pcre_compile.c

Scheme跨平台编译器

    case OP_TYPEMINUPTO:
    case OP_TYPEPOSUPTO:
    if (code[3] == OP_PROP || code[3] == OP_NOTPROP) code += 2;
    break;

    /* End of branch */

    case OP_KET:
    case OP_KETRMAX:
    case OP_KETRMIN:
    case OP_ALT:
    return TRUE;

    /* In UTF-8 mode, STAR, MINSTAR, POSSTAR, QUERY, MINQUERY, POSQUERY, UPTO,
    MINUPTO, and POSUPTO may be followed by a multibyte character */

#ifdef SUPPORT_UTF8
    case OP_STAR:
    case OP_MINSTAR:
    case OP_POSSTAR:
    case OP_QUERY:
    case OP_MINQUERY:
    case OP_POSQUERY:
    case OP_UPTO:
    case OP_MINUPTO:
    case OP_POSUPTO:
    if (utf8) while ((code[2] & 0xc0) == 0x80) code++;
    break;
#endif
    }
  }

return TRUE;
}



/*************************************************
*    Scan compiled regex for non-emptiness       *
*************************************************/

/* This function is called to check for left recursive calls. We want to check
the current branch of the current pattern to see if it could match the empty
string. If it could, we must look outwards for branches at other levels,
stopping when we pass beyond the bracket which is the subject of the recursion.

Arguments:
  code        points to start of the recursion
  endcode     points to where to stop (current RECURSE item)
  bcptr       points to the chain of current (unclosed) branch starts
  utf8        TRUE if in UTF-8 mode

Returns:      TRUE if what is matched could be empty
*/

static BOOL
could_be_empty(const uschar *code, const uschar *endcode, branch_chain *bcptr,
  BOOL utf8)
{
while (bcptr != NULL && bcptr->current >= code)
  {
  if (!could_be_empty_branch(bcptr->current, endcode, utf8)) return FALSE;
  bcptr = bcptr->outer;
  }
return TRUE;
}



/*************************************************
*           Check for POSIX class syntax         *
*************************************************/

/* This function is called when the sequence "[:" or "[." or "[=" is
encountered in a character class. It checks whether this is followed by a
sequence of characters terminated by a matching ":]" or ".]" or "=]". If we
reach an unescaped ']' without the special preceding character, return FALSE.

Originally, this function only recognized a sequence of letters between the
terminators, but it seems that Perl recognizes any sequence of characters,
though of course unknown POSIX names are subsequently rejected. Perl gives an
"Unknown POSIX class" error for [:f\oo:] for example, where previously PCRE
didn't consider this to be a POSIX class. Likewise for [:1234:].

The problem in trying to be exactly like Perl is in the handling of escapes. We
have to be sure that [abc[:x\]pqr] is *not* treated as containing a POSIX
class, but [abc[:x\]pqr:]] is (so that an error can be generated). The code
below handles the special case of \], but does not try to do any other escape
processing. This makes it different from Perl for cases such as [:l\ower:]
where Perl recognizes it as the POSIX class "lower" but PCRE does not recognize
"l\ower". This is a lesser evil that not diagnosing bad classes when Perl does,
I think.

Arguments:
  ptr      pointer to the initial [
  endptr   where to return the end pointer

Returns:   TRUE or FALSE
*/

static BOOL
check_posix_syntax(const uschar *ptr, const uschar **endptr)
{
int terminator;          /* Don't combine these lines; the Solaris cc */
terminator = *(++ptr);   /* compiler warns about "non-constant" initializer. */
for (++ptr; *ptr != 0; ptr++)
  {
  if (*ptr == '\\' && ptr[1] == ']') ptr++; else
    {
    if (*ptr == ']') return FALSE;
    if (*ptr == terminator && ptr[1] == ']')
      {
      *endptr = ptr;
      return TRUE;
      }
    }
  }
return FALSE;
}




/*************************************************
*          Check POSIX class name                *
*************************************************/

/* This function is called to check the name given in a POSIX-style class entry
such as [:alnum:].

Arguments:
  ptr        points to the first letter
  len        the length of the name

Returns:     a value representing the name, or -1 if unknown
*/

static int
check_posix_name(const uschar *ptr, int len)
{
const char *pn = posix_names;
register int yield = 0;
while (posix_name_lengths[yield] != 0)
  {
  if (len == posix_name_lengths[yield] &&
    strncmp((const char *)ptr, pn, len) == 0) return yield;
  pn += posix_name_lengths[yield] + 1;
  yield++;
  }
return -1;
}


/*************************************************
*    Adjust OP_RECURSE items in repeated group   *
*************************************************/

/* OP_RECURSE items contain an offset from the start of the regex to the group
that is referenced. This means that groups can be replicated for fixed
repetition simply by copying (because the recursion is allowed to refer to
earlier groups that are outside the current group). However, when a group is
optional (i.e. the minimum quantifier is zero), OP_BRAZERO or OP_SKIPZERO is
inserted before it, after it has been compiled. This means that any OP_RECURSE
items within it that refer to the group itself or any contained groups have to
have their offsets adjusted. That one of the jobs of this function. Before it
is called, the partially compiled regex must be temporarily terminated with
OP_END.

This function has been extended with the possibility of forward references for
recursions and subroutine calls. It must also check the list of such references
for the group we are dealing with. If it finds that one of the recursions in
the current group is on this list, it adjusts the offset in the list, not the
value in the reference (which is a group number).

Arguments:
  group      points to the start of the group
  adjust     the amount by which the group is to be moved
  utf8       TRUE in UTF-8 mode
  cd         contains pointers to tables etc.
  save_hwm   the hwm forward reference pointer at the start of the group

Returns:     nothing
*/

static void
adjust_recurse(uschar *group, int adjust, BOOL utf8, compile_data *cd,
  uschar *save_hwm)
{
uschar *ptr = group;

while ((ptr = (uschar *)find_recurse(ptr, utf8)) != NULL)
  {
  int offset;
  uschar *hc;

  /* See if this recursion is on the forward reference list. If so, adjust the
  reference. */

  for (hc = save_hwm; hc < cd->hwm; hc += LINK_SIZE)
    {
    offset = GET(hc, 0);
    if (cd->start_code + offset == ptr + 1)
      {
      PUT(hc, 0, offset + adjust);
      break;
      }
    }

  /* Otherwise, adjust the recursion offset if it's after the start of this
  group. */

  if (hc >= cd->hwm)
    {
    offset = GET(ptr, 1);
    if (cd->start_code + offset >= group) PUT(ptr, 1, offset + adjust);
    }

  ptr += 1 + LINK_SIZE;
  }
}



/*************************************************
*        Insert an automatic callout point       *
*************************************************/

/* This function is called when the PCRE_AUTO_CALLOUT option is set, to insert
callout points before each pattern item.

Arguments:
  code           current code pointer
  ptr            current pattern pointer
  cd             pointers to tables etc

Returns:         new code pointer
*/

static uschar *
auto_callout(uschar *code, const uschar *ptr, compile_data *cd)
{
*code++ = OP_CALLOUT;
*code++ = 255;
PUT(code, 0, ptr - cd->start_pattern);  /* Pattern offset */
PUT(code, LINK_SIZE, 0);                /* Default length */
return code + 2*LINK_SIZE;
}



/*************************************************
*         Complete a callout item                *
*************************************************/

/* A callout item contains the length of the next item in the pattern, which
we can't fill in till after we have reached the relevant point. This is used
for both automatic and manual callouts.

Arguments:
  previous_callout   points to previous callout item
  ptr                current pattern pointer
  cd                 pointers to tables etc

Returns:             nothing
*/

static void
complete_callout(uschar *previous_callout, const uschar *ptr, compile_data *cd)
{
int length = ptr - cd->start_pattern - GET(previous_callout, 2);
PUT(previous_callout, 2 + LINK_SIZE, length);
}



#ifdef SUPPORT_UCP
/*************************************************
*           Get othercase range                  *
*************************************************/

/* This function is passed the start and end of a class range, in UTF-8 mode
with UCP support. It searches up the characters, looking for internal ranges of
characters in the "other" case. Each call returns the next one, updating the
start address.

Arguments:
  cptr        points to starting character value; updated
  d           end value
  ocptr       where to put start of othercase range
  odptr       where to put end of othercase range

Yield:        TRUE when range returned; FALSE when no more
*/

static BOOL
get_othercase_range(unsigned int *cptr, unsigned int d, unsigned int *ocptr,
  unsigned int *odptr)
{
unsigned int c, othercase, next;

for (c = *cptr; c <= d; c++)
  { if ((othercase = _pcre_ucp_othercase(c)) != NOTACHAR) break; }

if (c > d) return FALSE;

*ocptr = othercase;
next = othercase + 1;

for (++c; c <= d; c++)
  {
  if (_pcre_ucp_othercase(c) != next) break;
  next++;
  }

*odptr = next - 1;
*cptr = c;

return TRUE;
}
#endif  /* SUPPORT_UCP */



/*************************************************
*     Check if auto-possessifying is possible    *
*************************************************/

/* This function is called for unlimited repeats of certain items, to see
whether the next thing could possibly match the repeated item. If not, it makes
sense to automatically possessify the repeated item.

Arguments:
  op_code       the repeated op code
  this          data for this item, depends on the opcode
  utf8          TRUE in UTF-8 mode
  utf8_char     used for utf8 character bytes, NULL if not relevant
  ptr           next character in pattern
  options       options bits
  cd            contains pointers to tables etc.

Returns:        TRUE if possessifying is wanted
*/

static BOOL
check_auto_possessive(int op_code, int item, BOOL utf8, uschar *utf8_char,
  const uschar *ptr, int options, compile_data *cd)
{
int next;

/* Skip whitespace and comments in extended mode */

if ((options & PCRE_EXTENDED) != 0)
  {
  for (;;)
    {
    while ((cd->ctypes[*ptr] & ctype_space) != 0) ptr++;
    if (*ptr == '#')
      {
      while (*(++ptr) != 0)
        if (IS_NEWLINE(ptr)) { ptr += cd->nllen; break; }
      }
    else break;
    }
  }

/* If the next item is one that we can handle, get its value. A non-negative
value is a character, a negative value is an escape value. */

if (*ptr == '\\')
  {
  int temperrorcode = 0;
  next = check_escape(&ptr, &temperrorcode, cd->bracount, options, FALSE);
  if (temperrorcode != 0) return FALSE;
  ptr++;    /* Point after the escape sequence */
  }

else if ((cd->ctypes[*ptr] & ctype_meta) == 0)
  {
#ifdef SUPPORT_UTF8
  if (utf8) { GETCHARINC(next, ptr); } else
#endif
  next = *ptr++;
  }

else return FALSE;

/* Skip whitespace and comments in extended mode */

if ((options & PCRE_EXTENDED) != 0)
  {
  for (;;)
    {
    while ((cd->ctypes[*ptr] & ctype_space) != 0) ptr++;
    if (*ptr == '#')
      {
      while (*(++ptr) != 0)
        if (IS_NEWLINE(ptr)) { ptr += cd->nllen; break; }
      }
    else break;
    }
  }

/* If the next thing is itself optional, we have to give up. */

if (*ptr == '*' || *ptr == '?' || strncmp((char *)ptr, "{0,", 3) == 0)
  return FALSE;

/* Now compare the next item with the previous opcode. If the previous is a
positive single character match, "item" either contains the character or, if
"item" is greater than 127 in utf8 mode, the character's bytes are in
utf8_char. */


/* Handle cases when the next item is a character. */

if (next >= 0) switch(op_code)
  {
  case OP_CHAR:
#ifdef SUPPORT_UTF8
  if (utf8 && item > 127) { GETCHAR(item, utf8_char); }
#endif
  return item != next;

  /* For CHARNC (caseless character) we must check the other case. If we have
  Unicode property support, we can use it to