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

Diff of /code/trunk/pcre_compile.c

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 507 by ph10, Wed Mar 10 16:08:01 2010 UTC revision 530 by ph10, Tue Jun 1 13:42:06 2010 UTC
# Line 124  static const short int escapes[] = { Line 124  static const short int escapes[] = {
124       -ESC_H,                  0,       -ESC_H,                  0,
125       0,                       -ESC_K,       0,                       -ESC_K,
126       0,                       0,       0,                       0,
127       0,                       0,       -ESC_N,                  0,
128       -ESC_P,                  -ESC_Q,       -ESC_P,                  -ESC_Q,
129       -ESC_R,                  -ESC_S,       -ESC_R,                  -ESC_S,
130       0,                       0,       0,                       0,
# Line 171  static const short int escapes[] = { Line 171  static const short int escapes[] = {
171  /*  B8 */     0,     0,      0,       0,      0,   ']',    '=',    '-',  /*  B8 */     0,     0,      0,       0,      0,   ']',    '=',    '-',
172  /*  C0 */   '{',-ESC_A, -ESC_B,  -ESC_C, -ESC_D,-ESC_E,      0, -ESC_G,  /*  C0 */   '{',-ESC_A, -ESC_B,  -ESC_C, -ESC_D,-ESC_E,      0, -ESC_G,
173  /*  C8 */-ESC_H,     0,      0,       0,      0,     0,      0,      0,  /*  C8 */-ESC_H,     0,      0,       0,      0,     0,      0,      0,
174  /*  D0 */   '}',     0, -ESC_K,       0,      0,     0,      0, -ESC_P,  /*  D0 */   '}',     0, -ESC_K,       0,      0,-ESC_N,      0, -ESC_P,
175  /*  D8 */-ESC_Q,-ESC_R,      0,       0,      0,     0,      0,      0,  /*  D8 */-ESC_Q,-ESC_R,      0,       0,      0,     0,      0,      0,
176  /*  E0 */  '\\',     0, -ESC_S,       0,      0,-ESC_V, -ESC_W, -ESC_X,  /*  E0 */  '\\',     0, -ESC_S,       0,      0,-ESC_V, -ESC_W, -ESC_X,
177  /*  E8 */     0,-ESC_Z,      0,       0,      0,     0,      0,      0,  /*  E8 */     0,-ESC_Z,      0,       0,      0,     0,      0,      0,
# Line 188  string is built from string macros so th Line 188  string is built from string macros so th
188  platforms. */  platforms. */
189    
190  typedef struct verbitem {  typedef struct verbitem {
191    int   len;    int   len;                 /* Length of verb name */
192    int   op;    int   op;                  /* Op when no arg, or -1 if arg mandatory */
193      int   op_arg;              /* Op when arg present, or -1 if not allowed */
194  } verbitem;  } verbitem;
195    
196  static const char verbnames[] =  static const char verbnames[] =
197      "\0"                       /* Empty name is a shorthand for MARK */
198      STRING_MARK0
199    STRING_ACCEPT0    STRING_ACCEPT0
200    STRING_COMMIT0    STRING_COMMIT0
201    STRING_F0    STRING_F0
# Line 202  static const char verbnames[] = Line 205  static const char verbnames[] =
205    STRING_THEN;    STRING_THEN;
206    
207  static const verbitem verbs[] = {  static const verbitem verbs[] = {
208    { 6, OP_ACCEPT },    { 0, -1,        OP_MARK },
209    { 6, OP_COMMIT },    { 4, -1,        OP_MARK },
210    { 1, OP_FAIL },    { 6, OP_ACCEPT, -1 },
211    { 4, OP_FAIL },    { 6, OP_COMMIT, -1 },
212    { 5, OP_PRUNE },    { 1, OP_FAIL,   -1 },
213    { 4, OP_SKIP  },    { 4, OP_FAIL,   -1 },
214    { 4, OP_THEN  }    { 5, OP_PRUNE,  OP_PRUNE_ARG },
215      { 4, OP_SKIP,   OP_SKIP_ARG  },
216      { 4, OP_THEN,   OP_THEN_ARG  }
217  };  };
218    
219  static const int verbcount = sizeof(verbs)/sizeof(verbitem);  static const int verbcount = sizeof(verbs)/sizeof(verbitem);
# Line 256  static const int posix_class_maps[] = { Line 261  static const int posix_class_maps[] = {
261    cbit_xdigit,-1,          0              /* xdigit */    cbit_xdigit,-1,          0              /* xdigit */
262  };  };
263    
264    /* Table of substitutes for \d etc when PCRE_UCP is set. The POSIX class
265    substitutes must be in the order of the names, defined above, and there are
266    both positive and negative cases. NULL means no substitute. */
267    
268    #ifdef SUPPORT_UCP
269    static const uschar *substitutes[] = {
270      (uschar *)"\\P{Nd}",    /* \D */
271      (uschar *)"\\p{Nd}",    /* \d */
272      (uschar *)"\\P{Xsp}",   /* \S */       /* NOTE: Xsp is Perl space */
273      (uschar *)"\\p{Xsp}",   /* \s */
274      (uschar *)"\\P{Xwd}",   /* \W */
275      (uschar *)"\\p{Xwd}"    /* \w */
276    };
277    
278    static const uschar *posix_substitutes[] = {
279      (uschar *)"\\p{L}",     /* alpha */
280      (uschar *)"\\p{Ll}",    /* lower */
281      (uschar *)"\\p{Lu}",    /* upper */
282      (uschar *)"\\p{Xan}",   /* alnum */
283      NULL,                   /* ascii */
284      (uschar *)"\\h",        /* blank */
285      NULL,                   /* cntrl */
286      (uschar *)"\\p{Nd}",    /* digit */
287      NULL,                   /* graph */
288      NULL,                   /* print */
289      NULL,                   /* punct */
290      (uschar *)"\\p{Xps}",   /* space */    /* NOTE: Xps is POSIX space */
291      (uschar *)"\\p{Xwd}",   /* word */
292      NULL,                   /* xdigit */
293      /* Negated cases */
294      (uschar *)"\\P{L}",     /* ^alpha */
295      (uschar *)"\\P{Ll}",    /* ^lower */
296      (uschar *)"\\P{Lu}",    /* ^upper */
297      (uschar *)"\\P{Xan}",   /* ^alnum */
298      NULL,                   /* ^ascii */
299      (uschar *)"\\H",        /* ^blank */
300      NULL,                   /* ^cntrl */
301      (uschar *)"\\P{Nd}",    /* ^digit */
302      NULL,                   /* ^graph */
303      NULL,                   /* ^print */
304      NULL,                   /* ^punct */
305      (uschar *)"\\P{Xps}",   /* ^space */   /* NOTE: Xps is POSIX space */
306      (uschar *)"\\P{Xwd}",   /* ^word */
307      NULL                    /* ^xdigit */
308    };
309    #define POSIX_SUBSIZE (sizeof(posix_substitutes)/sizeof(uschar *))
310    #endif
311    
312  #define STRING(a)  # a  #define STRING(a)  # a
313  #define XSTRING(s) STRING(s)  #define XSTRING(s) STRING(s)
# Line 319  static const char error_texts[] = Line 371  static const char error_texts[] =
371    /* 35 */    /* 35 */
372    "invalid condition (?(0)\0"    "invalid condition (?(0)\0"
373    "\\C not allowed in lookbehind assertion\0"    "\\C not allowed in lookbehind assertion\0"
374    "PCRE does not support \\L, \\l, \\N, \\U, or \\u\0"    "PCRE does not support \\L, \\l, \\N{name}, \\U, or \\u\0"
375    "number after (?C is > 255\0"    "number after (?C is > 255\0"
376    "closing ) for (?C expected\0"    "closing ) for (?C expected\0"
377    /* 40 */    /* 40 */
# Line 345  static const char error_texts[] = Line 397  static const char error_texts[] =
397    "inconsistent NEWLINE options\0"    "inconsistent NEWLINE options\0"
398    "\\g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number\0"    "\\g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number\0"
399    "a numbered reference must not be zero\0"    "a numbered reference must not be zero\0"
400    "(*VERB) with an argument is not supported\0"    "an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)\0"
401    /* 60 */    /* 60 */
402    "(*VERB) not recognized\0"    "(*VERB) not recognized\0"
403    "number is too big\0"    "number is too big\0"
# Line 353  static const char error_texts[] = Line 405  static const char error_texts[] =
405    "digit expected after (?+\0"    "digit expected after (?+\0"
406    "] is an invalid data character in JavaScript compatibility mode\0"    "] is an invalid data character in JavaScript compatibility mode\0"
407    /* 65 */    /* 65 */
408    "different names for subpatterns of the same number are not allowed\0";    "different names for subpatterns of the same number are not allowed\0"
409      "(*MARK) must have an argument\0"
410      "this version of PCRE is not compiled with PCRE_UCP support\0"
411      ;
412    
413  /* Table to identify digits and hex digits. This is used when compiling  /* Table to identify digits and hex digits. This is used when compiling
414  patterns. Note that the tables in chartables are dependent on the locale, and  patterns. Note that the tables in chartables are dependent on the locale, and
# Line 586  else Line 641  else
641    
642      case CHAR_l:      case CHAR_l:
643      case CHAR_L:      case CHAR_L:
     case CHAR_N:  
644      case CHAR_u:      case CHAR_u:
645      case CHAR_U:      case CHAR_U:
646      *errorcodeptr = ERR37;      *errorcodeptr = ERR37;
# Line 824  else Line 878  else
878      }      }
879    }    }
880    
881    /* Perl supports \N{name} for character names, as well as plain \N for "not
882    newline". PCRE does not support \N{name}. */
883    
884    if (c == -ESC_N && ptr[1] == CHAR_LEFT_CURLY_BRACKET)
885      *errorcodeptr = ERR37;
886    
887    /* If PCRE_UCP is set, we change the values for \d etc. */
888    
889    if ((options & PCRE_UCP) != 0 && c <= -ESC_D && c >= -ESC_w)
890      c -= (ESC_DU - ESC_D);
891    
892    /* Set the pointer to the final character before returning. */
893    
894  *ptrptr = ptr;  *ptrptr = ptr;
895  return c;  return c;
896  }  }
# Line 1615  for (;;) Line 1682  for (;;)
1682    
1683    /* Otherwise, we can get the item's length from the table, except that for    /* Otherwise, we can get the item's length from the table, except that for
1684    repeated character types, we have to test for \p and \P, which have an extra    repeated character types, we have to test for \p and \P, which have an extra
1685    two bytes of parameters. */    two bytes of parameters, and for MARK/PRUNE/SKIP/THEN with an argument, we
1686      must add in its length. */
1687    
1688    else    else
1689      {      {
# Line 1639  for (;;) Line 1707  for (;;)
1707        case OP_TYPEPOSUPTO:        case OP_TYPEPOSUPTO:
1708        if (code[3] == OP_PROP || code[3] == OP_NOTPROP) code += 2;        if (code[3] == OP_PROP || code[3] == OP_NOTPROP) code += 2;
1709        break;        break;
1710    
1711          case OP_MARK:
1712          case OP_PRUNE_ARG:
1713          case OP_SKIP_ARG:
1714          case OP_THEN_ARG:
1715          code += code[1];
1716          break;
1717        }        }
1718    
1719      /* Add in the fixed length from the table */      /* Add in the fixed length from the table */
# Line 1710  for (;;) Line 1785  for (;;)
1785    
1786    /* Otherwise, we can get the item's length from the table, except that for    /* Otherwise, we can get the item's length from the table, except that for
1787    repeated character types, we have to test for \p and \P, which have an extra    repeated character types, we have to test for \p and \P, which have an extra
1788    two bytes of parameters. */    two bytes of parameters, and for MARK/PRUNE/SKIP/THEN with an argument, we
1789      must add in its length. */
1790    
1791    else    else
1792      {      {
# Line 1734  for (;;) Line 1810  for (;;)
1810        case OP_TYPEEXACT:        case OP_TYPEEXACT:
1811        if (code[3] == OP_PROP || code[3] == OP_NOTPROP) code += 2;        if (code[3] == OP_PROP || code[3] == OP_NOTPROP) code += 2;
1812        break;        break;
1813    
1814          case OP_MARK:
1815          case OP_PRUNE_ARG:
1816          case OP_SKIP_ARG:
1817          case OP_THEN_ARG:
1818          code += code[1];
1819          break;
1820        }        }
1821    
1822      /* Add in the fixed length from the table */      /* Add in the fixed length from the table */
# Line 2003  for (code = first_significant_code(code Line 2086  for (code = first_significant_code(code
2086      break;      break;
2087  #endif  #endif
2088    
2089        /* MARK, and PRUNE/SKIP/THEN with an argument must skip over the argument
2090        string. */
2091    
2092        case OP_MARK:
2093        case OP_PRUNE_ARG:
2094        case OP_SKIP_ARG:
2095        case OP_THEN_ARG:
2096        code += code[1];
2097        break;
2098    
2099      /* None of the remaining opcodes are required to match a character. */      /* None of the remaining opcodes are required to match a character. */
2100    
2101      default:      default:
# Line 2223  auto_callout(uschar *code, const uschar Line 2316  auto_callout(uschar *code, const uschar
2316  {  {
2317  *code++ = OP_CALLOUT;  *code++ = OP_CALLOUT;
2318  *code++ = 255;  *code++ = 255;
2319  PUT(code, 0, ptr - cd->start_pattern);  /* Pattern offset */  PUT(code, 0, (int)(ptr - cd->start_pattern));  /* Pattern offset */
2320  PUT(code, LINK_SIZE, 0);                /* Default length */  PUT(code, LINK_SIZE, 0);                       /* Default length */
2321  return code + 2*LINK_SIZE;  return code + 2*LINK_SIZE;
2322  }  }
2323    
# Line 2249  Returns: nothing Line 2342  Returns: nothing
2342  static void  static void
2343  complete_callout(uschar *previous_callout, const uschar *ptr, compile_data *cd)  complete_callout(uschar *previous_callout, const uschar *ptr, compile_data *cd)
2344  {  {
2345  int length = ptr - cd->start_pattern - GET(previous_callout, 2);  int length = (int)(ptr - cd->start_pattern - GET(previous_callout, 2));
2346  PUT(previous_callout, 2 + LINK_SIZE, length);  PUT(previous_callout, 2 + LINK_SIZE, length);
2347  }  }
2348    
# Line 2451  if (next >= 0) switch(op_code) Line 2544  if (next >= 0) switch(op_code)
2544    else    else
2545  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF8 */
2546    return (item == cd->fcc[next]);  /* Non-UTF-8 mode */    return (item == cd->fcc[next]);  /* Non-UTF-8 mode */
2547    
2548      /* Note that OP_DIGIT etc. are generated only when PCRE_UCP is *not* set.
2549      When it is set, \d etc. are converted into OP_(NOT_)PROP codes. */
2550    
2551    case OP_DIGIT:    case OP_DIGIT:
2552    return next > 127 || (cd->ctypes[next] & ctype_digit) == 0;    return next > 127 || (cd->ctypes[next] & ctype_digit) == 0;
# Line 2493  if (next >= 0) switch(op_code) Line 2589  if (next >= 0) switch(op_code)
2589      case 0x202f:      case 0x202f:
2590      case 0x205f:      case 0x205f:
2591      case 0x3000:      case 0x3000:
2592      return op_code != OP_HSPACE;      return op_code == OP_NOT_HSPACE;
2593      default:      default:
2594      return op_code == OP_HSPACE;      return op_code != OP_NOT_HSPACE;
2595      }      }
2596    
2597      case OP_ANYNL:
2598    case OP_VSPACE:    case OP_VSPACE:
2599    case OP_NOT_VSPACE:    case OP_NOT_VSPACE:
2600    switch(next)    switch(next)
# Line 2509  if (next >= 0) switch(op_code) Line 2606  if (next >= 0) switch(op_code)
2606      case 0x85:      case 0x85:
2607      case 0x2028:      case 0x2028:
2608      case 0x2029:      case 0x2029:
2609      return op_code != OP_VSPACE;      return op_code == OP_NOT_VSPACE;
2610      default:      default:
2611      return op_code == OP_VSPACE;      return op_code != OP_NOT_VSPACE;
2612      }      }
2613    
2614    default:    default:
# Line 2519  if (next >= 0) switch(op_code) Line 2616  if (next >= 0) switch(op_code)
2616    }    }
2617    
2618    
2619  /* Handle the case when the next item is \d, \s, etc. */  /* Handle the case when the next item is \d, \s, etc. Note that when PCRE_UCP
2620    is set, \d turns into ESC_du rather than ESC_d, etc., so ESC_d etc. are
2621    generated only when PCRE_UCP is *not* set, that is, when only ASCII
2622    characteristics are recognized. */
2623    
2624  switch(op_code)  switch(op_code)
2625    {    {
# Line 2598  switch(op_code) Line 2698  switch(op_code)
2698    
2699    case OP_DIGIT:    case OP_DIGIT:
2700    return next == -ESC_D || next == -ESC_s || next == -ESC_W ||    return next == -ESC_D || next == -ESC_s || next == -ESC_W ||
2701           next == -ESC_h || next == -ESC_v;           next == -ESC_h || next == -ESC_v || next == -ESC_R;
2702    
2703    case OP_NOT_DIGIT:    case OP_NOT_DIGIT:
2704    return next == -ESC_d;    return next == -ESC_d;
2705    
2706    case OP_WHITESPACE:    case OP_WHITESPACE:
2707    return next == -ESC_S || next == -ESC_d || next == -ESC_w;    return next == -ESC_S || next == -ESC_d || next == -ESC_w || next == -ESC_R;
2708    
2709    case OP_NOT_WHITESPACE:    case OP_NOT_WHITESPACE:
2710    return next == -ESC_s || next == -ESC_h || next == -ESC_v;    return next == -ESC_s || next == -ESC_h || next == -ESC_v;
2711    
2712    case OP_HSPACE:    case OP_HSPACE:
2713    return next == -ESC_S || next == -ESC_H || next == -ESC_d || next == -ESC_w;    return next == -ESC_S || next == -ESC_H || next == -ESC_d ||
2714             next == -ESC_w || next == -ESC_v || next == -ESC_R;
2715    
2716    case OP_NOT_HSPACE:    case OP_NOT_HSPACE:
2717    return next == -ESC_h;    return next == -ESC_h;
2718    
2719    /* Can't have \S in here because VT matches \S (Perl anomaly) */    /* Can't have \S in here because VT matches \S (Perl anomaly) */
2720      case OP_ANYNL:
2721    case OP_VSPACE:    case OP_VSPACE:
2722    return next == -ESC_V || next == -ESC_d || next == -ESC_w;    return next == -ESC_V || next == -ESC_d || next == -ESC_w;
2723    
2724    case OP_NOT_VSPACE:    case OP_NOT_VSPACE:
2725    return next == -ESC_v;    return next == -ESC_v || next == -ESC_R;
2726    
2727    case OP_WORDCHAR:    case OP_WORDCHAR:
2728    return next == -ESC_W || next == -ESC_s || next == -ESC_h || next == -ESC_v;    return next == -ESC_W || next == -ESC_s || next == -ESC_h ||
2729             next == -ESC_v || next == -ESC_R;
2730    
2731    case OP_NOT_WORDCHAR:    case OP_NOT_WORDCHAR:
2732    return next == -ESC_w || next == -ESC_d;    return next == -ESC_w || next == -ESC_d;
# Line 2687  BOOL inescq = FALSE; Line 2790  BOOL inescq = FALSE;
2790  BOOL groupsetfirstbyte = FALSE;  BOOL groupsetfirstbyte = FALSE;
2791  const uschar *ptr = *ptrptr;  const uschar *ptr = *ptrptr;
2792  const uschar *tempptr;  const uschar *tempptr;
2793    const uschar *nestptr = NULL;
2794  uschar *previous = NULL;  uschar *previous = NULL;
2795  uschar *previous_callout = NULL;  uschar *previous_callout = NULL;
2796  uschar *save_hwm = NULL;  uschar *save_hwm = NULL;
# Line 2757  for (;; ptr++) Line 2861  for (;; ptr++)
2861    
2862    c = *ptr;    c = *ptr;
2863    
2864      /* If we are at the end of a nested substitution, revert to the outer level
2865      string. Nesting only happens one level deep. */
2866    
2867      if (c == 0 && nestptr != NULL)
2868        {
2869        ptr = nestptr;
2870        nestptr = NULL;
2871        c = *ptr;
2872        }
2873    
2874    /* If we are in the pre-compile phase, accumulate the length used for the    /* If we are in the pre-compile phase, accumulate the length used for the
2875    previous cycle of this loop. */    previous cycle of this loop. */
2876    
# Line 2787  for (;; ptr++) Line 2901  for (;; ptr++)
2901        goto FAILED;        goto FAILED;
2902        }        }
2903    
2904      *lengthptr += code - last_code;      *lengthptr += (int)(code - last_code);
2905      DPRINTF(("length=%d added %d c=%c\n", *lengthptr, code - last_code, c));      DPRINTF(("length=%d added %d c=%c\n", *lengthptr, code - last_code, c));
2906    
2907      /* If "previous" is set and it is not at the start of the work space, move      /* If "previous" is set and it is not at the start of the work space, move
# Line 2905  for (;; ptr++) Line 3019  for (;; ptr++)
3019          *errorcodeptr = ERR20;          *errorcodeptr = ERR20;
3020          goto FAILED;          goto FAILED;
3021          }          }
3022        *lengthptr += code - last_code;   /* To include callout length */        *lengthptr += (int)(code - last_code);   /* To include callout length */
3023        DPRINTF((">> end branch\n"));        DPRINTF((">> end branch\n"));
3024        }        }
3025      return TRUE;      return TRUE;
# Line 3054  for (;; ptr++) Line 3168  for (;; ptr++)
3168          {                           /* Braces are required because the */          {                           /* Braces are required because the */
3169          GETCHARLEN(c, ptr, ptr);    /* macro generates multiple statements */          GETCHARLEN(c, ptr, ptr);    /* macro generates multiple statements */
3170          }          }
3171    
3172        /* In the pre-compile phase, accumulate the length of any UTF-8 extra        /* In the pre-compile phase, accumulate the length of any UTF-8 extra
3173        data and reset the pointer. This is so that very large classes that        data and reset the pointer. This is so that very large classes that
3174        contain a zillion UTF-8 characters no longer overwrite the work space        contain a zillion UTF-8 characters no longer overwrite the work space
# Line 3110  for (;; ptr++) Line 3224  for (;; ptr++)
3224            ptr++;            ptr++;
3225            }            }
3226    
3227          posix_class = check_posix_name(ptr, tempptr - ptr);          posix_class = check_posix_name(ptr, (int)(tempptr - ptr));
3228          if (posix_class < 0)          if (posix_class < 0)
3229            {            {
3230            *errorcodeptr = ERR30;            *errorcodeptr = ERR30;
# Line 3123  for (;; ptr++) Line 3237  for (;; ptr++)
3237    
3238          if ((options & PCRE_CASELESS) != 0 && posix_class <= 2)          if ((options & PCRE_CASELESS) != 0 && posix_class <= 2)
3239            posix_class = 0;            posix_class = 0;
3240    
3241          /* We build the bit map for the POSIX class in a chunk of local store          /* When PCRE_UCP is set, some of the POSIX classes are converted to
3242          because we may be adding and subtracting from it, and we don't want to          different escape sequences that use Unicode properties. */
3243          subtract bits that may be in the main map already. At the end we or the  
3244          result into the bit map that is being built. */  #ifdef SUPPORT_UCP
3245            if ((options & PCRE_UCP) != 0)
3246              {
3247              int pc = posix_class + ((local_negate)? POSIX_SUBSIZE/2 : 0);
3248              if (posix_substitutes[pc] != NULL)
3249                {
3250                nestptr = tempptr + 1;
3251                ptr = posix_substitutes[pc] - 1;
3252                continue;
3253                }
3254              }
3255    #endif
3256            /* In the non-UCP case, we build the bit map for the POSIX class in a
3257            chunk of local store because we may be adding and subtracting from it,
3258            and we don't want to subtract bits that may be in the main map already.
3259            At the end we or the result into the bit map that is being built. */
3260    
3261          posix_class *= 3;          posix_class *= 3;
3262    
# Line 3171  for (;; ptr++) Line 3300  for (;; ptr++)
3300    
3301        /* Backslash may introduce a single character, or it may introduce one        /* Backslash may introduce a single character, or it may introduce one
3302        of the specials, which just set a flag. The sequence \b is a special        of the specials, which just set a flag. The sequence \b is a special
3303        case. Inside a class (and only there) it is treated as backspace.        case. Inside a class (and only there) it is treated as backspace. We
3304        Elsewhere it marks a word boundary. Other escapes have preset maps ready        assume that other escapes have more than one character in them, so set
3305        to 'or' into the one we are building. We assume they have more than one        class_charcount bigger than one. Unrecognized escapes fall through and
3306        character in them, so set class_charcount bigger than one. */        are either treated as literal characters (by default), or are faulted if
3307          PCRE_EXTRA is set. */
3308    
3309        if (c == CHAR_BACKSLASH)        if (c == CHAR_BACKSLASH)
3310          {          {
3311          c = check_escape(&ptr, errorcodeptr, cd->bracount, options, TRUE);          c = check_escape(&ptr, errorcodeptr, cd->bracount, options, TRUE);
3312          if (*errorcodeptr != 0) goto FAILED;          if (*errorcodeptr != 0) goto FAILED;
3313    
3314          if (-c == ESC_b) c = CHAR_BS;       /* \b is backspace in a class */          if (-c == ESC_b) c = CHAR_BS;    /* \b is backspace in a class */
         else if (-c == ESC_X) c = CHAR_X;   /* \X is literal X in a class */  
         else if (-c == ESC_R) c = CHAR_R;   /* \R is literal R in a class */  
3315          else if (-c == ESC_Q)            /* Handle start of quoted string */          else if (-c == ESC_Q)            /* Handle start of quoted string */
3316            {            {
3317            if (ptr[1] == CHAR_BACKSLASH && ptr[2] == CHAR_E)            if (ptr[1] == CHAR_BACKSLASH && ptr[2] == CHAR_E)
# Line 3200  for (;; ptr++) Line 3328  for (;; ptr++)
3328            register const uschar *cbits = cd->cbits;            register const uschar *cbits = cd->cbits;
3329            class_charcount += 2;     /* Greater than 1 is what matters */            class_charcount += 2;     /* Greater than 1 is what matters */
3330    
3331            /* Save time by not doing this in the pre-compile phase. */            switch (-c)
   
           if (lengthptr == NULL) switch (-c)  
3332              {              {
3333    #ifdef SUPPORT_UCP
3334                case ESC_du:     /* These are the values given for \d etc */
3335                case ESC_DU:     /* when PCRE_UCP is set. We replace the */
3336                case ESC_wu:     /* escape sequence with an appropriate \p */
3337                case ESC_WU:     /* or \P to test Unicode properties instead */
3338                case ESC_su:     /* of the default ASCII testing. */
3339                case ESC_SU:
3340                nestptr = ptr;
3341                ptr = substitutes[-c - ESC_DU] - 1;  /* Just before substitute */
3342                class_charcount -= 2;                /* Undo! */
3343                continue;
3344    #endif
3345              case ESC_d:              case ESC_d:
3346              for (c = 0; c < 32; c++) classbits[c] |= cbits[c+cbit_digit];              for (c = 0; c < 32; c++) classbits[c] |= cbits[c+cbit_digit];
3347              continue;              continue;
# Line 3233  for (;; ptr++) Line 3371  for (;; ptr++)
3371              classbits[1] |= 0x08;    /* Perl 5.004 onwards omits VT from \s */              classbits[1] |= 0x08;    /* Perl 5.004 onwards omits VT from \s */
3372              continue;              continue;
3373    
3374              default:    /* Not recognized; fall through */              case ESC_h:
             break;      /* Need "default" setting to stop compiler warning. */  
             }  
   
           /* In the pre-compile phase, just do the recognition. */  
   
           else if (c == -ESC_d || c == -ESC_D || c == -ESC_w ||  
                    c == -ESC_W || c == -ESC_s || c == -ESC_S) continue;  
   
           /* We need to deal with \H, \h, \V, and \v in both phases because  
           they use extra memory. */  
   
           if (-c == ESC_h)  
             {  
3375              SETBIT(classbits, 0x09); /* VT */              SETBIT(classbits, 0x09); /* VT */
3376              SETBIT(classbits, 0x20); /* SPACE */              SETBIT(classbits, 0x20); /* SPACE */
3377              SETBIT(classbits, 0xa0); /* NSBP */              SETBIT(classbits, 0xa0); /* NSBP */
# Line 3270  for (;; ptr++) Line 3395  for (;; ptr++)
3395                }                }
3396  #endif  #endif
3397              continue;              continue;
             }  
3398    
3399            if (-c == ESC_H)              case ESC_H:
             {  
3400              for (c = 0; c < 32; c++)              for (c = 0; c < 32; c++)
3401                {                {
3402                int x = 0xff;                int x = 0xff;
# Line 3315  for (;; ptr++) Line 3438  for (;; ptr++)
3438                }                }
3439  #endif  #endif
3440              continue;              continue;
             }  
3441    
3442            if (-c == ESC_v)              case ESC_v:
             {  
3443              SETBIT(classbits, 0x0a); /* LF */              SETBIT(classbits, 0x0a); /* LF */
3444              SETBIT(classbits, 0x0b); /* VT */              SETBIT(classbits, 0x0b); /* VT */
3445              SETBIT(classbits, 0x0c); /* FF */              SETBIT(classbits, 0x0c); /* FF */
# Line 3334  for (;; ptr++) Line 3455  for (;; ptr++)
3455                }                }
3456  #endif  #endif
3457              continue;              continue;
             }  
3458    
3459            if (-c == ESC_V)              case ESC_V:
             {  
3460              for (c = 0; c < 32; c++)              for (c = 0; c < 32; c++)
3461                {                {
3462                int x = 0xff;                int x = 0xff;
# Line 3367  for (;; ptr++) Line 3486  for (;; ptr++)
3486                }                }
3487  #endif  #endif
3488              continue;              continue;
             }  
   
           /* We need to deal with \P and \p in both phases. */  
3489    
3490  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3491            if (-c == ESC_p || -c == ESC_P)              case ESC_p:
3492              {              case ESC_P:
3493              BOOL negated;                {
3494              int pdata;                BOOL negated;
3495              int ptype = get_ucp(&ptr, &negated, &pdata, errorcodeptr);                int pdata;
3496              if (ptype < 0) goto FAILED;                int ptype = get_ucp(&ptr, &negated, &pdata, errorcodeptr);
3497              class_utf8 = TRUE;                if (ptype < 0) goto FAILED;
3498              *class_utf8data++ = ((-c == ESC_p) != negated)?                class_utf8 = TRUE;
3499                XCL_PROP : XCL_NOTPROP;                *class_utf8data++ = ((-c == ESC_p) != negated)?
3500              *class_utf8data++ = ptype;                  XCL_PROP : XCL_NOTPROP;
3501              *class_utf8data++ = pdata;                *class_utf8data++ = ptype;
3502              class_charcount -= 2;   /* Not a < 256 character */                *class_utf8data++ = pdata;
3503              continue;                class_charcount -= 2;   /* Not a < 256 character */
3504              }                continue;
3505                  }
3506  #endif  #endif
3507            /* Unrecognized escapes are faulted if PCRE is running in its              /* Unrecognized escapes are faulted if PCRE is running in its
3508            strict mode. By default, for compatibility with Perl, they are              strict mode. By default, for compatibility with Perl, they are
3509            treated as literals. */              treated as literals. */
3510    
3511            if ((options & PCRE_EXTRA) != 0)              default:
3512              {              if ((options & PCRE_EXTRA) != 0)
3513              *errorcodeptr = ERR7;                {
3514              goto FAILED;                *errorcodeptr = ERR7;
3515                  goto FAILED;
3516                  }
3517                class_charcount -= 2;  /* Undo the default count from above */
3518                c = *ptr;              /* Get the final character and fall through */
3519                break;
3520              }              }
   
           class_charcount -= 2;  /* Undo the default count from above */  
           c = *ptr;              /* Get the final character and fall through */  
3521            }            }
3522    
3523          /* Fall through if we have a single character (c >= 0). This may be          /* Fall through if we have a single character (c >= 0). This may be
# Line 3468  for (;; ptr++) Line 3587  for (;; ptr++)
3587            d = check_escape(&ptr, errorcodeptr, cd->bracount, options, TRUE);            d = check_escape(&ptr, errorcodeptr, cd->bracount, options, TRUE);
3588            if (*errorcodeptr != 0) goto FAILED;            if (*errorcodeptr != 0) goto FAILED;
3589    
3590            /* \b is backspace; \X is literal X; \R is literal R; any other            /* \b is backspace; any other special means the '-' was literal */
           special means the '-' was literal */  
3591    
3592            if (d < 0)            if (d < 0)
3593              {              {
3594              if (d == -ESC_b) d = CHAR_BS;              if (d == -ESC_b) d = CHAR_BS; else
             else if (d == -ESC_X) d = CHAR_X;  
             else if (d == -ESC_R) d = CHAR_R; else  
3595                {                {
3596                ptr = oldptr;                ptr = oldptr;
3597                goto LONE_SINGLE_CHARACTER;  /* A few lines below */                goto LONE_SINGLE_CHARACTER;  /* A few lines below */
# Line 3641  for (;; ptr++) Line 3757  for (;; ptr++)
3757          }          }
3758        }        }
3759    
3760      /* Loop until ']' reached. This "while" is the end of the "do" above. */      /* Loop until ']' reached. This "while" is the end of the "do" far above.
3761        If we are at the end of an internal nested string, revert to the outer
3762        string. */
3763    
3764        while (((c = *(++ptr)) != 0 ||
3765               (nestptr != NULL &&
3766                 (ptr = nestptr, nestptr = NULL, c = *(++ptr)) != 0)) &&
3767               (c != CHAR_RIGHT_SQUARE_BRACKET || inescq));
3768    
3769      while ((c = *(++ptr)) != 0 && (c != CHAR_RIGHT_SQUARE_BRACKET || inescq));      /* Check for missing terminating ']' */
3770    
3771      if (c == 0)                          /* Missing terminating ']' */      if (c == 0)
3772        {        {
3773        *errorcodeptr = ERR6;        *errorcodeptr = ERR6;
3774        goto FAILED;        goto FAILED;
3775        }        }
3776    
   
 /* This code has been disabled because it would mean that \s counts as  
 an explicit \r or \n reference, and that's not really what is wanted. Now  
 we set the flag only if there is a literal "\r" or "\n" in the class. */  
   
 #if 0  
     /* Remember whether \r or \n are in this class */  
   
     if (negate_class)  
       {  
       if ((classbits[1] & 0x24) != 0x24) cd->external_flags |= PCRE_HASCRORLF;  
       }  
     else  
       {  
       if ((classbits[1] & 0x24) != 0) cd->external_flags |= PCRE_HASCRORLF;  
       }  
 #endif  
   
   
3777      /* If class_charcount is 1, we saw precisely one character whose value is      /* If class_charcount is 1, we saw precisely one character whose value is
3778      less than 256. As long as there were no characters >= 128 and there was no      less than 256. As long as there were no characters >= 128 and there was no
3779      use of \p or \P, in other words, no use of any XCLASS features, we can      use of \p or \P, in other words, no use of any XCLASS features, we can
# Line 3686  we set the flag only if there is a liter Line 3790  we set the flag only if there is a liter
3790      can cause firstbyte to be set. Otherwise, there can be no first char if      can cause firstbyte to be set. Otherwise, there can be no first char if
3791      this item is first, whatever repeat count may follow. In the case of      this item is first, whatever repeat count may follow. In the case of
3792      reqbyte, save the previous value for reinstating. */      reqbyte, save the previous value for reinstating. */
3793    
3794  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
3795      if (class_charcount == 1 && !class_utf8 &&      if (class_charcount == 1 && !class_utf8 &&
3796        (!utf8 || !negate_class || class_lastchar < 128))        (!utf8 || !negate_class || class_lastchar < 128))
# Line 3733  we set the flag only if there is a liter Line 3837  we set the flag only if there is a liter
3837    
3838      /* If there are characters with values > 255, we have to compile an      /* If there are characters with values > 255, we have to compile an
3839      extended class, with its own opcode, unless there was a negated special      extended class, with its own opcode, unless there was a negated special
3840      such as \S in the class, because in that case all characters > 255 are in      such as \S in the class, and PCRE_UCP is not set, because in that case all
3841      the class, so any that were explicitly given as well can be ignored. If      characters > 255 are in the class, so any that were explicitly given as
3842      (when there are explicit characters > 255 that must be listed) there are no      well can be ignored. If (when there are explicit characters > 255 that must
3843      characters < 256, we can omit the bitmap in the actual compiled code. */      be listed) there are no characters < 256, we can omit the bitmap in the
3844        actual compiled code. */
3845    
3846  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
3847      if (class_utf8 && !should_flip_negation)      if (class_utf8 && (!should_flip_negation || (options & PCRE_UCP) != 0))
3848        {        {
3849        *class_utf8data++ = XCL_END;    /* Marks the end of extra data */        *class_utf8data++ = XCL_END;    /* Marks the end of extra data */
3850        *code++ = OP_XCLASS;        *code++ = OP_XCLASS;
# Line 3765  we set the flag only if there is a liter Line 3870  we set the flag only if there is a liter
3870        }        }
3871  #endif  #endif
3872    
3873      /* If there are no characters > 255, set the opcode to OP_CLASS or      /* If there are no characters > 255, or they are all to be included or
3874      OP_NCLASS, depending on whether the whole class was negated and whether      excluded, set the opcode to OP_CLASS or OP_NCLASS, depending on whether the
3875      there were negative specials such as \S in the class. Then copy the 32-byte      whole class was negated and whether there were negative specials such as \S
3876      map into the code vector, negating it if necessary. */      (non-UCP) in the class. Then copy the 32-byte map into the code vector,
3877        negating it if necessary. */
3878    
3879      *code++ = (negate_class == should_flip_negation) ? OP_CLASS : OP_NCLASS;      *code++ = (negate_class == should_flip_negation) ? OP_CLASS : OP_NCLASS;
3880      if (negate_class)      if (negate_class)
# Line 4148  we set the flag only if there is a liter Line 4254  we set the flag only if there is a liter
4254        {        {
4255        register int i;        register int i;
4256        int ketoffset = 0;        int ketoffset = 0;
4257        int len = code - previous;        int len = (int)(code - previous);
4258        uschar *bralink = NULL;        uschar *bralink = NULL;
4259    
4260        /* Repeating a DEFINE group is pointless */        /* Repeating a DEFINE group is pointless */
# Line 4169  we set the flag only if there is a liter Line 4275  we set the flag only if there is a liter
4275          {          {
4276          register uschar *ket = previous;          register uschar *ket = previous;
4277          do ket += GET(ket, 1); while (*ket != OP_KET);          do ket += GET(ket, 1); while (*ket != OP_KET);
4278          ketoffset = code - ket;          ketoffset = (int)(code - ket);
4279          }          }
4280    
4281        /* The case of a zero minimum is special because of the need to stick        /* The case of a zero minimum is special because of the need to stick
# Line 4237  we set the flag only if there is a liter Line 4343  we set the flag only if there is a liter
4343            /* We chain together the bracket offset fields that have to be            /* We chain together the bracket offset fields that have to be
4344            filled in later when the ends of the brackets are reached. */            filled in later when the ends of the brackets are reached. */
4345    
4346            offset = (bralink == NULL)? 0 : previous - bralink;            offset = (bralink == NULL)? 0 : (int)(previous - bralink);
4347            bralink = previous;            bralink = previous;
4348            PUTINC(previous, 0, offset);            PUTINC(previous, 0, offset);
4349            }            }
# Line 4346  we set the flag only if there is a liter Line 4452  we set the flag only if there is a liter
4452              {              {
4453              int offset;              int offset;
4454              *code++ = OP_BRA;              *code++ = OP_BRA;
4455              offset = (bralink == NULL)? 0 : code - bralink;              offset = (bralink == NULL)? 0 : (int)(code - bralink);
4456              bralink = code;              bralink = code;
4457              PUTINC(code, 0, offset);              PUTINC(code, 0, offset);
4458              }              }
# Line 4367  we set the flag only if there is a liter Line 4473  we set the flag only if there is a liter
4473          while (bralink != NULL)          while (bralink != NULL)
4474            {            {
4475            int oldlinkoffset;            int oldlinkoffset;
4476            int offset = code - bralink + 1;            int offset = (int)(code - bralink + 1);
4477            uschar *bra = code - offset;            uschar *bra = code - offset;
4478            oldlinkoffset = GET(bra, 1);            oldlinkoffset = GET(bra, 1);
4479            bralink = (oldlinkoffset == 0)? NULL : bralink - oldlinkoffset;            bralink = (oldlinkoffset == 0)? NULL : bralink - oldlinkoffset;
# Line 4455  we set the flag only if there is a liter Line 4561  we set the flag only if there is a liter
4561  #endif  #endif
4562          }          }
4563    
4564        len = code - tempcode;        len = (int)(code - tempcode);
4565        if (len > 0) switch (*tempcode)        if (len > 0) switch (*tempcode)
4566          {          {
4567          case OP_STAR:  *tempcode = OP_POSSTAR; break;          case OP_STAR:  *tempcode = OP_POSSTAR; break;
# Line 4514  we set the flag only if there is a liter Line 4620  we set the flag only if there is a liter
4620    
4621      /* First deal with various "verbs" that can be introduced by '*'. */      /* First deal with various "verbs" that can be introduced by '*'. */
4622    
4623      if (*(++ptr) == CHAR_ASTERISK && (cd->ctypes[ptr[1]] & ctype_letter) != 0)      if (*(++ptr) == CHAR_ASTERISK &&
4624             ((cd->ctypes[ptr[1]] & ctype_letter) != 0 || ptr[1] == ':'))
4625        {        {
4626        int i, namelen;        int i, namelen;
4627          int arglen = 0;
4628        const char *vn = verbnames;        const char *vn = verbnames;
4629        const uschar *name = ++ptr;        const uschar *name = ptr + 1;
4630          const uschar *arg = NULL;
4631        previous = NULL;        previous = NULL;
4632        while ((cd->ctypes[*++ptr] & ctype_letter) != 0) {};        while ((cd->ctypes[*++ptr] & ctype_letter) != 0) {};
4633          namelen = (int)(ptr - name);
4634    
4635        if (*ptr == CHAR_COLON)        if (*ptr == CHAR_COLON)
4636          {          {
4637          *errorcodeptr = ERR59;   /* Not supported */          arg = ++ptr;
4638          goto FAILED;          while ((cd->ctypes[*ptr] & (ctype_letter|ctype_digit)) != 0
4639              || *ptr == '_') ptr++;
4640            arglen = ptr - arg;
4641          }          }
4642    
4643        if (*ptr != CHAR_RIGHT_PARENTHESIS)        if (*ptr != CHAR_RIGHT_PARENTHESIS)
4644          {          {
4645          *errorcodeptr = ERR60;          *errorcodeptr = ERR60;
4646          goto FAILED;          goto FAILED;
4647          }          }
4648        namelen = ptr - name;  
4649          /* Scan the table of verb names */
4650    
4651        for (i = 0; i < verbcount; i++)        for (i = 0; i < verbcount; i++)
4652          {          {
4653          if (namelen == verbs[i].len &&          if (namelen == verbs[i].len &&
# Line 4549  we set the flag only if there is a liter Line 4665  we set the flag only if there is a liter
4665                PUT2INC(code, 0, oc->number);                PUT2INC(code, 0, oc->number);
4666                }                }
4667              }              }
4668            *code++ = verbs[i].op;  
4669            break;            /* Handle the cases with/without an argument */
4670    
4671              if (arglen == 0)
4672                {
4673                if (verbs[i].op < 0)   /* Argument is mandatory */
4674                  {
4675                  *errorcodeptr = ERR66;
4676                  goto FAILED;
4677                  }
4678                *code++ = verbs[i].op;
4679                }
4680    
4681              else
4682                {
4683                if (verbs[i].op_arg < 0)   /* Argument is forbidden */
4684                  {
4685                  *errorcodeptr = ERR59;
4686                  goto FAILED;
4687                  }
4688                *code++ = verbs[i].op_arg;
4689                *code++ = arglen;
4690                memcpy(code, arg, arglen);
4691                code += arglen;
4692                *code++ = 0;
4693                }
4694    
4695              break;  /* Found verb, exit loop */
4696            }            }
4697    
4698          vn += verbs[i].len + 1;          vn += verbs[i].len + 1;
4699          }          }
4700        if (i < verbcount) continue;  
4701        *errorcodeptr = ERR60;        if (i < verbcount) continue;    /* Successfully handled a verb */
4702          *errorcodeptr = ERR60;          /* Verb not recognized */
4703        goto FAILED;        goto FAILED;
4704        }        }
4705    
# Line 4674  we set the flag only if there is a liter Line 4818  we set the flag only if there is a liter
4818                recno * 10 + *ptr - CHAR_0 : -1;                recno * 10 + *ptr - CHAR_0 : -1;
4819            ptr++;            ptr++;
4820            }            }
4821          namelen = ptr - name;          namelen = (int)(ptr - name);
4822    
4823          if ((terminator > 0 && *ptr++ != terminator) ||          if ((terminator > 0 && *ptr++ != terminator) ||
4824              *ptr++ != CHAR_RIGHT_PARENTHESIS)              *ptr++ != CHAR_RIGHT_PARENTHESIS)
# Line 4870  we set the flag only if there is a liter Line 5014  we set the flag only if there is a liter
5014              goto FAILED;              goto FAILED;
5015              }              }
5016            *code++ = n;            *code++ = n;
5017            PUT(code, 0, ptr - cd->start_pattern + 1);  /* Pattern offset */            PUT(code, 0, (int)(ptr - cd->start_pattern + 1)); /* Pattern offset */
5018            PUT(code, LINK_SIZE, 0);                    /* Default length */            PUT(code, LINK_SIZE, 0);                          /* Default length */
5019            code += 2 * LINK_SIZE;            code += 2 * LINK_SIZE;
5020            }            }
5021          previous = NULL;          previous = NULL;
# Line 4904  we set the flag only if there is a liter Line 5048  we set the flag only if there is a liter
5048            name = ++ptr;            name = ++ptr;
5049    
5050            while ((cd->ctypes[*ptr] & ctype_word) != 0) ptr++;            while ((cd->ctypes[*ptr] & ctype_word) != 0) ptr++;
5051            namelen = ptr - name;            namelen = (int)(ptr - name);
5052    
5053            /* In the pre-compile phase, just do a syntax check. */            /* In the pre-compile phase, just do a syntax check. */
5054    
# Line 5034  we set the flag only if there is a liter Line 5178  we set the flag only if there is a liter
5178          NAMED_REF_OR_RECURSE:          NAMED_REF_OR_RECURSE:
5179          name = ++ptr;          name = ++ptr;
5180          while ((cd->ctypes[*ptr] & ctype_word) != 0) ptr++;          while ((cd->ctypes[*ptr] & ctype_word) != 0) ptr++;
5181          namelen = ptr - name;          namelen = (int)(ptr - name);
5182    
5183          /* In the pre-compile phase, do a syntax check and set a dummy          /* In the pre-compile phase, do a syntax check and set a dummy
5184          reference number. */          reference number. */
# Line 5203  we set the flag only if there is a liter Line 5347  we set the flag only if there is a liter
5347                of the group. */                of the group. */
5348    
5349                called = cd->start_code + recno;                called = cd->start_code + recno;
5350                PUTINC(cd->hwm, 0, code + 2 + LINK_SIZE - cd->start_code);                PUTINC(cd->hwm, 0, (int)(code + 2 + LINK_SIZE - cd->start_code));
5351                }                }
5352    
5353              /* If not a forward reference, and the subpattern is still open,              /* If not a forward reference, and the subpattern is still open,
# Line 5227  we set the flag only if there is a liter Line 5371  we set the flag only if there is a liter
5371            code += 1 + LINK_SIZE;            code += 1 + LINK_SIZE;
5372    
5373            *code = OP_RECURSE;            *code = OP_RECURSE;
5374            PUT(code, 1, called - cd->start_code);            PUT(code, 1, (int)(called - cd->start_code));
5375            code += 1 + LINK_SIZE;            code += 1 + LINK_SIZE;
5376    
5377            *code = OP_KET;            *code = OP_KET;
# Line 5338  we set the flag only if there is a liter Line 5482  we set the flag only if there is a liter
5482          }     /* End of switch for character following (? */          }     /* End of switch for character following (? */
5483        }       /* End of (? handling */        }       /* End of (? handling */
5484    
5485      /* Opening parenthesis not followed by '?'. If PCRE_NO_AUTO_CAPTURE is set,      /* Opening parenthesis not followed by '*' or '?'. If PCRE_NO_AUTO_CAPTURE
5486      all unadorned brackets become non-capturing and behave like (?:...)      is set, all unadorned brackets become non-capturing and behave like (?:...)
5487      brackets. */      brackets. */
5488    
5489      else if ((options & PCRE_NO_AUTO_CAPTURE) != 0)      else if ((options & PCRE_NO_AUTO_CAPTURE) != 0)
# Line 5531  we set the flag only if there is a liter Line 5675  we set the flag only if there is a liter
5675    
5676      /* ===================================================================*/      /* ===================================================================*/
5677      /* Handle metasequences introduced by \. For ones like \d, the ESC_ values      /* Handle metasequences introduced by \. For ones like \d, the ESC_ values
5678      are arranged to be the negation of the corresponding OP_values. For the      are arranged to be the negation of the corresponding OP_values in the
5679      back references, the values are ESC_REF plus the reference number. Only      default case when PCRE_UCP is not set. For the back references, the values
5680      back references and those types that consume a character may be repeated.      are ESC_REF plus the reference number. Only back references and those types
5681      We can test for values between ESC_b and ESC_Z for the latter; this may      that consume a character may be repeated. We can test for values between
5682      have to change if any new ones are ever created. */      ESC_b and ESC_Z for the latter; this may have to change if any new ones are
5683        ever created. */
5684    
5685      case CHAR_BACKSLASH:      case CHAR_BACKSLASH:
5686      tempptr = ptr;      tempptr = ptr;
# Line 5695  we set the flag only if there is a liter Line 5840  we set the flag only if there is a liter
5840  #endif  #endif
5841    
5842        /* For the rest (including \X when Unicode properties are supported), we        /* For the rest (including \X when Unicode properties are supported), we
5843        can obtain the OP value by negating the escape value. */        can obtain the OP value by negating the escape value in the default
5844          situation when PCRE_UCP is not set. When it *is* set, we substitute
5845          Unicode property tests. */
5846    
5847        else        else
5848          {          {
5849          previous = (-c > ESC_b && -c < ESC_Z)? code : NULL;  #ifdef SUPPORT_UCP
5850          *code++ = -c;          if (-c >= ESC_DU && -c <= ESC_wu)
5851              {
5852              nestptr = ptr + 1;                   /* Where to resume */
5853              ptr = substitutes[-c - ESC_DU] - 1;  /* Just before substitute */
5854              }
5855            else
5856    #endif
5857              {
5858              previous = (-c > ESC_b && -c < ESC_Z)? code : NULL;
5859              *code++ = -c;
5860              }
5861          }          }
5862        continue;        continue;
5863        }        }
# Line 6032  for (;;) Line 6189  for (;;)
6189      {      {
6190      if (lengthptr == NULL)      if (lengthptr == NULL)
6191        {        {
6192        int branch_length = code - last_branch;        int branch_length = (int)(code - last_branch);
6193        do        do
6194          {          {
6195          int prev_length = GET(last_branch, 1);          int prev_length = GET(last_branch, 1);
# Line 6046  for (;;) Line 6203  for (;;)
6203      /* Fill in the ket */      /* Fill in the ket */
6204    
6205      *code = OP_KET;      *code = OP_KET;
6206      PUT(code, 1, code - start_bracket);      PUT(code, 1, (int)(code - start_bracket));
6207      code += 1 + LINK_SIZE;      code += 1 + LINK_SIZE;
6208    
6209      /* If it was a capturing subpattern, check to see if it contained any      /* If it was a capturing subpattern, check to see if it contained any
# Line 6061  for (;;) Line 6218  for (;;)
6218            code - start_bracket);            code - start_bracket);
6219          *start_bracket = OP_ONCE;          *start_bracket = OP_ONCE;
6220          code += 1 + LINK_SIZE;          code += 1 + LINK_SIZE;
6221          PUT(start_bracket, 1, code - start_bracket);          PUT(start_bracket, 1, (int)(code - start_bracket));
6222          *code = OP_KET;          *code = OP_KET;
6223          PUT(code, 1, code - start_bracket);          PUT(code, 1, (int)(code - start_bracket));
6224          code += 1 + LINK_SIZE;          code += 1 + LINK_SIZE;
6225          length += 2 + 2*LINK_SIZE;          length += 2 + 2*LINK_SIZE;
6226          }          }
# Line 6118  for (;;) Line 6275  for (;;)
6275    else    else
6276      {      {
6277      *code = OP_ALT;      *code = OP_ALT;
6278      PUT(code, 1, code - last_branch);      PUT(code, 1, (int)(code - last_branch));
6279      bc.current_branch = last_branch = code;      bc.current_branch = last_branch = code;
6280      code += 1 + LINK_SIZE;      code += 1 + LINK_SIZE;
6281      }      }
# Line 6437  int length = 1; /* For final END opcode Line 6594  int length = 1; /* For final END opcode
6594  int firstbyte, reqbyte, newline;  int firstbyte, reqbyte, newline;
6595  int errorcode = 0;  int errorcode = 0;
6596  int skipatstart = 0;  int skipatstart = 0;
6597  BOOL utf8 = (options & PCRE_UTF8) != 0;  BOOL utf8;
6598  size_t size;  size_t size;
6599  uschar *code;  uschar *code;
6600  const uschar *codestart;  const uschar *codestart;
# Line 6507  while (ptr[skipatstart] == CHAR_LEFT_PAR Line 6664  while (ptr[skipatstart] == CHAR_LEFT_PAR
6664    
6665    if (strncmp((char *)(ptr+skipatstart+2), STRING_UTF8_RIGHTPAR, 5) == 0)    if (strncmp((char *)(ptr+skipatstart+2), STRING_UTF8_RIGHTPAR, 5) == 0)
6666      { skipatstart += 7; options |= PCRE_UTF8; continue; }      { skipatstart += 7; options |= PCRE_UTF8; continue; }
6667      else if (strncmp((char *)(ptr+skipatstart+2), STRING_UCP_RIGHTPAR, 4) == 0)
6668        { skipatstart += 6; options |= PCRE_UCP; continue; }
6669    
6670    if (strncmp((char *)(ptr+skipatstart+2), STRING_CR_RIGHTPAR, 3) == 0)    if (strncmp((char *)(ptr+skipatstart+2), STRING_CR_RIGHTPAR, 3) == 0)
6671      { skipatstart += 5; newnl = PCRE_NEWLINE_CR; }      { skipatstart += 5; newnl = PCRE_NEWLINE_CR; }
# Line 6530  while (ptr[skipatstart] == CHAR_LEFT_PAR Line 6689  while (ptr[skipatstart] == CHAR_LEFT_PAR
6689      options = (options & ~(PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) | newbsr;      options = (options & ~(PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) | newbsr;
6690    else break;    else break;
6691    }    }
6692    
6693    utf8 = (options & PCRE_UTF8) != 0;
6694    
6695  /* Can't support UTF8 unless PCRE has been compiled to include the code. */  /* Can't support UTF8 unless PCRE has been compiled to include the code. */
6696    
# Line 6548  if (utf8) Line 6709  if (utf8)
6709    }    }
6710  #endif  #endif
6711    
6712    /* Can't support UCP unless PCRE has been compiled to include the code. */
6713    
6714    #ifndef SUPPORT_UCP
6715    if ((options & PCRE_UCP) != 0)
6716      {
6717      errorcode = ERR67;
6718      goto PCRE_EARLY_ERROR_RETURN;
6719      }
6720    #endif
6721    
6722  /* Check validity of \R options. */  /* Check validity of \R options. */
6723    
6724  switch (options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE))  switch (options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE))
# Line 6676  regex compiled on a system with 4-byte p Line 6847  regex compiled on a system with 4-byte p
6847  pointers. */  pointers. */
6848    
6849  re->magic_number = MAGIC_NUMBER;  re->magic_number = MAGIC_NUMBER;
6850  re->size = size;  re->size = (int)size;
6851  re->options = cd->external_options;  re->options = cd->external_options;
6852  re->flags = cd->external_flags;  re->flags = cd->external_flags;
6853  re->dummy1 = 0;  re->dummy1 = 0;
# Line 6747  while (errorcode == 0 && cd->hwm > cwork Line 6918  while (errorcode == 0 && cd->hwm > cwork
6918    recno = GET(codestart, offset);    recno = GET(codestart, offset);
6919    groupptr = _pcre_find_bracket(codestart, utf8, recno);    groupptr = _pcre_find_bracket(codestart, utf8, recno);
6920    if (groupptr == NULL) errorcode = ERR53;    if (groupptr == NULL) errorcode = ERR53;
6921      else PUT(((uschar *)codestart), offset, groupptr - codestart);      else PUT(((uschar *)codestart), offset, (int)(groupptr - codestart));
6922    }    }
6923    
6924  /* Give an error if there's back reference to a non-existent capturing  /* Give an error if there's back reference to a non-existent capturing
# Line 6802  if (errorcode != 0) Line 6973  if (errorcode != 0)
6973    {    {
6974    (pcre_free)(re);    (pcre_free)(re);
6975    PCRE_EARLY_ERROR_RETURN:    PCRE_EARLY_ERROR_RETURN:
6976    *erroroffset = ptr - (const uschar *)pattern;    *erroroffset = (int)(ptr - (const uschar *)pattern);
6977    PCRE_EARLY_ERROR_RETURN2:    PCRE_EARLY_ERROR_RETURN2:
6978    *errorptr = find_error_text(errorcode);    *errorptr = find_error_text(errorcode);
6979    if (errorcodeptr != NULL) *errorcodeptr = errorcode;    if (errorcodeptr != NULL) *errorcodeptr = errorcode;

Legend:
Removed from v.507  
changed lines
  Added in v.530

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12