1 | #ifndef REGENC_H
|
---|
2 | #define 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 |
|
---|
33 | #ifndef PACKAGE
|
---|
34 | /* PACKAGE is defined in config.h */
|
---|
35 | #include "config.h"
|
---|
36 | #endif
|
---|
37 |
|
---|
38 | #ifdef ONIG_ESCAPE_UCHAR_COLLISION
|
---|
39 | #undef ONIG_ESCAPE_UCHAR_COLLISION
|
---|
40 | #endif
|
---|
41 |
|
---|
42 | #include "oniguruma.h"
|
---|
43 |
|
---|
44 | typedef struct {
|
---|
45 | OnigCodePoint from;
|
---|
46 | OnigCodePoint to;
|
---|
47 | } OnigPairCaseFoldCodes;
|
---|
48 |
|
---|
49 |
|
---|
50 | #ifndef NULL
|
---|
51 | #define NULL ((void* )0)
|
---|
52 | #endif
|
---|
53 |
|
---|
54 | #ifndef TRUE
|
---|
55 | #define TRUE 1
|
---|
56 | #endif
|
---|
57 |
|
---|
58 | #ifndef FALSE
|
---|
59 | #define FALSE 0
|
---|
60 | #endif
|
---|
61 |
|
---|
62 | #ifndef ARG_UNUSED
|
---|
63 | #if defined(__GNUC__)
|
---|
64 | # define ARG_UNUSED __attribute__ ((unused))
|
---|
65 | #else
|
---|
66 | # define ARG_UNUSED
|
---|
67 | #endif
|
---|
68 | #endif
|
---|
69 |
|
---|
70 | #define ONIG_IS_NULL(p) (((void*)(p)) == (void*)0)
|
---|
71 | #define ONIG_IS_NOT_NULL(p) (((void*)(p)) != (void*)0)
|
---|
72 | #define ONIG_CHECK_NULL_RETURN(p) if (ONIG_IS_NULL(p)) return NULL
|
---|
73 | #define ONIG_CHECK_NULL_RETURN_VAL(p,val) if (ONIG_IS_NULL(p)) return (val)
|
---|
74 |
|
---|
75 | #define enclen(enc,p) ONIGENC_MBC_ENC_LEN(enc,p)
|
---|
76 |
|
---|
77 | /* character types bit flag */
|
---|
78 | #define BIT_CTYPE_NEWLINE (1<< ONIGENC_CTYPE_NEWLINE)
|
---|
79 | #define BIT_CTYPE_ALPHA (1<< ONIGENC_CTYPE_ALPHA)
|
---|
80 | #define BIT_CTYPE_BLANK (1<< ONIGENC_CTYPE_BLANK)
|
---|
81 | #define BIT_CTYPE_CNTRL (1<< ONIGENC_CTYPE_CNTRL)
|
---|
82 | #define BIT_CTYPE_DIGIT (1<< ONIGENC_CTYPE_DIGIT)
|
---|
83 | #define BIT_CTYPE_GRAPH (1<< ONIGENC_CTYPE_GRAPH)
|
---|
84 | #define BIT_CTYPE_LOWER (1<< ONIGENC_CTYPE_LOWER)
|
---|
85 | #define BIT_CTYPE_PRINT (1<< ONIGENC_CTYPE_PRINT)
|
---|
86 | #define BIT_CTYPE_PUNCT (1<< ONIGENC_CTYPE_PUNCT)
|
---|
87 | #define BIT_CTYPE_SPACE (1<< ONIGENC_CTYPE_SPACE)
|
---|
88 | #define BIT_CTYPE_UPPER (1<< ONIGENC_CTYPE_UPPER)
|
---|
89 | #define BIT_CTYPE_XDIGIT (1<< ONIGENC_CTYPE_XDIGIT)
|
---|
90 | #define BIT_CTYPE_WORD (1<< ONIGENC_CTYPE_WORD)
|
---|
91 | #define BIT_CTYPE_ALNUM (1<< ONIGENC_CTYPE_ALNUM)
|
---|
92 | #define BIT_CTYPE_ASCII (1<< ONIGENC_CTYPE_ASCII)
|
---|
93 |
|
---|
94 | #define CTYPE_TO_BIT(ctype) (1<<(ctype))
|
---|
95 | #define CTYPE_IS_WORD_GRAPH_PRINT(ctype) \
|
---|
96 | ((ctype) == ONIGENC_CTYPE_WORD || (ctype) == ONIGENC_CTYPE_GRAPH ||\
|
---|
97 | (ctype) == ONIGENC_CTYPE_PRINT)
|
---|
98 |
|
---|
99 |
|
---|
100 | typedef struct {
|
---|
101 | const UChar *name;
|
---|
102 | int ctype;
|
---|
103 | short int len;
|
---|
104 | } PosixBracketEntryType;
|
---|
105 |
|
---|
106 | #define POSIX_BRACKET_ENTRY_INIT(name, ctype) \
|
---|
107 | {(const UChar* )(name), (ctype), (short int )(sizeof(name) - 1)}
|
---|
108 |
|
---|
109 | #ifndef numberof
|
---|
110 | #define numberof(array) (int )(sizeof(array) / sizeof((array)[0]))
|
---|
111 | #endif
|
---|
112 |
|
---|
113 |
|
---|
114 | #define USE_CRNL_AS_LINE_TERMINATOR
|
---|
115 | #define USE_UNICODE_PROPERTIES
|
---|
116 | #define USE_UNICODE_AGE_PROPERTIES
|
---|
117 | /* #define USE_UNICODE_CASE_FOLD_TURKISH_AZERI */
|
---|
118 | /* #define USE_UNICODE_ALL_LINE_TERMINATORS */ /* see Unicode.org UTS #18 */
|
---|
119 |
|
---|
120 |
|
---|
121 | #define ONIG_ENCODING_INIT_DEFAULT ONIG_ENCODING_UTF8
|
---|
122 |
|
---|
123 | /* for encoding system implementation (internal) */
|
---|
124 | ONIG_EXTERN int onigenc_ascii_apply_all_case_fold P_((OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg));
|
---|
125 | ONIG_EXTERN int onigenc_ascii_get_case_fold_codes_by_str P_((OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[]));
|
---|
126 | 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));
|
---|
127 | 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[]));
|
---|
128 | ONIG_EXTERN int onigenc_not_support_get_ctype_code_range P_((OnigCtype ctype, OnigCodePoint* sb_out, const OnigCodePoint* ranges[]));
|
---|
129 | ONIG_EXTERN int onigenc_is_mbc_newline_0x0a P_((const UChar* p, const UChar* end));
|
---|
130 |
|
---|
131 |
|
---|
132 | /* methods for single byte encoding */
|
---|
133 | ONIG_EXTERN int onigenc_ascii_mbc_case_fold P_((OnigCaseFoldType flag, const UChar** p, const UChar* end, UChar* lower));
|
---|
134 | ONIG_EXTERN int onigenc_single_byte_mbc_enc_len P_((const UChar* p));
|
---|
135 | ONIG_EXTERN OnigCodePoint onigenc_single_byte_mbc_to_code P_((const UChar* p, const UChar* end));
|
---|
136 | ONIG_EXTERN int onigenc_single_byte_code_to_mbclen P_((OnigCodePoint code));
|
---|
137 | ONIG_EXTERN int onigenc_single_byte_code_to_mbc P_((OnigCodePoint code, UChar *buf));
|
---|
138 | ONIG_EXTERN UChar* onigenc_single_byte_left_adjust_char_head P_((const UChar* start, const UChar* s));
|
---|
139 | ONIG_EXTERN int onigenc_always_true_is_allowed_reverse_match P_((const UChar* s, const UChar* end));
|
---|
140 | ONIG_EXTERN int onigenc_always_false_is_allowed_reverse_match P_((const UChar* s, const UChar* end));
|
---|
141 | ONIG_EXTERN int onigenc_ascii_is_code_ctype P_((OnigCodePoint code, unsigned int ctype));
|
---|
142 |
|
---|
143 | /* methods for multi byte encoding */
|
---|
144 | ONIG_EXTERN OnigCodePoint onigenc_mbn_mbc_to_code P_((OnigEncoding enc, const UChar* p, const UChar* end));
|
---|
145 | ONIG_EXTERN int onigenc_mbn_mbc_case_fold P_((OnigEncoding enc, OnigCaseFoldType flag, const UChar** p, const UChar* end, UChar* lower));
|
---|
146 | ONIG_EXTERN int onigenc_mb2_code_to_mbclen P_((OnigCodePoint code));
|
---|
147 | ONIG_EXTERN int onigenc_mb2_code_to_mbc P_((OnigEncoding enc, OnigCodePoint code, UChar *buf));
|
---|
148 | ONIG_EXTERN int onigenc_minimum_property_name_to_ctype P_((OnigEncoding enc, UChar* p, UChar* end));
|
---|
149 | ONIG_EXTERN int onigenc_unicode_property_name_to_ctype P_((OnigEncoding enc, UChar* p, UChar* end));
|
---|
150 | ONIG_EXTERN int onigenc_mb2_is_code_ctype P_((OnigEncoding enc, OnigCodePoint code, unsigned int ctype));
|
---|
151 | ONIG_EXTERN int onigenc_mb4_code_to_mbclen P_((OnigCodePoint code));
|
---|
152 | ONIG_EXTERN int onigenc_mb4_code_to_mbc P_((OnigEncoding enc, OnigCodePoint code, UChar *buf));
|
---|
153 | ONIG_EXTERN int onigenc_mb4_is_code_ctype P_((OnigEncoding enc, OnigCodePoint code, unsigned int ctype));
|
---|
154 |
|
---|
155 |
|
---|
156 | /* in enc/unicode.c */
|
---|
157 | ONIG_EXTERN int onigenc_unicode_is_code_ctype P_((OnigCodePoint code, unsigned int ctype));
|
---|
158 | ONIG_EXTERN int onigenc_utf16_32_get_ctype_code_range P_((OnigCtype ctype, OnigCodePoint *sb_out, const OnigCodePoint* ranges[]));
|
---|
159 | ONIG_EXTERN int onigenc_unicode_ctype_code_range P_((int ctype, const OnigCodePoint* ranges[]));
|
---|
160 | ONIG_EXTERN int onigenc_unicode_get_case_fold_codes_by_str P_((OnigEncoding enc, OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[]));
|
---|
161 | ONIG_EXTERN int onigenc_unicode_mbc_case_fold P_((OnigEncoding enc, OnigCaseFoldType flag, const UChar** pp, const UChar* end, UChar* fold));
|
---|
162 | ONIG_EXTERN int onigenc_unicode_apply_all_case_fold P_((OnigCaseFoldType flag, OnigApplyAllCaseFoldFunc f, void* arg));
|
---|
163 |
|
---|
164 |
|
---|
165 | #define UTF16_IS_SURROGATE_FIRST(c) (((c) & 0xfc) == 0xd8)
|
---|
166 | #define UTF16_IS_SURROGATE_SECOND(c) (((c) & 0xfc) == 0xdc)
|
---|
167 |
|
---|
168 | #define ONIGENC_ISO_8859_1_TO_LOWER_CASE(c) \
|
---|
169 | OnigEncISO_8859_1_ToLowerCaseTable[c]
|
---|
170 | #define ONIGENC_ISO_8859_1_TO_UPPER_CASE(c) \
|
---|
171 | OnigEncISO_8859_1_ToUpperCaseTable[c]
|
---|
172 |
|
---|
173 | ONIG_EXTERN const UChar OnigEncISO_8859_1_ToLowerCaseTable[];
|
---|
174 | ONIG_EXTERN const UChar OnigEncISO_8859_1_ToUpperCaseTable[];
|
---|
175 |
|
---|
176 | ONIG_EXTERN int
|
---|
177 | onigenc_with_ascii_strncmp P_((OnigEncoding enc, const UChar* p, const UChar* end, const UChar* sascii /* ascii */, int n));
|
---|
178 | ONIG_EXTERN int
|
---|
179 | onigenc_with_ascii_strnicmp P_((OnigEncoding enc, const UChar* p, const UChar* end, const UChar* sascii /* ascii */, int n));
|
---|
180 | ONIG_EXTERN UChar*
|
---|
181 | onigenc_step P_((OnigEncoding enc, const UChar* p, const UChar* end, int n));
|
---|
182 |
|
---|
183 | /* defined in regexec.c, but used in enc/xxx.c */
|
---|
184 | extern int onig_is_in_code_range P_((const UChar* p, OnigCodePoint code));
|
---|
185 |
|
---|
186 | ONIG_EXTERN OnigEncoding OnigEncDefaultCharEncoding;
|
---|
187 | ONIG_EXTERN const UChar OnigEncAsciiToLowerCaseTable[];
|
---|
188 | ONIG_EXTERN const UChar OnigEncAsciiToUpperCaseTable[];
|
---|
189 | ONIG_EXTERN const unsigned short OnigEncAsciiCtypeTable[];
|
---|
190 |
|
---|
191 | #define ONIGENC_IS_ASCII_CODE(code) ((code) < 0x80)
|
---|
192 | #define ONIGENC_ASCII_CODE_TO_LOWER_CASE(c) OnigEncAsciiToLowerCaseTable[c]
|
---|
193 | #define ONIGENC_ASCII_CODE_TO_UPPER_CASE(c) OnigEncAsciiToUpperCaseTable[c]
|
---|
194 | #define ONIGENC_IS_ASCII_CODE_CTYPE(code,ctype) \
|
---|
195 | ((OnigEncAsciiCtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
|
---|
196 | #define ONIGENC_IS_ASCII_CODE_CASE_AMBIG(code) \
|
---|
197 | (ONIGENC_IS_ASCII_CODE_CTYPE(code, ONIGENC_CTYPE_UPPER) ||\
|
---|
198 | ONIGENC_IS_ASCII_CODE_CTYPE(code, ONIGENC_CTYPE_LOWER))
|
---|
199 |
|
---|
200 | /* Check if the code is in the range. (from <= code && code <= to) */
|
---|
201 | #define ONIGENC_IS_IN_RANGE(code, from, to) \
|
---|
202 | ((OnigCodePoint )((code) - (from)) <= (OnigCodePoint )((to) - (from)))
|
---|
203 |
|
---|
204 |
|
---|
205 | #endif /* REGENC_H */
|
---|