1#ifndef ONIGURUMA_REGENC_H 2#define ONIGURUMA_REGENC_H 3/********************************************************************** 4 regenc.h - Onigmo (Oniguruma-mod) (regular expression library) 5**********************************************************************/ 6/*- 7 * Copyright (c) 2002-2008 K.Kosako <sndgk393 AT ybb DOT ne DOT jp> 8 * Copyright (c) 2011 K.Takata <kentkt AT csc DOT jp> 9 * All rights reserved. 10 * 11 * Redistribution and use in source and binary forms, with or without 12 * modification, are permitted provided that the following conditions 13 * are met: 14 * 1. Redistributions of source code must retain the above copyright 15 * notice, this list of conditions and the following disclaimer. 16 * 2. Redistributions in binary form must reproduce the above copyright 17 * notice, this list of conditions and the following disclaimer in the 18 * documentation and/or other materials provided with the distribution. 19 * 20 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 23 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 24 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 25 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 26 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 27 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 28 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 29 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 30 * SUCH DAMAGE. 31 */ 32#ifndef REGINT_H 33#ifndef RUBY_EXTERN 34#include "ruby/config.h" 35#include "ruby/defines.h" 36#endif 37#ifdef ONIG_ESCAPE_UCHAR_COLLISION 38#undef ONIG_ESCAPE_UCHAR_COLLISION 39#endif 40#endif 41#include "ruby/oniguruma.h" 42 43#if defined __GNUC__ && __GNUC__ >= 4 44#pragma GCC visibility push(default) 45#endif 46 47typedef struct { 48 OnigCodePoint from; 49 OnigCodePoint to; 50} OnigPairCaseFoldCodes; 51 52 53#ifndef NULL 54#define NULL ((void* )0) 55#endif 56 57#ifndef TRUE 58#define TRUE 1 59#endif 60 61#ifndef FALSE 62#define FALSE 0 63#endif 64 65#ifndef ARG_UNUSED 66#if defined(__GNUC__) 67# define ARG_UNUSED __attribute__ ((unused)) 68#else 69# define ARG_UNUSED 70#endif 71#endif 72 73#define ONIG_IS_NULL(p) (((void*)(p)) == (void*)0) 74#define ONIG_IS_NOT_NULL(p) (((void*)(p)) != (void*)0) 75#define ONIG_CHECK_NULL_RETURN(p) if (ONIG_IS_NULL(p)) return NULL 76#define ONIG_CHECK_NULL_RETURN_VAL(p,val) if (ONIG_IS_NULL(p)) return (val) 77 78#define enclen(enc,p,e) ((enc->max_enc_len == enc->min_enc_len) ? enc->min_enc_len : ONIGENC_MBC_ENC_LEN(enc,p,e)) 79 80/* character types bit flag */ 81#define BIT_CTYPE_NEWLINE (1<< ONIGENC_CTYPE_NEWLINE) 82#define BIT_CTYPE_ALPHA (1<< ONIGENC_CTYPE_ALPHA) 83#define BIT_CTYPE_BLANK (1<< ONIGENC_CTYPE_BLANK) 84#define BIT_CTYPE_CNTRL (1<< ONIGENC_CTYPE_CNTRL) 85#define BIT_CTYPE_DIGIT (1<< ONIGENC_CTYPE_DIGIT) 86#define BIT_CTYPE_GRAPH (1<< ONIGENC_CTYPE_GRAPH) 87#define BIT_CTYPE_LOWER (1<< ONIGENC_CTYPE_LOWER) 88#define BIT_CTYPE_PRINT (1<< ONIGENC_CTYPE_PRINT) 89#define BIT_CTYPE_PUNCT (1<< ONIGENC_CTYPE_PUNCT) 90#define BIT_CTYPE_SPACE (1<< ONIGENC_CTYPE_SPACE) 91#define BIT_CTYPE_UPPER (1<< ONIGENC_CTYPE_UPPER) 92#define BIT_CTYPE_XDIGIT (1<< ONIGENC_CTYPE_XDIGIT) 93#define BIT_CTYPE_WORD (1<< ONIGENC_CTYPE_WORD) 94#define BIT_CTYPE_ALNUM (1<< ONIGENC_CTYPE_ALNUM) 95#define BIT_CTYPE_ASCII (1<< ONIGENC_CTYPE_ASCII) 96 97#define CTYPE_TO_BIT(ctype) (1<<(ctype)) 98#define CTYPE_IS_WORD_GRAPH_PRINT(ctype) \ 99 ((ctype) == ONIGENC_CTYPE_WORD || (ctype) == ONIGENC_CTYPE_GRAPH ||\ 100 (ctype) == ONIGENC_CTYPE_PRINT) 101 102 103typedef struct { 104 const UChar *name; 105 int ctype; 106 short int len; 107} PosixBracketEntryType; 108 109#define PosixBracketEntryInit(name, ctype) {(const UChar *)name, ctype, (short int)(sizeof(name) - 1)} 110 111#define USE_CRNL_AS_LINE_TERMINATOR 112#define USE_UNICODE_PROPERTIES 113/* #define USE_UNICODE_CASE_FOLD_TURKISH_AZERI */ 114/* #define USE_UNICODE_ALL_LINE_TERMINATORS */ /* see Unicode.org UTS #18 */ 115 116 117#define ONIG_ENCODING_INIT_DEFAULT ONIG_ENCODING_ASCII 118 119/* for encoding system implementation (internal) */ 120ONIG_EXTERN int onigenc_ascii_apply_all_case_fold P_((OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg, OnigEncoding enc)); 121ONIG_EXTERN int onigenc_ascii_get_case_fold_codes_by_str P_((OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[], OnigEncoding enc)); 122ONIG_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)); 123ONIG_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[])); 124ONIG_EXTERN int onigenc_not_support_get_ctype_code_range P_((OnigCtype ctype, OnigCodePoint* sb_out, const OnigCodePoint* ranges[], OnigEncoding enc)); 125ONIG_EXTERN int onigenc_is_mbc_newline_0x0a P_((const UChar* p, const UChar* end, OnigEncoding enc)); 126 127 128/* methods for single byte encoding */ 129ONIG_EXTERN int onigenc_ascii_mbc_case_fold P_((OnigCaseFoldType flag, const UChar** p, const UChar* end, UChar* lower, OnigEncoding enc)); 130ONIG_EXTERN int onigenc_single_byte_mbc_enc_len P_((const UChar* p, const UChar* e, OnigEncoding enc)); 131ONIG_EXTERN OnigCodePoint onigenc_single_byte_mbc_to_code P_((const UChar* p, const UChar* end, OnigEncoding enc)); 132ONIG_EXTERN int onigenc_single_byte_code_to_mbclen P_((OnigCodePoint code, OnigEncoding enc)); 133ONIG_EXTERN int onigenc_single_byte_code_to_mbc P_((OnigCodePoint code, UChar *buf, OnigEncoding enc)); 134ONIG_EXTERN UChar* onigenc_single_byte_left_adjust_char_head P_((const UChar* start, const UChar* s, const OnigUChar* end, OnigEncoding enc)); 135ONIG_EXTERN int onigenc_always_true_is_allowed_reverse_match P_((const UChar* s, const UChar* end, OnigEncoding enc)); 136ONIG_EXTERN int onigenc_always_false_is_allowed_reverse_match P_((const UChar* s, const UChar* end, OnigEncoding enc)); 137ONIG_EXTERN int onigenc_ascii_is_code_ctype P_((OnigCodePoint code, unsigned int ctype, OnigEncoding enc)); 138 139/* methods for multi byte encoding */ 140ONIG_EXTERN OnigCodePoint onigenc_mbn_mbc_to_code P_((OnigEncoding enc, const UChar* p, const UChar* end)); 141ONIG_EXTERN int onigenc_mbn_mbc_case_fold P_((OnigEncoding enc, OnigCaseFoldType flag, const UChar** p, const UChar* end, UChar* lower)); 142ONIG_EXTERN int onigenc_mb2_code_to_mbclen P_((OnigCodePoint code, OnigEncoding enc)); 143ONIG_EXTERN int onigenc_mb2_code_to_mbc P_((OnigEncoding enc, OnigCodePoint code, UChar *buf)); 144ONIG_EXTERN int onigenc_minimum_property_name_to_ctype P_((OnigEncoding enc, UChar* p, UChar* end)); 145ONIG_EXTERN int onigenc_unicode_property_name_to_ctype P_((OnigEncoding enc, UChar* p, UChar* end)); 146ONIG_EXTERN int onigenc_mb2_is_code_ctype P_((OnigEncoding enc, OnigCodePoint code, unsigned int ctype)); 147ONIG_EXTERN int onigenc_mb4_code_to_mbclen P_((OnigCodePoint code, OnigEncoding enc)); 148ONIG_EXTERN int onigenc_mb4_code_to_mbc P_((OnigEncoding enc, OnigCodePoint code, UChar *buf)); 149ONIG_EXTERN int onigenc_mb4_is_code_ctype P_((OnigEncoding enc, OnigCodePoint code, unsigned int ctype)); 150 151 152/* in enc/unicode.c */ 153ONIG_EXTERN int onigenc_unicode_is_code_ctype P_((OnigCodePoint code, unsigned int ctype, OnigEncoding enc)); 154ONIG_EXTERN int onigenc_utf16_32_get_ctype_code_range P_((OnigCtype ctype, OnigCodePoint *sb_out, const OnigCodePoint* ranges[], OnigEncoding enc)); 155ONIG_EXTERN int onigenc_unicode_ctype_code_range P_((int ctype, const OnigCodePoint* ranges[])); 156ONIG_EXTERN int onigenc_unicode_get_case_fold_codes_by_str P_((OnigEncoding enc, OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[])); 157ONIG_EXTERN int onigenc_unicode_mbc_case_fold P_((OnigEncoding enc, OnigCaseFoldType flag, const UChar** pp, const UChar* end, UChar* fold)); 158ONIG_EXTERN int onigenc_unicode_apply_all_case_fold P_((OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg, OnigEncoding enc)); 159 160 161#define UTF16_IS_SURROGATE_FIRST(c) (((c) & 0xfc) == 0xd8) 162#define UTF16_IS_SURROGATE_SECOND(c) (((c) & 0xfc) == 0xdc) 163 164#define ONIGENC_ISO_8859_1_TO_LOWER_CASE(c) \ 165 OnigEncISO_8859_1_ToLowerCaseTable[c] 166#define ONIGENC_ISO_8859_1_TO_UPPER_CASE(c) \ 167 OnigEncISO_8859_1_ToUpperCaseTable[c] 168 169ONIG_EXTERN const UChar OnigEncISO_8859_1_ToLowerCaseTable[]; 170ONIG_EXTERN const UChar OnigEncISO_8859_1_ToUpperCaseTable[]; 171 172ONIG_EXTERN int 173onigenc_with_ascii_strncmp P_((OnigEncoding enc, const UChar* p, const UChar* end, const UChar* sascii /* ascii */, int n)); 174ONIG_EXTERN int 175onigenc_with_ascii_strnicmp P_((OnigEncoding enc, const UChar* p, const UChar* end, const UChar* sascii /* ascii */, int n)); 176ONIG_EXTERN UChar* 177onigenc_step P_((OnigEncoding enc, const UChar* p, const UChar* end, int n)); 178 179/* defined in regexec.c, but used in enc/xxx.c */ 180extern int onig_is_in_code_range P_((const UChar* p, OnigCodePoint code)); 181 182ONIG_EXTERN OnigEncoding OnigEncDefaultCharEncoding; 183ONIG_EXTERN const UChar OnigEncAsciiToLowerCaseTable[]; 184ONIG_EXTERN const UChar OnigEncAsciiToUpperCaseTable[]; 185ONIG_EXTERN const unsigned short OnigEncAsciiCtypeTable[]; 186 187#define ONIGENC_IS_ASCII_CODE(code) ((code) < 0x80) 188#define ONIGENC_ASCII_CODE_TO_LOWER_CASE(c) OnigEncAsciiToLowerCaseTable[c] 189#define ONIGENC_ASCII_CODE_TO_UPPER_CASE(c) OnigEncAsciiToUpperCaseTable[c] 190#define ONIGENC_IS_ASCII_CODE_CTYPE(code,ctype) \ 191 ((OnigEncAsciiCtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0) 192#define ONIGENC_IS_ASCII_CODE_CASE_AMBIG(code) \ 193 (ONIGENC_IS_ASCII_CODE_CTYPE(code, ONIGENC_CTYPE_UPPER) ||\ 194 ONIGENC_IS_ASCII_CODE_CTYPE(code, ONIGENC_CTYPE_LOWER)) 195 196/* Check if the code is in the range. (from <= code && code <= to) */ 197#define ONIGENC_IS_IN_RANGE(code, from, to) \ 198 ((OnigCodePoint )((code) - (from)) <= (OnigCodePoint )((to) - (from))) 199 200 201#ifdef ONIG_ENC_REGISTER 202extern int ONIG_ENC_REGISTER(const char *, OnigEncodingType*); 203#define OnigEncodingName(n) encoding_##n 204#define OnigEncodingDeclare(n) static OnigEncodingType OnigEncodingName(n) 205#define OnigEncodingDefine(f,n) \ 206 OnigEncodingDeclare(n); \ 207 void Init_##f(void) { \ 208 ONIG_ENC_REGISTER(OnigEncodingName(n).name, \ 209 &OnigEncodingName(n)); \ 210 } \ 211 OnigEncodingDeclare(n) 212#else 213#define OnigEncodingName(n) OnigEncoding##n 214#define OnigEncodingDeclare(n) OnigEncodingType OnigEncodingName(n) 215#define OnigEncodingDefine(f,n) OnigEncodingDeclare(n) 216#endif 217 218/* macros for define replica encoding and encoding alias */ 219#define ENC_REPLICATE(name, orig) 220#define ENC_ALIAS(name, orig) 221#define ENC_DUMMY(name) 222 223#if defined __GNUC__ && __GNUC__ >= 4 224#pragma GCC visibility pop 225#endif 226 227#endif /* ONIGURUMA_REGENC_H */ 228