enc/windows_1251.c

   1 /**********************************************************************
   2   cp1251.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
  33 #define ENC_CP1251_TO_LOWER_CASE(c) EncCP1251_ToLowerCaseTable[c]
  34 #define ENC_IS_CP1251_CTYPE(code,ctype) \
  35   ((EncCP1251_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
  36
  37 static const UChar EncCP1251_ToLowerCaseTable[256] = {
  38   '\000', '\001', '\002', '\003', '\004', '\005', '\006', '\007',
  39   '\010', '\011', '\012', '\013', '\014', '\015', '\016', '\017',
  40   '\020', '\021', '\022', '\023', '\024', '\025', '\026', '\027',
  41   '\030', '\031', '\032', '\033', '\034', '\035', '\036', '\037',
  42   '\040', '\041', '\042', '\043', '\044', '\045', '\046', '\047',
  43   '\050', '\051', '\052', '\053', '\054', '\055', '\056', '\057',
  44   '\060', '\061', '\062', '\063', '\064', '\065', '\066', '\067',
  45   '\070', '\071', '\072', '\073', '\074', '\075', '\076', '\077',
  46   '\100', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
  47   '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
  48   '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
  49   '\170', '\171', '\172', '\133', '\134', '\135', '\136', '\137',
  50   '\140', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
  51   '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
  52   '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
  53   '\170', '\171', '\172', '\173', '\174', '\175', '\176', '\177',
  54   '\220', '\203', '\202', '\203', '\204', '\205', '\206', '\207',
  55   '\210', '\211', '\232', '\213', '\234', '\235', '\236', '\237',
  56   '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227',
  57   '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\237',
  58   '\240', '\242', '\242', '\274', '\244', '\264', '\246', '\247',
  59   '\270', '\251', '\272', '\253', '\254', '\255', '\256', '\277',
  60   '\260', '\261', '\263', '\263', '\264', '\265', '\266', '\267',
  61   '\270', '\271', '\272', '\273', '\274', '\276', '\276', '\277',
  62   '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
  63   '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
  64   '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
  65   '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\377',
  66   '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
  67   '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
  68   '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
  69   '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\377'
  70 };
  71
  72 static const unsigned short EncCP1251_CtypeTable[256] = {
  73   0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
  74   0x4008, 0x428c, 0x4209, 0x4208, 0x4208, 0x4208, 0x4008, 0x4008,
  75   0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
  76   0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
  77   0x4284, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
  78   0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
  79   0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0,
  80   0x78b0, 0x78b0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
  81   0x41a0, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x74a2,
  82   0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
  83   0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
  84   0x74a2, 0x74a2, 0x74a2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x51a0,
  85   0x41a0, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x70e2,
  86   0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
  87   0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
  88   0x70e2, 0x70e2, 0x70e2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x4008,
  89   0x34a2, 0x34a2, 0x01a0, 0x30e2, 0x01a0, 0x01a0, 0x01a0, 0x01a0,
  90   0x0000, 0x01a0, 0x34a2, 0x01a0, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
  91   0x30e2, 0x01a0, 0x01a0, 0x01a0, 0x01a0, 0x01a0, 0x01a0, 0x01a0,
  92   0x0008, 0x0000, 0x30e2, 0x01a0, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
  93   0x0280, 0x34a2, 0x30e2, 0x34a2, 0x01a0, 0x34a2, 0x01a0, 0x01a0,
  94   0x34a2, 0x01a0, 0x34a2, 0x01a0, 0x01a0, 0x01a0, 0x01a0, 0x34a2,
  95   0x01a0, 0x01a0, 0x34a2, 0x30e2, 0x30e2, 0x31e2, 0x01a0, 0x01a0,
  96   0x30e2, 0x0000, 0x30e2, 0x01a0, 0x30e2, 0x34a2, 0x30e2, 0x30e2,
  97   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
  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, 0x34a2,
 101   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 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, 0x30e2
 105 };
 106
 107 static int
 108 cp1251_mbc_case_fold(OnigCaseFoldType flag ARG_UNUSED,
 109              const UChar** pp, const UChar* end ARG_UNUSED, UChar* lower, OnigEncoding enc ARG_UNUSED)
 110 {
 111   const UChar* p = *pp;
 112
 113   *lower = ENC_CP1251_TO_LOWER_CASE(*p);
 114   (*pp)++;
 115   return 1;
 116 }
 117
 118 static int
 119 cp1251_is_code_ctype(OnigCodePoint code, unsigned int ctype, OnigEncoding enc ARG_UNUSED)
 120 {
 121   if (code < 256)
 122     return ENC_IS_CP1251_CTYPE(code, ctype);
 123   else
 124     return FALSE;
 125 }
 126
 127 static const OnigPairCaseFoldCodes CaseFoldMap[] = {
 128   { 0xb8, 0xa8 },
 129
 130   { 0xe0, 0xc0 },
 131   { 0xe1, 0xc1 },
 132   { 0xe2, 0xc2 },
 133   { 0xe3, 0xc3 },
 134   { 0xe4, 0xc4 },
 135   { 0xe5, 0xc5 },
 136   { 0xe6, 0xc6 },
 137   { 0xe7, 0xc7 },
 138   { 0xe8, 0xc8 },
 139   { 0xe9, 0xc9 },
 140   { 0xea, 0xca },
 141   { 0xeb, 0xcb },
 142   { 0xec, 0xcc },
 143   { 0xed, 0xcd },
 144   { 0xee, 0xce },
 145   { 0xef, 0xcf },
 146
 147   { 0xf0, 0xd0 },
 148   { 0xf1, 0xd1 },
 149   { 0xf2, 0xd2 },
 150   { 0xf3, 0xd3 },
 151   { 0xf4, 0xd4 },
 152   { 0xf5, 0xd5 },
 153   { 0xf6, 0xd6 },
 154   { 0xf7, 0xd7 },
 155   { 0xf8, 0xd8 },
 156   { 0xf9, 0xd9 },
 157   { 0xfa, 0xda },
 158   { 0xfb, 0xdb },
 159   { 0xfc, 0xdc },
 160   { 0xfd, 0xdd },
 161   { 0xfe, 0xde },
 162   { 0xff, 0xdf }
 163 };
 164
 165 static int
 166 cp1251_apply_all_case_fold(OnigCaseFoldType flag,
 167                                OnigApplyAllCaseFoldFunc f, void* arg, OnigEncoding enc ARG_UNUSED)
 168 {
 169   return onigenc_apply_all_case_fold_with_map(
 170              sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 0,
 171              flag, f, arg);
 172 }
 173
 174 static int
 175 cp1251_get_case_fold_codes_by_str(OnigCaseFoldType flag,
 176     const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[], OnigEncoding enc ARG_UNUSED)
 177 {
 178   return onigenc_get_case_fold_codes_by_str_with_map(
 179              sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 0,
 180              flag, p, end, items);
 181 }
 182
 183 OnigEncodingDefine(windows_1251, Windows_1251) = {
 184   onigenc_single_byte_mbc_enc_len,
 185   "Windows-1251",      /* name */
 186   1,             /* max enc length */
 187   1,             /* min enc length */
 188   onigenc_is_mbc_newline_0x0a,
 189   onigenc_single_byte_mbc_to_code,
 190   onigenc_single_byte_code_to_mbclen,
 191   onigenc_single_byte_code_to_mbc,
 192   cp1251_mbc_case_fold,
 193   cp1251_apply_all_case_fold,
 194   cp1251_get_case_fold_codes_by_str,
 195   onigenc_minimum_property_name_to_ctype,
 196   cp1251_is_code_ctype,
 197   onigenc_not_support_get_ctype_code_range,
 198   onigenc_single_byte_left_adjust_char_head,
 199   onigenc_always_true_is_allowed_reverse_match
 200 };
 201 /*
 202  * Name: windows-1251
 203  * MIBenum: 2251
 204  * Link: http://www.iana.org/assignments/character-sets
 205  * Link: http://www.microsoft.com/globaldev/reference/sbcs/1251.mspx
 206  * Link: http://en.wikipedia.org/wiki/Windows-1251
 207  */
 208 ENC_ALIAS("CP1251", "Windows-1251")