1 : /**********************************************************************
2 : regerror.c - Oniguruma (regular expression library)
3 : **********************************************************************/
4 : /*-
5 : * Copyright (c) 2002-2009 K.Kosako <sndgk393 AT ybb DOT ne DOT jp>
6 : * All rights reserved.
7 : *
8 : * Redistribution and use in source and binary forms, with or without
9 : * modification, are permitted provided that the following conditions
10 : * are met:
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.
16 : *
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
27 : * SUCH DAMAGE.
28 : */
29 :
30 : #include "regint.h"
31 : #include <stdio.h> /* for vsnprintf() */
32 :
33 : #ifdef HAVE_STDARG_PROTOTYPES
34 : #include <stdarg.h>
35 : #define va_init_list(a,b) va_start(a,b)
36 : #else
37 : #include <varargs.h>
38 : #define va_init_list(a,b) va_start(a)
39 : #endif
40 :
41 : extern UChar*
42 : onig_error_code_to_format(int code)
43 0 : {
44 : char *p;
45 :
46 0 : if (code >= 0) return (UChar* )0;
47 :
48 0 : switch (code) {
49 : case ONIG_MISMATCH:
50 0 : p = "mismatch"; break;
51 : case ONIG_NO_SUPPORT_CONFIG:
52 0 : p = "no support in this configuration"; break;
53 : case ONIGERR_MEMORY:
54 0 : p = "fail to memory allocation"; break;
55 : case ONIGERR_MATCH_STACK_LIMIT_OVER:
56 0 : p = "match-stack limit over"; break;
57 : case ONIGERR_TYPE_BUG:
58 0 : p = "undefined type (bug)"; break;
59 : case ONIGERR_PARSER_BUG:
60 0 : p = "internal parser error (bug)"; break;
61 : case ONIGERR_STACK_BUG:
62 0 : p = "stack error (bug)"; break;
63 : case ONIGERR_UNDEFINED_BYTECODE:
64 0 : p = "undefined bytecode (bug)"; break;
65 : case ONIGERR_UNEXPECTED_BYTECODE:
66 0 : p = "unexpected bytecode (bug)"; break;
67 : case ONIGERR_DEFAULT_ENCODING_IS_NOT_SETTED:
68 0 : p = "default multibyte-encoding is not setted"; break;
69 : case ONIGERR_SPECIFIED_ENCODING_CANT_CONVERT_TO_WIDE_CHAR:
70 0 : p = "can't convert to wide-char on specified multibyte-encoding"; break;
71 : case ONIGERR_INVALID_ARGUMENT:
72 0 : p = "invalid argument"; break;
73 : case ONIGERR_END_PATTERN_AT_LEFT_BRACE:
74 0 : p = "end pattern at left brace"; break;
75 : case ONIGERR_END_PATTERN_AT_LEFT_BRACKET:
76 0 : p = "end pattern at left bracket"; break;
77 : case ONIGERR_EMPTY_CHAR_CLASS:
78 0 : p = "empty char-class"; break;
79 : case ONIGERR_PREMATURE_END_OF_CHAR_CLASS:
80 0 : p = "premature end of char-class"; break;
81 : case ONIGERR_END_PATTERN_AT_ESCAPE:
82 0 : p = "end pattern at escape"; break;
83 : case ONIGERR_END_PATTERN_AT_META:
84 0 : p = "end pattern at meta"; break;
85 : case ONIGERR_END_PATTERN_AT_CONTROL:
86 0 : p = "end pattern at control"; break;
87 : case ONIGERR_META_CODE_SYNTAX:
88 0 : p = "illegal meta-code syntax"; break;
89 : case ONIGERR_CONTROL_CODE_SYNTAX:
90 0 : p = "illegal control-code syntax"; break;
91 : case ONIGERR_CHAR_CLASS_VALUE_AT_END_OF_RANGE:
92 0 : p = "char-class value at end of range"; break;
93 : case ONIGERR_CHAR_CLASS_VALUE_AT_START_OF_RANGE:
94 0 : p = "char-class value at start of range"; break;
95 : case ONIGERR_UNMATCHED_RANGE_SPECIFIER_IN_CHAR_CLASS:
96 0 : p = "unmatched range specifier in char-class"; break;
97 : case ONIGERR_TARGET_OF_REPEAT_OPERATOR_NOT_SPECIFIED:
98 0 : p = "target of repeat operator is not specified"; break;
99 : case ONIGERR_TARGET_OF_REPEAT_OPERATOR_INVALID:
100 0 : p = "target of repeat operator is invalid"; break;
101 : case ONIGERR_NESTED_REPEAT_OPERATOR:
102 0 : p = "nested repeat operator"; break;
103 : case ONIGERR_UNMATCHED_CLOSE_PARENTHESIS:
104 0 : p = "unmatched close parenthesis"; break;
105 : case ONIGERR_END_PATTERN_WITH_UNMATCHED_PARENTHESIS:
106 0 : p = "end pattern with unmatched parenthesis"; break;
107 : case ONIGERR_END_PATTERN_IN_GROUP:
108 0 : p = "end pattern in group"; break;
109 : case ONIGERR_UNDEFINED_GROUP_OPTION:
110 0 : p = "undefined group option"; break;
111 : case ONIGERR_INVALID_POSIX_BRACKET_TYPE:
112 0 : p = "invalid POSIX bracket type"; break;
113 : case ONIGERR_INVALID_LOOK_BEHIND_PATTERN:
114 0 : p = "invalid pattern in look-behind"; break;
115 : case ONIGERR_INVALID_REPEAT_RANGE_PATTERN:
116 0 : p = "invalid repeat range {lower,upper}"; break;
117 : case ONIGERR_TOO_BIG_NUMBER:
118 0 : p = "too big number"; break;
119 : case ONIGERR_TOO_BIG_NUMBER_FOR_REPEAT_RANGE:
120 0 : p = "too big number for repeat range"; break;
121 : case ONIGERR_UPPER_SMALLER_THAN_LOWER_IN_REPEAT_RANGE:
122 0 : p = "upper is smaller than lower in repeat range"; break;
123 : case ONIGERR_EMPTY_RANGE_IN_CHAR_CLASS:
124 0 : p = "empty range in char class"; break;
125 : case ONIGERR_MISMATCH_CODE_LENGTH_IN_CLASS_RANGE:
126 0 : p = "mismatch multibyte code length in char-class range"; break;
127 : case ONIGERR_TOO_MANY_MULTI_BYTE_RANGES:
128 0 : p = "too many multibyte code ranges are specified"; break;
129 : case ONIGERR_TOO_SHORT_MULTI_BYTE_STRING:
130 0 : p = "too short multibyte code string"; break;
131 : case ONIGERR_TOO_BIG_BACKREF_NUMBER:
132 0 : p = "too big backref number"; break;
133 : case ONIGERR_INVALID_BACKREF:
134 : #ifdef USE_NAMED_GROUP
135 0 : p = "invalid backref number/name"; break;
136 : #else
137 : p = "invalid backref number"; break;
138 : #endif
139 : case ONIGERR_NUMBERED_BACKREF_OR_CALL_NOT_ALLOWED:
140 0 : p = "numbered backref/call is not allowed. (use name)"; break;
141 : case ONIGERR_TOO_BIG_WIDE_CHAR_VALUE:
142 0 : p = "too big wide-char value"; break;
143 : case ONIGERR_TOO_LONG_WIDE_CHAR_VALUE:
144 0 : p = "too long wide-char value"; break;
145 : case ONIGERR_INVALID_WIDE_CHAR_VALUE:
146 0 : p = "invalid wide-char value"; break;
147 : case ONIGERR_EMPTY_GROUP_NAME:
148 0 : p = "group name is empty"; break;
149 : case ONIGERR_INVALID_GROUP_NAME:
150 0 : p = "invalid group name <%n>"; break;
151 : case ONIGERR_INVALID_CHAR_IN_GROUP_NAME:
152 : #ifdef USE_NAMED_GROUP
153 0 : p = "invalid char in group name <%n>"; break;
154 : #else
155 : p = "invalid char in group number <%n>"; break;
156 : #endif
157 : case ONIGERR_UNDEFINED_NAME_REFERENCE:
158 0 : p = "undefined name <%n> reference"; break;
159 : case ONIGERR_UNDEFINED_GROUP_REFERENCE:
160 0 : p = "undefined group <%n> reference"; break;
161 : case ONIGERR_MULTIPLEX_DEFINED_NAME:
162 0 : p = "multiplex defined name <%n>"; break;
163 : case ONIGERR_MULTIPLEX_DEFINITION_NAME_CALL:
164 0 : p = "multiplex definition name <%n> call"; break;
165 : case ONIGERR_NEVER_ENDING_RECURSION:
166 0 : p = "never ending recursion"; break;
167 : case ONIGERR_GROUP_NUMBER_OVER_FOR_CAPTURE_HISTORY:
168 0 : p = "group number is too big for capture history"; break;
169 : case ONIGERR_INVALID_CHAR_PROPERTY_NAME:
170 0 : p = "invalid character property name {%n}"; break;
171 : case ONIGERR_NOT_SUPPORTED_ENCODING_COMBINATION:
172 0 : p = "not supported encoding combination"; break;
173 : case ONIGERR_INVALID_COMBINATION_OF_OPTIONS:
174 0 : p = "invalid combination of options"; break;
175 : case ONIGERR_OVER_THREAD_PASS_LIMIT_COUNT:
176 0 : p = "over thread pass limit count"; break;
177 :
178 : default:
179 0 : p = "undefined error code"; break;
180 : }
181 :
182 0 : return (UChar* )p;
183 : }
184 :
185 :
186 : /* for ONIG_MAX_ERROR_MESSAGE_LEN */
187 : #define MAX_ERROR_PAR_LEN 30
188 :
189 : extern int
190 : #ifdef HAVE_STDARG_PROTOTYPES
191 : onig_error_code_to_str(UChar* s, int code, ...)
192 : #else
193 : onig_error_code_to_str(s, code, va_alist)
194 : UChar* s;
195 : int code;
196 : va_dcl
197 : #endif
198 0 : {
199 : UChar *p, *q;
200 : OnigErrorInfo* einfo;
201 : int len;
202 : va_list vargs;
203 :
204 0 : va_init_list(vargs, code);
205 :
206 0 : switch (code) {
207 : case ONIGERR_UNDEFINED_NAME_REFERENCE:
208 : case ONIGERR_UNDEFINED_GROUP_REFERENCE:
209 : case ONIGERR_MULTIPLEX_DEFINED_NAME:
210 : case ONIGERR_MULTIPLEX_DEFINITION_NAME_CALL:
211 : case ONIGERR_INVALID_GROUP_NAME:
212 : case ONIGERR_INVALID_CHAR_IN_GROUP_NAME:
213 : case ONIGERR_INVALID_CHAR_PROPERTY_NAME:
214 0 : einfo = va_arg(vargs, OnigErrorInfo*);
215 0 : len = einfo->par_end - einfo->par;
216 0 : q = onig_error_code_to_format(code);
217 0 : p = s;
218 0 : while (*q != '\0') {
219 0 : if (*q == '%') {
220 0 : q++;
221 0 : if (*q == 'n') { /* '%n': name */
222 0 : if (len > MAX_ERROR_PAR_LEN) {
223 0 : xmemcpy(p, einfo->par, MAX_ERROR_PAR_LEN - 3);
224 0 : p += (MAX_ERROR_PAR_LEN - 3);
225 0 : xmemcpy(p, "...", 3);
226 0 : p += 3;
227 : }
228 : else {
229 0 : xmemcpy(p, einfo->par, len);
230 0 : p += len;
231 : }
232 0 : q++;
233 : }
234 : else
235 0 : goto normal_char;
236 : }
237 : else {
238 0 : normal_char:
239 0 : *p++ = *q++;
240 : }
241 : }
242 0 : *p = '\0';
243 0 : len = p - s;
244 0 : break;
245 :
246 : default:
247 0 : q = onig_error_code_to_format(code);
248 0 : len = onigenc_str_bytelen_null(ONIG_ENCODING_ASCII, q);
249 0 : xmemcpy(s, q, len);
250 0 : s[len] = '\0';
251 : break;
252 : }
253 :
254 0 : va_end(vargs);
255 0 : return len;
256 : }
257 :
258 :
259 : void
260 : #ifdef HAVE_STDARG_PROTOTYPES
261 : onig_snprintf_with_pattern(UChar buf[], int bufsize, OnigEncoding enc,
262 : UChar* pat, UChar* pat_end, const UChar *fmt, ...)
263 : #else
264 : onig_snprintf_with_pattern(buf, bufsize, enc, pat, pat_end, fmt, va_alist)
265 : UChar buf[];
266 : int bufsize;
267 : OnigEncoding enc;
268 : UChar* pat;
269 : UChar* pat_end;
270 : const UChar *fmt;
271 : va_dcl
272 : #endif
273 0 : {
274 : int n, need, len;
275 : UChar *p, *s, *bp;
276 : UChar bs[6];
277 : va_list args;
278 :
279 0 : va_init_list(args, fmt);
280 0 : n = vsnprintf((char* )buf, bufsize, (const char* )fmt, args);
281 0 : if (n < 0 || n >= bufsize) {
282 0 : n = bufsize - 1;
283 : }
284 0 : va_end(args);
285 :
286 0 : need = (pat_end - pat) * 4 + 4;
287 :
288 0 : if (n + need < bufsize) {
289 0 : strcat((char* )buf, ": /");
290 0 : s = buf + onigenc_str_bytelen_null(ONIG_ENCODING_ASCII, buf);
291 :
292 0 : p = pat;
293 0 : while (p < pat_end) {
294 0 : if (*p == MC_ESC(enc)) {
295 0 : *s++ = *p++;
296 0 : len = enc_len(enc, p);
297 0 : while (len-- > 0) *s++ = *p++;
298 : }
299 0 : else if (*p == '/') {
300 0 : *s++ = (unsigned char )MC_ESC(enc);
301 0 : *s++ = *p++;
302 : }
303 0 : else if (ONIGENC_IS_MBC_HEAD(enc, p)) {
304 0 : len = enc_len(enc, p);
305 0 : if (ONIGENC_MBC_MINLEN(enc) == 1) {
306 0 : while (len-- > 0) *s++ = *p++;
307 : }
308 : else { /* for UTF16 */
309 : int blen;
310 :
311 0 : while (len-- > 0) {
312 0 : sprintf((char* )bs, "\\%03o", *p++ & 0377);
313 0 : blen = onigenc_str_bytelen_null(ONIG_ENCODING_ASCII, bs);
314 0 : bp = bs;
315 0 : while (blen-- > 0) *s++ = *bp++;
316 : }
317 : }
318 : }
319 0 : else if (!ONIGENC_IS_CODE_PRINT(enc, *p) &&
320 : !ONIGENC_IS_CODE_SPACE(enc, *p)) {
321 0 : sprintf((char* )bs, "\\%03o", *p++ & 0377);
322 0 : len = onigenc_str_bytelen_null(ONIG_ENCODING_ASCII, bs);
323 0 : bp = bs;
324 0 : while (len-- > 0) *s++ = *bp++;
325 : }
326 : else {
327 0 : *s++ = *p++;
328 : }
329 : }
330 :
331 0 : *s++ = '/';
332 0 : *s = '\0';
333 : }
334 0 : }
|