regerror.c

   1 /**********************************************************************
   2   regerror.c -  Oniguruma (regular expression library)
   3 **********************************************************************/
   4 /*-
   5  * Copyright (c) 2002-2007  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 {
  44   const char *p;
  45
  46   if (code >= 0) return (UChar* )0;
  47
  48   switch (code) {
  49   case ONIG_MISMATCH:
  50     p = "mismatch"; break;
  51   case ONIG_NO_SUPPORT_CONFIG:
  52     p = "no support in this configuration"; break;
  53   case ONIGERR_MEMORY:
  54     p = "fail to memory allocation"; break;
  55   case ONIGERR_MATCH_STACK_LIMIT_OVER:
  56     p = "match-stack limit over"; break;
  57   case ONIGERR_TYPE_BUG:
  58     p = "undefined type (bug)"; break;
  59   case ONIGERR_PARSER_BUG:
  60     p = "internal parser error (bug)"; break;
  61   case ONIGERR_STACK_BUG:
  62     p = "stack error (bug)"; break;
  63   case ONIGERR_UNDEFINED_BYTECODE:
  64     p = "undefined bytecode (bug)"; break;
  65   case ONIGERR_UNEXPECTED_BYTECODE:
  66     p = "unexpected bytecode (bug)"; break;
  67   case ONIGERR_DEFAULT_ENCODING_IS_NOT_SETTED:
  68     p = "default multibyte-encoding is not setted"; break;
  69   case ONIGERR_SPECIFIED_ENCODING_CANT_CONVERT_TO_WIDE_CHAR:
  70     p = "can't convert to wide-char on specified multibyte-encoding"; break;
  71   case ONIGERR_INVALID_ARGUMENT:
  72     p = "invalid argument"; break;
  73   case ONIGERR_END_PATTERN_AT_LEFT_BRACE:
  74     p = "end pattern at left brace"; break;
  75   case ONIGERR_END_PATTERN_AT_LEFT_BRACKET:
  76     p = "end pattern at left bracket"; break;
  77   case ONIGERR_EMPTY_CHAR_CLASS:
  78     p = "empty char-class"; break;
  79   case ONIGERR_PREMATURE_END_OF_CHAR_CLASS:
  80     p = "premature end of char-class"; break;
  81   case ONIGERR_END_PATTERN_AT_ESCAPE:
  82     p = "end pattern at escape"; break;
  83   case ONIGERR_END_PATTERN_AT_META:
  84     p = "end pattern at meta"; break;
  85   case ONIGERR_END_PATTERN_AT_CONTROL:
  86     p = "end pattern at control"; break;
  87   case ONIGERR_META_CODE_SYNTAX:
  88     p = "invalid meta-code syntax"; break;
  89   case ONIGERR_CONTROL_CODE_SYNTAX:
  90     p = "invalid control-code syntax"; break;
  91   case ONIGERR_CHAR_CLASS_VALUE_AT_END_OF_RANGE:
  92     p = "char-class value at end of range"; break;
  93   case ONIGERR_CHAR_CLASS_VALUE_AT_START_OF_RANGE:
  94     p = "char-class value at start of range"; break;
  95   case ONIGERR_UNMATCHED_RANGE_SPECIFIER_IN_CHAR_CLASS:
  96     p = "unmatched range specifier in char-class"; break;
  97   case ONIGERR_TARGET_OF_REPEAT_OPERATOR_NOT_SPECIFIED:
  98     p = "target of repeat operator is not specified"; break;
  99   case ONIGERR_TARGET_OF_REPEAT_OPERATOR_INVALID:
 100     p = "target of repeat operator is invalid"; break;
 101   case ONIGERR_NESTED_REPEAT_OPERATOR:
 102     p = "nested repeat operator"; break;
 103   case ONIGERR_UNMATCHED_CLOSE_PARENTHESIS:
 104     p = "unmatched close parenthesis"; break;
 105   case ONIGERR_END_PATTERN_WITH_UNMATCHED_PARENTHESIS:
 106     p = "end pattern with unmatched parenthesis"; break;
 107   case ONIGERR_END_PATTERN_IN_GROUP:
 108     p = "end pattern in group"; break;
 109   case ONIGERR_UNDEFINED_GROUP_OPTION:
 110     p = "undefined group option"; break;
 111   case ONIGERR_INVALID_POSIX_BRACKET_TYPE:
 112     p = "invalid POSIX bracket type"; break;
 113   case ONIGERR_INVALID_LOOK_BEHIND_PATTERN:
 114     p = "invalid pattern in look-behind"; break;
 115   case ONIGERR_INVALID_REPEAT_RANGE_PATTERN:
 116     p = "invalid repeat range {lower,upper}"; break;
 117   case ONIGERR_TOO_BIG_NUMBER:
 118     p = "too big number"; break;
 119   case ONIGERR_TOO_BIG_NUMBER_FOR_REPEAT_RANGE:
 120     p = "too big number for repeat range"; break;
 121   case ONIGERR_UPPER_SMALLER_THAN_LOWER_IN_REPEAT_RANGE:
 122     p = "upper is smaller than lower in repeat range"; break;
 123   case ONIGERR_EMPTY_RANGE_IN_CHAR_CLASS:
 124     p = "empty range in char class"; break;
 125   case ONIGERR_MISMATCH_CODE_LENGTH_IN_CLASS_RANGE:
 126     p = "mismatch multibyte code length in char-class range"; break;
 127   case ONIGERR_TOO_MANY_MULTI_BYTE_RANGES:
 128     p = "too many multibyte code ranges are specified"; break;
 129   case ONIGERR_TOO_SHORT_MULTI_BYTE_STRING:
 130     p = "too short multibyte code string"; break;
 131   case ONIGERR_TOO_BIG_BACKREF_NUMBER:
 132     p = "too big backref number"; break;
 133   case ONIGERR_INVALID_BACKREF:
 134 #ifdef USE_NAMED_GROUP
 135     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     p = "numbered backref/call is not allowed. (use name)"; break;
 141   case ONIGERR_TOO_BIG_WIDE_CHAR_VALUE:
 142     p = "too big wide-char value"; break;
 143   case ONIGERR_TOO_LONG_WIDE_CHAR_VALUE:
 144     p = "too long wide-char value"; break;
 145   case ONIGERR_INVALID_CODE_POINT_VALUE:
 146     p = "invalid code point value"; break;
 147   case ONIGERR_EMPTY_GROUP_NAME:
 148     p = "group name is empty"; break;
 149   case ONIGERR_INVALID_GROUP_NAME:
 150     p = "invalid group name <%n>"; break;
 151   case ONIGERR_INVALID_CHAR_IN_GROUP_NAME:
 152 #ifdef USE_NAMED_GROUP
 153     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     p = "undefined name <%n> reference"; break;
 159   case ONIGERR_UNDEFINED_GROUP_REFERENCE:
 160     p = "undefined group <%n> reference"; break;
 161   case ONIGERR_MULTIPLEX_DEFINED_NAME:
 162     p = "multiplex defined name <%n>"; break;
 163   case ONIGERR_MULTIPLEX_DEFINITION_NAME_CALL:
 164     p = "multiplex definition name <%n> call"; break;
 165   case ONIGERR_NEVER_ENDING_RECURSION:
 166     p = "never ending recursion"; break;
 167   case ONIGERR_GROUP_NUMBER_OVER_FOR_CAPTURE_HISTORY:
 168     p = "group number is too big for capture history"; break;
 169   case ONIGERR_INVALID_CHAR_PROPERTY_NAME:
 170     p = "invalid character property name {%n}"; break;
 171   case ONIGERR_NOT_SUPPORTED_ENCODING_COMBINATION:
 172     p = "not supported encoding combination"; break;
 173   case ONIGERR_INVALID_COMBINATION_OF_OPTIONS:
 174     p = "invalid combination of options"; break;
 175   case ONIGERR_OVER_THREAD_PASS_LIMIT_COUNT:
 176     p = "over thread pass limit count"; break;
 177
 178   default:
 179     p = "undefined error code"; break;
 180   }
 181
 182   return (UChar* )p;
 183 }
 184
 185 static void sprint_byte(char* s, unsigned int v)
 186 {
 187   sprintf(s, "%02x", (v & 0377));
 188 }
 189
 190 static void sprint_byte_with_x(char* s, unsigned int v)
 191 {
 192   sprintf(s, "\\x%02x", (v & 0377));
 193 }
 194
 195 static int to_ascii(OnigEncoding enc, UChar *s, UChar *end,
 196                     UChar buf[], int buf_size, int *is_over)
 197 {
 198   int len;
 199   UChar *p;
 200   OnigCodePoint code;
 201
 202   if (ONIGENC_MBC_MINLEN(enc) > 1) {
 203     p = s;
 204     len = 0;
 205     while (p < end) {
 206       code = ONIGENC_MBC_TO_CODE(enc, p, end);
 207       if (code >= 0x80) {
 208         if (code > 0xffff && len + 10 <= buf_size) {
 209           sprint_byte_with_x((char*)(&(buf[len])), (unsigned int)(code >> 24));
 210           sprint_byte((char*)(&(buf[len+4])),      (unsigned int)(code >> 16));
 211           sprint_byte((char*)(&(buf[len+6])),      (unsigned int)(code >>  8));
 212           sprint_byte((char*)(&(buf[len+8])),      (unsigned int)code);
 213           len += 10;
 214         }
 215         else if (len + 6 <= buf_size) {
 216           sprint_byte_with_x((char*)(&(buf[len])), (unsigned int)(code >> 8));
 217           sprint_byte((char*)(&(buf[len+4])),      (unsigned int)code);
 218           len += 6;
 219         }
 220         else {
 221           break;
 222         }
 223       }
 224       else {
 225         buf[len++] = (UChar )code;
 226       }
 227
 228       p += enclen(enc, p, end);
 229       if (len >= buf_size) break;
 230     }
 231
 232     *is_over = ((p < end) ? 1 : 0);
 233   }
 234   else {
 235     len = MIN((end - s), buf_size);
 236     xmemcpy(buf, s, (size_t )len);
 237     *is_over = ((buf_size < (end - s)) ? 1 : 0);
 238   }
 239
 240   return len;
 241 }
 242
 243
 244 /* for ONIG_MAX_ERROR_MESSAGE_LEN */
 245 #define MAX_ERROR_PAR_LEN   30
 246
 247 extern int
 248 #ifdef HAVE_STDARG_PROTOTYPES
 249 onig_error_code_to_str(UChar* s, int code, ...)
 250 #else
 251 onig_error_code_to_str(s, code, va_alist)
 252   UChar* s;
 253   int code;
 254   va_dcl
 255 #endif
 256 {
 257   UChar *p, *q;
 258   OnigErrorInfo* einfo;
 259   int len, is_over;
 260   UChar parbuf[MAX_ERROR_PAR_LEN];
 261   va_list vargs;
 262
 263   va_init_list(vargs, code);
 264
 265   switch (code) {
 266   case ONIGERR_UNDEFINED_NAME_REFERENCE:
 267   case ONIGERR_UNDEFINED_GROUP_REFERENCE:
 268   case ONIGERR_MULTIPLEX_DEFINED_NAME:
 269   case ONIGERR_MULTIPLEX_DEFINITION_NAME_CALL:
 270   case ONIGERR_INVALID_GROUP_NAME:
 271   case ONIGERR_INVALID_CHAR_IN_GROUP_NAME:
 272   case ONIGERR_INVALID_CHAR_PROPERTY_NAME:
 273     einfo = va_arg(vargs, OnigErrorInfo*);
 274     len = to_ascii(einfo->enc, einfo->par, einfo->par_end,
 275                    parbuf, MAX_ERROR_PAR_LEN - 3, &is_over);
 276     q = onig_error_code_to_format(code);
 277     p = s;
 278     while (*q != '\0') {
 279       if (*q == '%') {
 280         q++;
 281         if (*q == 'n') { /* '%n': name */
 282           xmemcpy(p, parbuf, len);
 283           p += len;
 284           if (is_over != 0) {
 285             xmemcpy(p, "...", 3);
 286             p += 3;
 287           }
 288           q++;
 289         }
 290         else
 291           goto normal_char;
 292       }
 293       else {
 294       normal_char:
 295         *p++ = *q++;
 296       }
 297     }
 298     *p = '\0';
 299     len = p - s;
 300     break;
 301
 302   default:
 303     q = onig_error_code_to_format(code);
 304     len = onigenc_str_bytelen_null(ONIG_ENCODING_ASCII, q);
 305     xmemcpy(s, q, len);
 306     s[len] = '\0';
 307     break;
 308   }
 309
 310   va_end(vargs);
 311   return len;
 312 }
 313
 314
 315 void
 316 #ifdef HAVE_STDARG_PROTOTYPES
 317 onig_snprintf_with_pattern(UChar buf[], int bufsize, OnigEncoding enc,
 318                            UChar* pat, UChar* pat_end, const UChar *fmt, ...)
 319 #else
 320 onig_snprintf_with_pattern(buf, bufsize, enc, pat, pat_end, fmt, va_alist)
 321     UChar buf[];
 322     int bufsize;
 323     OnigEncoding enc;
 324     UChar* pat;
 325     UChar* pat_end;
 326     const UChar *fmt;
 327     va_dcl
 328 #endif
 329 {
 330   int n, need, len;
 331   UChar *p, *s, *bp;
 332   UChar bs[6];
 333   va_list args;
 334
 335   va_init_list(args, fmt);
 336   n = xvsnprintf((char* )buf, bufsize, (const char* )fmt, args);
 337   va_end(args);
 338
 339   need = (pat_end - pat) * 4 + 4;
 340
 341   if (n + need < bufsize) {
 342     strcat((char* )buf, ": /");
 343     s = buf + onigenc_str_bytelen_null(ONIG_ENCODING_ASCII, buf);
 344
 345     p = pat;
 346     while (p < pat_end) {
 347       if (*p == '\\') {
 348         *s++ = *p++;
 349         len = enclen(enc, p, pat_end);
 350         while (len-- > 0) *s++ = *p++;
 351       }
 352       else if (*p == '/') {
 353         *s++ = (unsigned char )'\\';
 354         *s++ = *p++;
 355       }
 356       else if (ONIGENC_IS_MBC_HEAD(enc, p, pat_end)) {
 357         len = enclen(enc, p, pat_end);
 358         if (ONIGENC_MBC_MINLEN(enc) == 1) {
 359           while (len-- > 0) *s++ = *p++;
 360         }
 361         else { /* for UTF16 */
 362           int blen;
 363
 364           while (len-- > 0) {
 365             sprint_byte_with_x((char* )bs, (unsigned int )(*p++));
 366             blen = onigenc_str_bytelen_null(ONIG_ENCODING_ASCII, bs);
 367             bp = bs;
 368             while (blen-- > 0) *s++ = *bp++;
 369           }
 370         }
 371       }
 372       else if (!ONIGENC_IS_CODE_PRINT(enc, *p) &&
 373                !ONIGENC_IS_CODE_SPACE(enc, *p)) {
 374         sprint_byte_with_x((char* )bs, (unsigned int )(*p++));
 375         len = onigenc_str_bytelen_null(ONIG_ENCODING_ASCII, bs);
 376         bp = bs;
 377         while (len-- > 0) *s++ = *bp++;
 378       }
 379       else {
 380         *s++ = *p++;
 381       }
 382     }
 383
 384     *s++ = '/';
 385     *s   = '\0';
 386   }
 387 }