enc/windows_1252.c

   1 /**********************************************************************
   2   windows_1252.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_CP1252_TO_LOWER_CASE(c) EncCP1252_ToLowerCaseTable[c]
  35 #define ENC_IS_CP1252_CTYPE(code,ctype) \
  36   ((EncCP1252_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
  37
  38 static const UChar EncCP1252_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', '\215', '\236', '\217',
  57   '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227',
  58   '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\377',
  59   '\240', '\241', '\242', '\243', '\244', '\245', '\246', '\247',
  60   '\250', '\251', '\252', '\253', '\254', '\255', '\256', '\277',
  61   '\260', '\261', '\262', '\263', '\264', '\265', '\266', '\267',
  62   '\270', '\271', '\272', '\273', '\274', '\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 EncCP1252_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,
  91   0x0008, 0x0008, 0x34a2, 0x0008, 0x34a2, 0x0000, 0x34a2, 0x0000,
  92   0x0000, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
  93   0x0008, 0x0008, 0x30e2, 0x0008, 0x30e2, 0x0000, 0x30e2, 0x34a2,
  94   0x0284, 0x01a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
  95   0x00a0, 0x00a0, 0x30e2, 0x01a0, 0x00a0, 0x01a0, 0x00a0, 0x00a0,
  96   0x00a0, 0x00a0, 0x10a0, 0x10a0, 0x00a0, 0x30e2, 0x00a0, 0x01a0,
  97   0x00a0, 0x10a0, 0x30e2, 0x01a0, 0x10a0, 0x10a0, 0x10a0, 0x01a0,
  98   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
  99   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
 100   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x00a0,
 101   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x30e2,
 102   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
 103   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
 104   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x00a0,
 105   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
 106 };
 107
 108 static int
 109 cp1252_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_CP1252_TO_LOWER_CASE(*p);
 115   (*pp)++;
 116   return 1;
 117 }
 118 static int
 119 cp1252_is_code_ctype(OnigCodePoint code, unsigned int ctype, OnigEncoding enc ARG_UNUSED)
 120 {
 121   if (code < 256)
 122     return ENC_IS_CP1252_CTYPE(code, ctype);
 123   else
 124     return FALSE;
 125 }
 126
 127 static const OnigPairCaseFoldCodes CaseFoldMap[] = {
 128   { 0x8a, 0x9a },
 129   { 0x8c, 0x9c },
 130   { 0x8e, 0x9e },
 131   { 0x9f, 0xff },
 132
 133   { 0xc0, 0xe0 },
 134   { 0xc1, 0xe1 },
 135   { 0xc2, 0xe2 },
 136   { 0xc3, 0xe3 },
 137   { 0xc4, 0xe4 },
 138   { 0xc5, 0xe5 },
 139   { 0xc6, 0xe6 },
 140   { 0xc7, 0xe7 },
 141   { 0xc8, 0xe8 },
 142   { 0xc9, 0xe9 },
 143   { 0xca, 0xea },
 144   { 0xcb, 0xeb },
 145   { 0xcc, 0xec },
 146   { 0xcd, 0xed },
 147   { 0xce, 0xee },
 148   { 0xcf, 0xef },
 149
 150   { 0xd0, 0xf0 },
 151   { 0xd1, 0xf1 },
 152   { 0xd2, 0xf2 },
 153   { 0xd3, 0xf3 },
 154   { 0xd4, 0xf4 },
 155   { 0xd5, 0xf5 },
 156   { 0xd6, 0xf6 },
 157   { 0xd8, 0xf8 },
 158   { 0xd9, 0xf9 },
 159   { 0xda, 0xfa },
 160   { 0xdb, 0xfb },
 161   { 0xdc, 0xfc },
 162   { 0xdd, 0xfd },
 163   { 0xde, 0xfe },
 164 };
 165
 166 static int
 167 cp1252_apply_all_case_fold(OnigCaseFoldType flag,
 168                                OnigApplyAllCaseFoldFunc f, void* arg, OnigEncoding enc ARG_UNUSED)
 169 {
 170   return onigenc_apply_all_case_fold_with_map(
 171              sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 1,
 172              flag, f, arg);
 173 }
 174
 175 static int
 176 cp1252_get_case_fold_codes_by_str(OnigCaseFoldType flag,
 177     const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[], OnigEncoding enc ARG_UNUSED)
 178 {
 179   return onigenc_get_case_fold_codes_by_str_with_map(
 180              sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 1,
 181              flag, p, end, items);
 182 }
 183
 184 static int
 185 case_map(OnigCaseFoldType* flagP, const OnigUChar** pp,
 186          const OnigUChar* end, OnigUChar* to, OnigUChar* to_end,
 187          const struct OnigEncodingTypeST* enc)
 188 {
 189   OnigCodePoint code;
 190   OnigUChar *to_start = to;
 191   OnigCaseFoldType flags = *flagP;
 192
 193   while (*pp < end && to < to_end) {
 194     code = *(*pp)++;
 195     if (code == SHARP_s) {
 196       if (flags & ONIGENC_CASE_UPCASE) {
 197         flags |= ONIGENC_CASE_MODIFIED;
 198         *to++ = 'S';
 199         code = (flags & ONIGENC_CASE_TITLECASE) ? 's' : 'S';
 200       }
 201       else if (flags & ONIGENC_CASE_FOLD) {
 202         flags |= ONIGENC_CASE_MODIFIED;
 203         *to++ = 's';
 204         code = 's';
 205       }
 206     }
 207     else if ((EncCP1252_CtypeTable[code] & BIT_CTYPE_UPPER)
 208              && (flags & (ONIGENC_CASE_DOWNCASE | ONIGENC_CASE_FOLD))) {
 209       flags |= ONIGENC_CASE_MODIFIED;
 210       code = ENC_CP1252_TO_LOWER_CASE(code);
 211     }
 212     else if (code == 0x83 || code == 0xAA || code == 0xBA || code == 0xB5)
 213       ;
 214     else if ((EncCP1252_CtypeTable[code]&BIT_CTYPE_LOWER)
 215              && (flags & ONIGENC_CASE_UPCASE)) {
 216       flags |= ONIGENC_CASE_MODIFIED;
 217       if (code == 0x9A || code == 0x9C || code == 0x9E)
 218         code -= 0x10;
 219       else if (code == 0xFF)
 220         code -= 0x60;
 221       else
 222         code -= 0x20;
 223     }
 224     *to++ = code;
 225     if (flags & ONIGENC_CASE_TITLECASE)  /* switch from titlecase to lowercase for capitalize */
 226       flags ^= (ONIGENC_CASE_UPCASE | ONIGENC_CASE_DOWNCASE | ONIGENC_CASE_TITLECASE);
 227   }
 228   *flagP = flags;
 229   return (int )(to - to_start);
 230 }
 231
 232 OnigEncodingDefine(windows_1252, Windows_1252) = {
 233   onigenc_single_byte_mbc_enc_len,
 234   "Windows-1252",      /* name */
 235   1,             /* max enc length */
 236   1,             /* min enc length */
 237   onigenc_is_mbc_newline_0x0a,
 238   onigenc_single_byte_mbc_to_code,
 239   onigenc_single_byte_code_to_mbclen,
 240   onigenc_single_byte_code_to_mbc,
 241   cp1252_mbc_case_fold,
 242   cp1252_apply_all_case_fold,
 243   cp1252_get_case_fold_codes_by_str,
 244   onigenc_minimum_property_name_to_ctype,
 245   cp1252_is_code_ctype,
 246   onigenc_not_support_get_ctype_code_range,
 247   onigenc_single_byte_left_adjust_char_head,
 248   onigenc_always_true_is_allowed_reverse_match,
 249   case_map,
 250   0,
 251   ONIGENC_FLAG_NONE,
 252 };
 253 /*
 254  * Name: windows-1252
 255  * MIBenum: 2251
 256  * Link: http://www.iana.org/assignments/character-sets
 257  * Link: http://www.microsoft.com/globaldev/reference/sbcs/1252.mspx
 258  * Link: https://en.wikipedia.org/wiki/Windows-1252
 259  */
 260 ENC_ALIAS("CP1252", "Windows-1252")