enc/iso_8859_14.c

   1 /**********************************************************************
   2   iso8859_14.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 #include "iso_8859.h"
  32
  33 #define ENC_ISO_8859_14_TO_LOWER_CASE(c) EncISO_8859_14_ToLowerCaseTable[c]
  34 #define ENC_IS_ISO_8859_14_CTYPE(code,ctype) \
  35   ((EncISO_8859_14_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
  36
  37 static const UChar EncISO_8859_14_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   '\200', '\201', '\202', '\203', '\204', '\205', '\206', '\207',
  55   '\210', '\211', '\212', '\213', '\214', '\215', '\216', '\217',
  56   '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227',
  57   '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\237',
  58   '\240', '\242', '\242', '\243', '\245', '\245', '\253', '\247',
  59   '\270', '\251', '\272', '\253', '\274', '\255', '\256', '\377',
  60   '\261', '\261', '\263', '\263', '\265', '\265', '\266', '\271',
  61   '\270', '\271', '\272', '\277', '\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', '\337',
  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 EncISO_8859_14_CtypeTable[256] = {
  73   0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
  74   0x4008, 0x420c, 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   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   0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
  93   0x0284, 0x34a2, 0x30e2, 0x00a0, 0x34a2, 0x30e2, 0x34a2, 0x00a0,
  94   0x34a2, 0x00a0, 0x34a2, 0x30e2, 0x34a2, 0x01a0, 0x00a0, 0x34a2,
  95   0x34a2, 0x30e2, 0x34a2, 0x30e2, 0x34a2, 0x30e2, 0x00a0, 0x34a2,
  96   0x30e2, 0x30e2, 0x30e2, 0x34a2, 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, 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, 0x30e2,
 104   0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2
 105 };
 106
 107 static int
 108 mbc_case_fold(OnigCaseFoldType flag,
 109                           const UChar** pp, const UChar* end ARG_UNUSED,
 110                           UChar* lower, OnigEncoding enc ARG_UNUSED)
 111 {
 112   const UChar* p = *pp;
 113
 114   if (*p == SHARP_s && (flag & INTERNAL_ONIGENC_CASE_FOLD_MULTI_CHAR) != 0) {
 115     *lower++ = 's';
 116     *lower   = 's';
 117     (*pp)++;
 118     return 2;
 119   }
 120
 121   *lower = ENC_ISO_8859_14_TO_LOWER_CASE(*p);
 122   (*pp)++;
 123   return 1; /* return byte length of converted char to lower */
 124 }
 125
 126 #if 0
 127 static int
 128 is_mbc_ambiguous(OnigCaseFoldType flag,
 129                  const UChar** pp, const UChar* end)
 130 {
 131   int v;
 132   const UChar* p = *pp;
 133
 134   if (*p == SHARP_s && (flag & INTERNAL_ONIGENC_CASE_FOLD_MULTI_CHAR) != 0) {
 135     (*pp)++;
 136     return TRUE;
 137   }
 138
 139   (*pp)++;
 140   v = (EncISO_8859_14_CtypeTable[*p] & (BIT_CTYPE_UPPER | BIT_CTYPE_LOWER));
 141   if ((v | BIT_CTYPE_LOWER) != 0) {
 142     return TRUE;
 143   }
 144
 145   return (v != 0 ? TRUE : FALSE);
 146 }
 147 #endif
 148
 149 static int
 150 is_code_ctype(OnigCodePoint code, unsigned int ctype, OnigEncoding enc ARG_UNUSED)
 151 {
 152   if (code < 256)
 153     return ENC_IS_ISO_8859_14_CTYPE(code, ctype);
 154   else
 155     return FALSE;
 156 }
 157
 158 static const OnigPairCaseFoldCodes CaseFoldMap[] = {
 159  { 0xa1, 0xa2 },
 160  { 0xa4, 0xa5 },
 161  { 0xa6, 0xab },
 162  { 0xa8, 0xb8 },
 163  { 0xaa, 0xba },
 164  { 0xac, 0xbc },
 165  { 0xaf, 0xff },
 166
 167  { 0xb0, 0xb1 },
 168  { 0xb2, 0xb3 },
 169  { 0xb4, 0xb5 },
 170  { 0xb7, 0xb9 },
 171  { 0xbb, 0xbf },
 172  { 0xbd, 0xbe },
 173
 174  { 0xc0, 0xe0 },
 175  { 0xc1, 0xe1 },
 176  { 0xc2, 0xe2 },
 177  { 0xc3, 0xe3 },
 178  { 0xc4, 0xe4 },
 179  { 0xc5, 0xe5 },
 180  { 0xc6, 0xe6 },
 181  { 0xc7, 0xe7 },
 182  { 0xc8, 0xe8 },
 183  { 0xc9, 0xe9 },
 184  { 0xca, 0xea },
 185  { 0xcb, 0xeb },
 186  { 0xcc, 0xec },
 187  { 0xcd, 0xed },
 188  { 0xce, 0xee },
 189  { 0xcf, 0xef },
 190
 191  { 0xd0, 0xf0 },
 192  { 0xd1, 0xf1 },
 193  { 0xd2, 0xf2 },
 194  { 0xd3, 0xf3 },
 195  { 0xd4, 0xf4 },
 196  { 0xd5, 0xf5 },
 197  { 0xd6, 0xf6 },
 198  { 0xd7, 0xf7 },
 199  { 0xd8, 0xf8 },
 200  { 0xd9, 0xf9 },
 201  { 0xda, 0xfa },
 202  { 0xdb, 0xfb },
 203  { 0xdc, 0xfc },
 204  { 0xdd, 0xfd },
 205  { 0xde, 0xfe }
 206 };
 207
 208 static int
 209 apply_all_case_fold(OnigCaseFoldType flag,
 210                                 OnigApplyAllCaseFoldFunc f, void* arg,
 211                                 OnigEncoding enc ARG_UNUSED)
 212 {
 213   return onigenc_apply_all_case_fold_with_map(
 214              numberof(CaseFoldMap), CaseFoldMap, 1,
 215              flag, f, arg);
 216 }
 217
 218 static int
 219 get_case_fold_codes_by_str(OnigCaseFoldType flag,
 220                            const OnigUChar* p, const OnigUChar* end,
 221                            OnigCaseFoldCodeItem items[],
 222                            OnigEncoding enc ARG_UNUSED)
 223 {
 224   return onigenc_get_case_fold_codes_by_str_with_map(
 225              numberof(CaseFoldMap), CaseFoldMap, 1,
 226              flag, p, end, items);
 227 }
 228
 229 static int
 230 case_map(OnigCaseFoldType* flagP, const OnigUChar** pp,
 231          const OnigUChar* end, OnigUChar* to, OnigUChar* to_end,
 232          const struct OnigEncodingTypeST* enc)
 233 {
 234   OnigCodePoint code;
 235   OnigUChar *to_start = to;
 236   OnigCaseFoldType flags = *flagP;
 237
 238   while (*pp < end && to < to_end) {
 239     code = *(*pp)++;
 240     if (code == SHARP_s) {
 241       if (flags & ONIGENC_CASE_UPCASE) {
 242         flags |= ONIGENC_CASE_MODIFIED;
 243         *to++ = 'S';
 244         code = (flags & ONIGENC_CASE_TITLECASE) ? 's' : 'S';
 245       }
 246       else if (flags & ONIGENC_CASE_FOLD) {
 247         flags |= ONIGENC_CASE_MODIFIED;
 248         *to++ = 's';
 249         code = 's';
 250       }
 251     }
 252     /* else if (code == 0xAA || code == 0xBA) ; */
 253     else if ((EncISO_8859_14_CtypeTable[code] & BIT_CTYPE_UPPER)
 254              && (flags & (ONIGENC_CASE_DOWNCASE | ONIGENC_CASE_FOLD))) {
 255       flags |= ONIGENC_CASE_MODIFIED;
 256       code = ENC_ISO_8859_14_TO_LOWER_CASE(code);
 257     }
 258     else if ((EncISO_8859_14_CtypeTable[code] & BIT_CTYPE_LOWER)
 259              && (flags & ONIGENC_CASE_UPCASE)) {
 260       flags |= ONIGENC_CASE_MODIFIED;
 261       if (code == 0xA2 || code == 0xA5 || code == 0xB1 || code == 0xB3 || code == 0xB5 || code == 0xBE)
 262         code -= 0x1;
 263       else if (code == 0xAB)
 264         code -= 0x5;
 265       else if (code == 0xFF)
 266         code -= 0x50;
 267       else if (code == 0xB9)
 268         code -= 0x2;
 269       else if (code == 0xBF)
 270         code -= 0x4;
 271       else if (code == 0xB8 || code == 0xBA || code == 0xBC)
 272         code -= 0x10;
 273       else
 274         code -= 0x20;
 275     }
 276     *to++ = code;
 277     if (flags & ONIGENC_CASE_TITLECASE)  /* switch from titlecase to lowercase for capitalize */
 278       flags ^= (ONIGENC_CASE_UPCASE | ONIGENC_CASE_DOWNCASE | ONIGENC_CASE_TITLECASE);
 279   }
 280   *flagP = flags;
 281   return (int )(to - to_start);
 282 }
 283
 284 OnigEncodingDefine(iso_8859_14, ISO_8859_14) = {
 285   onigenc_single_byte_mbc_enc_len,
 286   "ISO-8859-14",  /* name */
 287   1,             /* max enc length */
 288   1,             /* min enc length */
 289   onigenc_is_mbc_newline_0x0a,
 290   onigenc_single_byte_mbc_to_code,
 291   onigenc_single_byte_code_to_mbclen,
 292   onigenc_single_byte_code_to_mbc,
 293   mbc_case_fold,
 294   apply_all_case_fold,
 295   get_case_fold_codes_by_str,
 296   onigenc_minimum_property_name_to_ctype,
 297   is_code_ctype,
 298   onigenc_not_support_get_ctype_code_range,
 299   onigenc_single_byte_left_adjust_char_head,
 300   onigenc_always_true_is_allowed_reverse_match,
 301   case_map,
 302   0,
 303   ONIGENC_FLAG_NONE,
 304 };
 305 ENC_ALIAS("ISO8859-14", "ISO-8859-14")