enc/windows_1250.c

   1 /**********************************************************************
   2   windows_1250.c -  Oniguruma (regular expression library)
   3 **********************************************************************/
   4 /*-
   5  * Copyright (c) 2006-2007  Byte      <byte AT mail DOT kna DOT ru>
   6  *                          K.Kosako  <sndgk393 AT ybb DOT ne DOT jp>
   7  * All rights reserved.
   8  *
   9  * Redistribution and use in source and binary forms, with or without
  10  * modification, are permitted provided that the following conditions
  11  * are met:
  12  * 1. Redistributions of source code must retain the above copyright
  13  *    notice, this list of conditions and the following disclaimer.
  14  * 2. Redistributions in binary form must reproduce the above copyright
  15  *    notice, this list of conditions and the following disclaimer in the
  16  *    documentation and/or other materials provided with the distribution.
  17  *
  18  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
  19  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  20  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  21  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
  22  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  23  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  24  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  25  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  26  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  27  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  28  * SUCH DAMAGE.
  29  */
  30
  31 #include "regenc.h"
  32 #include "iso_8859.h"
  33
  34 #define ENC_CP1250_TO_LOWER_CASE(c) EncCP1250_ToLowerCaseTable[c]
  35 #define ENC_IS_CP1250_CTYPE(code,ctype) \
  36   ((EncCP1250_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
  37
  38 static const UChar EncCP1250_ToLowerCaseTable[256] = {
  39   '\000', '\001', '\002', '\003', '\004', '\005', '\006', '\007',
  40   '\010', '\011', '\012', '\013', '\014', '\015', '\016', '\017',
  41   '\020', '\021', '\022', '\023', '\024', '\025', '\026', '\027',
  42   '\030', '\031', '\032', '\033', '\034', '\035', '\036', '\037',
  43   '\040', '\041', '\042', '\043', '\044', '\045', '\046', '\047',
  44   '\050', '\051', '\052', '\053', '\054', '\055', '\056', '\057',
  45   '\060', '\061', '\062', '\063', '\064', '\065', '\066', '\067',
  46   '\070', '\071', '\072', '\073', '\074', '\075', '\076', '\077',
  47   '\100', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
  48   '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
  49   '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
  50   '\170', '\171', '\172', '\133', '\134', '\135', '\136', '\137',
  51   '\140', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
  52   '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
  53   '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
  54   '\170', '\171', '\172', '\173', '\174', '\175', '\176', '\177',
  55   '\200', '\201', '\202', '\203', '\204', '\205', '\206', '\207',
  56   '\210', '\211', '\232', '\213', '\234', '\235', '\236', '\237',
  57   '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227',
  58   '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\237',
  59   '\240', '\241', '\242', '\263', '\244', '\271', '\246', '\247',
  60   '\250', '\251', '\272', '\253', '\254', '\255', '\256', '\277',
  61   '\260', '\261', '\262', '\263', '\264', '\265', '\266', '\267',
  62   '\270', '\271', '\272', '\273', '\276', '\275', '\276', '\277',
  63   '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
  64   '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
  65   '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
  66   '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\337',
  67   '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
  68   '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
  69   '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
  70   '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\377',
  71 };
  72
  73 static const unsigned short EncCP1250_CtypeTable[256] = {
  74   0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
  75   0x4008, 0x420c, 0x4209, 0x4208, 0x4208, 0x4208, 0x4008, 0x4008,
  76   0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
  77   0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
  78   0x4284, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
  79   0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
  80   0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0,
  81   0x78b0, 0x78b0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
  82   0x41a0, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x74a2,
  83   0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
  84   0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
  85   0x74a2, 0x74a2, 0x74a2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x51a0,
  86   0x41a0, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x70e2,
  87   0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
  88   0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
  89   0x70e2, 0x70e2, 0x70e2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x4008,
  90   0x00a0, 0x0000, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,/* 8 */
  91   0x0008, 0x0008, 0x34a2, 0x0008, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
  92   0x0000, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,/* 9 */
  93   0x0008, 0x0008, 0x30e2, 0x0008, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
  94   0x0284, 0x01a0, 0x00a0, 0x34a2, 0x00a0, 0x34a2, 0x00a0, 0x00a0,/* A */
  95   0x00a0, 0x00a0, 0x34a2, 0x01a0, 0x00a0, 0x01a0, 0x00a0, 0x34a2,
  96   0x00a0, 0x00a0, 0x10a0, 0x30e2, 0x00a0, 0x30e2, 0x00a0, 0x01a0,/* B */
  97   0x00a0, 0x30e2, 0x30e2, 0x01a0, 0x34a2, 0x10a0, 0x30e2, 0x30e2,
  98   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,/* C */
  99   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
 100   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x00a0,/* D */
 101   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x30e2,
 102   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,/* E */
 103   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
 104   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x00a0,/* F */
 105   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x00a0,
 106 };
 107
 108 static int
 109 cp1250_mbc_case_fold(OnigCaseFoldType flag ARG_UNUSED,
 110              const UChar** pp, const UChar* end ARG_UNUSED, UChar* lower, OnigEncoding enc ARG_UNUSED)
 111 {
 112   const UChar* p = *pp;
 113
 114   *lower = ENC_CP1250_TO_LOWER_CASE(*p);
 115   (*pp)++;
 116   return 1;
 117 }
 118
 119 static int
 120 cp1250_is_code_ctype(OnigCodePoint code, unsigned int ctype, OnigEncoding enc ARG_UNUSED)
 121 {
 122   if (code < 256)
 123     return ENC_IS_CP1250_CTYPE(code, ctype);
 124   else
 125     return FALSE;
 126 }
 127
 128 static const OnigPairCaseFoldCodes CaseFoldMap[] = {
 129   { 0x8a, 0x9a },
 130   { 0x8c, 0x9c },
 131   { 0x8d, 0x9d },
 132   { 0x8e, 0x9e },
 133   { 0x8f, 0x9f },
 134
 135   { 0xa3, 0xb3 },
 136   { 0xa5, 0xb9 },
 137   { 0xaa, 0xba },
 138   { 0xaf, 0xbf },
 139
 140   { 0xbc, 0xbe },
 141
 142   { 0xc0, 0xe0 },
 143   { 0xc1, 0xe1 },
 144   { 0xc2, 0xe2 },
 145   { 0xc3, 0xe3 },
 146   { 0xc4, 0xe4 },
 147   { 0xc5, 0xe5 },
 148   { 0xc6, 0xe6 },
 149   { 0xc7, 0xe7 },
 150   { 0xc8, 0xe8 },
 151   { 0xc9, 0xe9 },
 152   { 0xca, 0xea },
 153   { 0xcb, 0xeb },
 154   { 0xcc, 0xec },
 155   { 0xcd, 0xed },
 156   { 0xce, 0xee },
 157   { 0xcf, 0xef },
 158
 159   { 0xd0, 0xf0 },
 160   { 0xd1, 0xf1 },
 161   { 0xd2, 0xf2 },
 162   { 0xd3, 0xf3 },
 163   { 0xd4, 0xf4 },
 164   { 0xd5, 0xf5 },
 165   { 0xd6, 0xf6 },
 166   { 0xd8, 0xf8 },
 167   { 0xd9, 0xf9 },
 168   { 0xda, 0xfa },
 169   { 0xdb, 0xfb },
 170   { 0xdc, 0xfc },
 171   { 0xdd, 0xfd },
 172   { 0xde, 0xfe },
 173 };
 174
 175 static int
 176 cp1250_apply_all_case_fold(OnigCaseFoldType flag,
 177                                OnigApplyAllCaseFoldFunc f, void* arg, OnigEncoding enc ARG_UNUSED)
 178 {
 179   return onigenc_apply_all_case_fold_with_map(
 180              sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 1,
 181              flag, f, arg);
 182 }
 183
 184 static int
 185 cp1250_get_case_fold_codes_by_str(OnigCaseFoldType flag,
 186     const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[], OnigEncoding enc ARG_UNUSED)
 187 {
 188   return onigenc_get_case_fold_codes_by_str_with_map(
 189              sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 1,
 190              flag, p, end, items);
 191 }
 192
 193 static int
 194 case_map(OnigCaseFoldType* flagP, const OnigUChar** pp,
 195          const OnigUChar* end, OnigUChar* to, OnigUChar* to_end,
 196          const struct OnigEncodingTypeST* enc)
 197 {
 198   OnigCodePoint code;
 199   OnigUChar *to_start = to;
 200   OnigCaseFoldType flags = *flagP;
 201
 202   while (*pp < end && to < to_end) {
 203     code = *(*pp)++;
 204     if (code == SHARP_s) {
 205       if (flags & ONIGENC_CASE_UPCASE) {
 206         flags |= ONIGENC_CASE_MODIFIED;
 207         *to++ = 'S';
 208         code = (flags & ONIGENC_CASE_TITLECASE) ? 's' : 'S';
 209       }
 210       else if (flags & ONIGENC_CASE_FOLD) {
 211         flags |= ONIGENC_CASE_MODIFIED;
 212         *to++ = 's';
 213         code = 's';
 214       }
 215     }
 216     else if ((EncCP1250_CtypeTable[code] & BIT_CTYPE_UPPER)
 217              && (flags & (ONIGENC_CASE_DOWNCASE | ONIGENC_CASE_FOLD))) {
 218       flags |= ONIGENC_CASE_MODIFIED;
 219       code = ENC_CP1250_TO_LOWER_CASE(code);
 220     }
 221     else if (code == 0xB5)
 222       ;
 223     else if ((EncCP1250_CtypeTable[code]&BIT_CTYPE_LOWER)
 224              && (flags & ONIGENC_CASE_UPCASE)) {
 225       flags |= ONIGENC_CASE_MODIFIED;
 226       if (code == 0xB9)
 227         code = 0xA5;
 228       else if (code == 0xBE)
 229         code = 0xBC;
 230       else if (code >= 0x8A && code <= 0xBF && code!=0xB9)
 231         code -= 0x10;
 232       else
 233         code -= 0x20;
 234     }
 235     *to++ = code;
 236     if (flags & ONIGENC_CASE_TITLECASE)  /* switch from titlecase to lowercase for capitalize */
 237       flags ^= (ONIGENC_CASE_UPCASE | ONIGENC_CASE_DOWNCASE | ONIGENC_CASE_TITLECASE);
 238   }
 239   *flagP = flags;
 240   return (int )(to - to_start);
 241 }
 242
 243 OnigEncodingDefine(windows_1250, Windows_1250) = {
 244   onigenc_single_byte_mbc_enc_len,
 245   "Windows-1250",      /* name */
 246   1,             /* max enc length */
 247   1,             /* min enc length */
 248   onigenc_is_mbc_newline_0x0a,
 249   onigenc_single_byte_mbc_to_code,
 250   onigenc_single_byte_code_to_mbclen,
 251   onigenc_single_byte_code_to_mbc,
 252   cp1250_mbc_case_fold,
 253   cp1250_apply_all_case_fold,
 254   cp1250_get_case_fold_codes_by_str,
 255   onigenc_minimum_property_name_to_ctype,
 256   cp1250_is_code_ctype,
 257   onigenc_not_support_get_ctype_code_range,
 258   onigenc_single_byte_left_adjust_char_head,
 259   onigenc_always_true_is_allowed_reverse_match,
 260   case_map,
 261   0,
 262   ONIGENC_FLAG_NONE,
 263 };
 264 /*
 265  * Name: windows-1250
 266  * MIBenum: 2251
 267  * Link: http://www.iana.org/assignments/character-sets
 268  * Link: http://www.microsoft.com/globaldev/reference/sbcs/1250.mspx
 269  * Link: https://en.wikipedia.org/wiki/Windows-1250
 270  */
 271 ENC_ALIAS("CP1250", "Windows-1250")