1 /**********************************************************************
2 regerror.c - Oniguruma (regular expression library)
3 **********************************************************************/
5 * Copyright (c) 2002-2018 K.Kosako <sndgk393 AT ybb DOT ne DOT jp>
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 #include <stdio.h> /* for vsnprintf() */
34 #ifdef HAVE_STDARG_PROTOTYPES
36 #define va_init_list(a,b) va_start(a,b)
39 #define va_init_list(a,b) va_start(a)
44 onig_error_code_to_format(int code
)
50 p
= "mismatch"; break;
51 case ONIG_NO_SUPPORT_CONFIG
:
52 p
= "no support in this configuration"; break;
56 p
= "fail to memory allocation"; break;
57 case ONIGERR_MATCH_STACK_LIMIT_OVER
:
58 p
= "match-stack limit over"; break;
59 case ONIGERR_PARSE_DEPTH_LIMIT_OVER
:
60 p
= "parse depth limit over"; break;
61 case ONIGERR_RETRY_LIMIT_IN_MATCH_OVER
:
62 p
= "retry-limit-in-match over"; break;
63 case ONIGERR_TYPE_BUG
:
64 p
= "undefined type (bug)"; break;
65 case ONIGERR_PARSER_BUG
:
66 p
= "internal parser error (bug)"; break;
67 case ONIGERR_STACK_BUG
:
68 p
= "stack error (bug)"; break;
69 case ONIGERR_UNDEFINED_BYTECODE
:
70 p
= "undefined bytecode (bug)"; break;
71 case ONIGERR_UNEXPECTED_BYTECODE
:
72 p
= "unexpected bytecode (bug)"; break;
73 case ONIGERR_DEFAULT_ENCODING_IS_NOT_SETTED
:
74 p
= "default multibyte-encoding is not setted"; break;
75 case ONIGERR_SPECIFIED_ENCODING_CANT_CONVERT_TO_WIDE_CHAR
:
76 p
= "can't convert to wide-char on specified multibyte-encoding"; break;
77 case ONIGERR_FAIL_TO_INITIALIZE
:
78 p
= "fail to initialize"; break;
79 case ONIGERR_INVALID_ARGUMENT
:
80 p
= "invalid argument"; break;
81 case ONIGERR_END_PATTERN_AT_LEFT_BRACE
:
82 p
= "end pattern at left brace"; break;
83 case ONIGERR_END_PATTERN_AT_LEFT_BRACKET
:
84 p
= "end pattern at left bracket"; break;
85 case ONIGERR_EMPTY_CHAR_CLASS
:
86 p
= "empty char-class"; break;
87 case ONIGERR_PREMATURE_END_OF_CHAR_CLASS
:
88 p
= "premature end of char-class"; break;
89 case ONIGERR_END_PATTERN_AT_ESCAPE
:
90 p
= "end pattern at escape"; break;
91 case ONIGERR_END_PATTERN_AT_META
:
92 p
= "end pattern at meta"; break;
93 case ONIGERR_END_PATTERN_AT_CONTROL
:
94 p
= "end pattern at control"; break;
95 case ONIGERR_META_CODE_SYNTAX
:
96 p
= "invalid meta-code syntax"; break;
97 case ONIGERR_CONTROL_CODE_SYNTAX
:
98 p
= "invalid control-code syntax"; break;
99 case ONIGERR_CHAR_CLASS_VALUE_AT_END_OF_RANGE
:
100 p
= "char-class value at end of range"; break;
101 case ONIGERR_CHAR_CLASS_VALUE_AT_START_OF_RANGE
:
102 p
= "char-class value at start of range"; break;
103 case ONIGERR_UNMATCHED_RANGE_SPECIFIER_IN_CHAR_CLASS
:
104 p
= "unmatched range specifier in char-class"; break;
105 case ONIGERR_TARGET_OF_REPEAT_OPERATOR_NOT_SPECIFIED
:
106 p
= "target of repeat operator is not specified"; break;
107 case ONIGERR_TARGET_OF_REPEAT_OPERATOR_INVALID
:
108 p
= "target of repeat operator is invalid"; break;
109 case ONIGERR_NESTED_REPEAT_OPERATOR
:
110 p
= "nested repeat operator"; break;
111 case ONIGERR_UNMATCHED_CLOSE_PARENTHESIS
:
112 p
= "unmatched close parenthesis"; break;
113 case ONIGERR_END_PATTERN_WITH_UNMATCHED_PARENTHESIS
:
114 p
= "end pattern with unmatched parenthesis"; break;
115 case ONIGERR_END_PATTERN_IN_GROUP
:
116 p
= "end pattern in group"; break;
117 case ONIGERR_UNDEFINED_GROUP_OPTION
:
118 p
= "undefined group option"; break;
119 case ONIGERR_INVALID_POSIX_BRACKET_TYPE
:
120 p
= "invalid POSIX bracket type"; break;
121 case ONIGERR_INVALID_LOOK_BEHIND_PATTERN
:
122 p
= "invalid pattern in look-behind"; break;
123 case ONIGERR_INVALID_REPEAT_RANGE_PATTERN
:
124 p
= "invalid repeat range {lower,upper}"; break;
125 case ONIGERR_TOO_BIG_NUMBER
:
126 p
= "too big number"; break;
127 case ONIGERR_TOO_BIG_NUMBER_FOR_REPEAT_RANGE
:
128 p
= "too big number for repeat range"; break;
129 case ONIGERR_UPPER_SMALLER_THAN_LOWER_IN_REPEAT_RANGE
:
130 p
= "upper is smaller than lower in repeat range"; break;
131 case ONIGERR_EMPTY_RANGE_IN_CHAR_CLASS
:
132 p
= "empty range in char class"; break;
133 case ONIGERR_MISMATCH_CODE_LENGTH_IN_CLASS_RANGE
:
134 p
= "mismatch multibyte code length in char-class range"; break;
135 case ONIGERR_TOO_MANY_MULTI_BYTE_RANGES
:
136 p
= "too many multibyte code ranges are specified"; break;
137 case ONIGERR_TOO_SHORT_MULTI_BYTE_STRING
:
138 p
= "too short multibyte code string"; break;
139 case ONIGERR_TOO_BIG_BACKREF_NUMBER
:
140 p
= "too big backref number"; break;
141 case ONIGERR_INVALID_BACKREF
:
142 p
= "invalid backref number/name"; break;
143 case ONIGERR_NUMBERED_BACKREF_OR_CALL_NOT_ALLOWED
:
144 p
= "numbered backref/call is not allowed. (use name)"; break;
145 case ONIGERR_TOO_MANY_CAPTURES
:
146 p
= "too many captures"; break;
147 case ONIGERR_TOO_BIG_WIDE_CHAR_VALUE
:
148 p
= "too big wide-char value"; break;
149 case ONIGERR_TOO_LONG_WIDE_CHAR_VALUE
:
150 p
= "too long wide-char value"; break;
151 case ONIGERR_INVALID_CODE_POINT_VALUE
:
152 p
= "invalid code point value"; break;
153 case ONIGERR_EMPTY_GROUP_NAME
:
154 p
= "group name is empty"; break;
155 case ONIGERR_INVALID_GROUP_NAME
:
156 p
= "invalid group name <%n>"; break;
157 case ONIGERR_INVALID_CHAR_IN_GROUP_NAME
:
158 p
= "invalid char in group name <%n>"; break;
159 case ONIGERR_UNDEFINED_NAME_REFERENCE
:
160 p
= "undefined name <%n> reference"; break;
161 case ONIGERR_UNDEFINED_GROUP_REFERENCE
:
162 p
= "undefined group <%n> reference"; break;
163 case ONIGERR_MULTIPLEX_DEFINED_NAME
:
164 p
= "multiplex defined name <%n>"; break;
165 case ONIGERR_MULTIPLEX_DEFINITION_NAME_CALL
:
166 p
= "multiplex definition name <%n> call"; break;
167 case ONIGERR_NEVER_ENDING_RECURSION
:
168 p
= "never ending recursion"; break;
169 case ONIGERR_GROUP_NUMBER_OVER_FOR_CAPTURE_HISTORY
:
170 p
= "group number is too big for capture history"; break;
171 case ONIGERR_INVALID_CHAR_PROPERTY_NAME
:
172 p
= "invalid character property name {%n}"; break;
173 case ONIGERR_INVALID_IF_ELSE_SYNTAX
:
174 p
= "invalid if-else syntax"; break;
175 case ONIGERR_INVALID_ABSENT_GROUP_PATTERN
:
176 p
= "invalid absent group pattern"; break;
177 case ONIGERR_INVALID_ABSENT_GROUP_GENERATOR_PATTERN
:
178 p
= "invalid absent group generator pattern"; break;
179 case ONIGERR_INVALID_CALLOUT_PATTERN
:
180 p
= "invalid callout pattern"; break;
181 case ONIGERR_INVALID_CALLOUT_NAME
:
182 p
= "invalid callout name"; break;
183 case ONIGERR_UNDEFINED_CALLOUT_NAME
:
184 p
= "undefined callout name"; break;
185 case ONIGERR_INVALID_CALLOUT_BODY
:
186 p
= "invalid callout body"; break;
187 case ONIGERR_INVALID_CALLOUT_TAG_NAME
:
188 p
= "invalid callout tag name"; break;
189 case ONIGERR_INVALID_CALLOUT_ARG
:
190 p
= "invalid callout arg"; break;
191 case ONIGERR_NOT_SUPPORTED_ENCODING_COMBINATION
:
192 p
= "not supported encoding combination"; break;
193 case ONIGERR_INVALID_COMBINATION_OF_OPTIONS
:
194 p
= "invalid combination of options"; break;
195 case ONIGERR_LIBRARY_IS_NOT_INITIALIZED
:
196 p
= "library is not initialized"; break;
199 p
= "undefined error code"; break;
205 static void sprint_byte(char* s
, unsigned int v
)
207 sprintf_s(s
, sizeof("00"), "%02x", (v
& 0377));
210 static void sprint_byte_with_x(char* s
, unsigned int v
)
212 sprintf_s(s
, sizeof("\\x00"), "\\x%02x", (v
& 0377));
215 static int to_ascii(OnigEncoding enc
, UChar
*s
, UChar
*end
,
216 UChar buf
[], int buf_size
, int *is_over
)
222 if (ONIGENC_MBC_MINLEN(enc
) > 1) {
226 code
= ONIGENC_MBC_TO_CODE(enc
, p
, end
);
228 if (code
> 0xffff && len
+ 10 <= buf_size
) {
229 sprint_byte_with_x((char*)(&(buf
[len
])), (unsigned int)(code
>> 24));
230 sprint_byte((char*)(&(buf
[len
+4])), (unsigned int)(code
>> 16));
231 sprint_byte((char*)(&(buf
[len
+6])), (unsigned int)(code
>> 8));
232 sprint_byte((char*)(&(buf
[len
+8])), (unsigned int)code
);
235 else if (len
+ 6 <= buf_size
) {
236 sprint_byte_with_x((char*)(&(buf
[len
])), (unsigned int)(code
>> 8));
237 sprint_byte((char*)(&(buf
[len
+4])), (unsigned int)code
);
245 buf
[len
++] = (UChar
)code
;
249 if (len
>= buf_size
) break;
252 *is_over
= ((p
< end
) ? 1 : 0);
255 len
= MIN((int )(end
- s
), buf_size
);
256 xmemcpy(buf
, s
, (size_t )len
);
257 *is_over
= ((buf_size
< (end
- s
)) ? 1 : 0);
264 /* for ONIG_MAX_ERROR_MESSAGE_LEN */
265 #define MAX_ERROR_PAR_LEN 30
269 #ifdef HAVE_STDARG_PROTOTYPES
270 onig_error_code_to_str(UChar
* s
, int code
, ...)
272 onig_error_code_to_str(s
, code
, va_alist
)
279 OnigErrorInfo
* einfo
;
281 UChar parbuf
[MAX_ERROR_PAR_LEN
];
284 va_init_list(vargs
, code
);
287 case ONIGERR_UNDEFINED_NAME_REFERENCE
:
288 case ONIGERR_UNDEFINED_GROUP_REFERENCE
:
289 case ONIGERR_MULTIPLEX_DEFINED_NAME
:
290 case ONIGERR_MULTIPLEX_DEFINITION_NAME_CALL
:
291 case ONIGERR_INVALID_GROUP_NAME
:
292 case ONIGERR_INVALID_CHAR_IN_GROUP_NAME
:
293 case ONIGERR_INVALID_CHAR_PROPERTY_NAME
:
294 einfo
= va_arg(vargs
, OnigErrorInfo
*);
295 len
= to_ascii(einfo
->enc
, einfo
->par
, einfo
->par_end
,
296 parbuf
, MAX_ERROR_PAR_LEN
- 3, &is_over
);
297 q
= onig_error_code_to_format(code
);
302 if (*q
== 'n') { /* '%n': name */
303 xmemcpy(p
, parbuf
, len
);
306 xmemcpy(p
, "...", 3);
324 q
= onig_error_code_to_format(code
);
325 len
= onigenc_str_bytelen_null(ONIG_ENCODING_ASCII
, q
);
338 #ifdef HAVE_STDARG_PROTOTYPES
339 onig_snprintf_with_pattern(UChar buf
[], int bufsize
, OnigEncoding enc
,
340 UChar
* pat
, UChar
* pat_end
, const UChar
*fmt
, ...)
342 onig_snprintf_with_pattern(buf
, bufsize
, enc
, pat
, pat_end
, fmt
, va_alist
)
357 va_init_list(args
, fmt
);
358 n
= xvsnprintf((char* )buf
, bufsize
, (const char* )fmt
, args
);
361 need
= (int )(pat_end
- pat
) * 4 + 4;
363 if (n
+ need
< bufsize
) {
364 strcat_s((char* )buf
, bufsize
, ": /");
365 s
= buf
+ onigenc_str_bytelen_null(ONIG_ENCODING_ASCII
, buf
);
368 while (p
< pat_end
) {
369 if (ONIGENC_IS_MBC_HEAD(enc
, p
)) {
370 len
= enclen(enc
, p
);
371 if (ONIGENC_MBC_MINLEN(enc
) == 1) {
372 while (len
-- > 0) *s
++ = *p
++;
374 else { /* for UTF16/32 */
378 sprint_byte_with_x((char* )bs
, (unsigned int )(*p
++));
379 blen
= onigenc_str_bytelen_null(ONIG_ENCODING_ASCII
, bs
);
381 while (blen
-- > 0) *s
++ = *bp
++;
385 else if (*p
== '\\') {
387 len
= enclen(enc
, p
);
388 while (len
-- > 0) *s
++ = *p
++;
390 else if (*p
== '/') {
391 *s
++ = (unsigned char )'\\';
394 else if (!ONIGENC_IS_CODE_PRINT(enc
, *p
) &&
395 !ONIGENC_IS_CODE_SPACE(enc
, *p
)) {
396 sprint_byte_with_x((char* )bs
, (unsigned int )(*p
++));
397 len
= onigenc_str_bytelen_null(ONIG_ENCODING_ASCII
, bs
);
399 while (len
-- > 0) *s
++ = *bp
++;