/[pcre]/code/trunk/pcre_compile.c

Diff of /code/trunk/pcre_compile.c

Parent Directory | Revision Log | View Patch Patch

-revision 182 by ph10,
Wed Jun 13 15:09:54 2007 UTC
+revision 216 by ph10,
Wed Aug 15 14:35:57 2007 UTC
 Line 42 
 POSSIBILITY OF SUCH DAMAGE.
  supporting internal functions that are not used by other modules. */
+ #ifdef HAVE_CONFIG_H
+ #include <config.h>
+ #endif
  #define NLBLOCK cd             /* Block containing newline information */
  #define PSSTART start_pattern  /* Field containing processed string start */
  #define PSEND   end_pattern    /* Field containing processed string end */
  #include "pcre_internal.h"
-Line 62 
 used by pcretest. DEBUG is not defined w
+Line 65 
 used by pcretest. DEBUG is not defined w
  #define SETBIT(a,b) a[b/8] |= (1 << (b%8))
+ /* Maximum length value to check against when making sure that the integer that
+ holds the compiled pattern length does not overflow. We make it a bit less than
+ INT_MAX to allow for adding in group terminating bytes, so that we don't have
+ to check them every time. */
+ #define OFLOW_MAX (INT_MAX - 20)
  /*************************************************
  *      Code parameters and static tables         *
-Line 120 
 static const short int escapes[] = {
+Line 130 
 static const short int escapes[] = {
  /*  B8 */     0,     0,      0,       0,      0,   ']',    '=',    '-',
  /*  C0 */   '{',-ESC_A, -ESC_B,  -ESC_C, -ESC_D,-ESC_E,      0, -ESC_G,
  /*  C8 */-ESC_H,     0,      0,       0,      0,     0,      0,      0,
- /*  D0 */   '}',     0,      0,       0,      0,     0,      0, -ESC_P,
+ /*  D0 */   '}',     0, -ESC_K,       0,      0,     0,      0, -ESC_P,
  /*  D8 */-ESC_Q,-ESC_R,      0,       0,      0,     0,      0,      0,
  /*  E0 */  '\\',     0, -ESC_S,       0,      0,-ESC_V, -ESC_W, -ESC_X,
  /*  E8 */     0,-ESC_Z,      0,       0,      0,     0,      0,      0,
-Line 130 
 static const short int escapes[] = {
+Line 140 
 static const short int escapes[] = {
  #endif
+ /* Table of special "verbs" like (*PRUNE) */
+ typedef struct verbitem {
+   const char *name;
+   int   len;
+   int   op;
+ } verbitem;
+ static verbitem verbs[] = {
+   { "ACCEPT", 6, OP_ACCEPT },
+   { "COMMIT", 6, OP_COMMIT },
+   { "F",      1, OP_FAIL },
+   { "FAIL",   4, OP_FAIL },
+   { "PRUNE",  5, OP_PRUNE },
+   { "SKIP",   4, OP_SKIP  },
+   { "THEN",   4, OP_THEN  }
+ };
+ static int verbcount = sizeof(verbs)/sizeof(verbitem);
  /* Tables of names of POSIX character classes and their lengths. The list is
  terminated by a zero length entry. The first three must be alpha, lower, upper,
  as this is assumed for handling case independence. */
-Line 203 
 static const char *error_texts[] = {
+Line 234 
 static const char *error_texts[] = {
    "missing ) after comment",
    "parentheses nested too deeply",  /** DEAD **/
    /* 20 */
-   "regular expression too large",
+   "regular expression is too large",
    "failed to get memory",
    "unmatched parentheses",
    "internal error: code overflow",
-Line 239 
 static const char *error_texts[] = {
+Line 270 
 static const char *error_texts[] = {
    "subpattern name is too long (maximum " XSTRING(MAX_NAME_SIZE) " characters)",
    "too many named subpatterns (maximum " XSTRING(MAX_NAME_COUNT) ")",
    /* 50 */
-   "repeated subpattern is too long",
+   "repeated subpattern is too long",    /** DEAD **/
    "octal value is greater than \\377 (not in UTF-8 mode)",
    "internal error: overran compiling workspace",
    "internal error: previously-checked referenced subpattern not found",
-Line 248 
 static const char *error_texts[] = {
+Line 279 
 static const char *error_texts[] = {
    "repeating a DEFINE group is not allowed",
    "inconsistent NEWLINE options",
    "\\g is not followed by a braced name or an optionally braced non-zero number",
-   "(?+ or (?- or (?(+ or (?(- must be followed by a non-zero number"
+   "(?+ or (?- or (?(+ or (?(- must be followed by a non-zero number",
+   "(*VERB) with an argument is not supported",
+   /* 60 */
+   "(*VERB) not recognized",
+   "number is too big"
  };
-Line 405 
 Arguments:
+Line 440 
 Arguments:
  Returns:         zero or positive => a data character
                   negative => a special escape sequence
-                  on error, errorptr is set
+                  on error, errorcodeptr is set
  */
  static int
-Line 489 
 else
+Line 524 
 else
      c = 0;
      while ((digitab[ptr[1]] & ctype_digit) != 0)
        c = c * 10 + *(++ptr) - '0';
+     if (c < 0)
+       {
+       *errorcodeptr = ERR61;
+       break;
+       }
      if (c == 0 || (braced && *(++ptr) != '}'))
        {
        *errorcodeptr = ERR57;
-       return 0;
+       break;
        }
      if (negated)
-Line 501 
 else
+Line 542 
 else
        if (c > bracount)
          {
          *errorcodeptr = ERR15;
-         return 0;
+         break;
          }
        c = bracount - (c - 1);
        }
-Line 530 
 else
+Line 571 
 else
        c -= '0';
        while ((digitab[ptr[1]] & ctype_digit) != 0)
          c = c * 10 + *(++ptr) - '0';
+       if (c < 0)
+         {
+         *errorcodeptr = ERR61;
+         break;
+         }
        if (c < 10 || c <= bracount)
          {
          c = -(ESC_REF + c);
-Line 625 
 else
+Line 671 
 else
      if (c == 0)
        {
        *errorcodeptr = ERR2;
-       return 0;
+       break;
        }
  #ifndef EBCDIC  /* ASCII coding */
-Line 701 
 if (c == '{')
+Line 747 
 if (c == '{')
      *negptr = TRUE;
      ptr++;
      }
-   for (i = 0; i < sizeof(name) - 1; i++)
+   for (i = 0; i < (int)sizeof(name) - 1; i++)
      {
      c = *(++ptr);
      if (c == 0) goto ERROR_RETURN;
-Line 904 
 for (; *ptr != 0; ptr++)
+Line 950 
 for (; *ptr != 0; ptr++)
      {
      while (*(++ptr) != ']')
        {
+       if (*ptr == 0) return -1;
        if (*ptr == '\\')
          {
          if (*(++ptr) == 0) return -1;
-Line 931 
 for (; *ptr != 0; ptr++)
+Line 978 
 for (; *ptr != 0; ptr++)
    /* An opening parens must now be a real metacharacter */
    if (*ptr != '(') continue;
-   if (ptr[1] != '?')
+   if (ptr[1] != '?' && ptr[1] != '*')
      {
      count++;
      if (name == NULL && count == lorn) return count;
-Line 1399 
 for (code = first_significant_code(code
+Line 1446 
 for (code = first_significant_code(code
    /* For other groups, scan the branches. */
-   if (c == OP_BRA || c == OP_CBRA || c == OP_ONCE)
+   if (c == OP_BRA || c == OP_CBRA || c == OP_ONCE || c == OP_COND)
      {
      BOOL empty_branch;
      if (GET(code, 1) == 0) return TRUE;    /* Hit unclosed bracket */
-Line 1423 
 for (code = first_significant_code(code
+Line 1470 
 for (code = first_significant_code(code
    switch (c)
      {
-     /* Check for quantifiers after a class */
+     /* Check for quantifiers after a class. XCLASS is used for classes that
+     cannot be represented just by a bit map. This includes negated single
+     high-valued characters. The length in _pcre_OP_lengths[] is zero; the
+     actual length is stored in the compiled code, so we must update "code"
+     here. */
  #ifdef SUPPORT_UTF8
      case OP_XCLASS:
-     ccode = code + GET(code, 1);
+     ccode = code += GET(code, 1);
      goto CHECK_CLASS_REPEAT;
  #endif
-Line 2255 
 for (;; ptr++)
+Line 2306 
 for (;; ptr++)
      */
      if (code < last_code) code = last_code;
+     /* Paranoid check for integer overflow */
+     if (OFLOW_MAX - *lengthptr < code - last_code)
+       {
+       *errorcodeptr = ERR20;
+       goto FAILED;
+       }
      *lengthptr += code - last_code;
      DPRINTF(("length=%d added %d c=%c\n", *lengthptr, code - last_code, c));
-Line 2367 
 for (;; ptr++)
+Line 2427 
 for (;; ptr++)
      *ptrptr = ptr;
      if (lengthptr != NULL)
        {
+       if (OFLOW_MAX - *lengthptr < code - last_code)
+         {
+         *errorcodeptr = ERR20;
+         goto FAILED;
+         }
        *lengthptr += code - last_code;   /* To include callout length */
        DPRINTF((">> end branch\n"));
        }
-Line 2429 
 for (;; ptr++)
+Line 2494 
 for (;; ptr++)
        goto FAILED;
        }
-     /* If the first character is '^', set the negation flag and skip it. */
+     /* If the first character is '^', set the negation flag and skip it. Also,
+     if the first few characters (either before or after ^) are \Q\E or \E we
+     skip them too. This makes for compatibility with Perl. */
-     if ((c = *(++ptr)) == '^')
+     negate_class = FALSE;
+     for (;;)
        {
-       negate_class = TRUE;
        c = *(++ptr);
-       }
+       if (c == '\\')
-     else
+         {
-       {
+         if (ptr[1] == 'E') ptr++;
-       negate_class = FALSE;
+           else if (strncmp((const char *)ptr+1, "Q\\E", 3) == 0) ptr += 3;
+             else break;
+         }
+       else if (!negate_class && c == '^')
+         negate_class = TRUE;
+       else break;
        }
      /* Keep a count of chars with values < 256 so that we can optimize the case
-Line 2579 
 for (;; ptr++)
+Line 2651 
 for (;; ptr++)
        of the specials, which just set a flag. The sequence \b is a special
        case. Inside a class (and only there) it is treated as backspace.
        Elsewhere it marks a word boundary. Other escapes have preset maps ready
-       to or into the one we are building. We assume they have more than one
+       to 'or' into the one we are building. We assume they have more than one
        character in them, so set class_charcount bigger than one. */
        if (c == '\\')
-Line 3521 
 for (;; ptr++)
+Line 3593 
 for (;; ptr++)
          goto FAILED;
          }
-       /* This is a paranoid check to stop integer overflow later on */
-       if (len > MAX_DUPLENGTH)
-         {
-         *errorcodeptr = ERR50;
-         goto FAILED;
-         }
        /* If the maximum repeat count is unlimited, find the end of the bracket
        by scanning through from the start, and compute the offset back to it
        from the current code pointer. There may be an OP_OPT setting following
-Line 3617 
 for (;; ptr++)
+Line 3681 
 for (;; ptr++)
          if (repeat_min > 1)
            {
            /* In the pre-compile phase, we don't actually do the replication. We
-           just adjust the length as if we had. */
+           just adjust the length as if we had. Do some paranoid checks for
+           potential integer overflow. */
            if (lengthptr != NULL)
-             *lengthptr += (repeat_min - 1)*length_prevgroup;
+             {
+             int delta = (repeat_min - 1)*length_prevgroup;
+             if ((double)(repeat_min - 1)*(double)length_prevgroup >
+                                                             (double)INT_MAX ||
+                 OFLOW_MAX - *lengthptr < delta)
+               {
+               *errorcodeptr = ERR20;
+               goto FAILED;
+               }
+             *lengthptr += delta;
+             }
            /* This is compiling for real */
-Line 3658 
 for (;; ptr++)
+Line 3733 
 for (;; ptr++)
          /* In the pre-compile phase, we don't actually do the replication. We
          just adjust the length as if we had. For each repetition we must add 1
          to the length for BRAZERO and for all but the last repetition we must
-         add 2 + 2*LINKSIZE to allow for the nesting that occurs. */
+         add 2 + 2*LINKSIZE to allow for the nesting that occurs. Do some
+         paranoid checks to avoid integer overflow. */
          if (lengthptr != NULL && repeat_max > 0)
-           *lengthptr += repeat_max * (length_prevgroup + 1 + 2 + 2*LINK_SIZE) -
+           {
-- 2*LINK_SIZE;  /* Last one doesn't nest */
+           int delta = repeat_max * (length_prevgroup + 1 + 2 + 2*LINK_SIZE) -
+- 2*LINK_SIZE;   /* Last one doesn't nest */
+           if ((double)repeat_max *
+                 (double)(length_prevgroup + 1 + 2 + 2*LINK_SIZE)
+                   > (double)INT_MAX ||
+               OFLOW_MAX - *lengthptr < delta)
+             {
+             *errorcodeptr = ERR20;
+             goto FAILED;
+             }
+           *lengthptr += delta;
+           }
          /* This is compiling for real */
-Line 3814 
 for (;; ptr++)
+Line 3901 
 for (;; ptr++)
      /* ===================================================================*/
      /* Start of nested parenthesized sub-expression, or comment or lookahead or
      lookbehind or option setting or condition or all the other extended
-     parenthesis forms. First deal with the specials; all are introduced by ?,
+     parenthesis forms.  */
-     and the appearance of any of them means that this is not a capturing
-     group. */
      case '(':
      newoptions = options;
-Line 3825 
 for (;; ptr++)
+Line 3910 
 for (;; ptr++)
      save_hwm = cd->hwm;
      reset_bracount = FALSE;
-     if (*(++ptr) == '?')
+     /* First deal with various "verbs" that can be introduced by '*'. */
+     if (*(++ptr) == '*' && (cd->ctypes[ptr[1]] & ctype_letter) != 0)
+       {
+       int i, namelen;
+       const uschar *name = ++ptr;
+       previous = NULL;
+       while ((cd->ctypes[*++ptr] & ctype_letter) != 0);
+       if (*ptr == ':')
+         {
+         *errorcodeptr = ERR59;   /* Not supported */
+         goto FAILED;
+         }
+       if (*ptr != ')')
+         {
+         *errorcodeptr = ERR60;
+         goto FAILED;
+         }
+       namelen = ptr - name;
+       for (i = 0; i < verbcount; i++)
+         {
+         if (namelen == verbs[i].len &&
+             strncmp((char *)name, verbs[i].name, namelen) == 0)
+           {
+           *code = verbs[i].op;
+           if (*code++ == OP_ACCEPT) cd->had_accept = TRUE;
+           break;
+           }
+         }
+       if (i < verbcount) continue;
+       *errorcodeptr = ERR60;
+       goto FAILED;
+       }
+     /* Deal with the extended parentheses; all are introduced by '?', and the
+     appearance of any of them means that this is not a capturing group. */
+     else if (*ptr == '?')
        {
        int i, set, unset, namelen;
        int *optset;
-Line 4067 
 for (;; ptr++)
+Line 4189 
 for (;; ptr++)
          /* ------------------------------------------------------------ */
          case '!':                 /* Negative lookahead */
-         bravalue = OP_ASSERT_NOT;
          ptr++;
+         if (*ptr == ')')          /* Optimize (?!) */
+           {
+           *code++ = OP_FAIL;
+           previous = NULL;
+           continue;
+           }
+         bravalue = OP_ASSERT_NOT;
          break;
-Line 4624 
 for (;; ptr++)
+Line 4752 
 for (;; ptr++)
      if (lengthptr != NULL)
        {
+       if (OFLOW_MAX - *lengthptr < length_prevgroup - 2 - 2*LINK_SIZE)
+         {
+         *errorcodeptr = ERR20;
+         goto FAILED;
+         }
        *lengthptr += length_prevgroup - 2 - 2*LINK_SIZE;
        code++;
        PUTINC(code, 0, 1 + LINK_SIZE);
-Line 5119 
 for (;;)
+Line 5252 
 for (;;)
      *ptrptr = ptr;
      *firstbyteptr = firstbyte;
      *reqbyteptr = reqbyte;
-     if (lengthptr != NULL) *lengthptr += length;
+     if (lengthptr != NULL)
+       {
+       if (OFLOW_MAX - *lengthptr < length)
+         {
+         *errorcodeptr = ERR20;
+         return FALSE;
+         }
+       *lengthptr += length;
+       }
      return TRUE;
      }
-Line 5647 
 cd->start_code = codestart;
+Line 5788 
 cd->start_code = codestart;
  cd->hwm = cworkspace;
  cd->req_varyopt = 0;
  cd->nopartial = FALSE;
+ cd->had_accept = FALSE;
  /* Set up a starting, non-extracting bracket, then compile the expression. On
  error, errorcode will be set non-zero, so we don't need to look at the result
-Line 5661 
 re->top_bracket = cd->bracount;
+Line 5803 
 re->top_bracket = cd->bracount;
  re->top_backref = cd->top_backref;
  if (cd->nopartial) re->options |= PCRE_NOPARTIAL;
+ if (cd->had_accept) reqbyte = -1;   /* Must disable after (*ACCEPT) */
  /* If not reached end of pattern on success, there's an excess bracket. */

 Legend:



Removed from v.182
 


changed lines


 
Added in v.216
 Legend:



Removed from v.182
 


changed lines


 
Added in v.216
-Removed from v.182
+Added in v.216

	ViewVC Help
Powered by ViewVC 1.1.5