regenc.h

   1 #ifndef ONIGURUMA_REGENC_H
   2 #define ONIGURUMA_REGENC_H
   3 /**********************************************************************
   4   regenc.h -  Oniguruma (regular expression library)
   5 **********************************************************************/
   6 /*-
   7  * Copyright (c) 2002-2007  K.Kosako  <sndgk393 AT ybb DOT ne DOT jp>
   8  * All rights reserved.
   9  *
  10  * Redistribution and use in source and binary forms, with or without
  11  * modification, are permitted provided that the following conditions
  12  * are met:
  13  * 1. Redistributions of source code must retain the above copyright
  14  *    notice, this list of conditions and the following disclaimer.
  15  * 2. Redistributions in binary form must reproduce the above copyright
  16  *    notice, this list of conditions and the following disclaimer in the
  17  *    documentation and/or other materials provided with the distribution.
  18  *
  19  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
  20  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  22  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
  23  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  24  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  25  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  26  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  27  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  28  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  29  * SUCH DAMAGE.
  30  */
  31 #ifndef REGINT_H
  32 #ifndef RUBY_EXTERN
  33 #include "ruby/config.h"
  34 #include "ruby/defines.h"
  35 #endif
  36 #ifdef ONIG_ESCAPE_UCHAR_COLLISION
  37 #undef ONIG_ESCAPE_UCHAR_COLLISION
  38 #endif
  39 #endif
  40 #include "ruby/oniguruma.h"
  41
  42 typedef struct {
  43   OnigCodePoint from;
  44   OnigCodePoint to;
  45 } OnigPairCaseFoldCodes;
  46
  47
  48 #ifndef NULL
  49 #define NULL   ((void* )0)
  50 #endif
  51
  52 #ifndef TRUE
  53 #define TRUE    1
  54 #endif
  55
  56 #ifndef FALSE
  57 #define FALSE   0
  58 #endif
  59
  60 #ifndef ARG_UNUSED
  61 #if defined(__GNUC__)
  62 #  define ARG_UNUSED  __attribute__ ((unused))
  63 #else
  64 #  define ARG_UNUSED
  65 #endif
  66 #endif
  67
  68 #define ONIG_IS_NULL(p)                    (((void*)(p)) == (void*)0)
  69 #define ONIG_IS_NOT_NULL(p)                (((void*)(p)) != (void*)0)
  70 #define ONIG_CHECK_NULL_RETURN(p)          if (ONIG_IS_NULL(p)) return NULL
  71 #define ONIG_CHECK_NULL_RETURN_VAL(p,val)  if (ONIG_IS_NULL(p)) return (val)
  72
  73 #define enclen(enc,p,e)      ONIGENC_MBC_ENC_LEN(enc,p,e)
  74
  75 /* character types bit flag */
  76 #define BIT_CTYPE_NEWLINE  (1<< ONIGENC_CTYPE_NEWLINE)
  77 #define BIT_CTYPE_ALPHA    (1<< ONIGENC_CTYPE_ALPHA)
  78 #define BIT_CTYPE_BLANK    (1<< ONIGENC_CTYPE_BLANK)
  79 #define BIT_CTYPE_CNTRL    (1<< ONIGENC_CTYPE_CNTRL)
  80 #define BIT_CTYPE_DIGIT    (1<< ONIGENC_CTYPE_DIGIT)
  81 #define BIT_CTYPE_GRAPH    (1<< ONIGENC_CTYPE_GRAPH)
  82 #define BIT_CTYPE_LOWER    (1<< ONIGENC_CTYPE_LOWER)
  83 #define BIT_CTYPE_PRINT    (1<< ONIGENC_CTYPE_PRINT)
  84 #define BIT_CTYPE_PUNCT    (1<< ONIGENC_CTYPE_PUNCT)
  85 #define BIT_CTYPE_SPACE    (1<< ONIGENC_CTYPE_SPACE)
  86 #define BIT_CTYPE_UPPER    (1<< ONIGENC_CTYPE_UPPER)
  87 #define BIT_CTYPE_XDIGIT   (1<< ONIGENC_CTYPE_XDIGIT)
  88 #define BIT_CTYPE_WORD     (1<< ONIGENC_CTYPE_WORD)
  89 #define BIT_CTYPE_ALNUM    (1<< ONIGENC_CTYPE_ALNUM)
  90 #define BIT_CTYPE_ASCII    (1<< ONIGENC_CTYPE_ASCII)
  91
  92 #define CTYPE_TO_BIT(ctype)  (1<<(ctype))
  93 #define CTYPE_IS_WORD_GRAPH_PRINT(ctype) \
  94   ((ctype) == ONIGENC_CTYPE_WORD || (ctype) == ONIGENC_CTYPE_GRAPH ||\
  95    (ctype) == ONIGENC_CTYPE_PRINT)
  96
  97
  98 typedef struct {
  99   UChar    *name;
 100   int       ctype;
 101   short int len;
 102 } PosixBracketEntryType;
 103
 104
 105 /* #define USE_CRNL_AS_LINE_TERMINATOR */
 106 #define USE_UNICODE_PROPERTIES
 107 /* #define USE_UNICODE_CASE_FOLD_TURKISH_AZERI */
 108 /* #define USE_UNICODE_ALL_LINE_TERMINATORS */  /* see Unicode.org UTF#18 */
 109
 110
 111 #define ONIG_ENCODING_INIT_DEFAULT           ONIG_ENCODING_ASCII
 112
 113 /* for encoding system implementation (internal) */
 114 ONIG_EXTERN int onigenc_ascii_apply_all_case_fold P_((OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg, OnigEncoding enc));
 115 ONIG_EXTERN int onigenc_ascii_get_case_fold_codes_by_str P_((OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[], OnigEncoding enc));
 116 ONIG_EXTERN int onigenc_apply_all_case_fold_with_map P_((int map_size, const OnigPairCaseFoldCodes map[], int ess_tsett_flag, OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg));
 117 ONIG_EXTERN int onigenc_get_case_fold_codes_by_str_with_map P_((int map_size, const OnigPairCaseFoldCodes map[], int ess_tsett_flag, OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[]));
 118 ONIG_EXTERN int onigenc_not_support_get_ctype_code_range P_((OnigCtype ctype, OnigCodePoint* sb_out, const OnigCodePoint* ranges[], OnigEncoding enc));
 119 ONIG_EXTERN int onigenc_is_mbc_newline_0x0a P_((const UChar* p, const UChar* end, OnigEncoding enc));
 120
 121
 122 /* methods for single byte encoding */
 123 ONIG_EXTERN int onigenc_ascii_mbc_case_fold P_((OnigCaseFoldType flag, const UChar** p, const UChar* end, UChar* lower, OnigEncoding enc));
 124 ONIG_EXTERN int onigenc_single_byte_mbc_enc_len P_((const UChar* p, const UChar* e, OnigEncoding enc));
 125 ONIG_EXTERN OnigCodePoint onigenc_single_byte_mbc_to_code P_((const UChar* p, const UChar* end, OnigEncoding enc));
 126 ONIG_EXTERN int onigenc_single_byte_code_to_mbclen P_((OnigCodePoint code, OnigEncoding enc));
 127 ONIG_EXTERN int onigenc_single_byte_code_to_mbc P_((OnigCodePoint code, UChar *buf, OnigEncoding enc));
 128 ONIG_EXTERN UChar* onigenc_single_byte_left_adjust_char_head P_((const UChar* start, const UChar* s, OnigEncoding enc));
 129 ONIG_EXTERN int onigenc_always_true_is_allowed_reverse_match P_((const UChar* s, const UChar* end, OnigEncoding enc));
 130 ONIG_EXTERN int onigenc_always_false_is_allowed_reverse_match P_((const UChar* s, const UChar* end, OnigEncoding enc));
 131 ONIG_EXTERN int onigenc_ascii_is_code_ctype P_((OnigCodePoint code, unsigned int ctype, OnigEncoding enc));
 132
 133 /* methods for multi byte encoding */
 134 ONIG_EXTERN OnigCodePoint onigenc_mbn_mbc_to_code P_((OnigEncoding enc, const UChar* p, const UChar* end));
 135 ONIG_EXTERN int onigenc_mbn_mbc_case_fold P_((OnigEncoding enc, OnigCaseFoldType flag, const UChar** p, const UChar* end, UChar* lower));
 136 ONIG_EXTERN int onigenc_mb2_code_to_mbclen P_((OnigCodePoint code, OnigEncoding enc));
 137 ONIG_EXTERN int onigenc_mb2_code_to_mbc P_((OnigEncoding enc, OnigCodePoint code, UChar *buf));
 138 ONIG_EXTERN int onigenc_minimum_property_name_to_ctype P_((OnigEncoding enc, UChar* p, UChar* end));
 139 ONIG_EXTERN int onigenc_unicode_property_name_to_ctype P_((OnigEncoding enc, UChar* p, UChar* end));
 140 ONIG_EXTERN int onigenc_mb2_is_code_ctype P_((OnigEncoding enc, OnigCodePoint code, unsigned int ctype));
 141 ONIG_EXTERN int onigenc_mb4_code_to_mbclen P_((OnigCodePoint code, OnigEncoding enc));
 142 ONIG_EXTERN int onigenc_mb4_code_to_mbc P_((OnigEncoding enc, OnigCodePoint code, UChar *buf));
 143 ONIG_EXTERN int onigenc_mb4_is_code_ctype P_((OnigEncoding enc, OnigCodePoint code, unsigned int ctype));
 144
 145
 146 /* in enc/unicode.c */
 147 ONIG_EXTERN int onigenc_unicode_is_code_ctype P_((OnigCodePoint code, unsigned int ctype, OnigEncoding enc));
 148 ONIG_EXTERN int onigenc_utf16_32_get_ctype_code_range P_((OnigCtype ctype, OnigCodePoint *sb_out, const OnigCodePoint* ranges[], OnigEncoding enc));
 149 ONIG_EXTERN int onigenc_unicode_ctype_code_range P_((int ctype, const OnigCodePoint* ranges[]));
 150 ONIG_EXTERN int onigenc_unicode_get_case_fold_codes_by_str P_((OnigEncoding enc, OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[]));
 151 ONIG_EXTERN int onigenc_unicode_mbc_case_fold P_((OnigEncoding enc, OnigCaseFoldType flag, const UChar** pp, const UChar* end, UChar* fold));
 152 ONIG_EXTERN int onigenc_unicode_apply_all_case_fold P_((OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg, OnigEncoding enc));
 153
 154
 155 #define ONIGENC_ISO_8859_1_TO_LOWER_CASE(c) \
 156   OnigEncISO_8859_1_ToLowerCaseTable[c]
 157 #define ONIGENC_ISO_8859_1_TO_UPPER_CASE(c) \
 158   OnigEncISO_8859_1_ToUpperCaseTable[c]
 159
 160 ONIG_EXTERN const UChar OnigEncISO_8859_1_ToLowerCaseTable[];
 161 ONIG_EXTERN const UChar OnigEncISO_8859_1_ToUpperCaseTable[];
 162
 163 ONIG_EXTERN int
 164 onigenc_with_ascii_strncmp P_((OnigEncoding enc, const UChar* p, const UChar* end, const UChar* sascii /* ascii */, int n));
 165 ONIG_EXTERN UChar*
 166 onigenc_step P_((OnigEncoding enc, const UChar* p, const UChar* end, int n));
 167
 168 /* defined in regexec.c, but used in enc/xxx.c */
 169 extern int  onig_is_in_code_range P_((const UChar* p, OnigCodePoint code));
 170
 171 ONIG_EXTERN OnigEncoding  OnigEncDefaultCharEncoding;
 172 ONIG_EXTERN const UChar  OnigEncAsciiToLowerCaseTable[];
 173 ONIG_EXTERN const UChar  OnigEncAsciiToUpperCaseTable[];
 174 ONIG_EXTERN const unsigned short OnigEncAsciiCtypeTable[];
 175
 176 #define ONIGENC_IS_ASCII_CODE(code)  ((code) < 0x80)
 177 #define ONIGENC_ASCII_CODE_TO_LOWER_CASE(c) OnigEncAsciiToLowerCaseTable[c]
 178 #define ONIGENC_ASCII_CODE_TO_UPPER_CASE(c) OnigEncAsciiToUpperCaseTable[c]
 179 #define ONIGENC_IS_ASCII_CODE_CTYPE(code,ctype) \
 180   ((OnigEncAsciiCtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
 181 #define ONIGENC_IS_ASCII_CODE_CASE_AMBIG(code) \
 182  (ONIGENC_IS_ASCII_CODE_CTYPE(code, ONIGENC_CTYPE_UPPER) ||\
 183   ONIGENC_IS_ASCII_CODE_CTYPE(code, ONIGENC_CTYPE_LOWER))
 184
 185 #ifdef ONIG_ENC_REGISTER
 186 extern int ONIG_ENC_REGISTER(const char *, OnigEncodingType*);
 187 #define OnigEncodingName(n) encoding_##n
 188 #define OnigEncodingDeclare(n) static OnigEncodingType OnigEncodingName(n)
 189 #define OnigEncodingDefine(f,n)                      \
 190     OnigEncodingDeclare(n);                          \
 191     void Init_##f(void) {                            \
 192         ONIG_ENC_REGISTER(OnigEncodingName(n).name,  \
 193                           &OnigEncodingName(n));     \
 194     }                                                \
 195     OnigEncodingDeclare(n)
 196 #else
 197 #define OnigEncodingName(n) OnigEncoding##n
 198 #define OnigEncodingDeclare(n) OnigEncodingType OnigEncodingName(n)
 199 #define OnigEncodingDefine(f,n) OnigEncodingDeclare(n)
 200 #endif
 201
 202 /* macros for define replica encoding and encoding alias */
 203 #define ENC_REPLICATE(name, orig)
 204 #define ENC_ALIAS(name, orig)
 205 #define ENC_DUMMY(name)
 206
 207 #endif /* ONIGURUMA_REGENC_H */