Ruby 3.5.0dev (2025-05-17 revision aa0f689bf45352c4a592e7f1a044912c40435266)
regenc.h (aa0f689bf45352c4a592e7f1a044912c40435266)
1#ifndef ONIGMO_REGENC_H
2#define ONIGMO_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) 2011-2016 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
33#if !defined(RUBY) && (defined(RUBY_EXPORT) || defined(ONIG_ENC_REGISTER))
34# define RUBY
35#endif
36#ifdef RUBY
37# ifndef ONIGMO_REGINT_H
38# ifndef RUBY_EXTERN
39# include "ruby/config.h"
40# include "ruby/defines.h"
41# endif
42# endif
43# include "ruby.h"
44#else /* RUBY */
45# ifndef PACKAGE
46/* PACKAGE is defined in config.h */
47# include "config.h"
48# endif
49#endif /* RUBY */
50
51#ifdef ONIG_ESCAPE_UCHAR_COLLISION
52# undef ONIG_ESCAPE_UCHAR_COLLISION
53#endif
54
55#ifdef RUBY
56# include "ruby/onigmo.h"
57#else
58# include "onigmo.h"
59#endif
60
61RUBY_SYMBOL_EXPORT_BEGIN
62
63typedef struct {
64 OnigCodePoint from;
65 OnigCodePoint to;
66} OnigPairCaseFoldCodes;
67
68
69#ifndef NULL
70# define NULL ((void* )0)
71#endif
72
73#ifndef TRUE
74# define TRUE 1
75#endif
76
77#ifndef FALSE
78# define FALSE 0
79#endif
80
81#ifndef ARG_UNUSED
82# if defined(__GNUC__)
83# define ARG_UNUSED __attribute__ ((unused))
84# else
85# define ARG_UNUSED
86# endif
87#endif
88
89#define ONIG_IS_NULL(p) (((void*)(p)) == (void*)0)
90#define ONIG_IS_NOT_NULL(p) (((void*)(p)) != (void*)0)
91#define ONIG_CHECK_NULL_RETURN(p) if (ONIG_IS_NULL(p)) return NULL
92#define ONIG_CHECK_NULL_RETURN_VAL(p,val) if (ONIG_IS_NULL(p)) return (val)
93
94#define enclen(enc,p,e) ((enc->max_enc_len == enc->min_enc_len) ? (p < e ? enc->min_enc_len : 0) : ONIGENC_MBC_ENC_LEN(enc,p,e))
95
96/* character types bit flag */
97#define BIT_CTYPE_NEWLINE (1<< ONIGENC_CTYPE_NEWLINE)
98#define BIT_CTYPE_ALPHA (1<< ONIGENC_CTYPE_ALPHA)
99#define BIT_CTYPE_BLANK (1<< ONIGENC_CTYPE_BLANK)
100#define BIT_CTYPE_CNTRL (1<< ONIGENC_CTYPE_CNTRL)
101#define BIT_CTYPE_DIGIT (1<< ONIGENC_CTYPE_DIGIT)
102#define BIT_CTYPE_GRAPH (1<< ONIGENC_CTYPE_GRAPH)
103#define BIT_CTYPE_LOWER (1<< ONIGENC_CTYPE_LOWER)
104#define BIT_CTYPE_PRINT (1<< ONIGENC_CTYPE_PRINT)
105#define BIT_CTYPE_PUNCT (1<< ONIGENC_CTYPE_PUNCT)
106#define BIT_CTYPE_SPACE (1<< ONIGENC_CTYPE_SPACE)
107#define BIT_CTYPE_UPPER (1<< ONIGENC_CTYPE_UPPER)
108#define BIT_CTYPE_XDIGIT (1<< ONIGENC_CTYPE_XDIGIT)
109#define BIT_CTYPE_WORD (1<< ONIGENC_CTYPE_WORD)
110#define BIT_CTYPE_ALNUM (1<< ONIGENC_CTYPE_ALNUM)
111#define BIT_CTYPE_ASCII (1<< ONIGENC_CTYPE_ASCII)
112
113#define CTYPE_TO_BIT(ctype) (1<<(ctype))
114#define CTYPE_IS_WORD_GRAPH_PRINT(ctype) \
115 ((ctype) == ONIGENC_CTYPE_WORD || (ctype) == ONIGENC_CTYPE_GRAPH ||\
116 (ctype) == ONIGENC_CTYPE_PRINT)
117
118
119typedef struct {
120 short int len;
121#if defined(__has_attribute)
122# if __has_attribute(nonstring)
123 __attribute__((nonstring))
124# endif
125#endif
126 const UChar name[6];
127 int ctype;
128} PosixBracketEntryType;
129
130#define POSIX_BRACKET_ENTRY_INIT(name, ctype) \
131 {(short int )(sizeof(name) - 1), name, (ctype)}
132
133#define numberof(array) ((int)(sizeof(array) / sizeof((array)[0])))
134#define roomof(x, y) (((x) + (y) - 1) / (y))
135#define type_roomof(x, y) roomof(sizeof(x), sizeof(y))
136
137#define USE_CRNL_AS_LINE_TERMINATOR
138#define USE_UNICODE_PROPERTIES
139#define USE_UNICODE_AGE_PROPERTIES
140/* #define USE_UNICODE_CASE_FOLD_TURKISH_AZERI */
141/* #define USE_UNICODE_ALL_LINE_TERMINATORS */ /* see Unicode.org UTS #18 */
142
143
144#define ONIG_ENCODING_INIT_DEFAULT ONIG_ENCODING_ASCII
145
146/* for encoding system implementation (internal) */
147ONIG_EXTERN int onigenc_ascii_apply_all_case_fold(OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg, OnigEncoding enc);
148ONIG_EXTERN int onigenc_ascii_get_case_fold_codes_by_str(OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[], OnigEncoding enc);
149ONIG_EXTERN int onigenc_apply_all_case_fold_with_map(int map_size, const OnigPairCaseFoldCodes map[], int ess_tsett_flag, OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg);
150ONIG_EXTERN int onigenc_get_case_fold_codes_by_str_with_map(int map_size, const OnigPairCaseFoldCodes map[], int ess_tsett_flag, OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[]);
151ONIG_EXTERN int onigenc_not_support_get_ctype_code_range(OnigCtype ctype, OnigCodePoint* sb_out, const OnigCodePoint* ranges[], OnigEncoding enc);
152ONIG_EXTERN int onigenc_is_mbc_newline_0x0a(const UChar* p, const UChar* end, OnigEncoding enc);
153ONIG_EXTERN int onigenc_single_byte_ascii_only_case_map(OnigCaseFoldType* flagP, const OnigUChar** pp, const OnigUChar* end, OnigUChar* to, OnigUChar* to_end, const struct OnigEncodingTypeST* enc);
154
155
156/* methods for single byte encoding */
157ONIG_EXTERN int onigenc_ascii_mbc_case_fold(OnigCaseFoldType flag, const UChar** p, const UChar* end, UChar* lower, OnigEncoding enc);
158ONIG_EXTERN int onigenc_single_byte_mbc_enc_len(const UChar* p, const UChar* e, OnigEncoding enc);
159ONIG_EXTERN OnigCodePoint onigenc_single_byte_mbc_to_code(const UChar* p, const UChar* end, OnigEncoding enc);
160ONIG_EXTERN int onigenc_single_byte_code_to_mbclen(OnigCodePoint code, OnigEncoding enc);
161ONIG_EXTERN int onigenc_single_byte_code_to_mbc(OnigCodePoint code, UChar *buf, OnigEncoding enc);
162ONIG_EXTERN UChar* onigenc_single_byte_left_adjust_char_head(const UChar* start, const UChar* s, const OnigUChar* end, OnigEncoding enc);
163ONIG_EXTERN int onigenc_always_true_is_allowed_reverse_match(const UChar* s, const UChar* end, OnigEncoding enc);
164ONIG_EXTERN int onigenc_always_false_is_allowed_reverse_match(const UChar* s, const UChar* end, OnigEncoding enc);
165ONIG_EXTERN int onigenc_ascii_is_code_ctype(OnigCodePoint code, unsigned int ctype, OnigEncoding enc);
166
167/* methods for multi byte encoding */
168ONIG_EXTERN OnigCodePoint onigenc_mbn_mbc_to_code(OnigEncoding enc, const UChar* p, const UChar* end);
169ONIG_EXTERN int onigenc_mbn_mbc_case_fold(OnigEncoding enc, OnigCaseFoldType flag, const UChar** p, const UChar* end, UChar* lower);
170ONIG_EXTERN int onigenc_mb2_code_to_mbclen(OnigCodePoint code, OnigEncoding enc);
171ONIG_EXTERN int onigenc_mb2_code_to_mbc(OnigEncoding enc, OnigCodePoint code, UChar *buf);
172ONIG_EXTERN int onigenc_minimum_property_name_to_ctype(OnigEncoding enc, const UChar* p, const UChar* end);
173ONIG_EXTERN int onigenc_unicode_property_name_to_ctype(OnigEncoding enc, const UChar* p, const UChar* end);
174ONIG_EXTERN int onigenc_mb2_is_code_ctype(OnigEncoding enc, OnigCodePoint code, unsigned int ctype);
175ONIG_EXTERN int onigenc_mb4_code_to_mbclen(OnigCodePoint code, OnigEncoding enc);
176ONIG_EXTERN int onigenc_mb4_code_to_mbc(OnigEncoding enc, OnigCodePoint code, UChar *buf);
177ONIG_EXTERN int onigenc_mb4_is_code_ctype(OnigEncoding enc, OnigCodePoint code, unsigned int ctype);
178
179ONIG_EXTERN int onigenc_unicode_case_map(OnigCaseFoldType* flagP, const OnigUChar** pp, const OnigUChar* end, OnigUChar* to, OnigUChar* to_end, const struct OnigEncodingTypeST* enc);
180
181
182/* in enc/unicode.c */
183ONIG_EXTERN int onigenc_unicode_is_code_ctype(OnigCodePoint code, unsigned int ctype, OnigEncoding enc);
184ONIG_EXTERN int onigenc_utf16_32_get_ctype_code_range(OnigCtype ctype, OnigCodePoint *sb_out, const OnigCodePoint* ranges[], OnigEncoding enc);
185ONIG_EXTERN int onigenc_unicode_ctype_code_range(int ctype, const OnigCodePoint* ranges[]);
186ONIG_EXTERN int onigenc_unicode_get_case_fold_codes_by_str(OnigEncoding enc, OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[]);
187ONIG_EXTERN int onigenc_unicode_mbc_case_fold(OnigEncoding enc, OnigCaseFoldType flag, const UChar** pp, const UChar* end, UChar* fold);
188ONIG_EXTERN int onigenc_unicode_apply_all_case_fold(OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg, OnigEncoding enc);
189
190
191#define UTF16_IS_SURROGATE_FIRST(c) (((c) & 0xfc) == 0xd8)
192#define UTF16_IS_SURROGATE_SECOND(c) (((c) & 0xfc) == 0xdc)
193#define UTF16_IS_SURROGATE(c) (((c) & 0xf8) == 0xd8)
194#define UNICODE_VALID_CODEPOINT_P(c) ( \
195 ((c) <= 0x10ffff) && \
196 !((c) < 0x10000 && UTF16_IS_SURROGATE((c) >> 8)))
197
198#define ONIGENC_ISO_8859_1_TO_LOWER_CASE(c) \
199 OnigEncISO_8859_1_ToLowerCaseTable[c]
200#define ONIGENC_ISO_8859_1_TO_UPPER_CASE(c) \
201 OnigEncISO_8859_1_ToUpperCaseTable[c]
202
203ONIG_EXTERN const UChar OnigEncISO_8859_1_ToLowerCaseTable[];
204ONIG_EXTERN const UChar OnigEncISO_8859_1_ToUpperCaseTable[];
205
206ONIG_EXTERN int
207onigenc_with_ascii_strncmp(OnigEncoding enc, const UChar* p, const UChar* end, const UChar* sascii /* ascii */, int n);
208ONIG_EXTERN int
209onigenc_with_ascii_strnicmp(OnigEncoding enc, const UChar* p, const UChar* end, const UChar* sascii /* ascii */, int n);
210ONIG_EXTERN UChar*
211onigenc_step(OnigEncoding enc, const UChar* p, const UChar* end, int n);
212
213/* defined in regexec.c, but used in enc/xxx.c */
214extern int onig_is_in_code_range(const UChar* p, OnigCodePoint code);
215
216ONIG_EXTERN OnigEncoding OnigEncDefaultCharEncoding;
217ONIG_EXTERN const UChar OnigEncAsciiToLowerCaseTable[];
218ONIG_EXTERN const UChar OnigEncAsciiToUpperCaseTable[];
219ONIG_EXTERN const unsigned short OnigEncAsciiCtypeTable[];
220
221#define ONIGENC_IS_ASCII_CODE(code) ((code) < 0x80)
222#define ONIGENC_ASCII_CODE_TO_LOWER_CASE(c) OnigEncAsciiToLowerCaseTable[c]
223#define ONIGENC_ASCII_CODE_TO_UPPER_CASE(c) OnigEncAsciiToUpperCaseTable[c]
224#define ONIGENC_IS_ASCII_CODE_CTYPE(code,ctype) \
225 ((OnigEncAsciiCtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
226#define ONIGENC_IS_ASCII_CODE_CASE_AMBIG(code) \
227 (ONIGENC_IS_ASCII_CODE_CTYPE(code, ONIGENC_CTYPE_UPPER) ||\
228 ONIGENC_IS_ASCII_CODE_CTYPE(code, ONIGENC_CTYPE_LOWER))
229
230/* Check if the code is in the range. (from <= code && code <= to) */
231#define ONIGENC_IS_IN_RANGE(code, from, to) \
232 ((OnigCodePoint )((code) - (from)) <= (OnigCodePoint )((to) - (from)))
233
234
235#ifdef ONIG_ENC_REGISTER
236extern int ONIG_ENC_REGISTER(const char *, OnigEncoding);
237# define OnigEncodingName(n) encoding_##n
238# define OnigEncodingDeclare(n) static const OnigEncodingType OnigEncodingName(n)
239# define OnigEncodingDefine(f,n) \
240 OnigEncodingDeclare(n); \
241 void Init_##f(void) { \
242 ONIG_ENC_REGISTER(OnigEncodingName(n).name, \
243 &OnigEncodingName(n)); \
244 } \
245 OnigEncodingDeclare(n)
246#else
247# define OnigEncodingName(n) OnigEncoding##n
248# define OnigEncodingDeclare(n) const OnigEncodingType OnigEncodingName(n)
249# define OnigEncodingDefine(f,n) OnigEncodingDeclare(n)
250#endif
251
252/* macros for define replica encoding and encoding alias */
253#define ENC_REPLICATE(name, orig)
254#define ENC_ALIAS(name, orig)
255#define ENC_DUMMY(name)
256
257RUBY_SYMBOL_EXPORT_END
258
259#endif /* ONIGMO_REGENC_H */
len
int len
Length of the buffer.
Definition io.h:8
PosixBracketEntryType
Definition regenc.h:119
__attribute__
Definition vm_insnhelper.c:2288
Generated by doxygen 1.9.8