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\377'
  69 };
  70
  71 static const unsigned short EncISO_8859_4_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   0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
  89   0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
  90   0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
  91   0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
  92   0x0284, 0x34a2, 0x30e2, 0x34a2, 0x00a0, 0x34a2, 0x34a2, 0x00a0,
  93   0x00a0, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x01a0, 0x34a2, 0x00a0,
  94   0x00a0, 0x30e2, 0x00a0, 0x30e2, 0x00a0, 0x30e2, 0x30e2, 0x00a0,
  95   0x00a0, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x34a2, 0x30e2, 0x30e2,
  96   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
  97   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
  98   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x00a0,
  99   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x30e2,
 100   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
 101   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
 102   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x00a0,
 103   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x00a0
 104 };
 105
 106 static int
 107 mbc_case_fold(OnigCaseFoldType flag,
 108               const UChar** pp, const UChar* end ARG_UNUSED, UChar* lower,
 109               OnigEncoding enc ARG_UNUSED)
 110 {
 111   const UChar* p = *pp;
 112
 113   if (*p == 0xdf && (flag & INTERNAL_ONIGENC_CASE_FOLD_MULTI_CHAR) != 0) {
 114     *lower++ = 's';
 115     *lower   = 's';
 116     (*pp)++;
 117     return 2;
 118   }
 119
 120   *lower = ENC_ISO_8859_4_TO_LOWER_CASE(*p);
 121   (*pp)++;
 122   return 1; /* return byte length of converted char to lower */
 123 }
 124
 125 #if 0
 126 static int
 127 is_mbc_ambiguous(OnigCaseFoldType flag, const UChar** pp, const UChar* end)
 128 {
 129   int v;
 130   const UChar* p = *pp;
 131
 132   if (*p == 0xdf && (flag & INTERNAL_ONIGENC_CASE_FOLD_MULTI_CHAR) != 0) {
 133     (*pp)++;
 134     return TRUE;
 135   }
 136
 137   (*pp)++;
 138   v = (EncISO_8859_4_CtypeTable[*p] & (BIT_CTYPE_UPPER | BIT_CTYPE_LOWER));
 139   if ((v | BIT_CTYPE_LOWER) != 0) {
 140     if (*p == 0xa2)
 141       return FALSE;
 142     else
 143       return TRUE;
 144   }
 145
 146   return (v != 0 ? TRUE : FALSE);
 147 }
 148 #endif
 149
 150 static int
 151 is_code_ctype(OnigCodePoint code, unsigned int ctype, OnigEncoding enc ARG_UNUSED)
 152 {
 153   if (code < 256)
 154     return ENC_IS_ISO_8859_4_CTYPE(code, ctype);
 155   else
 156     return FALSE;
 157 }
 158
 159 static const OnigPairCaseFoldCodes CaseFoldMap[] = {
 160  { 0xa1, 0xb1 },
 161  { 0xa3, 0xb3 },
 162  { 0xa5, 0xb5 },
 163  { 0xa6, 0xb6 },
 164  { 0xa9, 0xb9 },
 165  { 0xaa, 0xba },
 166  { 0xab, 0xbb },
 167  { 0xac, 0xbc },
 168  { 0xae, 0xbe },
 169
 170  { 0xc0, 0xe0 },
 171  { 0xc1, 0xe1 },
 172  { 0xc2, 0xe2 },
 173  { 0xc3, 0xe3 },
 174  { 0xc4, 0xe4 },
 175  { 0xc5, 0xe5 },
 176  { 0xc6, 0xe6 },
 177  { 0xc7, 0xe7 },
 178  { 0xc8, 0xe8 },
 179  { 0xc9, 0xe9 },
 180  { 0xca, 0xea },
 181  { 0xcb, 0xeb },
 182  { 0xcc, 0xec },
 183  { 0xcd, 0xed },
 184  { 0xce, 0xee },
 185  { 0xcf, 0xef },
 186
 187  { 0xd0, 0xf0 },
 188  { 0xd1, 0xf1 },
 189  { 0xd2, 0xf2 },
 190  { 0xd3, 0xf3 },
 191  { 0xd4, 0xf4 },
 192  { 0xd5, 0xf5 },
 193  { 0xd6, 0xf6 },
 194  { 0xd8, 0xf8 },
 195  { 0xd9, 0xf9 },
 196  { 0xda, 0xfa },
 197  { 0xdb, 0xfb },
 198  { 0xdc, 0xfc },
 199  { 0xdd, 0xfd },
 200  { 0xde, 0xfe }
 201 };
 202
 203 static int
 204 apply_all_case_fold(OnigCaseFoldType flag,
 205                     OnigApplyAllCaseFoldFunc f, void* arg,
 206                     OnigEncoding enc ARG_UNUSED)
 207 {
 208   return onigenc_apply_all_case_fold_with_map(
 209              sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 1,
 210              flag, f, arg);
 211 }
 212
 213 static int
 214 get_case_fold_codes_by_str(OnigCaseFoldType flag,
 215                            const OnigUChar* p, const OnigUChar* end,
 216                            OnigCaseFoldCodeItem items[],
 217                            OnigEncoding enc ARG_UNUSED)
 218 {
 219   return onigenc_get_case_fold_codes_by_str_with_map(
 220              sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 1,
 221              flag, p, end, items);
 222 }
 223
 224 OnigEncodingDefine(iso_8859_4, ISO_8859_4) = {
 225   onigenc_single_byte_mbc_enc_len,
 226   "ISO-8859-4",  /* name */
 227   1,             /* max enc length */
 228   1,             /* min enc length */
 229   onigenc_is_mbc_newline_0x0a,
 230   onigenc_single_byte_mbc_to_code,
 231   onigenc_single_byte_code_to_mbclen,
 232   onigenc_single_byte_code_to_mbc,
 233   mbc_case_fold,
 234   apply_all_case_fold,
 235   get_case_fold_codes_by_str,
 236   onigenc_minimum_property_name_to_ctype,
 237   is_code_ctype,
 238   onigenc_not_support_get_ctype_code_range,
 239   onigenc_single_byte_left_adjust_char_head,
 240   onigenc_always_true_is_allowed_reverse_match
 241 };
 242 ENC_ALIAS("ISO8859-4", "ISO-8859-4");