enc/iso_8859_7.c

   1 /**********************************************************************
   2   iso8859_7.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_7_TO_LOWER_CASE(c) EncISO_8859_7_ToLowerCaseTable[c]
  33 #define ENC_IS_ISO_8859_7_CTYPE(code,ctype) \
  34   ((EncISO_8859_7_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
  35
  36 static const UChar EncISO_8859_7_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', '\263', '\264', '\265', '\334', '\267',
  60   '\335', '\336', '\337', '\273', '\374', '\275', '\375', '\376',
  61   '\300', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
  62   '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
  63   '\360', '\361', '\322', '\363', '\364', '\365', '\366', '\367',
  64   '\370', '\371', '\372', '\373', '\334', '\335', '\336', '\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_7_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, 0x01a0, 0x01a0, 0x00a0, 0x0000, 0x0000, 0x00a0, 0x00a0,
  93   0x00a0, 0x00a0, 0x0000, 0x01a0, 0x00a0, 0x01a0, 0x0000, 0x01a0,
  94   0x00a0, 0x00a0, 0x10a0, 0x10a0, 0x00a0, 0x00a0, 0x34a2, 0x01a0,
  95   0x34a2, 0x34a2, 0x34a2, 0x01a0, 0x34a2, 0x10a0, 0x34a2, 0x34a2,
  96   0x30e2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
  97   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
  98   0x34a2, 0x34a2, 0x0000, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
  99   0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x30e2, 0x30e2, 0x30e2, 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, 0x30e2,
 103   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x0000
 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   *lower = ENC_ISO_8859_7_TO_LOWER_CASE(*p);
 114   (*pp)++;
 115   return 1;
 116 }
 117
 118 #if 0
 119 static int
 120 is_mbc_ambiguous(OnigCaseFoldType flag,
 121                  const UChar** pp, const UChar* end)
 122 {
 123   int v;
 124   const UChar* p = *pp;
 125
 126   (*pp)++;
 127   v = (EncISO_8859_7_CtypeTable[*p] & (BIT_CTYPE_UPPER | BIT_CTYPE_LOWER));
 128   if ((v | BIT_CTYPE_LOWER) != 0) {
 129     if (*p == 0xc0 || *p == 0xe0)
 130       return FALSE;
 131     else
 132       return TRUE;
 133   }
 134
 135   return (v != 0 ? TRUE : FALSE);
 136 }
 137 #endif
 138
 139 static int
 140 is_code_ctype(OnigCodePoint code, unsigned int ctype, OnigEncoding enc ARG_UNUSED)
 141 {
 142   if (code < 256)
 143     return ENC_IS_ISO_8859_7_CTYPE(code, ctype);
 144   else
 145     return FALSE;
 146 }
 147
 148 static const OnigPairCaseFoldCodes CaseFoldMap[] = {
 149  { 0xb6, 0xdc },
 150  { 0xb8, 0xdd },
 151  { 0xb9, 0xde },
 152  { 0xba, 0xdf },
 153  { 0xbc, 0xfc },
 154  { 0xbe, 0xfd },
 155  { 0xbf, 0xfe },
 156
 157  { 0xc1, 0xe1 },
 158  { 0xc2, 0xe2 },
 159  { 0xc3, 0xe3 },
 160  { 0xc4, 0xe4 },
 161  { 0xc5, 0xe5 },
 162  { 0xc6, 0xe6 },
 163  { 0xc7, 0xe7 },
 164  { 0xc8, 0xe8 },
 165  { 0xc9, 0xe9 },
 166  { 0xca, 0xea },
 167  { 0xcb, 0xeb },
 168  { 0xcc, 0xec },
 169  { 0xcd, 0xed },
 170  { 0xce, 0xee },
 171  { 0xcf, 0xef },
 172
 173  { 0xd0, 0xf0 },
 174  { 0xd1, 0xf1 },
 175  { 0xd2, 0xf2 },
 176  { 0xd3, 0xf3 },
 177  { 0xd4, 0xf4 },
 178  { 0xd5, 0xf5 },
 179  { 0xd6, 0xf6 },
 180  { 0xd7, 0xf7 },
 181  { 0xd8, 0xf8 },
 182  { 0xd9, 0xf9 },
 183  { 0xda, 0xfa },
 184  { 0xdb, 0xfb }
 185 };
 186
 187 static int
 188 apply_all_case_fold(OnigCaseFoldType flag,
 189                     OnigApplyAllCaseFoldFunc f, void* arg,
 190                     OnigEncoding enc ARG_UNUSED)
 191 {
 192   return onigenc_apply_all_case_fold_with_map(
 193              numberof(CaseFoldMap), CaseFoldMap, 0,
 194              flag, f, arg);
 195 }
 196
 197 static int
 198 get_case_fold_codes_by_str(OnigCaseFoldType flag,
 199                            const OnigUChar* p, const OnigUChar* end,
 200                            OnigCaseFoldCodeItem items[],
 201                            OnigEncoding enc ARG_UNUSED)
 202 {
 203   return onigenc_get_case_fold_codes_by_str_with_map(
 204              numberof(CaseFoldMap), CaseFoldMap, 0,
 205              flag, p, end, items);
 206 }
 207
 208 static int
 209 case_map(OnigCaseFoldType* flagP, const OnigUChar** pp,
 210          const OnigUChar* end, OnigUChar* to, OnigUChar* to_end,
 211          const struct OnigEncodingTypeST* enc)
 212 {
 213   OnigCodePoint code;
 214   OnigUChar *to_start = to;
 215   OnigCaseFoldType flags = *flagP;
 216
 217   while (*pp < end && to < to_end) {
 218     code = *(*pp)++;
 219     if (code == 0xF2) {
 220       if (flags & ONIGENC_CASE_UPCASE) {
 221         flags |= ONIGENC_CASE_MODIFIED;
 222         code = 0xD3;
 223       }
 224       else if (flags & ONIGENC_CASE_FOLD) {
 225         flags |= ONIGENC_CASE_MODIFIED;
 226         code = 0xF3;
 227       }
 228     }
 229     else if ((EncISO_8859_7_CtypeTable[code] & BIT_CTYPE_UPPER)
 230              && (flags & (ONIGENC_CASE_DOWNCASE | ONIGENC_CASE_FOLD))) {
 231       flags |= ONIGENC_CASE_MODIFIED;
 232       code = ENC_ISO_8859_7_TO_LOWER_CASE(code);
 233     }
 234     else if (code == 0xC0 || code == 0xE0)
 235       ;
 236     else if ((EncISO_8859_7_CtypeTable[code]&BIT_CTYPE_LOWER)
 237              && (flags & ONIGENC_CASE_UPCASE)) {
 238       flags |= ONIGENC_CASE_MODIFIED;
 239       if (code == 0xDC) {
 240         code -= 0x26;
 241       }
 242       else if (code >= 0xDD && code <= 0xDF) {
 243         code -= 0x25;
 244       }
 245       else if (code == 0xFC) {
 246         code -= 0x40;
 247       }
 248       else if (code == 0xFD || code == 0xFE) {
 249         code -= 0x3F;
 250       }
 251       else {
 252         code -= 0x20;
 253       }
 254     }
 255     *to++ = code;
 256     if (flags & ONIGENC_CASE_TITLECASE)  /* switch from titlecase to lowercase for capitalize */
 257       flags ^= (ONIGENC_CASE_UPCASE | ONIGENC_CASE_DOWNCASE | ONIGENC_CASE_TITLECASE);
 258   }
 259   *flagP = flags;
 260   return (int )(to - to_start);
 261 }
 262
 263 OnigEncodingDefine(iso_8859_7, ISO_8859_7) = {
 264   onigenc_single_byte_mbc_enc_len,
 265   "ISO-8859-7",  /* name */
 266   1,             /* max enc length */
 267   1,             /* min enc length */
 268   onigenc_is_mbc_newline_0x0a,
 269   onigenc_single_byte_mbc_to_code,
 270   onigenc_single_byte_code_to_mbclen,
 271   onigenc_single_byte_code_to_mbc,
 272   mbc_case_fold,
 273   apply_all_case_fold,
 274   get_case_fold_codes_by_str,
 275   onigenc_minimum_property_name_to_ctype,
 276   is_code_ctype,
 277   onigenc_not_support_get_ctype_code_range,
 278   onigenc_single_byte_left_adjust_char_head,
 279   onigenc_always_true_is_allowed_reverse_match,
 280   case_map,
 281   0,
 282   ONIGENC_FLAG_NONE,
 283 };
 284 ENC_ALIAS("ISO8859-7", "ISO-8859-7")