pcre_compile.c 315 KB
Newer Older
1 2 3 4 5 6 7 8
/*************************************************
*      Perl-Compatible Regular Expressions       *
*************************************************/

/* PCRE is a library of functions to support regular expressions whose syntax
and semantics are as close as possible to those of the Perl 5 language.

                       Written by Philip Hazel
9
           Copyright (c) 1997-2016 University of Cambridge
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:

    * Redistributions of source code must retain the above copyright notice,
      this list of conditions and the following disclaimer.

    * Redistributions in binary form must reproduce the above copyright
      notice, this list of conditions and the following disclaimer in the
      documentation and/or other materials provided with the distribution.

    * Neither the name of the University of Cambridge nor the names of its
      contributors may be used to endorse or promote products derived from
      this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
POSSIBILITY OF SUCH DAMAGE.
-----------------------------------------------------------------------------
*/


/* This module contains the external function pcre_compile(), along with
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 */
50 51
#define PSSTART start_pattern  /* Field containing pattern start */
#define PSEND   end_pattern    /* Field containing pattern end */
52 53 54 55

#include "pcre_internal.h"


56 57 58 59
/* When PCRE_DEBUG is defined, we need the pcre(16|32)_printint() function, which
is also used by pcretest. PCRE_DEBUG is not defined when building a production
library. We do not need to select pcre16_printint.c specially, because the
COMPILE_PCREx macro will already be appropriately set. */
60

61 62 63 64 65
#ifdef PCRE_DEBUG
/* pcre_printint.c should not include any headers */
#define PCRE_INCLUDED
#include "pcre_printint.c"
#undef PCRE_INCLUDED
66 67 68 69 70
#endif


/* Macro for setting individual bits in class bitmaps. */

71
#define SETBIT(a,b) a[(b)/8] |= (1 << ((b)&7))
72 73 74 75 76 77 78 79

/* 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)

80 81 82 83 84 85 86 87 88 89 90 91
/* Definitions to allow mutual recursion */

static int
  add_list_to_class(pcre_uint8 *, pcre_uchar **, int, compile_data *,
    const pcre_uint32 *, unsigned int);

static BOOL
  compile_regex(int, pcre_uchar **, const pcre_uchar **, int *, BOOL, BOOL, int, int,
    pcre_uint32 *, pcre_int32 *, pcre_uint32 *, pcre_int32 *, branch_chain *,
    compile_data *, int *);


92 93 94 95 96 97 98 99 100 101 102 103 104 105 106

/*************************************************
*      Code parameters and static tables         *
*************************************************/

/* This value specifies the size of stack workspace that is used during the
first pre-compile phase that determines how much memory is required. The regex
is partly compiled into this space, but the compiled parts are discarded as
soon as they can be, so that hopefully there will never be an overrun. The code
does, however, check for an overrun. The largest amount I've seen used is 218,
so this number is very generous.

The same workspace is used during the second, actual compile phase for
remembering forward references to groups so that they can be filled in at the
end. Each entry in this list occupies LINK_SIZE bytes, so even when LINK_SIZE
107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
is 4 there is plenty of room for most patterns. However, the memory can get
filled up by repetitions of forward references, for example patterns like
/(?1){0,1999}(b)/, and one user did hit the limit. The code has been changed so
that the workspace is expanded using malloc() in this situation. The value
below is therefore a minimum, and we put a maximum on it for safety. The
minimum is now also defined in terms of LINK_SIZE so that the use of malloc()
kicks in at the same number of forward references in all cases. */

#define COMPILE_WORK_SIZE (2048*LINK_SIZE)
#define COMPILE_WORK_SIZE_MAX (100*COMPILE_WORK_SIZE)

/* This value determines the size of the initial vector that is used for
remembering named groups during the pre-compile. It is allocated on the stack,
but if it is too small, it is expanded using malloc(), in a similar way to the
workspace. The value is the number of slots in the list. */
122

123
#define NAMED_GROUP_LIST_SIZE  20
124

125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140
/* The overrun tests check for a slightly smaller size so that they detect the
overrun before it actually does run off the end of the data block. */

#define WORK_SIZE_SAFETY_MARGIN (100)

/* Private flags added to firstchar and reqchar. */

#define REQ_CASELESS    (1 << 0)        /* Indicates caselessness */
#define REQ_VARY        (1 << 1)        /* Reqchar followed non-literal item */
/* Negative values for the firstchar and reqchar flags */
#define REQ_UNSET       (-2)
#define REQ_NONE        (-1)

/* Repeated character flags. */

#define UTF_LENGTH     0x10000000l      /* The char contains its length. */
141 142 143 144 145 146

/* Table for handling escaped characters in the range '0'-'z'. Positive returns
are simple data values; negative values are for special things like \d and so
on. Zero means further processing is needed (for things like \x), or the escape
is invalid. */

147 148 149 150 151
#ifndef EBCDIC

/* This is the "normal" table for ASCII systems or for EBCDIC systems running
in UTF-8 mode. */

152
static const short int escapes[] = {
153 154 155 156 157 158 159 160 161 162 163 164 165 166 167
     0,                       0,
     0,                       0,
     0,                       0,
     0,                       0,
     0,                       0,
     CHAR_COLON,              CHAR_SEMICOLON,
     CHAR_LESS_THAN_SIGN,     CHAR_EQUALS_SIGN,
     CHAR_GREATER_THAN_SIGN,  CHAR_QUESTION_MARK,
     CHAR_COMMERCIAL_AT,      -ESC_A,
     -ESC_B,                  -ESC_C,
     -ESC_D,                  -ESC_E,
     0,                       -ESC_G,
     -ESC_H,                  0,
     0,                       -ESC_K,
     0,                       0,
168
     -ESC_N,                  0,
169 170 171 172 173 174 175 176
     -ESC_P,                  -ESC_Q,
     -ESC_R,                  -ESC_S,
     0,                       0,
     -ESC_V,                  -ESC_W,
     -ESC_X,                  0,
     -ESC_Z,                  CHAR_LEFT_SQUARE_BRACKET,
     CHAR_BACKSLASH,          CHAR_RIGHT_SQUARE_BRACKET,
     CHAR_CIRCUMFLEX_ACCENT,  CHAR_UNDERSCORE,
177
     CHAR_GRAVE_ACCENT,       ESC_a,
178 179 180 181 182 183 184 185 186 187 188 189 190
     -ESC_b,                  0,
     -ESC_d,                  ESC_e,
     ESC_f,                   0,
     -ESC_h,                  0,
     0,                       -ESC_k,
     0,                       0,
     ESC_n,                   0,
     -ESC_p,                  0,
     ESC_r,                   -ESC_s,
     ESC_tee,                 0,
     -ESC_v,                  -ESC_w,
     0,                       0,
     -ESC_z
191 192
};

193 194 195 196
#else

/* This is the "abnormal" table for EBCDIC systems without UTF-8 support. */

197 198 199 200 201 202 203 204
static const short int escapes[] = {
/*  48 */     0,     0,      0,     '.',    '<',   '(',    '+',    '|',
/*  50 */   '&',     0,      0,       0,      0,     0,      0,      0,
/*  58 */     0,     0,    '!',     '$',    '*',   ')',    ';',    '~',
/*  60 */   '-',   '/',      0,       0,      0,     0,      0,      0,
/*  68 */     0,     0,    '|',     ',',    '%',   '_',    '>',    '?',
/*  70 */     0,     0,      0,       0,      0,     0,      0,      0,
/*  78 */     0,   '`',    ':',     '#',    '@',  '\'',    '=',    '"',
205
/*  80 */     0, ESC_a, -ESC_b,       0, -ESC_d, ESC_e,  ESC_f,      0,
206
/*  88 */-ESC_h,     0,      0,     '{',      0,     0,      0,      0,
207
/*  90 */     0,     0, -ESC_k,       0,      0, ESC_n,      0, -ESC_p,
208 209 210 211 212 213 214
/*  98 */     0, ESC_r,      0,     '}',      0,     0,      0,      0,
/*  A0 */     0,   '~', -ESC_s, ESC_tee,      0,-ESC_v, -ESC_w,      0,
/*  A8 */     0,-ESC_z,      0,       0,      0,   '[',      0,      0,
/*  B0 */     0,     0,      0,       0,      0,     0,      0,      0,
/*  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,
215
/*  D0 */   '}',     0, -ESC_K,       0,      0,-ESC_N,      0, -ESC_P,
216 217 218 219 220 221
/*  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,
/*  F0 */     0,     0,      0,       0,      0,     0,      0,      0,
/*  F8 */     0,     0,      0,       0,      0,     0,      0,      0
};
222 223 224 225 226 227

/* We also need a table of characters that may follow \c in an EBCDIC
environment for characters 0-31. */

static unsigned char ebcdic_escape_c[] = "@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_";

228 229 230 231 232
#endif


/* Table of special "verbs" like (*PRUNE). This is a short table, so it is
searched linearly. Put all the names into a single string, in order to reduce
233 234 235
the number of relocations when a shared library is dynamically linked. The
string is built from string macros so that it works in UTF-8 mode on EBCDIC
platforms. */
236 237

typedef struct verbitem {
238 239 240
  int   len;                 /* Length of verb name */
  int   op;                  /* Op when no arg, or -1 if arg mandatory */
  int   op_arg;              /* Op when arg present, or -1 if not allowed */
241 242 243
} verbitem;

static const char verbnames[] =
244 245
  "\0"                       /* Empty name is a shorthand for MARK */
  STRING_MARK0
246 247 248 249 250 251 252 253 254
  STRING_ACCEPT0
  STRING_COMMIT0
  STRING_F0
  STRING_FAIL0
  STRING_PRUNE0
  STRING_SKIP0
  STRING_THEN;

static const verbitem verbs[] = {
255 256 257 258 259 260 261 262 263
  { 0, -1,        OP_MARK },
  { 4, -1,        OP_MARK },
  { 6, OP_ACCEPT, -1 },
  { 6, OP_COMMIT, -1 },
  { 1, OP_FAIL,   -1 },
  { 4, OP_FAIL,   -1 },
  { 5, OP_PRUNE,  OP_PRUNE_ARG },
  { 4, OP_SKIP,   OP_SKIP_ARG  },
  { 4, OP_THEN,   OP_THEN_ARG  }
264 265
};

266
static const int verbcount = sizeof(verbs)/sizeof(verbitem);
267 268


269 270 271 272 273 274 275 276 277 278 279 280 281
/* Substitutes for [[:<:]] and [[:>:]], which mean start and end of word in
another regex library. */

static const pcre_uchar sub_start_of_word[] = {
  CHAR_BACKSLASH, CHAR_b, CHAR_LEFT_PARENTHESIS, CHAR_QUESTION_MARK,
  CHAR_EQUALS_SIGN, CHAR_BACKSLASH, CHAR_w, CHAR_RIGHT_PARENTHESIS, '\0' };

static const pcre_uchar sub_end_of_word[] = {
  CHAR_BACKSLASH, CHAR_b, CHAR_LEFT_PARENTHESIS, CHAR_QUESTION_MARK,
  CHAR_LESS_THAN_SIGN, CHAR_EQUALS_SIGN, CHAR_BACKSLASH, CHAR_w,
  CHAR_RIGHT_PARENTHESIS, '\0' };


282 283 284 285
/* Tables of names of POSIX character classes and their lengths. The names are
now all in a single string, to reduce the number of relocations when a shared
library is dynamically loaded. The list of lengths is terminated by a zero
length entry. The first three must be alpha, lower, upper, as this is assumed
286 287
for handling case independence. The indices for graph, print, and punct are
needed, so identify them. */
288 289

static const char posix_names[] =
290 291 292 293
  STRING_alpha0 STRING_lower0 STRING_upper0 STRING_alnum0
  STRING_ascii0 STRING_blank0 STRING_cntrl0 STRING_digit0
  STRING_graph0 STRING_print0 STRING_punct0 STRING_space0
  STRING_word0  STRING_xdigit;
294

295
static const pcre_uint8 posix_name_lengths[] = {
296 297
  5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 4, 6, 0 };

298 299 300 301 302
#define PC_GRAPH  8
#define PC_PRINT  9
#define PC_PUNCT 10


303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329
/* Table of class bit maps for each POSIX class. Each class is formed from a
base map, with an optional addition or removal of another map. Then, for some
classes, there is some additional tweaking: for [:blank:] the vertical space
characters are removed, and for [:alpha:] and [:alnum:] the underscore
character is removed. The triples in the table consist of the base map offset,
second map offset or -1 if no second map, and a non-negative value for map
addition or a negative value for map subtraction (if there are two maps). The
absolute value of the third field has these meanings: 0 => no tweaking, 1 =>
remove vertical space characters, 2 => remove underscore. */

static const int posix_class_maps[] = {
  cbit_word,  cbit_digit, -2,             /* alpha */
  cbit_lower, -1,          0,             /* lower */
  cbit_upper, -1,          0,             /* upper */
  cbit_word,  -1,          2,             /* alnum - word without underscore */
  cbit_print, cbit_cntrl,  0,             /* ascii */
  cbit_space, -1,          1,             /* blank - a GNU extension */
  cbit_cntrl, -1,          0,             /* cntrl */
  cbit_digit, -1,          0,             /* digit */
  cbit_graph, -1,          0,             /* graph */
  cbit_print, -1,          0,             /* print */
  cbit_punct, -1,          0,             /* punct */
  cbit_space, -1,          0,             /* space */
  cbit_word,  -1,          0,             /* word - a Perl extension */
  cbit_xdigit,-1,          0              /* xdigit */
};

330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435
/* Table of substitutes for \d etc when PCRE_UCP is set. They are replaced by
Unicode property escapes. */

#ifdef SUPPORT_UCP
static const pcre_uchar string_PNd[]  = {
  CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
  CHAR_N, CHAR_d, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_pNd[]  = {
  CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
  CHAR_N, CHAR_d, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_PXsp[] = {
  CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
  CHAR_X, CHAR_s, CHAR_p, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_pXsp[] = {
  CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
  CHAR_X, CHAR_s, CHAR_p, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_PXwd[] = {
  CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
  CHAR_X, CHAR_w, CHAR_d, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_pXwd[] = {
  CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
  CHAR_X, CHAR_w, CHAR_d, CHAR_RIGHT_CURLY_BRACKET, '\0' };

static const pcre_uchar *substitutes[] = {
  string_PNd,           /* \D */
  string_pNd,           /* \d */
  string_PXsp,          /* \S */   /* Xsp is Perl space, but from 8.34, Perl */
  string_pXsp,          /* \s */   /* space and POSIX space are the same. */
  string_PXwd,          /* \W */
  string_pXwd           /* \w */
};

/* The POSIX class substitutes must be in the order of the POSIX class names,
defined above, and there are both positive and negative cases. NULL means no
general substitute of a Unicode property escape (\p or \P). However, for some
POSIX classes (e.g. graph, print, punct) a special property code is compiled
directly. */

static const pcre_uchar string_pL[] =   {
  CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
  CHAR_L, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_pLl[] =  {
  CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
  CHAR_L, CHAR_l, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_pLu[] =  {
  CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
  CHAR_L, CHAR_u, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_pXan[] = {
  CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
  CHAR_X, CHAR_a, CHAR_n, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_h[] =    {
  CHAR_BACKSLASH, CHAR_h, '\0' };
static const pcre_uchar string_pXps[] = {
  CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
  CHAR_X, CHAR_p, CHAR_s, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_PL[] =   {
  CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
  CHAR_L, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_PLl[] =  {
  CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
  CHAR_L, CHAR_l, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_PLu[] =  {
  CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
  CHAR_L, CHAR_u, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_PXan[] = {
  CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
  CHAR_X, CHAR_a, CHAR_n, CHAR_RIGHT_CURLY_BRACKET, '\0' };
static const pcre_uchar string_H[] =    {
  CHAR_BACKSLASH, CHAR_H, '\0' };
static const pcre_uchar string_PXps[] = {
  CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
  CHAR_X, CHAR_p, CHAR_s, CHAR_RIGHT_CURLY_BRACKET, '\0' };

static const pcre_uchar *posix_substitutes[] = {
  string_pL,            /* alpha */
  string_pLl,           /* lower */
  string_pLu,           /* upper */
  string_pXan,          /* alnum */
  NULL,                 /* ascii */
  string_h,             /* blank */
  NULL,                 /* cntrl */
  string_pNd,           /* digit */
  NULL,                 /* graph */
  NULL,                 /* print */
  NULL,                 /* punct */
  string_pXps,          /* space */   /* Xps is POSIX space, but from 8.34 */
  string_pXwd,          /* word  */   /* Perl and POSIX space are the same */
  NULL,                 /* xdigit */
  /* Negated cases */
  string_PL,            /* ^alpha */
  string_PLl,           /* ^lower */
  string_PLu,           /* ^upper */
  string_PXan,          /* ^alnum */
  NULL,                 /* ^ascii */
  string_H,             /* ^blank */
  NULL,                 /* ^cntrl */
  string_PNd,           /* ^digit */
  NULL,                 /* ^graph */
  NULL,                 /* ^print */
  NULL,                 /* ^punct */
  string_PXps,          /* ^space */  /* Xps is POSIX space, but from 8.34 */
  string_PXwd,          /* ^word */   /* Perl and POSIX space are the same */
  NULL                  /* ^xdigit */
};
#define POSIX_SUBSIZE (sizeof(posix_substitutes) / sizeof(pcre_uchar *))
#endif
436 437 438 439 440 441 442 443 444 445 446 447

#define STRING(a)  # a
#define XSTRING(s) STRING(s)

/* The texts of compile-time error messages. These are "char *" because they
are passed to the outside world. Do not ever re-use any error number, because
they are documented. Always add a new error instead. Messages marked DEAD below
are no longer used. This used to be a table of strings, but in order to reduce
the number of relocations needed when a shared library is loaded dynamically,
it is now one long string. We cannot use a table of offsets, because the
lengths of inserts such as XSTRING(MAX_NAME_SIZE) are not known. Instead, we
simply count through to the one we want - this isn't a performance issue
448 449 450 451 452
because these strings are used only when there is a compilation error.

Each substring ends with \0 to insert a null character. This includes the final
substring, so that the whole string ends with \0\0, which can be detected when
counting through. */
453 454 455 456 457 458 459 460 461 462 463 464 465 466

static const char error_texts[] =
  "no error\0"
  "\\ at end of pattern\0"
  "\\c at end of pattern\0"
  "unrecognized character follows \\\0"
  "numbers out of order in {} quantifier\0"
  /* 5 */
  "number too big in {} quantifier\0"
  "missing terminating ] for character class\0"
  "invalid escape sequence in character class\0"
  "range out of order in character class\0"
  "nothing to repeat\0"
  /* 10 */
467
  "internal error: invalid forward reference offset\0"
468
  "internal error: unexpected repeat\0"
469
  "unrecognized character after (? or (?-\0"
470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487
  "POSIX named classes are supported only within a class\0"
  "missing )\0"
  /* 15 */
  "reference to non-existent subpattern\0"
  "erroffset passed as NULL\0"
  "unknown option bit(s) set\0"
  "missing ) after comment\0"
  "parentheses nested too deeply\0"  /** DEAD **/
  /* 20 */
  "regular expression is too large\0"
  "failed to get memory\0"
  "unmatched parentheses\0"
  "internal error: code overflow\0"
  "unrecognized character after (?<\0"
  /* 25 */
  "lookbehind assertion is not fixed length\0"
  "malformed number or name after (?(\0"
  "conditional group contains more than two branches\0"
488
  "assertion expected after (?( or (?(?C)\0"
489 490 491 492
  "(?R or (?[+-]digits must be followed by )\0"
  /* 30 */
  "unknown POSIX class name\0"
  "POSIX collating elements are not supported\0"
493
  "this version of PCRE is compiled without UTF support\0"
494
  "spare error\0"  /** DEAD **/
495
  "character value in \\x{} or \\o{} is too large\0"
496 497 498
  /* 35 */
  "invalid condition (?(0)\0"
  "\\C not allowed in lookbehind assertion\0"
499
  "PCRE does not support \\L, \\l, \\N{name}, \\U, or \\u\0"
500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
  "number after (?C is > 255\0"
  "closing ) for (?C expected\0"
  /* 40 */
  "recursive call could loop indefinitely\0"
  "unrecognized character after (?P\0"
  "syntax error in subpattern name (missing terminator)\0"
  "two named subpatterns have the same name\0"
  "invalid UTF-8 string\0"
  /* 45 */
  "support for \\P, \\p, and \\X has not been compiled\0"
  "malformed \\P or \\p sequence\0"
  "unknown property name after \\P or \\p\0"
  "subpattern name is too long (maximum " XSTRING(MAX_NAME_SIZE) " characters)\0"
  "too many named subpatterns (maximum " XSTRING(MAX_NAME_COUNT) ")\0"
  /* 50 */
  "repeated subpattern is too long\0"    /** DEAD **/
516
  "octal value is greater than \\377 in 8-bit non-UTF-8 mode\0"
517 518 519 520
  "internal error: overran compiling workspace\0"
  "internal error: previously-checked referenced subpattern not found\0"
  "DEFINE group contains more than one branch\0"
  /* 55 */
521
  "repeating a DEFINE group is not allowed\0"  /** DEAD **/
522
  "inconsistent NEWLINE options\0"
523 524
  "\\g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number\0"
  "a numbered reference must not be zero\0"
525
  "an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)\0"
526
  /* 60 */
527
  "(*VERB) not recognized or malformed\0"
528 529 530
  "number is too big\0"
  "subpattern name expected\0"
  "digit expected after (?+\0"
531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564
  "] is an invalid data character in JavaScript compatibility mode\0"
  /* 65 */
  "different names for subpatterns of the same number are not allowed\0"
  "(*MARK) must have an argument\0"
  "this version of PCRE is not compiled with Unicode property support\0"
#ifndef EBCDIC
  "\\c must be followed by an ASCII character\0"
#else
  "\\c must be followed by a letter or one of [\\]^_?\0"
#endif
  "\\k is not followed by a braced, angle-bracketed, or quoted name\0"
  /* 70 */
  "internal error: unknown opcode in find_fixedlength()\0"
  "\\N is not supported in a class\0"
  "too many forward references\0"
  "disallowed Unicode code point (>= 0xd800 && <= 0xdfff)\0"
  "invalid UTF-16 string\0"
  /* 75 */
  "name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)\0"
  "character value in \\u.... sequence is too large\0"
  "invalid UTF-32 string\0"
  "setting UTF is disabled by the application\0"
  "non-hex character in \\x{} (closing brace missing?)\0"
  /* 80 */
  "non-octal character in \\o{} (closing brace missing?)\0"
  "missing opening brace after \\o\0"
  "parentheses are too deeply nested\0"
  "invalid range in character class\0"
  "group name must start with a non-digit\0"
  /* 85 */
  "parentheses are too deeply nested (stack check)\0"
  "digits missing in \\x{} or \\o{}\0"
  "regular expression is too complicated\0"
  ;
565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581

/* Table to identify digits and hex digits. This is used when compiling
patterns. Note that the tables in chartables are dependent on the locale, and
may mark arbitrary characters as digits - but the PCRE compiling code expects
to handle only 0-9, a-z, and A-Z as digits when compiling. That is why we have
a private table here. It costs 256 bytes, but it is a lot faster than doing
character value tests (at least in some simple cases I timed), and in some
applications one wants PCRE to compile efficiently as well as match
efficiently.

For convenience, we use the same bit definitions as in chartables:

  0x04   decimal digit
  0x08   hexadecimal digit

Then we can use ctype_digit and ctype_xdigit in the code. */

582 583 584 585 586 587
/* Using a simple comparison for decimal numbers rather than a memory read
is much faster, and the resulting code is simpler (the compiler turns it
into a subtraction and unsigned comparison). */

#define IS_DIGIT(x) ((x) >= CHAR_0 && (x) <= CHAR_9)

588 589 590 591 592
#ifndef EBCDIC

/* This is the "normal" case, for ASCII systems, and EBCDIC systems running in
UTF-8 mode. */

593
static const pcre_uint8 digitab[] =
594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627
  {
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   0-  7 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   8- 15 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  16- 23 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  24- 31 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*    - '  */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  ( - /  */
  0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c, /*  0 - 7  */
  0x0c,0x0c,0x00,0x00,0x00,0x00,0x00,0x00, /*  8 - ?  */
  0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /*  @ - G  */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  H - O  */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  P - W  */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  X - _  */
  0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /*  ` - g  */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  h - o  */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  p - w  */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  x -127 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 128-135 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 136-143 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144-151 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 152-159 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160-167 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 168-175 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 176-183 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 192-199 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 200-207 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 208-215 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 216-223 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 224-231 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 232-239 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 240-247 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00};/* 248-255 */

628 629 630 631
#else

/* This is the "abnormal" case, for EBCDIC systems not running in UTF-8 mode. */

632
static const pcre_uint8 digitab[] =
633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666
  {
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   0-  7  0 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   8- 15    */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  16- 23 10 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  24- 31    */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  32- 39 20 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  40- 47    */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  48- 55 30 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  56- 63    */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*    - 71 40 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  72- |     */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  & - 87 50 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  88- 95    */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  - -103 60 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 104- ?     */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 112-119 70 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 120- "     */
  0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /* 128- g  80 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  h -143    */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144- p  90 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  q -159    */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160- x  A0 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  y -175    */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  ^ -183 B0 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191    */
  0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /*  { - G  C0 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  H -207    */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  } - P  D0 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  Q -223    */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  \ - X  E0 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  Y -239    */
  0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c, /*  0 - 7  F0 */
  0x0c,0x0c,0x00,0x00,0x00,0x00,0x00,0x00};/*  8 -255    */

667
static const pcre_uint8 ebcdic_chartab[] = { /* chartable partial dup */
668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702
  0x80,0x00,0x00,0x00,0x00,0x01,0x00,0x00, /*   0-  7 */
  0x00,0x00,0x00,0x00,0x01,0x01,0x00,0x00, /*   8- 15 */
  0x00,0x00,0x00,0x00,0x00,0x01,0x00,0x00, /*  16- 23 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  24- 31 */
  0x00,0x00,0x00,0x00,0x00,0x01,0x00,0x00, /*  32- 39 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  40- 47 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  48- 55 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  56- 63 */
  0x01,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*    - 71 */
  0x00,0x00,0x00,0x80,0x00,0x80,0x80,0x80, /*  72- |  */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  & - 87 */
  0x00,0x00,0x00,0x80,0x80,0x80,0x00,0x00, /*  88- 95 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  - -103 */
  0x00,0x00,0x00,0x00,0x00,0x10,0x00,0x80, /* 104- ?  */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 112-119 */
  0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 120- "  */
  0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /* 128- g  */
  0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /*  h -143 */
  0x00,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /* 144- p  */
  0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /*  q -159 */
  0x00,0x00,0x12,0x12,0x12,0x12,0x12,0x12, /* 160- x  */
  0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /*  y -175 */
  0x80,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  ^ -183 */
  0x00,0x00,0x80,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
  0x80,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  { - G  */
  0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /*  H -207 */
  0x00,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  } - P  */
  0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /*  Q -223 */
  0x00,0x00,0x12,0x12,0x12,0x12,0x12,0x12, /*  \ - X  */
  0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /*  Y -239 */
  0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c, /*  0 - 7  */
  0x1c,0x1c,0x00,0x00,0x00,0x00,0x00,0x00};/*  8 -255 */
#endif


703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738
/* This table is used to check whether auto-possessification is possible
between adjacent character-type opcodes. The left-hand (repeated) opcode is
used to select the row, and the right-hand opcode is use to select the column.
A value of 1 means that auto-possessification is OK. For example, the second
value in the first row means that \D+\d can be turned into \D++\d.

The Unicode property types (\P and \p) have to be present to fill out the table
because of what their opcode values are, but the table values should always be
zero because property types are handled separately in the code. The last four
columns apply to items that cannot be repeated, so there is no need to have
rows for them. Note that OP_DIGIT etc. are generated only when PCRE_UCP is
*not* set. When it is set, \d etc. are converted into OP_(NOT_)PROP codes. */

#define APTROWS (LAST_AUTOTAB_LEFT_OP - FIRST_AUTOTAB_OP + 1)
#define APTCOLS (LAST_AUTOTAB_RIGHT_OP - FIRST_AUTOTAB_OP + 1)

static const pcre_uint8 autoposstab[APTROWS][APTCOLS] = {
/* \D \d \S \s \W \w  . .+ \C \P \p \R \H \h \V \v \X \Z \z  $ $M */
  { 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0 },  /* \D */
  { 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1 },  /* \d */
  { 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1 },  /* \S */
  { 0, 1, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0 },  /* \s */
  { 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0 },  /* \W */
  { 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1 },  /* \w */
  { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0 },  /* .  */
  { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0 },  /* .+ */
  { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0 },  /* \C */
  { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },  /* \P */
  { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },  /* \p */
  { 0, 1, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0 },  /* \R */
  { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0 },  /* \H */
  { 0, 1, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 0, 0 },  /* \h */
  { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 1, 0, 0 },  /* \V */
  { 0, 1, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0 },  /* \v */
  { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0 }   /* \X */
};
739

740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877

/* This table is used to check whether auto-possessification is possible
between adjacent Unicode property opcodes (OP_PROP and OP_NOTPROP). The
left-hand (repeated) opcode is used to select the row, and the right-hand
opcode is used to select the column. The values are as follows:

  0   Always return FALSE (never auto-possessify)
  1   Character groups are distinct (possessify if both are OP_PROP)
  2   Check character categories in the same group (general or particular)
  3   TRUE if the two opcodes are not the same (PROP vs NOTPROP)

  4   Check left general category vs right particular category
  5   Check right general category vs left particular category

  6   Left alphanum vs right general category
  7   Left space vs right general category
  8   Left word vs right general category

  9   Right alphanum vs left general category
 10   Right space vs left general category
 11   Right word vs left general category

 12   Left alphanum vs right particular category
 13   Left space vs right particular category
 14   Left word vs right particular category

 15   Right alphanum vs left particular category
 16   Right space vs left particular category
 17   Right word vs left particular category
*/

static const pcre_uint8 propposstab[PT_TABSIZE][PT_TABSIZE] = {
/* ANY LAMP GC  PC  SC ALNUM SPACE PXSPACE WORD CLIST UCNC */
  { 0,  0,  0,  0,  0,    0,    0,      0,   0,    0,   0 },  /* PT_ANY */
  { 0,  3,  0,  0,  0,    3,    1,      1,   0,    0,   0 },  /* PT_LAMP */
  { 0,  0,  2,  4,  0,    9,   10,     10,  11,    0,   0 },  /* PT_GC */
  { 0,  0,  5,  2,  0,   15,   16,     16,  17,    0,   0 },  /* PT_PC */
  { 0,  0,  0,  0,  2,    0,    0,      0,   0,    0,   0 },  /* PT_SC */
  { 0,  3,  6, 12,  0,    3,    1,      1,   0,    0,   0 },  /* PT_ALNUM */
  { 0,  1,  7, 13,  0,    1,    3,      3,   1,    0,   0 },  /* PT_SPACE */
  { 0,  1,  7, 13,  0,    1,    3,      3,   1,    0,   0 },  /* PT_PXSPACE */
  { 0,  0,  8, 14,  0,    0,    1,      1,   3,    0,   0 },  /* PT_WORD */
  { 0,  0,  0,  0,  0,    0,    0,      0,   0,    0,   0 },  /* PT_CLIST */
  { 0,  0,  0,  0,  0,    0,    0,      0,   0,    0,   3 }   /* PT_UCNC */
};

/* This table is used to check whether auto-possessification is possible
between adjacent Unicode property opcodes (OP_PROP and OP_NOTPROP) when one
specifies a general category and the other specifies a particular category. The
row is selected by the general category and the column by the particular
category. The value is 1 if the particular category is not part of the general
category. */

static const pcre_uint8 catposstab[7][30] = {
/* Cc Cf Cn Co Cs Ll Lm Lo Lt Lu Mc Me Mn Nd Nl No Pc Pd Pe Pf Pi Po Ps Sc Sk Sm So Zl Zp Zs */
  { 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 },  /* C */
  { 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 },  /* L */
  { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 },  /* M */
  { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 },  /* N */
  { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1 },  /* P */
  { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1 },  /* S */
  { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0 }   /* Z */
};

/* This table is used when checking ALNUM, (PX)SPACE, SPACE, and WORD against
a general or particular category. The properties in each row are those
that apply to the character set in question. Duplication means that a little
unnecessary work is done when checking, but this keeps things much simpler
because they can all use the same code. For more details see the comment where
this table is used.

Note: SPACE and PXSPACE used to be different because Perl excluded VT from
"space", but from Perl 5.18 it's included, so both categories are treated the
same here. */

static const pcre_uint8 posspropstab[3][4] = {
  { ucp_L, ucp_N, ucp_N, ucp_Nl },  /* ALNUM, 3rd and 4th values redundant */
  { ucp_Z, ucp_Z, ucp_C, ucp_Cc },  /* SPACE and PXSPACE, 2nd value redundant */
  { ucp_L, ucp_N, ucp_P, ucp_Po }   /* WORD */
};

/* This table is used when converting repeating opcodes into possessified
versions as a result of an explicit possessive quantifier such as ++. A zero
value means there is no possessified version - in those cases the item in
question must be wrapped in ONCE brackets. The table is truncated at OP_CALLOUT
because all relevant opcodes are less than that. */

static const pcre_uint8 opcode_possessify[] = {
  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,   /* 0 - 15  */
  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,   /* 16 - 31 */

  0,                       /* NOTI */
  OP_POSSTAR, 0,           /* STAR, MINSTAR */
  OP_POSPLUS, 0,           /* PLUS, MINPLUS */
  OP_POSQUERY, 0,          /* QUERY, MINQUERY */
  OP_POSUPTO, 0,           /* UPTO, MINUPTO */
  0,                       /* EXACT */
  0, 0, 0, 0,              /* POS{STAR,PLUS,QUERY,UPTO} */

  OP_POSSTARI, 0,          /* STARI, MINSTARI */
  OP_POSPLUSI, 0,          /* PLUSI, MINPLUSI */
  OP_POSQUERYI, 0,         /* QUERYI, MINQUERYI */
  OP_POSUPTOI, 0,          /* UPTOI, MINUPTOI */
  0,                       /* EXACTI */
  0, 0, 0, 0,              /* POS{STARI,PLUSI,QUERYI,UPTOI} */

  OP_NOTPOSSTAR, 0,        /* NOTSTAR, NOTMINSTAR */
  OP_NOTPOSPLUS, 0,        /* NOTPLUS, NOTMINPLUS */
  OP_NOTPOSQUERY, 0,       /* NOTQUERY, NOTMINQUERY */
  OP_NOTPOSUPTO, 0,        /* NOTUPTO, NOTMINUPTO */
  0,                       /* NOTEXACT */
  0, 0, 0, 0,              /* NOTPOS{STAR,PLUS,QUERY,UPTO} */

  OP_NOTPOSSTARI, 0,       /* NOTSTARI, NOTMINSTARI */
  OP_NOTPOSPLUSI, 0,       /* NOTPLUSI, NOTMINPLUSI */
  OP_NOTPOSQUERYI, 0,      /* NOTQUERYI, NOTMINQUERYI */
  OP_NOTPOSUPTOI, 0,       /* NOTUPTOI, NOTMINUPTOI */
  0,                       /* NOTEXACTI */
  0, 0, 0, 0,              /* NOTPOS{STARI,PLUSI,QUERYI,UPTOI} */

  OP_TYPEPOSSTAR, 0,       /* TYPESTAR, TYPEMINSTAR */
  OP_TYPEPOSPLUS, 0,       /* TYPEPLUS, TYPEMINPLUS */
  OP_TYPEPOSQUERY, 0,      /* TYPEQUERY, TYPEMINQUERY */
  OP_TYPEPOSUPTO, 0,       /* TYPEUPTO, TYPEMINUPTO */
  0,                       /* TYPEEXACT */
  0, 0, 0, 0,              /* TYPEPOS{STAR,PLUS,QUERY,UPTO} */

  OP_CRPOSSTAR, 0,         /* CRSTAR, CRMINSTAR */
  OP_CRPOSPLUS, 0,         /* CRPLUS, CRMINPLUS */
  OP_CRPOSQUERY, 0,        /* CRQUERY, CRMINQUERY */
  OP_CRPOSRANGE, 0,        /* CRRANGE, CRMINRANGE */
  0, 0, 0, 0,              /* CRPOS{STAR,PLUS,QUERY,RANGE} */

  0, 0, 0,                 /* CLASS, NCLASS, XCLASS */
  0, 0,                    /* REF, REFI */
  0, 0,                    /* DNREF, DNREFI */
  0, 0                     /* RECURSE, CALLOUT */
};
878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897



/*************************************************
*            Find an error text                  *
*************************************************/

/* The error texts are now all in one long string, to save on relocations. As
some of the text is of unknown length, we can't use a table of offsets.
Instead, just count through the strings. This is not a performance issue
because it happens only when there has been a compilation error.

Argument:   the error number
Returns:    pointer to the error string
*/

static const char *
find_error_text(int n)
{
const char *s = error_texts;
898 899 900 901 902
for (; n > 0; n--)
  {
  while (*s++ != CHAR_NULL) {};
  if (*s == CHAR_NULL) return "Error text not found (please report)";
  }
903 904 905 906
return s;
}


907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979

/*************************************************
*           Expand the workspace                 *
*************************************************/

/* This function is called during the second compiling phase, if the number of
forward references fills the existing workspace, which is originally a block on
the stack. A larger block is obtained from malloc() unless the ultimate limit
has been reached or the increase will be rather small.

Argument: pointer to the compile data block
Returns:  0 if all went well, else an error number
*/

static int
expand_workspace(compile_data *cd)
{
pcre_uchar *newspace;
int newsize = cd->workspace_size * 2;

if (newsize > COMPILE_WORK_SIZE_MAX) newsize = COMPILE_WORK_SIZE_MAX;
if (cd->workspace_size >= COMPILE_WORK_SIZE_MAX ||
    newsize - cd->workspace_size < WORK_SIZE_SAFETY_MARGIN)
 return ERR72;

newspace = (PUBL(malloc))(IN_UCHARS(newsize));
if (newspace == NULL) return ERR21;
memcpy(newspace, cd->start_workspace, cd->workspace_size * sizeof(pcre_uchar));
cd->hwm = (pcre_uchar *)newspace + (cd->hwm - cd->start_workspace);
if (cd->workspace_size > COMPILE_WORK_SIZE)
  (PUBL(free))((void *)cd->start_workspace);
cd->start_workspace = newspace;
cd->workspace_size = newsize;
return 0;
}



/*************************************************
*            Check for counted repeat            *
*************************************************/

/* This function is called when a '{' is encountered in a place where it might
start a quantifier. It looks ahead to see if it really is a quantifier or not.
It is only a quantifier if it is one of the forms {ddd} {ddd,} or {ddd,ddd}
where the ddds are digits.

Arguments:
  p         pointer to the first char after '{'

Returns:    TRUE or FALSE
*/

static BOOL
is_counted_repeat(const pcre_uchar *p)
{
if (!IS_DIGIT(*p)) return FALSE;
p++;
while (IS_DIGIT(*p)) p++;
if (*p == CHAR_RIGHT_CURLY_BRACKET) return TRUE;

if (*p++ != CHAR_COMMA) return FALSE;
if (*p == CHAR_RIGHT_CURLY_BRACKET) return TRUE;

if (!IS_DIGIT(*p)) return FALSE;
p++;
while (IS_DIGIT(*p)) p++;

return (*p == CHAR_RIGHT_CURLY_BRACKET);
}



980 981 982 983 984
/*************************************************
*            Handle escapes                      *
*************************************************/

/* This function is called when a \ has been encountered. It either returns a
985 986 987 988 989
positive value for a simple escape such as \n, or 0 for a data character which
will be placed in chptr. A backreference to group n is returned as negative n.
When UTF-8 is enabled, a positive value greater than 255 may be returned in
chptr. On entry, ptr is pointing at the \. On exit, it is on the final
character of the escape sequence.
990 991 992

Arguments:
  ptrptr         points to the pattern position pointer
993
  chptr          points to a returned data character
994 995 996 997 998
  errorcodeptr   points to the errorcode variable
  bracount       number of previous extracting brackets
  options        the options bits
  isclass        TRUE if inside a character class

999 1000 1001
Returns:         zero => a data character
                 positive => a special escape sequence
                 negative => a back reference
1002 1003 1004 1005
                 on error, errorcodeptr is set
*/

static int
1006 1007
check_escape(const pcre_uchar **ptrptr, pcre_uint32 *chptr, int *errorcodeptr,
  int bracount, int options, BOOL isclass)
1008
{
1009 1010 1011 1012 1013 1014
/* PCRE_UTF16 has the same value as PCRE_UTF8. */
BOOL utf = (options & PCRE_UTF8) != 0;
const pcre_uchar *ptr = *ptrptr + 1;
pcre_uint32 c;
int escape = 0;
int i;
1015 1016 1017 1018 1019 1020

GETCHARINCTEST(c, ptr);           /* Get character value, increment pointer */
ptr--;                            /* Set pointer back to the last byte */

/* If backslash is at the end of the pattern, it's an error. */

1021
if (c == CHAR_NULL) *errorcodeptr = ERR1;
1022

1023 1024
/* Non-alphanumerics are literals. For digits or letters, do an initial lookup
in a table. A non-zero result is something that can be returned immediately.
1025 1026
Otherwise further processing may be required. */

1027
#ifndef EBCDIC  /* ASCII/UTF-8 coding */
1028 1029 1030 1031
/* Not alphanumeric */
else if (c < CHAR_0 || c > CHAR_z) {}
else if ((i = escapes[c - CHAR_0]) != 0)
  { if (i > 0) c = (pcre_uint32)i; else escape = -i; }
1032 1033

#else           /* EBCDIC coding */
1034 1035 1036
/* Not alphanumeric */
else if (c < CHAR_a || (!MAX_255(c) || (ebcdic_chartab[c] & 0x0E) == 0)) {}
else if ((i = escapes[c - 0x48]) != 0)  { if (i > 0) c = (pcre_uint32)i; else escape = -i; }
1037 1038 1039 1040 1041 1042
#endif

/* Escapes that need further processing, or are illegal. */

else
  {
1043 1044 1045
  const pcre_uchar *oldptr;
  BOOL braced, negated, overflow;
  int s;
1046 1047 1048 1049 1050 1051

  switch (c)
    {
    /* A number of Perl escapes are not handled by PCRE. We give an explicit
    error. */

1052 1053
    case CHAR_l:
    case CHAR_L:
1054 1055 1056
    *errorcodeptr = ERR37;
    break;

1057
    case CHAR_u:
1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096
    if ((options & PCRE_JAVASCRIPT_COMPAT) != 0)
      {
      /* In JavaScript, \u must be followed by four hexadecimal numbers.
      Otherwise it is a lowercase u letter. */
      if (MAX_255(ptr[1]) && (digitab[ptr[1]] & ctype_xdigit) != 0
        && MAX_255(ptr[2]) && (digitab[ptr[2]] & ctype_xdigit) != 0
        && MAX_255(ptr[3]) && (digitab[ptr[3]] & ctype_xdigit) != 0
        && MAX_255(ptr[4]) && (digitab[ptr[4]] & ctype_xdigit) != 0)
        {
        c = 0;
        for (i = 0; i < 4; ++i)
          {
          register pcre_uint32 cc = *(++ptr);
#ifndef EBCDIC  /* ASCII/UTF-8 coding */
          if (cc >= CHAR_a) cc -= 32;               /* Convert to upper case */
          c = (c << 4) + cc - ((cc < CHAR_A)? CHAR_0 : (CHAR_A - 10));
#else           /* EBCDIC coding */
          if (cc >= CHAR_a && cc <= CHAR_z) cc += 64;  /* Convert to upper case */
          c = (c << 4) + cc - ((cc >= CHAR_0)? CHAR_0 : (CHAR_A - 10));
#endif
          }

#if defined COMPILE_PCRE8
        if (c > (utf ? 0x10ffffU : 0xffU))
#elif defined COMPILE_PCRE16
        if (c > (utf ? 0x10ffffU : 0xffffU))
#elif defined COMPILE_PCRE32
        if (utf && c > 0x10ffffU)
#endif
          {
          *errorcodeptr = ERR76;
          }
        else if (utf && c >= 0xd800 && c <= 0xdfff) *errorcodeptr = ERR73;
        }
      }
    else
      *errorcodeptr = ERR37;
    break;

1097
    case CHAR_U:
1098 1099
    /* In JavaScript, \U is an uppercase U letter. */
    if ((options & PCRE_JAVASCRIPT_COMPAT) == 0) *errorcodeptr = ERR37;
1100 1101
    break;

1102 1103
    /* In a character class, \g is just a literal "g". Outside a character
    class, \g must be followed by one of a number of specific things:
1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116

    (1) A number, either plain or braced. If positive, it is an absolute
    backreference. If negative, it is a relative backreference. This is a Perl
    5.10 feature.

    (2) Perl 5.10 also supports \g{name} as a reference to a named group. This
    is part of Perl's movement towards a unified syntax for back references. As
    this is synonymous with \k{name}, we fudge it up by pretending it really
    was \k.

    (3) For Oniguruma compatibility we also support \g followed by a name or a
    number either in angle brackets or in single quotes. However, these are
    (possibly recursive) subroutine calls, _not_ backreferences. Just return
1117
    the ESC_g code (cf \k). */
1118 1119

    case CHAR_g:
1120
    if (isclass) break;
1121 1122
    if (ptr[1] == CHAR_LESS_THAN_SIGN || ptr[1] == CHAR_APOSTROPHE)
      {
1123
      escape = ESC_g;
1124 1125 1126 1127
      break;
      }

    /* Handle the Perl-compatible cases */
1128

1129
    if (ptr[1] == CHAR_LEFT_CURLY_BRACKET)
1130
      {
1131 1132 1133 1134
      const pcre_uchar *p;
      for (p = ptr+2; *p != CHAR_NULL && *p != CHAR_RIGHT_CURLY_BRACKET; p++)
        if (*p != CHAR_MINUS && !IS_DIGIT(*p)) break;
      if (*p != CHAR_NULL && *p != CHAR_RIGHT_CURLY_BRACKET)
1135
        {
1136
        escape = ESC_k;
1137 1138 1139 1140 1141 1142 1143
        break;
        }
      braced = TRUE;
      ptr++;
      }
    else braced = FALSE;

1144
    if (ptr[1] == CHAR_MINUS)
1145 1146 1147 1148 1149 1150
      {
      negated = TRUE;
      ptr++;
      }
    else negated = FALSE;

1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163
    /* The integer range is limited by the machine's int representation. */
    s = 0;
    overflow = FALSE;
    while (IS_DIGIT(ptr[1]))
      {
      if (s > INT_MAX / 10 - 1) /* Integer overflow */
        {
        overflow = TRUE;
        break;
        }
      s = s * 10 + (int)(*(++ptr) - CHAR_0);
      }
    if (overflow) /* Integer overflow */
1164
      {
1165 1166
      while (IS_DIGIT(ptr[1]))
        ptr++;
1167 1168 1169 1170
      *errorcodeptr = ERR61;
      break;
      }

1171
    if (braced && *(++ptr) != CHAR_RIGHT_CURLY_BRACKET)
1172 1173 1174 1175 1176
      {
      *errorcodeptr = ERR57;
      break;
      }

1177
    if (s == 0)
1178 1179 1180 1181 1182
      {
      *errorcodeptr = ERR58;
      break;
      }

1183 1184
    if (negated)
      {
1185
      if (s > bracount)
1186 1187 1188 1189
        {
        *errorcodeptr = ERR15;
        break;
        }
1190
      s = bracount - (s - 1);
1191 1192
      }

1193
    escape = -s;
1194 1195 1196
    break;

    /* The handling of escape sequences consisting of a string of digits
1197 1198 1199
    starting with one that is not zero is not straightforward. Perl has changed
    over the years. Nowadays \g{} for backreferences and \o{} for octal are
    recommended to avoid the ambiguities in the old syntax.
1200 1201

    Outside a character class, the digits are read as a decimal number. If the
1202 1203 1204 1205 1206 1207 1208 1209 1210
    number is less than 8 (used to be 10), or if there are that many previous
    extracting left brackets, then it is a back reference. Otherwise, up to
    three octal digits are read to form an escaped byte. Thus \123 is likely to
    be octal 123 (cf \0123, which is octal 012 followed by the literal 3). If
    the octal value is greater than 377, the least significant 8 bits are
    taken. \8 and \9 are treated as the literal characters 8 and 9.

    Inside a character class, \ followed by a digit is always either a literal
    8 or 9 or an octal number. */
1211

1212 1213
    case CHAR_1: case CHAR_2: case CHAR_3: case CHAR_4: case CHAR_5:
    case CHAR_6: case CHAR_7: case CHAR_8: case CHAR_9:
1214 1215 1216 1217

    if (!isclass)
      {
      oldptr = ptr;
1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230
      /* The integer range is limited by the machine's int representation. */
      s = (int)(c -CHAR_0);
      overflow = FALSE;
      while (IS_DIGIT(ptr[1]))
        {
        if (s > INT_MAX / 10 - 1) /* Integer overflow */
          {
          overflow = TRUE;
          break;
          }
        s = s * 10 + (int)(*(++ptr) - CHAR_0);
        }
      if (overflow) /* Integer overflow */
1231
        {
1232 1233
        while (IS_DIGIT(ptr[1]))
          ptr++;
1234 1235 1236
        *errorcodeptr = ERR61;
        break;
        }
1237
      if (s < 8 || s <= bracount)  /* Check for back reference */
1238
        {
1239
        escape = -s;
1240 1241 1242 1243 1244
        break;
        }
      ptr = oldptr;      /* Put the pointer back and fall through */
      }

1245 1246 1247 1248
    /* Handle a digit following \ when the number is not a back reference. If
    the first digit is 8 or 9, Perl used to generate a binary zero byte and
    then treat the digit as a following literal. At least by Perl 5.18 this
    changed so as not to insert the binary zero. */
1249

1250 1251 1252
    if ((c = *ptr) >= CHAR_8) break;

    /* Fall through with a digit less than 8 */
1253 1254 1255 1256

    /* \0 always starts an octal number, but we may drop through to here with a
    larger first octal digit. The original code used just to take the least
    significant 8 bits of octal numbers (I think this is what early Perls used
1257 1258
    to do). Nowadays we allow for larger numbers in UTF-8 mode and 16-bit mode,
    but no more than 3 octal digits. */
1259

1260 1261 1262 1263
    case CHAR_0:
    c -= CHAR_0;
    while(i++ < 2 && ptr[1] >= CHAR_0 && ptr[1] <= CHAR_7)
        c = c * 8 + *(++ptr) - CHAR_0;
1264 1265 1266
#ifdef COMPILE_PCRE8
    if (!utf && c > 0xff) *errorcodeptr = ERR51;
#endif
1267 1268
    break;

1269 1270
    /* \o is a relatively new Perl feature, supporting a more general way of
    specifying character codes in octal. The only supported form is \o{ddd}. */
1271

1272 1273 1274
    case CHAR_o:
    if (ptr[1] != CHAR_LEFT_CURLY_BRACKET) *errorcodeptr = ERR81; else
    if (ptr[2] == CHAR_RIGHT_CURLY_BRACKET) *errorcodeptr = ERR86; else
1275
      {
1276
      ptr += 2;
1277
      c = 0;
1278 1279
      overflow = FALSE;
      while (*ptr >= CHAR_0 && *ptr <= CHAR_7)
1280
        {
1281
        register pcre_uint32 cc = *ptr++;
1282
        if (c == 0 && cc == CHAR_0) continue;     /* Leading zeroes */
1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309
#ifdef COMPILE_PCRE32
        if (c >= 0x20000000l) { overflow = TRUE; break; }
#endif
        c = (c << 3) + cc - CHAR_0 ;
#if defined COMPILE_PCRE8
        if (c > (utf ? 0x10ffffU : 0xffU)) { overflow = TRUE; break; }
#elif defined COMPILE_PCRE16
        if (c > (utf ? 0x10ffffU : 0xffffU)) { overflow = TRUE; break; }
#elif defined COMPILE_PCRE32
        if (utf && c > 0x10ffffU) { overflow = TRUE; break; }
#endif
        }
      if (overflow)
        {
        while (*ptr >= CHAR_0 && *ptr <= CHAR_7) ptr++;
        *errorcodeptr = ERR34;
        }
      else if (*ptr == CHAR_RIGHT_CURLY_BRACKET)
        {
        if (utf && c >= 0xd800 && c <= 0xdfff) *errorcodeptr = ERR73;
        }
      else *errorcodeptr = ERR80;
      }
    break;

    /* \x is complicated. In JavaScript, \x must be followed by two hexadecimal
    numbers. Otherwise it is a lowercase x letter. */
1310

1311 1312 1313 1314 1315 1316 1317 1318 1319 1320
    case CHAR_x:
    if ((options & PCRE_JAVASCRIPT_COMPAT) != 0)
      {
      if (MAX_255(ptr[1]) && (digitab[ptr[1]] & ctype_xdigit) != 0
        && MAX_255(ptr[2]) && (digitab[ptr[2]] & ctype_xdigit) != 0)
        {
        c = 0;
        for (i = 0; i < 2; ++i)
          {
          register pcre_uint32 cc = *(++ptr);
1321
#ifndef EBCDIC  /* ASCII/UTF-8 coding */
1322 1323
          if (cc >= CHAR_a) cc -= 32;               /* Convert to upper case */
          c = (c << 4) + cc - ((cc < CHAR_A)? CHAR_0 : (CHAR_A - 10));
1324
#else           /* EBCDIC coding */
1325 1326
          if (cc >= CHAR_a && cc <= CHAR_z) cc += 64;  /* Convert to upper case */
          c = (c << 4) + cc - ((cc >= CHAR_0)? CHAR_0 : (CHAR_A - 10));
1327
#endif
1328
          }
1329
        }
1330 1331 1332 1333 1334 1335 1336 1337
      }    /* End JavaScript handling */

    /* Handle \x in Perl's style. \x{ddd} is a character number which can be
    greater than 0xff in utf or non-8bit mode, but only if the ddd are hex
    digits. If not, { used to be treated as a data character. However, Perl
    seems to read hex digits up to the first non-such, and ignore the rest, so
    that, for example \x{zz} matches a binary zero. This seems crazy, so PCRE
    now gives an error. */
1338

1339 1340 1341
    else
      {
      if (ptr[1] == CHAR_LEFT_CURLY_BRACKET)
1342
        {
1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354
        ptr += 2;
        if (*ptr == CHAR_RIGHT_CURLY_BRACKET)
          {
          *errorcodeptr = ERR86;
          break;
          }
        c = 0;
        overflow = FALSE;
        while (MAX_255(*ptr) && (digitab[*ptr] & ctype_xdigit) != 0)
          {
          register pcre_uint32 cc = *ptr++;
          if (c == 0 && cc == CHAR_0) continue;     /* Leading zeroes */
1355

1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366
#ifdef COMPILE_PCRE32
          if (c >= 0x10000000l) { overflow = TRUE; break; }
#endif

#ifndef EBCDIC  /* ASCII/UTF-8 coding */
          if (cc >= CHAR_a) cc -= 32;               /* Convert to upper case */
          c = (c << 4) + cc - ((cc < CHAR_A)? CHAR_0 : (CHAR_A - 10));
#else           /* EBCDIC coding */
          if (cc >= CHAR_a && cc <= CHAR_z) cc += 64;  /* Convert to upper case */
          c = (c << 4) + cc - ((cc >= CHAR_0)? CHAR_0 : (CHAR_A - 10));
#endif
1367

1368 1369 1370 1371 1372 1373 1374 1375
#if defined COMPILE_PCRE8
          if (c > (utf ? 0x10ffffU : 0xffU)) { overflow = TRUE; break; }
#elif defined COMPILE_PCRE16
          if (c > (utf ? 0x10ffffU : 0xffffU)) { overflow = TRUE; break; }
#elif defined COMPILE_PCRE32
          if (utf && c > 0x10ffffU) { overflow = TRUE; break; }
#endif
          }
1376

1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404
        if (overflow)
          {
          while (MAX_255(*ptr) && (digitab[*ptr] & ctype_xdigit) != 0) ptr++;
          *errorcodeptr = ERR34;
          }

        else if (*ptr == CHAR_RIGHT_CURLY_BRACKET)
          {
          if (utf && c >= 0xd800 && c <= 0xdfff) *errorcodeptr = ERR73;
          }

        /* If the sequence of hex digits does not end with '}', give an error.
        We used just to recognize this construct and fall through to the normal
        \x handling, but nowadays Perl gives an error, which seems much more
        sensible, so we do too. */

        else *errorcodeptr = ERR79;
        }   /* End of \x{} processing */

      /* Read a single-byte hex-defined char (up to two hex digits after \x) */

      else
        {
        c = 0;
        while (i++ < 2 && MAX_255(ptr[1]) && (digitab[ptr[1]] & ctype_xdigit) != 0)
          {
          pcre_uint32 cc;                          /* Some compilers don't like */
          cc = *(++ptr);                           /* ++ in initializers */
1405
#ifndef EBCDIC  /* ASCII/UTF-8 coding */
1406 1407
          if (cc >= CHAR_a) cc -= 32;              /* Convert to upper case */
          c = c * 16 + cc - ((cc < CHAR_A)? CHAR_0 : (CHAR_A - 10));
1408
#else           /* EBCDIC coding */
1409 1410
          if (cc <= CHAR_z) cc += 64;              /* Convert to upper case */
          c = c * 16 + cc - ((cc >= CHAR_0)? CHAR_0 : (CHAR_A - 10));
1411
#endif
1412 1413 1414
          }
        }     /* End of \xdd handling */
      }       /* End of Perl-style \x handling */
1415 1416 1417
    break;

    /* For \c, a following letter is upper-cased; then the 0x40 bit is flipped.
1418 1419
    An error is given if the byte following \c is not an ASCII character. This
    coding is ASCII-specific, but then the whole concept of \cx is
1420 1421
    ASCII-specific. (However, an EBCDIC equivalent has now been added.) */

1422
    case CHAR_c:
1423
    c = *(++ptr);
1424
    if (c == CHAR_NULL)
1425 1426 1427 1428
      {
      *errorcodeptr = ERR2;
      break;
      }
1429 1430 1431 1432 1433 1434
#ifndef EBCDIC    /* ASCII/UTF-8 coding */
    if (c > 127)  /* Excludes all non-ASCII in either mode */
      {
      *errorcodeptr = ERR68;
      break;
      }
1435
    if (c >= CHAR_a && c <= CHAR_z) c -= 32;
1436
    c ^= 0x40;
1437
#else             /* EBCDIC coding */
1438
    if (c >= CHAR_a && c <= CHAR_z) c += 64;
1439 1440 1441 1442 1443 1444 1445 1446 1447 1448
    if (c == CHAR_QUESTION_MARK)
      c = ('\\' == 188 && '`' == 74)? 0x5f : 0xff;
    else
      {
      for (i = 0; i < 32; i++)
        {
        if (c == ebcdic_escape_c[i]) break;
        }
      if (i < 32) c = i; else *errorcodeptr = ERR68;
      }
1449 1450 1451 1452
#endif
    break;

    /* PCRE_EXTRA enables extensions to Perl in the matter of escapes. Any
1453 1454 1455 1456
    other alphanumeric following \ is an error if PCRE_EXTRA was set;
    otherwise, for Perl compatibility, it is a literal. This code looks a bit
    odd, but there used to be some cases other than the default, and there may
    be again in future, so I haven't "optimized" it. */
1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468

    default:
    if ((options & PCRE_EXTRA) != 0) switch(c)
      {
      default:
      *errorcodeptr = ERR3;
      break;
      }
    break;
    }
  }

1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483
/* Perl supports \N{name} for character names, as well as plain \N for "not
newline". PCRE does not support \N{name}. However, it does support
quantification such as \N{2,3}. */

if (escape == ESC_N && ptr[1] == CHAR_LEFT_CURLY_BRACKET &&
     !is_counted_repeat(ptr+2))
  *errorcodeptr = ERR37;

/* If PCRE_UCP is set, we change the values for \d etc. */

if ((options & PCRE_UCP) != 0 && escape >= ESC_D && escape <= ESC_w)
  escape += (ESC_DU - ESC_D);

/* Set the pointer to the final character before returning. */

1484
*ptrptr = ptr;
1485 1486
*chptr = c;
return escape;
1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503
}



#ifdef SUPPORT_UCP
/*************************************************
*               Handle \P and \p                 *
*************************************************/

/* This function is called after \P or \p has been encountered, provided that
PCRE is compiled with support for Unicode properties. On entry, ptrptr is
pointing at the P or p. On exit, it is pointing at the final character of the
escape sequence.

Argument:
  ptrptr         points to the pattern position pointer
  negptr         points to a boolean that is set TRUE for negation else FALSE
1504 1505
  ptypeptr       points to an unsigned int that is set to the type value
  pdataptr       points to an unsigned int that is set to the detailed property value
1506 1507
  errorcodeptr   points to the error code variable

1508
Returns:         TRUE if the type value was found, or FALSE for an invalid type
1509 1510
*/

1511 1512 1513
static BOOL
get_ucp(const pcre_uchar **ptrptr, BOOL *negptr, unsigned int *ptypeptr,
  unsigned int *pdataptr, int *errorcodeptr)
1514
{
1515 1516 1517 1518
pcre_uchar c;
int i, bot, top;
const pcre_uchar *ptr = *ptrptr;
pcre_uchar name[32];
1519 1520

c = *(++ptr);
1521
if (c == CHAR_NULL) goto ERROR_RETURN;
1522 1523 1524 1525 1526 1527

*negptr = FALSE;

/* \P or \p can be followed by a name in {}, optionally preceded by ^ for
negation. */

1528
if (c == CHAR_LEFT_CURLY_BRACKET)
1529
  {
1530
  if (ptr[1] == CHAR_CIRCUMFLEX_ACCENT)
1531 1532 1533 1534
    {
    *negptr = TRUE;
    ptr++;
    }
1535
  for (i = 0; i < (int)(sizeof(name) / sizeof(pcre_uchar)) - 1; i++)
1536 1537
    {
    c = *(++ptr);
1538
    if (c == CHAR_NULL) goto ERROR_RETURN;
1539
    if (c == CHAR_RIGHT_CURLY_BRACKET) break;
1540 1541
    name[i] = c;
    }
1542
  if (c != CHAR_RIGHT_CURLY_BRACKET) goto ERROR_RETURN;
1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558
  name[i] = 0;
  }

/* Otherwise there is just one following character */

else
  {
  name[0] = c;
  name[1] = 0;
  }

*ptrptr = ptr;

/* Search for a recognized property name using binary chop */

bot = 0;
1559
top = PRIV(utt_size);
1560 1561 1562

while (bot < top)
  {
1563
  int r;
1564
  i = (bot + top) >> 1;
1565 1566
  r = STRCMP_UC_C8(name, PRIV(utt_names) + PRIV(utt)[i].name_offset);
  if (r == 0)
1567
    {
1568 1569 1570
    *ptypeptr = PRIV(utt)[i].type;
    *pdataptr = PRIV(utt)[i].value;
    return TRUE;
1571
    }
1572
  if (r > 0) bot = i + 1; else top = i;
1573 1574 1575 1576
  }

*errorcodeptr = ERR47;
*ptrptr = ptr;
1577
return FALSE;
1578 1579 1580 1581

ERROR_RETURN:
*errorcodeptr = ERR46;
*ptrptr = ptr;
1582
return FALSE;
1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606
}
#endif



/*************************************************
*         Read repeat counts                     *
*************************************************/

/* Read an item of the form {n,m} and return the values. This is called only
after is_counted_repeat() has confirmed that a repeat-count quantifier exists,
so the syntax is guaranteed to be correct, but we need to check the values.

Arguments:
  p              pointer to first char after '{'
  minp           pointer to int for min
  maxp           pointer to int for max
                 returned as -1 if no max
  errorcodeptr   points to error code variable

Returns:         pointer to '}' on success;
                 current ptr on error, with errorcodeptr set non-zero
*/

1607 1608
static const pcre_uchar *
read_repeat_counts(const pcre_uchar *p, int *minp, int *maxp, int *errorcodeptr)
1609 1610 1611 1612
{
int min = 0;
int max = -1;

1613
while (IS_DIGIT(*p))
1614
  {
1615 1616 1617 1618 1619 1620
  min = min * 10 + (int)(*p++ - CHAR_0);
  if (min > 65535)
    {
    *errorcodeptr = ERR5;
    return p;
    }
1621 1622
  }

1623
if (*p == CHAR_RIGHT_CURLY_BRACKET) max = min; else
1624
  {
1625
  if (*(++p) != CHAR_RIGHT_CURLY_BRACKET)
1626 1627
    {
    max = 0;
1628
    while(IS_DIGIT(*p))
1629
      {
1630 1631 1632 1633 1634 1635
      max = max * 10 + (int)(*p++ - CHAR_0);
      if (max > 65535)
        {
        *errorcodeptr = ERR5;
        return p;
        }
1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652
      }
    if (max < min)
      {
      *errorcodeptr = ERR4;
      return p;
      }
    }
  }

*minp = min;
*maxp = max;
return p;
}



/*************************************************
1653
*      Find first significant op code            *
1654 1655
*************************************************/

1656 1657 1658 1659 1660
/* This is called by several functions that scan a compiled expression looking
for a fixed first character, or an anchoring op code etc. It skips over things
that do not influence this. For some calls, it makes sense to skip negative
forward and all backward assertions, and also the \b assertion; for others it
does not.
1661 1662

Arguments:
1663 1664 1665 1666
  code         pointer to the start of the group
  skipassert   TRUE if certain assertions are to be skipped

Returns:       pointer to the first significant opcode
1667 1668
*/

1669 1670
static const pcre_uchar*
first_significant_code(const pcre_uchar *code, BOOL skipassert)
1671 1672 1673 1674 1675 1676 1677 1678 1679 1680
{
for (;;)
  {
  switch ((int)*code)
    {
    case OP_ASSERT_NOT:
    case OP_ASSERTBACK:
    case OP_ASSERTBACK_NOT:
    if (!skipassert) return code;
    do code += GET(code, 1); while (*code == OP_ALT);
1681
    code += PRIV(OP_lengths)[*code];
1682 1683 1684 1685 1686 1687 1688 1689 1690
    break;

    case OP_WORD_BOUNDARY:
    case OP_NOT_WORD_BOUNDARY:
    if (!skipassert) return code;
    /* Fall through */

    case OP_CALLOUT:
    case OP_CREF:
1691
    case OP_DNCREF:
1692
    case OP_RREF:
1693
    case OP_DNRREF:
1694
    case OP_DEF:
1695
    code += PRIV(OP_lengths)[*code];
1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707
    break;

    default:
    return code;
    }
  }
/* Control never reaches here */
}



/*************************************************
1708
*        Find the fixed length of a branch       *
1709 1710
*************************************************/

1711
/* Scan a branch and compute the fixed length of subject that will match it,
1712
if the length is fixed. This is needed for dealing with backward assertions.
1713 1714 1715 1716 1717 1718 1719 1720
In UTF8 mode, the result is in characters rather than bytes. The branch is
temporarily terminated with OP_END when this function is called.

This function is called when a backward assertion is encountered, so that if it
fails, the error message can point to the correct place in the pattern.
However, we cannot do this when the assertion contains subroutine calls,
because they can be forward references. We solve this by remembering this case
and doing the check at the end; a flag specifies which mode we are running in.
1721 1722 1723

Arguments:
  code     points to the start of the pattern (the bracket)
1724 1725 1726 1727 1728 1729 1730 1731 1732 1733
  utf      TRUE in UTF-8 / UTF-16 / UTF-32 mode
  atend    TRUE if called when the pattern is complete
  cd       the "compile data" structure
  recurses    chain of recurse_check to catch mutual recursion

Returns:   the fixed length,
             or -1 if there is no fixed length,
             or -2 if \C was encountered (in UTF-8 mode only)
             or -3 if an OP_RECURSE item was encountered and atend is FALSE
             or -4 if an unknown opcode was encountered (internal error)
1734 1735 1736
*/

static int
1737 1738
find_fixedlength(pcre_uchar *code, BOOL utf, BOOL atend, compile_data *cd,
  recurse_check *recurses)
1739 1740
{
int length = -1;
1741
recurse_check this_recurse;
1742
register int branchlength = 0;
1743
register pcre_uchar *cc = code + 1 + LINK_SIZE;
1744 1745 1746 1747 1748 1749 1750

/* Scan along the opcodes for this branch. If we get to the end of the
branch, check the length against that of the other branches. */

for (;;)
  {
  int d;
1751 1752 1753
  pcre_uchar *ce, *cs;
  register pcre_uchar op = *cc;

1754 1755
  switch (op)
    {
1756 1757 1758 1759 1760
    /* We only need to continue for OP_CBRA (normal capturing bracket) and
    OP_BRA (normal non-capturing bracket) because the other variants of these
    opcodes are all concerned with unlimited repeated groups, which of course
    are not of fixed length. */

1761 1762 1763
    case OP_CBRA:
    case OP_BRA:
    case OP_ONCE:
1764
    case OP_ONCE_NC:
1765
    case OP_COND:
1766 1767
    d = find_fixedlength(cc + ((op == OP_CBRA)? IMM2_SIZE : 0), utf, atend, cd,
      recurses);
1768 1769 1770 1771 1772 1773
    if (d < 0) return d;
    branchlength += d;
    do cc += GET(cc, 1); while (*cc == OP_ALT);
    cc += 1 + LINK_SIZE;
    break;

1774 1775 1776 1777 1778
    /* Reached end of a branch; if it's a ket it is the end of a nested call.
    If it's ALT it is an alternation in a nested call. An ACCEPT is effectively
    an ALT. If it is END it's the end of the outer call. All can be handled by
    the same code. Note that we must not include the OP_KETRxxx opcodes here,
    because they all imply an unlimited repeat. */
1779 1780 1781 1782

    case OP_ALT:
    case OP_KET:
    case OP_END:
1783 1784
    case OP_ACCEPT:
    case OP_ASSERT_ACCEPT:
1785 1786 1787 1788 1789 1790 1791
    if (length < 0) length = branchlength;
      else if (length != branchlength) return -1;
    if (*cc != OP_ALT) return length;
    cc += 1 + LINK_SIZE;
    branchlength = 0;
    break;

1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814
    /* A true recursion implies not fixed length, but a subroutine call may
    be OK. If the subroutine is a forward reference, we can't deal with
    it until the end of the pattern, so return -3. */

    case OP_RECURSE:
    if (!atend) return -3;
    cs = ce = (pcre_uchar *)cd->start_code + GET(cc, 1);  /* Start subpattern */
    do ce += GET(ce, 1); while (*ce == OP_ALT);           /* End subpattern */
    if (cc > cs && cc < ce) return -1;                    /* Recursion */
    else   /* Check for mutual recursion */
      {
      recurse_check *r = recurses;
      for (r = recurses; r != NULL; r = r->prev) if (r->group == cs) break;
      if (r != NULL) return -1;   /* Mutual recursion */
      }
    this_recurse.prev = recurses;
    this_recurse.group = cs;
    d = find_fixedlength(cs + IMM2_SIZE, utf, atend, cd, &this_recurse);
    if (d < 0) return d;
    branchlength += d;
    cc += 1 + LINK_SIZE;
    break;

1815 1816 1817 1818 1819 1820 1821
    /* Skip over assertive subpatterns */

    case OP_ASSERT:
    case OP_ASSERT_NOT:
    case OP_ASSERTBACK:
    case OP_ASSERTBACK_NOT:
    do cc += GET(cc, 1); while (*cc == OP_ALT);
1822 1823
    cc += 1 + LINK_SIZE;
    break;
1824 1825 1826