1 #ifndef REGENC_H 2 #define REGENC_H 3 /********************************************************************** 4 regenc.h - Oniguruma (regular expression library) 5 **********************************************************************/ 6 /*- 7 * Copyright (c) 2002-2016 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 32 #ifndef PACKAGE 33 /* PACKAGE is defined in config.h */ 34 #include "config.h" 35 #endif 36 37 #ifdef ONIG_ESCAPE_UCHAR_COLLISION 38 #undef ONIG_ESCAPE_UCHAR_COLLISION 39 #endif 40 41 #include "oniguruma.h" 42 43 typedef struct { 44 OnigCodePoint from; 45 OnigCodePoint to; 46 } OnigPairCaseFoldCodes; 47 48 49 #ifndef NULL 50 #define NULL ((void* )0) 51 #endif 52 53 #ifndef TRUE 54 #define TRUE 1 55 #endif 56 57 #ifndef FALSE 58 #define FALSE 0 59 #endif 60 61 #ifndef ARG_UNUSED 62 #if defined(__GNUC__) 63 # define ARG_UNUSED __attribute__ ((unused)) 64 #else 65 # define ARG_UNUSED 66 #endif 67 #endif 68 69 #define ONIG_IS_NULL(p) (((void*)(p)) == (void*)0) 70 #define ONIG_IS_NOT_NULL(p) (((void*)(p)) != (void*)0) 71 #define ONIG_CHECK_NULL_RETURN(p) if (ONIG_IS_NULL(p)) return NULL 72 #define ONIG_CHECK_NULL_RETURN_VAL(p,val) if (ONIG_IS_NULL(p)) return (val) 73 74 #define enclen(enc,p) ONIGENC_MBC_ENC_LEN(enc,p) 75 76 /* character types bit flag */ 77 #define BIT_CTYPE_NEWLINE (1<< ONIGENC_CTYPE_NEWLINE) 78 #define BIT_CTYPE_ALPHA (1<< ONIGENC_CTYPE_ALPHA) 79 #define BIT_CTYPE_BLANK (1<< ONIGENC_CTYPE_BLANK) 80 #define BIT_CTYPE_CNTRL (1<< ONIGENC_CTYPE_CNTRL) 81 #define BIT_CTYPE_DIGIT (1<< ONIGENC_CTYPE_DIGIT) 82 #define BIT_CTYPE_GRAPH (1<< ONIGENC_CTYPE_GRAPH) 83 #define BIT_CTYPE_LOWER (1<< ONIGENC_CTYPE_LOWER) 84 #define BIT_CTYPE_PRINT (1<< ONIGENC_CTYPE_PRINT) 85 #define BIT_CTYPE_PUNCT (1<< ONIGENC_CTYPE_PUNCT) 86 #define BIT_CTYPE_SPACE (1<< ONIGENC_CTYPE_SPACE) 87 #define BIT_CTYPE_UPPER (1<< ONIGENC_CTYPE_UPPER) 88 #define BIT_CTYPE_XDIGIT (1<< ONIGENC_CTYPE_XDIGIT) 89 #define BIT_CTYPE_WORD (1<< ONIGENC_CTYPE_WORD) 90 #define BIT_CTYPE_ALNUM (1<< ONIGENC_CTYPE_ALNUM) 91 #define BIT_CTYPE_ASCII (1<< ONIGENC_CTYPE_ASCII) 92 93 #define CTYPE_TO_BIT(ctype) (1<<(ctype)) 94 #define CTYPE_IS_WORD_GRAPH_PRINT(ctype) \ 95 ((ctype) == ONIGENC_CTYPE_WORD || (ctype) == ONIGENC_CTYPE_GRAPH ||\ 96 (ctype) == ONIGENC_CTYPE_PRINT) 97 98 99 typedef struct { 100 UChar *name; 101 int ctype; 102 short int len; 103 } PosixBracketEntryType; 104 105 struct PropertyNameCtype { 106 char *name; 107 int ctype; 108 }; 109 110 /* #define USE_CRNL_AS_LINE_TERMINATOR */ 111 #define USE_UNICODE_PROPERTIES 112 /* #define USE_UNICODE_CASE_FOLD_TURKISH_AZERI */ 113 /* #define USE_UNICODE_ALL_LINE_TERMINATORS */ /* see Unicode.org UTS #18 */ 114 115 116 #define ONIG_ENCODING_INIT_DEFAULT ONIG_ENCODING_ASCII 117 118 /* for encoding system implementation (internal) */ 119 ONIG_EXTERN int onigenc_ascii_apply_all_case_fold P_((OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg)); 120 ONIG_EXTERN int onigenc_ascii_get_case_fold_codes_by_str P_((OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[])); 121 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)); 122 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[])); 123 ONIG_EXTERN int onigenc_not_support_get_ctype_code_range P_((OnigCtype ctype, OnigCodePoint* sb_out, const OnigCodePoint* ranges[])); 124 ONIG_EXTERN int onigenc_is_mbc_newline_0x0a P_((const UChar* p, const UChar* end)); 125 126 127 /* methods for single byte encoding */ 128 ONIG_EXTERN int onigenc_ascii_mbc_case_fold P_((OnigCaseFoldType flag, const UChar** p, const UChar* end, UChar* lower)); 129 ONIG_EXTERN int onigenc_single_byte_mbc_enc_len P_((const UChar* p)); 130 ONIG_EXTERN OnigCodePoint onigenc_single_byte_mbc_to_code P_((const UChar* p, const UChar* end)); 131 ONIG_EXTERN int onigenc_single_byte_code_to_mbclen P_((OnigCodePoint code)); 132 ONIG_EXTERN int onigenc_single_byte_code_to_mbc P_((OnigCodePoint code, UChar *buf)); 133 ONIG_EXTERN UChar* onigenc_single_byte_left_adjust_char_head P_((const UChar* start, const UChar* s)); 134 ONIG_EXTERN int onigenc_always_true_is_allowed_reverse_match P_((const UChar* s, const UChar* end)); 135 ONIG_EXTERN int onigenc_always_false_is_allowed_reverse_match P_((const UChar* s, const UChar* end)); 136 ONIG_EXTERN int onigenc_always_true_is_valid_mbc_string P_((const UChar* s, const UChar* end)); 137 ONIG_EXTERN int onigenc_length_check_is_valid_mbc_string P_((OnigEncoding enc, const UChar* s, const UChar* end)); 138 139 /* methods for multi byte encoding */ 140 ONIG_EXTERN OnigCodePoint onigenc_mbn_mbc_to_code P_((OnigEncoding enc, const UChar* p, const UChar* end)); 141 ONIG_EXTERN int onigenc_mbn_mbc_case_fold P_((OnigEncoding enc, OnigCaseFoldType flag, const UChar** p, const UChar* end, UChar* lower)); 142 ONIG_EXTERN int onigenc_mb2_code_to_mbclen P_((OnigCodePoint code)); 143 ONIG_EXTERN int onigenc_mb2_code_to_mbc P_((OnigEncoding enc, OnigCodePoint code, UChar *buf)); 144 ONIG_EXTERN int onigenc_minimum_property_name_to_ctype P_((OnigEncoding enc, UChar* p, UChar* end)); 145 ONIG_EXTERN int onigenc_unicode_property_name_to_ctype P_((OnigEncoding enc, UChar* p, UChar* end)); 146 ONIG_EXTERN int onigenc_mb2_is_code_ctype P_((OnigEncoding enc, OnigCodePoint code, unsigned int ctype)); 147 ONIG_EXTERN int onigenc_mb4_code_to_mbclen P_((OnigCodePoint code)); 148 ONIG_EXTERN int onigenc_mb4_code_to_mbc P_((OnigEncoding enc, OnigCodePoint code, UChar *buf)); 149 ONIG_EXTERN int onigenc_mb4_is_code_ctype P_((OnigEncoding enc, OnigCodePoint code, unsigned int ctype)); 150 ONIG_EXTERN struct PropertyNameCtype* euc_jp_lookup_property_name P_((register const char *str, register unsigned int len)); 151 ONIG_EXTERN struct PropertyNameCtype* sjis_lookup_property_name P_((register const char *str, register unsigned int len)); 152 //ONIG_EXTERN const struct PropertyNameCtype* unicode_lookup_property_name P_((register const char *str, register unsigned int len)); 153 154 /* in enc/unicode.c */ 155 ONIG_EXTERN int onigenc_unicode_is_code_ctype P_((OnigCodePoint code, unsigned int ctype)); 156 ONIG_EXTERN int onigenc_utf16_32_get_ctype_code_range P_((OnigCtype ctype, OnigCodePoint *sb_out, const OnigCodePoint* ranges[])); 157 ONIG_EXTERN int onigenc_unicode_ctype_code_range P_((int ctype, const OnigCodePoint* ranges[])); 158 ONIG_EXTERN int onigenc_unicode_get_case_fold_codes_by_str P_((OnigEncoding enc, OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[])); 159 ONIG_EXTERN int onigenc_unicode_mbc_case_fold P_((OnigEncoding enc, OnigCaseFoldType flag, const UChar** pp, const UChar* end, UChar* fold)); 160 ONIG_EXTERN int onigenc_unicode_apply_all_case_fold P_((OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg)); 161 162 163 #define UTF16_IS_SURROGATE_FIRST(c) (((c) & 0xfc) == 0xd8) 164 #define UTF16_IS_SURROGATE_SECOND(c) (((c) & 0xfc) == 0xdc) 165 166 /* from unicode generated codes */ 167 #define FOLDS1_FOLD(i) (OnigUnicodeFolds1 + (i)) 168 #define FOLDS2_FOLD(i) (OnigUnicodeFolds2 + (i)) 169 #define FOLDS3_FOLD(i) (OnigUnicodeFolds3 + (i)) 170 #define FOLDS1_UNFOLDS_NUM(i) (OnigUnicodeFolds1[(i)+1]) 171 #define FOLDS2_UNFOLDS_NUM(i) (OnigUnicodeFolds2[(i)+2]) 172 #define FOLDS3_UNFOLDS_NUM(i) (OnigUnicodeFolds3[(i)+3]) 173 #define FOLDS1_UNFOLDS(i) (OnigUnicodeFolds1 + (i) + 2) 174 #define FOLDS2_UNFOLDS(i) (OnigUnicodeFolds2 + (i) + 3) 175 #define FOLDS3_UNFOLDS(i) (OnigUnicodeFolds3 + (i) + 4) 176 #define FOLDS1_NEXT_INDEX(i) ((i) + 2 + OnigUnicodeFolds1[(i)+1]) 177 #define FOLDS2_NEXT_INDEX(i) ((i) + 3 + OnigUnicodeFolds2[(i)+2]) 178 #define FOLDS3_NEXT_INDEX(i) ((i) + 4 + OnigUnicodeFolds3[(i)+3]) 179 180 #define FOLDS_FOLD_ADDR_BUK(buk, addr) do {\ 181 if ((buk)->fold_len == 1)\ 182 addr = OnigUnicodeFolds1 + (buk)->index;\ 183 else if ((buk)->fold_len == 2)\ 184 addr = OnigUnicodeFolds2 + (buk)->index;\ 185 else if ((buk)->fold_len == 3)\ 186 addr = OnigUnicodeFolds3 + (buk)->index;\ 187 else\ 188 addr = 0;\ 189 } while (0) 190 191 extern OnigCodePoint OnigUnicodeFolds1[]; 192 extern OnigCodePoint OnigUnicodeFolds2[]; 193 extern OnigCodePoint OnigUnicodeFolds3[]; 194 195 struct ByUnfoldKey { 196 OnigCodePoint code; 197 int index; 198 int fold_len; 199 }; 200 201 extern const struct ByUnfoldKey* unicode_unfold_key(OnigCodePoint code); 202 extern int unicode_fold1_key(OnigCodePoint code[]); 203 extern int unicode_fold2_key(OnigCodePoint code[]); 204 extern int unicode_fold3_key(OnigCodePoint code[]); 205 206 extern int onig_codes_cmp(OnigCodePoint a[], OnigCodePoint b[], int n); 207 extern int onig_codes_byte_at(OnigCodePoint code[], int at); 208 209 210 211 #define ONIGENC_ISO_8859_1_TO_LOWER_CASE(c) \ 212 OnigEncISO_8859_1_ToLowerCaseTable[c] 213 #define ONIGENC_ISO_8859_1_TO_UPPER_CASE(c) \ 214 OnigEncISO_8859_1_ToUpperCaseTable[c] 215 216 ONIG_EXTERN const UChar OnigEncISO_8859_1_ToLowerCaseTable[]; 217 ONIG_EXTERN const UChar OnigEncISO_8859_1_ToUpperCaseTable[]; 218 219 ONIG_EXTERN int 220 onigenc_with_ascii_strncmp P_((OnigEncoding enc, const UChar* p, const UChar* end, const UChar* sascii /* ascii */, int n)); 221 ONIG_EXTERN UChar* 222 onigenc_step P_((OnigEncoding enc, const UChar* p, const UChar* end, int n)); 223 224 /* defined in regexec.c, but used in enc/xxx.c */ 225 extern int onig_is_in_code_range P_((const UChar* p, OnigCodePoint code)); 226 227 ONIG_EXTERN OnigEncoding OnigEncDefaultCharEncoding; 228 ONIG_EXTERN const UChar OnigEncAsciiToLowerCaseTable[]; 229 ONIG_EXTERN const UChar OnigEncAsciiToUpperCaseTable[]; 230 ONIG_EXTERN const unsigned short OnigEncAsciiCtypeTable[]; 231 232 233 #define ONIGENC_IS_ASCII_CODE(code) ((code) < 0x80) 234 #define ONIGENC_ASCII_CODE_TO_LOWER_CASE(c) OnigEncAsciiToLowerCaseTable[c] 235 #define ONIGENC_ASCII_CODE_TO_UPPER_CASE(c) OnigEncAsciiToUpperCaseTable[c] 236 #define ONIGENC_IS_ASCII_CODE_CTYPE(code,ctype) \ 237 ((OnigEncAsciiCtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0) 238 #define ONIGENC_IS_ASCII_CODE_CASE_AMBIG(code) \ 239 (ONIGENC_IS_ASCII_CODE_CTYPE(code, ONIGENC_CTYPE_UPPER) ||\ 240 ONIGENC_IS_ASCII_CODE_CTYPE(code, ONIGENC_CTYPE_LOWER)) 241 242 243 #endif /* REGENC_H */ 244