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

Contents of /code/trunk/pcre_compile.c

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1059 - (show annotations)
Tue Oct 16 15:53:53 2012 UTC (6 years, 11 months ago) by chpe
File MIME type: text/plain
File size: 272074 byte(s)
pcre32: compile: Make check_escape return the data character in an out param

check_escape needs to return both the escape code and possibly a data
character. Return the data character in an out param instead of mixing
it with the escape code; this is in preparation to making the character
a pcre_uint32 to enable the full 32-bit range in pcre32 in non-UTF-32
mode.
1 /*************************************************
2 * Perl-Compatible Regular Expressions *
3 *************************************************/
4
5 /* PCRE is a library of functions to support regular expressions whose syntax
6 and semantics are as close as possible to those of the Perl 5 language.
7
8 Written by Philip Hazel
9 Copyright (c) 1997-2012 University of Cambridge
10
11 -----------------------------------------------------------------------------
12 Redistribution and use in source and binary forms, with or without
13 modification, are permitted provided that the following conditions are met:
14
15 * Redistributions of source code must retain the above copyright notice,
16 this list of conditions and the following disclaimer.
17
18 * Redistributions in binary form must reproduce the above copyright
19 notice, this list of conditions and the following disclaimer in the
20 documentation and/or other materials provided with the distribution.
21
22 * Neither the name of the University of Cambridge nor the names of its
23 contributors may be used to endorse or promote products derived from
24 this software without specific prior written permission.
25
26 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
27 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29 ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
30 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
31 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
32 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
33 INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
34 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
35 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
36 POSSIBILITY OF SUCH DAMAGE.
37 -----------------------------------------------------------------------------
38 */
39
40
41 /* This module contains the external function pcre_compile(), along with
42 supporting internal functions that are not used by other modules. */
43
44
45 #ifdef HAVE_CONFIG_H
46 #include "config.h"
47 #endif
48
49 #define NLBLOCK cd /* Block containing newline information */
50 #define PSSTART start_pattern /* Field containing processed string start */
51 #define PSEND end_pattern /* Field containing processed string end */
52
53 #include "pcre_internal.h"
54
55
56 /* When PCRE_DEBUG is defined, we need the pcre(16|32)_printint() function, which
57 is also used by pcretest. PCRE_DEBUG is not defined when building a production
58 library. We do not need to select pcre16_printint.c specially, because the
59 COMPILE_PCREx macro will already be appropriately set. */
60
61 #ifdef PCRE_DEBUG
62 /* pcre_printint.c should not include any headers */
63 #define PCRE_INCLUDED
64 #include "pcre_printint.c"
65 #undef PCRE_INCLUDED
66 #endif
67
68
69 /* Macro for setting individual bits in class bitmaps. */
70
71 #define SETBIT(a,b) a[(b)/8] |= (1 << ((b)&7))
72
73 /* Maximum length value to check against when making sure that the integer that
74 holds the compiled pattern length does not overflow. We make it a bit less than
75 INT_MAX to allow for adding in group terminating bytes, so that we don't have
76 to check them every time. */
77
78 #define OFLOW_MAX (INT_MAX - 20)
79
80 /* Definitions to allow mutual recursion */
81
82 static int
83 add_list_to_class(pcre_uint8 *, pcre_uchar **, int, compile_data *,
84 const pcre_uint32 *, unsigned int);
85
86 static BOOL
87 compile_regex(int, pcre_uchar **, const pcre_uchar **, int *, BOOL, BOOL,
88 int, int, int *, int *, branch_chain *, compile_data *, int *);
89
90
91
92 /*************************************************
93 * Code parameters and static tables *
94 *************************************************/
95
96 /* This value specifies the size of stack workspace that is used during the
97 first pre-compile phase that determines how much memory is required. The regex
98 is partly compiled into this space, but the compiled parts are discarded as
99 soon as they can be, so that hopefully there will never be an overrun. The code
100 does, however, check for an overrun. The largest amount I've seen used is 218,
101 so this number is very generous.
102
103 The same workspace is used during the second, actual compile phase for
104 remembering forward references to groups so that they can be filled in at the
105 end. Each entry in this list occupies LINK_SIZE bytes, so even when LINK_SIZE
106 is 4 there is plenty of room for most patterns. However, the memory can get
107 filled up by repetitions of forward references, for example patterns like
108 /(?1){0,1999}(b)/, and one user did hit the limit. The code has been changed so
109 that the workspace is expanded using malloc() in this situation. The value
110 below is therefore a minimum, and we put a maximum on it for safety. The
111 minimum is now also defined in terms of LINK_SIZE so that the use of malloc()
112 kicks in at the same number of forward references in all cases. */
113
114 #define COMPILE_WORK_SIZE (2048*LINK_SIZE)
115 #define COMPILE_WORK_SIZE_MAX (100*COMPILE_WORK_SIZE)
116
117 /* The overrun tests check for a slightly smaller size so that they detect the
118 overrun before it actually does run off the end of the data block. */
119
120 #define WORK_SIZE_SAFETY_MARGIN (100)
121
122 /* Private flags added to firstchar and reqchar. */
123
124 #define REQ_CASELESS 0x10000000l /* Indicates caselessness */
125 #define REQ_VARY 0x20000000l /* Reqchar followed non-literal item */
126 #define REQ_MASK (REQ_CASELESS | REQ_VARY)
127
128 /* Repeated character flags. */
129
130 #define UTF_LENGTH 0x10000000l /* The char contains its length. */
131
132 /* Table for handling escaped characters in the range '0'-'z'. Positive returns
133 are simple data values; negative values are for special things like \d and so
134 on. Zero means further processing is needed (for things like \x), or the escape
135 is invalid. */
136
137 #ifndef EBCDIC
138
139 /* This is the "normal" table for ASCII systems or for EBCDIC systems running
140 in UTF-8 mode. */
141
142 static const short int escapes[] = {
143 0, 0,
144 0, 0,
145 0, 0,
146 0, 0,
147 0, 0,
148 CHAR_COLON, CHAR_SEMICOLON,
149 CHAR_LESS_THAN_SIGN, CHAR_EQUALS_SIGN,
150 CHAR_GREATER_THAN_SIGN, CHAR_QUESTION_MARK,
151 CHAR_COMMERCIAL_AT, -ESC_A,
152 -ESC_B, -ESC_C,
153 -ESC_D, -ESC_E,
154 0, -ESC_G,
155 -ESC_H, 0,
156 0, -ESC_K,
157 0, 0,
158 -ESC_N, 0,
159 -ESC_P, -ESC_Q,
160 -ESC_R, -ESC_S,
161 0, 0,
162 -ESC_V, -ESC_W,
163 -ESC_X, 0,
164 -ESC_Z, CHAR_LEFT_SQUARE_BRACKET,
165 CHAR_BACKSLASH, CHAR_RIGHT_SQUARE_BRACKET,
166 CHAR_CIRCUMFLEX_ACCENT, CHAR_UNDERSCORE,
167 CHAR_GRAVE_ACCENT, 7,
168 -ESC_b, 0,
169 -ESC_d, ESC_e,
170 ESC_f, 0,
171 -ESC_h, 0,
172 0, -ESC_k,
173 0, 0,
174 ESC_n, 0,
175 -ESC_p, 0,
176 ESC_r, -ESC_s,
177 ESC_tee, 0,
178 -ESC_v, -ESC_w,
179 0, 0,
180 -ESC_z
181 };
182
183 #else
184
185 /* This is the "abnormal" table for EBCDIC systems without UTF-8 support. */
186
187 static const short int escapes[] = {
188 /* 48 */ 0, 0, 0, '.', '<', '(', '+', '|',
189 /* 50 */ '&', 0, 0, 0, 0, 0, 0, 0,
190 /* 58 */ 0, 0, '!', '$', '*', ')', ';', '~',
191 /* 60 */ '-', '/', 0, 0, 0, 0, 0, 0,
192 /* 68 */ 0, 0, '|', ',', '%', '_', '>', '?',
193 /* 70 */ 0, 0, 0, 0, 0, 0, 0, 0,
194 /* 78 */ 0, '`', ':', '#', '@', '\'', '=', '"',
195 /* 80 */ 0, 7, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0,
196 /* 88 */-ESC_h, 0, 0, '{', 0, 0, 0, 0,
197 /* 90 */ 0, 0, -ESC_k, 'l', 0, ESC_n, 0, -ESC_p,
198 /* 98 */ 0, ESC_r, 0, '}', 0, 0, 0, 0,
199 /* A0 */ 0, '~', -ESC_s, ESC_tee, 0,-ESC_v, -ESC_w, 0,
200 /* A8 */ 0,-ESC_z, 0, 0, 0, '[', 0, 0,
201 /* B0 */ 0, 0, 0, 0, 0, 0, 0, 0,
202 /* B8 */ 0, 0, 0, 0, 0, ']', '=', '-',
203 /* C0 */ '{',-ESC_A, -ESC_B, -ESC_C, -ESC_D,-ESC_E, 0, -ESC_G,
204 /* C8 */-ESC_H, 0, 0, 0, 0, 0, 0, 0,
205 /* D0 */ '}', 0, -ESC_K, 0, 0,-ESC_N, 0, -ESC_P,
206 /* D8 */-ESC_Q,-ESC_R, 0, 0, 0, 0, 0, 0,
207 /* E0 */ '\\', 0, -ESC_S, 0, 0,-ESC_V, -ESC_W, -ESC_X,
208 /* E8 */ 0,-ESC_Z, 0, 0, 0, 0, 0, 0,
209 /* F0 */ 0, 0, 0, 0, 0, 0, 0, 0,
210 /* F8 */ 0, 0, 0, 0, 0, 0, 0, 0
211 };
212 #endif
213
214
215 /* Table of special "verbs" like (*PRUNE). This is a short table, so it is
216 searched linearly. Put all the names into a single string, in order to reduce
217 the number of relocations when a shared library is dynamically linked. The
218 string is built from string macros so that it works in UTF-8 mode on EBCDIC
219 platforms. */
220
221 typedef struct verbitem {
222 int len; /* Length of verb name */
223 int op; /* Op when no arg, or -1 if arg mandatory */
224 int op_arg; /* Op when arg present, or -1 if not allowed */
225 } verbitem;
226
227 static const char verbnames[] =
228 "\0" /* Empty name is a shorthand for MARK */
229 STRING_MARK0
230 STRING_ACCEPT0
231 STRING_COMMIT0
232 STRING_F0
233 STRING_FAIL0
234 STRING_PRUNE0
235 STRING_SKIP0
236 STRING_THEN;
237
238 static const verbitem verbs[] = {
239 { 0, -1, OP_MARK },
240 { 4, -1, OP_MARK },
241 { 6, OP_ACCEPT, -1 },
242 { 6, OP_COMMIT, -1 },
243 { 1, OP_FAIL, -1 },
244 { 4, OP_FAIL, -1 },
245 { 5, OP_PRUNE, OP_PRUNE_ARG },
246 { 4, OP_SKIP, OP_SKIP_ARG },
247 { 4, OP_THEN, OP_THEN_ARG }
248 };
249
250 static const int verbcount = sizeof(verbs)/sizeof(verbitem);
251
252
253 /* Tables of names of POSIX character classes and their lengths. The names are
254 now all in a single string, to reduce the number of relocations when a shared
255 library is dynamically loaded. The list of lengths is terminated by a zero
256 length entry. The first three must be alpha, lower, upper, as this is assumed
257 for handling case independence. */
258
259 static const char posix_names[] =
260 STRING_alpha0 STRING_lower0 STRING_upper0 STRING_alnum0
261 STRING_ascii0 STRING_blank0 STRING_cntrl0 STRING_digit0
262 STRING_graph0 STRING_print0 STRING_punct0 STRING_space0
263 STRING_word0 STRING_xdigit;
264
265 static const pcre_uint8 posix_name_lengths[] = {
266 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 4, 6, 0 };
267
268 /* Table of class bit maps for each POSIX class. Each class is formed from a
269 base map, with an optional addition or removal of another map. Then, for some
270 classes, there is some additional tweaking: for [:blank:] the vertical space
271 characters are removed, and for [:alpha:] and [:alnum:] the underscore
272 character is removed. The triples in the table consist of the base map offset,
273 second map offset or -1 if no second map, and a non-negative value for map
274 addition or a negative value for map subtraction (if there are two maps). The
275 absolute value of the third field has these meanings: 0 => no tweaking, 1 =>
276 remove vertical space characters, 2 => remove underscore. */
277
278 static const int posix_class_maps[] = {
279 cbit_word, cbit_digit, -2, /* alpha */
280 cbit_lower, -1, 0, /* lower */
281 cbit_upper, -1, 0, /* upper */
282 cbit_word, -1, 2, /* alnum - word without underscore */
283 cbit_print, cbit_cntrl, 0, /* ascii */
284 cbit_space, -1, 1, /* blank - a GNU extension */
285 cbit_cntrl, -1, 0, /* cntrl */
286 cbit_digit, -1, 0, /* digit */
287 cbit_graph, -1, 0, /* graph */
288 cbit_print, -1, 0, /* print */
289 cbit_punct, -1, 0, /* punct */
290 cbit_space, -1, 0, /* space */
291 cbit_word, -1, 0, /* word - a Perl extension */
292 cbit_xdigit,-1, 0 /* xdigit */
293 };
294
295 /* Table of substitutes for \d etc when PCRE_UCP is set. The POSIX class
296 substitutes must be in the order of the names, defined above, and there are
297 both positive and negative cases. NULL means no substitute. */
298
299 #ifdef SUPPORT_UCP
300 static const pcre_uchar string_PNd[] = {
301 CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
302 CHAR_N, CHAR_d, CHAR_RIGHT_CURLY_BRACKET, '\0' };
303 static const pcre_uchar string_pNd[] = {
304 CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
305 CHAR_N, CHAR_d, CHAR_RIGHT_CURLY_BRACKET, '\0' };
306 static const pcre_uchar string_PXsp[] = {
307 CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
308 CHAR_X, CHAR_s, CHAR_p, CHAR_RIGHT_CURLY_BRACKET, '\0' };
309 static const pcre_uchar string_pXsp[] = {
310 CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
311 CHAR_X, CHAR_s, CHAR_p, CHAR_RIGHT_CURLY_BRACKET, '\0' };
312 static const pcre_uchar string_PXwd[] = {
313 CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
314 CHAR_X, CHAR_w, CHAR_d, CHAR_RIGHT_CURLY_BRACKET, '\0' };
315 static const pcre_uchar string_pXwd[] = {
316 CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
317 CHAR_X, CHAR_w, CHAR_d, CHAR_RIGHT_CURLY_BRACKET, '\0' };
318
319 static const pcre_uchar *substitutes[] = {
320 string_PNd, /* \D */
321 string_pNd, /* \d */
322 string_PXsp, /* \S */ /* NOTE: Xsp is Perl space */
323 string_pXsp, /* \s */
324 string_PXwd, /* \W */
325 string_pXwd /* \w */
326 };
327
328 static const pcre_uchar string_pL[] = {
329 CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
330 CHAR_L, CHAR_RIGHT_CURLY_BRACKET, '\0' };
331 static const pcre_uchar string_pLl[] = {
332 CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
333 CHAR_L, CHAR_l, CHAR_RIGHT_CURLY_BRACKET, '\0' };
334 static const pcre_uchar string_pLu[] = {
335 CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
336 CHAR_L, CHAR_u, CHAR_RIGHT_CURLY_BRACKET, '\0' };
337 static const pcre_uchar string_pXan[] = {
338 CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
339 CHAR_X, CHAR_a, CHAR_n, CHAR_RIGHT_CURLY_BRACKET, '\0' };
340 static const pcre_uchar string_h[] = {
341 CHAR_BACKSLASH, CHAR_h, '\0' };
342 static const pcre_uchar string_pXps[] = {
343 CHAR_BACKSLASH, CHAR_p, CHAR_LEFT_CURLY_BRACKET,
344 CHAR_X, CHAR_p, CHAR_s, CHAR_RIGHT_CURLY_BRACKET, '\0' };
345 static const pcre_uchar string_PL[] = {
346 CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
347 CHAR_L, CHAR_RIGHT_CURLY_BRACKET, '\0' };
348 static const pcre_uchar string_PLl[] = {
349 CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
350 CHAR_L, CHAR_l, CHAR_RIGHT_CURLY_BRACKET, '\0' };
351 static const pcre_uchar string_PLu[] = {
352 CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
353 CHAR_L, CHAR_u, CHAR_RIGHT_CURLY_BRACKET, '\0' };
354 static const pcre_uchar string_PXan[] = {
355 CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
356 CHAR_X, CHAR_a, CHAR_n, CHAR_RIGHT_CURLY_BRACKET, '\0' };
357 static const pcre_uchar string_H[] = {
358 CHAR_BACKSLASH, CHAR_H, '\0' };
359 static const pcre_uchar string_PXps[] = {
360 CHAR_BACKSLASH, CHAR_P, CHAR_LEFT_CURLY_BRACKET,
361 CHAR_X, CHAR_p, CHAR_s, CHAR_RIGHT_CURLY_BRACKET, '\0' };
362
363 static const pcre_uchar *posix_substitutes[] = {
364 string_pL, /* alpha */
365 string_pLl, /* lower */
366 string_pLu, /* upper */
367 string_pXan, /* alnum */
368 NULL, /* ascii */
369 string_h, /* blank */
370 NULL, /* cntrl */
371 string_pNd, /* digit */
372 NULL, /* graph */
373 NULL, /* print */
374 NULL, /* punct */
375 string_pXps, /* space */ /* NOTE: Xps is POSIX space */
376 string_pXwd, /* word */
377 NULL, /* xdigit */
378 /* Negated cases */
379 string_PL, /* ^alpha */
380 string_PLl, /* ^lower */
381 string_PLu, /* ^upper */
382 string_PXan, /* ^alnum */
383 NULL, /* ^ascii */
384 string_H, /* ^blank */
385 NULL, /* ^cntrl */
386 string_PNd, /* ^digit */
387 NULL, /* ^graph */
388 NULL, /* ^print */
389 NULL, /* ^punct */
390 string_PXps, /* ^space */ /* NOTE: Xps is POSIX space */
391 string_PXwd, /* ^word */
392 NULL /* ^xdigit */
393 };
394 #define POSIX_SUBSIZE (sizeof(posix_substitutes) / sizeof(pcre_uchar *))
395 #endif
396
397 #define STRING(a) # a
398 #define XSTRING(s) STRING(s)
399
400 /* The texts of compile-time error messages. These are "char *" because they
401 are passed to the outside world. Do not ever re-use any error number, because
402 they are documented. Always add a new error instead. Messages marked DEAD below
403 are no longer used. This used to be a table of strings, but in order to reduce
404 the number of relocations needed when a shared library is loaded dynamically,
405 it is now one long string. We cannot use a table of offsets, because the
406 lengths of inserts such as XSTRING(MAX_NAME_SIZE) are not known. Instead, we
407 simply count through to the one we want - this isn't a performance issue
408 because these strings are used only when there is a compilation error.
409
410 Each substring ends with \0 to insert a null character. This includes the final
411 substring, so that the whole string ends with \0\0, which can be detected when
412 counting through. */
413
414 static const char error_texts[] =
415 "no error\0"
416 "\\ at end of pattern\0"
417 "\\c at end of pattern\0"
418 "unrecognized character follows \\\0"
419 "numbers out of order in {} quantifier\0"
420 /* 5 */
421 "number too big in {} quantifier\0"
422 "missing terminating ] for character class\0"
423 "invalid escape sequence in character class\0"
424 "range out of order in character class\0"
425 "nothing to repeat\0"
426 /* 10 */
427 "operand of unlimited repeat could match the empty string\0" /** DEAD **/
428 "internal error: unexpected repeat\0"
429 "unrecognized character after (? or (?-\0"
430 "POSIX named classes are supported only within a class\0"
431 "missing )\0"
432 /* 15 */
433 "reference to non-existent subpattern\0"
434 "erroffset passed as NULL\0"
435 "unknown option bit(s) set\0"
436 "missing ) after comment\0"
437 "parentheses nested too deeply\0" /** DEAD **/
438 /* 20 */
439 "regular expression is too large\0"
440 "failed to get memory\0"
441 "unmatched parentheses\0"
442 "internal error: code overflow\0"
443 "unrecognized character after (?<\0"
444 /* 25 */
445 "lookbehind assertion is not fixed length\0"
446 "malformed number or name after (?(\0"
447 "conditional group contains more than two branches\0"
448 "assertion expected after (?(\0"
449 "(?R or (?[+-]digits must be followed by )\0"
450 /* 30 */
451 "unknown POSIX class name\0"
452 "POSIX collating elements are not supported\0"
453 "this version of PCRE is compiled without UTF support\0"
454 "spare error\0" /** DEAD **/
455 "character value in \\x{...} sequence is too large\0"
456 /* 35 */
457 "invalid condition (?(0)\0"
458 "\\C not allowed in lookbehind assertion\0"
459 "PCRE does not support \\L, \\l, \\N{name}, \\U, or \\u\0"
460 "number after (?C is > 255\0"
461 "closing ) for (?C expected\0"
462 /* 40 */
463 "recursive call could loop indefinitely\0"
464 "unrecognized character after (?P\0"
465 "syntax error in subpattern name (missing terminator)\0"
466 "two named subpatterns have the same name\0"
467 "invalid UTF-8 string\0"
468 /* 45 */
469 "support for \\P, \\p, and \\X has not been compiled\0"
470 "malformed \\P or \\p sequence\0"
471 "unknown property name after \\P or \\p\0"
472 "subpattern name is too long (maximum " XSTRING(MAX_NAME_SIZE) " characters)\0"
473 "too many named subpatterns (maximum " XSTRING(MAX_NAME_COUNT) ")\0"
474 /* 50 */
475 "repeated subpattern is too long\0" /** DEAD **/
476 "octal value is greater than \\377 in 8-bit non-UTF-8 mode\0"
477 "internal error: overran compiling workspace\0"
478 "internal error: previously-checked referenced subpattern not found\0"
479 "DEFINE group contains more than one branch\0"
480 /* 55 */
481 "repeating a DEFINE group is not allowed\0" /** DEAD **/
482 "inconsistent NEWLINE options\0"
483 "\\g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number\0"
484 "a numbered reference must not be zero\0"
485 "an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)\0"
486 /* 60 */
487 "(*VERB) not recognized\0"
488 "number is too big\0"
489 "subpattern name expected\0"
490 "digit expected after (?+\0"
491 "] is an invalid data character in JavaScript compatibility mode\0"
492 /* 65 */
493 "different names for subpatterns of the same number are not allowed\0"
494 "(*MARK) must have an argument\0"
495 "this version of PCRE is not compiled with Unicode property support\0"
496 "\\c must be followed by an ASCII character\0"
497 "\\k is not followed by a braced, angle-bracketed, or quoted name\0"
498 /* 70 */
499 "internal error: unknown opcode in find_fixedlength()\0"
500 "\\N is not supported in a class\0"
501 "too many forward references\0"
502 "disallowed Unicode code point (>= 0xd800 && <= 0xdfff)\0"
503 "invalid UTF-16 string\0"
504 /* 75 */
505 "name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)\0"
506 "character value in \\u.... sequence is too large\0"
507 "invalid UTF-32 string\0"
508 ;
509
510 /* Table to identify digits and hex digits. This is used when compiling
511 patterns. Note that the tables in chartables are dependent on the locale, and
512 may mark arbitrary characters as digits - but the PCRE compiling code expects
513 to handle only 0-9, a-z, and A-Z as digits when compiling. That is why we have
514 a private table here. It costs 256 bytes, but it is a lot faster than doing
515 character value tests (at least in some simple cases I timed), and in some
516 applications one wants PCRE to compile efficiently as well as match
517 efficiently.
518
519 For convenience, we use the same bit definitions as in chartables:
520
521 0x04 decimal digit
522 0x08 hexadecimal digit
523
524 Then we can use ctype_digit and ctype_xdigit in the code. */
525
526 /* Using a simple comparison for decimal numbers rather than a memory read
527 is much faster, and the resulting code is simpler (the compiler turns it
528 into a subtraction and unsigned comparison). */
529
530 #define IS_DIGIT(x) ((x) >= CHAR_0 && (x) <= CHAR_9)
531
532 #ifndef EBCDIC
533
534 /* This is the "normal" case, for ASCII systems, and EBCDIC systems running in
535 UTF-8 mode. */
536
537 static const pcre_uint8 digitab[] =
538 {
539 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 0- 7 */
540 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 8- 15 */
541 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 16- 23 */
542 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 24- 31 */
543 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* - ' */
544 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* ( - / */
545 0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c, /* 0 - 7 */
546 0x0c,0x0c,0x00,0x00,0x00,0x00,0x00,0x00, /* 8 - ? */
547 0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /* @ - G */
548 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* H - O */
549 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* P - W */
550 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* X - _ */
551 0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /* ` - g */
552 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* h - o */
553 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* p - w */
554 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* x -127 */
555 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 128-135 */
556 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 136-143 */
557 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144-151 */
558 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 152-159 */
559 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160-167 */
560 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 168-175 */
561 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 176-183 */
562 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
563 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 192-199 */
564 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 200-207 */
565 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 208-215 */
566 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 216-223 */
567 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 224-231 */
568 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 232-239 */
569 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 240-247 */
570 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00};/* 248-255 */
571
572 #else
573
574 /* This is the "abnormal" case, for EBCDIC systems not running in UTF-8 mode. */
575
576 static const pcre_uint8 digitab[] =
577 {
578 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 0- 7 0 */
579 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 8- 15 */
580 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 16- 23 10 */
581 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 24- 31 */
582 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 32- 39 20 */
583 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 40- 47 */
584 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 48- 55 30 */
585 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 56- 63 */
586 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* - 71 40 */
587 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 72- | */
588 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* & - 87 50 */
589 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 88- 95 */
590 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* - -103 60 */
591 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 104- ? */
592 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 112-119 70 */
593 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 120- " */
594 0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /* 128- g 80 */
595 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* h -143 */
596 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144- p 90 */
597 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* q -159 */
598 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160- x A0 */
599 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* y -175 */
600 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* ^ -183 B0 */
601 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
602 0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /* { - G C0 */
603 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* H -207 */
604 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* } - P D0 */
605 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* Q -223 */
606 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* \ - X E0 */
607 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* Y -239 */
608 0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c, /* 0 - 7 F0 */
609 0x0c,0x0c,0x00,0x00,0x00,0x00,0x00,0x00};/* 8 -255 */
610
611 static const pcre_uint8 ebcdic_chartab[] = { /* chartable partial dup */
612 0x80,0x00,0x00,0x00,0x00,0x01,0x00,0x00, /* 0- 7 */
613 0x00,0x00,0x00,0x00,0x01,0x01,0x00,0x00, /* 8- 15 */
614 0x00,0x00,0x00,0x00,0x00,0x01,0x00,0x00, /* 16- 23 */
615 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 24- 31 */
616 0x00,0x00,0x00,0x00,0x00,0x01,0x00,0x00, /* 32- 39 */
617 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 40- 47 */
618 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 48- 55 */
619 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 56- 63 */
620 0x01,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* - 71 */
621 0x00,0x00,0x00,0x80,0x00,0x80,0x80,0x80, /* 72- | */
622 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* & - 87 */
623 0x00,0x00,0x00,0x80,0x80,0x80,0x00,0x00, /* 88- 95 */
624 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* - -103 */
625 0x00,0x00,0x00,0x00,0x00,0x10,0x00,0x80, /* 104- ? */
626 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 112-119 */
627 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 120- " */
628 0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /* 128- g */
629 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* h -143 */
630 0x00,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /* 144- p */
631 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* q -159 */
632 0x00,0x00,0x12,0x12,0x12,0x12,0x12,0x12, /* 160- x */
633 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* y -175 */
634 0x80,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* ^ -183 */
635 0x00,0x00,0x80,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
636 0x80,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /* { - G */
637 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* H -207 */
638 0x00,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /* } - P */
639 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* Q -223 */
640 0x00,0x00,0x12,0x12,0x12,0x12,0x12,0x12, /* \ - X */
641 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* Y -239 */
642 0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c, /* 0 - 7 */
643 0x1c,0x1c,0x00,0x00,0x00,0x00,0x00,0x00};/* 8 -255 */
644 #endif
645
646
647
648
649 /*************************************************
650 * Find an error text *
651 *************************************************/
652
653 /* The error texts are now all in one long string, to save on relocations. As
654 some of the text is of unknown length, we can't use a table of offsets.
655 Instead, just count through the strings. This is not a performance issue
656 because it happens only when there has been a compilation error.
657
658 Argument: the error number
659 Returns: pointer to the error string
660 */
661
662 static const char *
663 find_error_text(int n)
664 {
665 const char *s = error_texts;
666 for (; n > 0; n--)
667 {
668 while (*s++ != 0) {};
669 if (*s == 0) return "Error text not found (please report)";
670 }
671 return s;
672 }
673
674
675 /*************************************************
676 * Expand the workspace *
677 *************************************************/
678
679 /* This function is called during the second compiling phase, if the number of
680 forward references fills the existing workspace, which is originally a block on
681 the stack. A larger block is obtained from malloc() unless the ultimate limit
682 has been reached or the increase will be rather small.
683
684 Argument: pointer to the compile data block
685 Returns: 0 if all went well, else an error number
686 */
687
688 static int
689 expand_workspace(compile_data *cd)
690 {
691 pcre_uchar *newspace;
692 int newsize = cd->workspace_size * 2;
693
694 if (newsize > COMPILE_WORK_SIZE_MAX) newsize = COMPILE_WORK_SIZE_MAX;
695 if (cd->workspace_size >= COMPILE_WORK_SIZE_MAX ||
696 newsize - cd->workspace_size < WORK_SIZE_SAFETY_MARGIN)
697 return ERR72;
698
699 newspace = (PUBL(malloc))(IN_UCHARS(newsize));
700 if (newspace == NULL) return ERR21;
701 memcpy(newspace, cd->start_workspace, cd->workspace_size * sizeof(pcre_uchar));
702 cd->hwm = (pcre_uchar *)newspace + (cd->hwm - cd->start_workspace);
703 if (cd->workspace_size > COMPILE_WORK_SIZE)
704 (PUBL(free))((void *)cd->start_workspace);
705 cd->start_workspace = newspace;
706 cd->workspace_size = newsize;
707 return 0;
708 }
709
710
711
712 /*************************************************
713 * Check for counted repeat *
714 *************************************************/
715
716 /* This function is called when a '{' is encountered in a place where it might
717 start a quantifier. It looks ahead to see if it really is a quantifier or not.
718 It is only a quantifier if it is one of the forms {ddd} {ddd,} or {ddd,ddd}
719 where the ddds are digits.
720
721 Arguments:
722 p pointer to the first char after '{'
723
724 Returns: TRUE or FALSE
725 */
726
727 static BOOL
728 is_counted_repeat(const pcre_uchar *p)
729 {
730 if (!IS_DIGIT(*p)) return FALSE;
731 p++;
732 while (IS_DIGIT(*p)) p++;
733 if (*p == CHAR_RIGHT_CURLY_BRACKET) return TRUE;
734
735 if (*p++ != CHAR_COMMA) return FALSE;
736 if (*p == CHAR_RIGHT_CURLY_BRACKET) return TRUE;
737
738 if (!IS_DIGIT(*p)) return FALSE;
739 p++;
740 while (IS_DIGIT(*p)) p++;
741
742 return (*p == CHAR_RIGHT_CURLY_BRACKET);
743 }
744
745
746
747 /*************************************************
748 * Handle escapes *
749 *************************************************/
750
751 /* This function is called when a \ has been encountered. It either returns a
752 positive value for a simple escape such as \n, or 0 for a data character
753 which will be placed in chptr. A backreference to group
754 n is returned as ESC_REF + n; ESC_REF is the highest ESC_xxx macro. When
755 UTF-8 is enabled, a positive value greater than 255 may be returned in chptr.
756 On entry,ptr is pointing at the \. On exit, it is on the final character of the
757 escape sequence.
758
759 Arguments:
760 ptrptr points to the pattern position pointer
761 chptr points to the data character
762 errorcodeptr points to the errorcode variable
763 bracount number of previous extracting brackets
764 options the options bits
765 isclass TRUE if inside a character class
766
767 Returns: zero => a data character
768 positive => a special escape sequence
769 on error, errorcodeptr is set
770 */
771
772 static int
773 check_escape(const pcre_uchar **ptrptr, int *chptr, int *errorcodeptr,
774 int bracount, int options, BOOL isclass)
775 {
776 /* PCRE_UTF16 has the same value as PCRE_UTF8. */
777 BOOL utf = (options & PCRE_UTF8) != 0;
778 const pcre_uchar *ptr = *ptrptr + 1;
779 pcre_int32 c;
780 int escape = 0;
781 int i;
782
783 GETCHARINCTEST(c, ptr); /* Get character value, increment pointer */
784 ptr--; /* Set pointer back to the last byte */
785
786 /* If backslash is at the end of the pattern, it's an error. */
787
788 if (c == 0) *errorcodeptr = ERR1;
789
790 /* Non-alphanumerics are literals. For digits or letters, do an initial lookup
791 in a table. A non-zero result is something that can be returned immediately.
792 Otherwise further processing may be required. */
793
794 #ifndef EBCDIC /* ASCII/UTF-8 coding */
795 /* Not alphanumeric */
796 else if (c < CHAR_0 || c > CHAR_z) {}
797 else if ((i = escapes[c - CHAR_0]) != 0) { if (i > 0) c = i; else escape = -i; }
798
799 #else /* EBCDIC coding */
800 /* Not alphanumeric */
801 else if (c < CHAR_a || (!MAX_255(c) || (ebcdic_chartab[c] & 0x0E) == 0)) {}
802 else if ((i = escapes[c - 0x48]) != 0) { if (i > 0) c = i; else escape = -i; }
803 #endif
804
805 /* Escapes that need further processing, or are illegal. */
806
807 else
808 {
809 const pcre_uchar *oldptr;
810 BOOL braced, negated;
811
812 switch (c)
813 {
814 /* A number of Perl escapes are not handled by PCRE. We give an explicit
815 error. */
816
817 case CHAR_l:
818 case CHAR_L:
819 *errorcodeptr = ERR37;
820 break;
821
822 case CHAR_u:
823 if ((options & PCRE_JAVASCRIPT_COMPAT) != 0)
824 {
825 /* In JavaScript, \u must be followed by four hexadecimal numbers.
826 Otherwise it is a lowercase u letter. */
827 if (MAX_255(ptr[1]) && (digitab[ptr[1]] & ctype_xdigit) != 0
828 && MAX_255(ptr[2]) && (digitab[ptr[2]] & ctype_xdigit) != 0
829 && MAX_255(ptr[3]) && (digitab[ptr[3]] & ctype_xdigit) != 0
830 && MAX_255(ptr[4]) && (digitab[ptr[4]] & ctype_xdigit) != 0)
831 {
832 c = 0;
833 for (i = 0; i < 4; ++i)
834 {
835 register int cc = *(++ptr);
836 #ifndef EBCDIC /* ASCII/UTF-8 coding */
837 if (cc >= CHAR_a) cc -= 32; /* Convert to upper case */
838 c = (c << 4) + cc - ((cc < CHAR_A)? CHAR_0 : (CHAR_A - 10));
839 #else /* EBCDIC coding */
840 if (cc >= CHAR_a && cc <= CHAR_z) cc += 64; /* Convert to upper case */
841 c = (c << 4) + cc - ((cc >= CHAR_0)? CHAR_0 : (CHAR_A - 10));
842 #endif
843 }
844
845 #if defined COMPILE_PCRE8
846 if (c > (utf ? 0x10ffff : 0xff))
847 #elif defined COMPILE_PCRE16
848 if (c > (utf ? 0x10ffff : 0xffff))
849 #elif defined COMPILE_PCRE32
850 if (utf && c > 0x10ffff)
851 #endif
852 {
853 *errorcodeptr = ERR76;
854 }
855 else if (utf && c >= 0xd800 && c <= 0xdfff) *errorcodeptr = ERR73;
856 }
857 }
858 else
859 *errorcodeptr = ERR37;
860 break;
861
862 case CHAR_U:
863 /* In JavaScript, \U is an uppercase U letter. */
864 if ((options & PCRE_JAVASCRIPT_COMPAT) == 0) *errorcodeptr = ERR37;
865 break;
866
867 /* In a character class, \g is just a literal "g". Outside a character
868 class, \g must be followed by one of a number of specific things:
869
870 (1) A number, either plain or braced. If positive, it is an absolute
871 backreference. If negative, it is a relative backreference. This is a Perl
872 5.10 feature.
873
874 (2) Perl 5.10 also supports \g{name} as a reference to a named group. This
875 is part of Perl's movement towards a unified syntax for back references. As
876 this is synonymous with \k{name}, we fudge it up by pretending it really
877 was \k.
878
879 (3) For Oniguruma compatibility we also support \g followed by a name or a
880 number either in angle brackets or in single quotes. However, these are
881 (possibly recursive) subroutine calls, _not_ backreferences. Just return
882 the ESC_g code (cf \k). */
883
884 case CHAR_g:
885 if (isclass) break;
886 if (ptr[1] == CHAR_LESS_THAN_SIGN || ptr[1] == CHAR_APOSTROPHE)
887 {
888 escape = ESC_g;
889 break;
890 }
891
892 /* Handle the Perl-compatible cases */
893
894 if (ptr[1] == CHAR_LEFT_CURLY_BRACKET)
895 {
896 const pcre_uchar *p;
897 for (p = ptr+2; *p != 0 && *p != CHAR_RIGHT_CURLY_BRACKET; p++)
898 if (*p != CHAR_MINUS && !IS_DIGIT(*p)) break;
899 if (*p != 0 && *p != CHAR_RIGHT_CURLY_BRACKET)
900 {
901 escape = ESC_k;
902 break;
903 }
904 braced = TRUE;
905 ptr++;
906 }
907 else braced = FALSE;
908
909 if (ptr[1] == CHAR_MINUS)
910 {
911 negated = TRUE;
912 ptr++;
913 }
914 else negated = FALSE;
915
916 /* The integer range is limited by the machine's int representation. */
917 c = 0;
918 while (IS_DIGIT(ptr[1]))
919 {
920 if (((unsigned int)c) > INT_MAX / 10) /* Integer overflow */
921 {
922 c = -1;
923 break;
924 }
925 c = c * 10 + *(++ptr) - CHAR_0;
926 }
927 if (((unsigned int)c) > INT_MAX) /* Integer overflow */
928 {
929 while (IS_DIGIT(ptr[1]))
930 ptr++;
931 *errorcodeptr = ERR61;
932 break;
933 }
934
935 if (braced && *(++ptr) != CHAR_RIGHT_CURLY_BRACKET)
936 {
937 *errorcodeptr = ERR57;
938 break;
939 }
940
941 if (c == 0)
942 {
943 *errorcodeptr = ERR58;
944 break;
945 }
946
947 if (negated)
948 {
949 if (c > bracount)
950 {
951 *errorcodeptr = ERR15;
952 break;
953 }
954 c = bracount - (c - 1);
955 }
956
957 escape = ESC_REF + c;
958 break;
959
960 /* The handling of escape sequences consisting of a string of digits
961 starting with one that is not zero is not straightforward. By experiment,
962 the way Perl works seems to be as follows:
963
964 Outside a character class, the digits are read as a decimal number. If the
965 number is less than 10, or if there are that many previous extracting
966 left brackets, then it is a back reference. Otherwise, up to three octal
967 digits are read to form an escaped byte. Thus \123 is likely to be octal
968 123 (cf \0123, which is octal 012 followed by the literal 3). If the octal
969 value is greater than 377, the least significant 8 bits are taken. Inside a
970 character class, \ followed by a digit is always an octal number. */
971
972 case CHAR_1: case CHAR_2: case CHAR_3: case CHAR_4: case CHAR_5:
973 case CHAR_6: case CHAR_7: case CHAR_8: case CHAR_9:
974
975 if (!isclass)
976 {
977 oldptr = ptr;
978 /* The integer range is limited by the machine's int representation. */
979 c -= CHAR_0;
980 while (IS_DIGIT(ptr[1]))
981 {
982 if (((unsigned int)c) > INT_MAX / 10) /* Integer overflow */
983 {
984 c = -1;
985 break;
986 }
987 c = c * 10 + *(++ptr) - CHAR_0;
988 }
989 if (((unsigned int)c) > INT_MAX) /* Integer overflow */
990 {
991 while (IS_DIGIT(ptr[1]))
992 ptr++;
993 *errorcodeptr = ERR61;
994 break;
995 }
996 if (c < 10 || c <= bracount)
997 {
998 escape = ESC_REF + c;
999 break;
1000 }
1001 ptr = oldptr; /* Put the pointer back and fall through */
1002 }
1003
1004 /* Handle an octal number following \. If the first digit is 8 or 9, Perl
1005 generates a binary zero byte and treats the digit as a following literal.
1006 Thus we have to pull back the pointer by one. */
1007
1008 if ((c = *ptr) >= CHAR_8)
1009 {
1010 ptr--;
1011 c = 0;
1012 break;
1013 }
1014
1015 /* \0 always starts an octal number, but we may drop through to here with a
1016 larger first octal digit. The original code used just to take the least
1017 significant 8 bits of octal numbers (I think this is what early Perls used
1018 to do). Nowadays we allow for larger numbers in UTF-8 mode and 16-bit mode,
1019 but no more than 3 octal digits. */
1020
1021 case CHAR_0:
1022 c -= CHAR_0;
1023 while(i++ < 2 && ptr[1] >= CHAR_0 && ptr[1] <= CHAR_7)
1024 c = c * 8 + *(++ptr) - CHAR_0;
1025 #ifdef COMPILE_PCRE8
1026 if (!utf && c > 0xff) *errorcodeptr = ERR51;
1027 #endif
1028 break;
1029
1030 /* \x is complicated. \x{ddd} is a character number which can be greater
1031 than 0xff in utf or non-8bit mode, but only if the ddd are hex digits.
1032 If not, { is treated as a data character. */
1033
1034 case CHAR_x:
1035 if ((options & PCRE_JAVASCRIPT_COMPAT) != 0)
1036 {
1037 /* In JavaScript, \x must be followed by two hexadecimal numbers.
1038 Otherwise it is a lowercase x letter. */
1039 if (MAX_255(ptr[1]) && (digitab[ptr[1]] & ctype_xdigit) != 0
1040 && MAX_255(ptr[2]) && (digitab[ptr[2]] & ctype_xdigit) != 0)
1041 {
1042 c = 0;
1043 for (i = 0; i < 2; ++i)
1044 {
1045 register int cc = *(++ptr);
1046 #ifndef EBCDIC /* ASCII/UTF-8 coding */
1047 if (cc >= CHAR_a) cc -= 32; /* Convert to upper case */
1048 c = (c << 4) + cc - ((cc < CHAR_A)? CHAR_0 : (CHAR_A - 10));
1049 #else /* EBCDIC coding */
1050 if (cc >= CHAR_a && cc <= CHAR_z) cc += 64; /* Convert to upper case */
1051 c = (c << 4) + cc - ((cc >= CHAR_0)? CHAR_0 : (CHAR_A - 10));
1052 #endif
1053 }
1054 }
1055 break;
1056 }
1057
1058 if (ptr[1] == CHAR_LEFT_CURLY_BRACKET)
1059 {
1060 const pcre_uchar *pt = ptr + 2;
1061
1062 c = 0;
1063 while (MAX_255(*pt) && (digitab[*pt] & ctype_xdigit) != 0)
1064 {
1065 register int cc = *pt++;
1066 if (c == 0 && cc == CHAR_0) continue; /* Leading zeroes */
1067
1068 #ifndef EBCDIC /* ASCII/UTF-8 coding */
1069 if (cc >= CHAR_a) cc -= 32; /* Convert to upper case */
1070 c = (c << 4) + cc - ((cc < CHAR_A)? CHAR_0 : (CHAR_A - 10));
1071 #else /* EBCDIC coding */
1072 if (cc >= CHAR_a && cc <= CHAR_z) cc += 64; /* Convert to upper case */
1073 c = (c << 4) + cc - ((cc >= CHAR_0)? CHAR_0 : (CHAR_A - 10));
1074 #endif
1075
1076 #if defined COMPILE_PCRE8
1077 if (c > (utf ? 0x10ffff : 0xff)) { c = -1; break; }
1078 #elif defined COMPILE_PCRE16
1079 if (c > (utf ? 0x10ffff : 0xffff)) { c = -1; break; }
1080 #elif defined COMPILE_PCRE32
1081 if (utf && c > 0x10ffff) { c = -1; break; }
1082 #endif
1083 }
1084
1085 if (c < 0)
1086 {
1087 while (MAX_255(*pt) && (digitab[*pt] & ctype_xdigit) != 0) pt++;
1088 *errorcodeptr = ERR34;
1089 }
1090
1091 if (*pt == CHAR_RIGHT_CURLY_BRACKET)
1092 {
1093 if (utf && c >= 0xd800 && c <= 0xdfff) *errorcodeptr = ERR73;
1094 ptr = pt;
1095 break;
1096 }
1097
1098 /* If the sequence of hex digits does not end with '}', then we don't
1099 recognize this construct; fall through to the normal \x handling. */
1100 }
1101
1102 /* Read just a single-byte hex-defined char */
1103
1104 c = 0;
1105 while (i++ < 2 && MAX_255(ptr[1]) && (digitab[ptr[1]] & ctype_xdigit) != 0)
1106 {
1107 int cc; /* Some compilers don't like */
1108 cc = *(++ptr); /* ++ in initializers */
1109 #ifndef EBCDIC /* ASCII/UTF-8 coding */
1110 if (cc >= CHAR_a) cc -= 32; /* Convert to upper case */
1111 c = c * 16 + cc - ((cc < CHAR_A)? CHAR_0 : (CHAR_A - 10));
1112 #else /* EBCDIC coding */
1113 if (cc <= CHAR_z) cc += 64; /* Convert to upper case */
1114 c = c * 16 + cc - ((cc >= CHAR_0)? CHAR_0 : (CHAR_A - 10));
1115 #endif
1116 }
1117 break;
1118
1119 /* For \c, a following letter is upper-cased; then the 0x40 bit is flipped.
1120 An error is given if the byte following \c is not an ASCII character. This
1121 coding is ASCII-specific, but then the whole concept of \cx is
1122 ASCII-specific. (However, an EBCDIC equivalent has now been added.) */
1123
1124 case CHAR_c:
1125 c = *(++ptr);
1126 if (c == 0)
1127 {
1128 *errorcodeptr = ERR2;
1129 break;
1130 }
1131 #ifndef EBCDIC /* ASCII/UTF-8 coding */
1132 if (c > 127) /* Excludes all non-ASCII in either mode */
1133 {
1134 *errorcodeptr = ERR68;
1135 break;
1136 }
1137 if (c >= CHAR_a && c <= CHAR_z) c -= 32;
1138 c ^= 0x40;
1139 #else /* EBCDIC coding */
1140 if (c >= CHAR_a && c <= CHAR_z) c += 64;
1141 c ^= 0xC0;
1142 #endif
1143 break;
1144
1145 /* PCRE_EXTRA enables extensions to Perl in the matter of escapes. Any
1146 other alphanumeric following \ is an error if PCRE_EXTRA was set;
1147 otherwise, for Perl compatibility, it is a literal. This code looks a bit
1148 odd, but there used to be some cases other than the default, and there may
1149 be again in future, so I haven't "optimized" it. */
1150
1151 default:
1152 if ((options & PCRE_EXTRA) != 0) switch(c)
1153 {
1154 default:
1155 *errorcodeptr = ERR3;
1156 break;
1157 }
1158 break;
1159 }
1160 }
1161
1162 /* Perl supports \N{name} for character names, as well as plain \N for "not
1163 newline". PCRE does not support \N{name}. However, it does support
1164 quantification such as \N{2,3}. */
1165
1166 if (escape == ESC_N && ptr[1] == CHAR_LEFT_CURLY_BRACKET &&
1167 !is_counted_repeat(ptr+2))
1168 *errorcodeptr = ERR37;
1169
1170 /* If PCRE_UCP is set, we change the values for \d etc. */
1171
1172 if ((options & PCRE_UCP) != 0 && escape >= ESC_D && escape <= ESC_w)
1173 escape += (ESC_DU - ESC_D);
1174
1175 /* Set the pointer to the final character before returning. */
1176
1177 *ptrptr = ptr;
1178 *chptr = c;
1179 return escape;
1180 }
1181
1182 #ifdef SUPPORT_UCP
1183 /*************************************************
1184 * Handle \P and \p *
1185 *************************************************/
1186
1187 /* This function is called after \P or \p has been encountered, provided that
1188 PCRE is compiled with support for Unicode properties. On entry, ptrptr is
1189 pointing at the P or p. On exit, it is pointing at the final character of the
1190 escape sequence.
1191
1192 Argument:
1193 ptrptr points to the pattern position pointer
1194 negptr points to a boolean that is set TRUE for negation else FALSE
1195 dptr points to an int that is set to the detailed property value
1196 errorcodeptr points to the error code variable
1197
1198 Returns: type value from ucp_type_table, or -1 for an invalid type
1199 */
1200
1201 static int
1202 get_ucp(const pcre_uchar **ptrptr, BOOL *negptr, int *dptr, int *errorcodeptr)
1203 {
1204 int c, i, bot, top;
1205 const pcre_uchar *ptr = *ptrptr;
1206 pcre_uchar name[32];
1207
1208 c = *(++ptr);
1209 if (c == 0) goto ERROR_RETURN;
1210
1211 *negptr = FALSE;
1212
1213 /* \P or \p can be followed by a name in {}, optionally preceded by ^ for
1214 negation. */
1215
1216 if (c == CHAR_LEFT_CURLY_BRACKET)
1217 {
1218 if (ptr[1] == CHAR_CIRCUMFLEX_ACCENT)
1219 {
1220 *negptr = TRUE;
1221 ptr++;
1222 }
1223 for (i = 0; i < (int)(sizeof(name) / sizeof(pcre_uchar)) - 1; i++)
1224 {
1225 c = *(++ptr);
1226 if (c == 0) goto ERROR_RETURN;
1227 if (c == CHAR_RIGHT_CURLY_BRACKET) break;
1228 name[i] = c;
1229 }
1230 if (c != CHAR_RIGHT_CURLY_BRACKET) goto ERROR_RETURN;
1231 name[i] = 0;
1232 }
1233
1234 /* Otherwise there is just one following character */
1235
1236 else
1237 {
1238 name[0] = c;
1239 name[1] = 0;
1240 }
1241
1242 *ptrptr = ptr;
1243
1244 /* Search for a recognized property name using binary chop */
1245
1246 bot = 0;
1247 top = PRIV(utt_size);
1248
1249 while (bot < top)
1250 {
1251 i = (bot + top) >> 1;
1252 c = STRCMP_UC_C8(name, PRIV(utt_names) + PRIV(utt)[i].name_offset);
1253 if (c == 0)
1254 {
1255 *dptr = PRIV(utt)[i].value;
1256 return PRIV(utt)[i].type;
1257 }
1258 if (c > 0) bot = i + 1; else top = i;
1259 }
1260
1261 *errorcodeptr = ERR47;
1262 *ptrptr = ptr;
1263 return -1;
1264
1265 ERROR_RETURN:
1266 *errorcodeptr = ERR46;
1267 *ptrptr = ptr;
1268 return -1;
1269 }
1270 #endif
1271
1272
1273
1274
1275 /*************************************************
1276 * Read repeat counts *
1277 *************************************************/
1278
1279 /* Read an item of the form {n,m} and return the values. This is called only
1280 after is_counted_repeat() has confirmed that a repeat-count quantifier exists,
1281 so the syntax is guaranteed to be correct, but we need to check the values.
1282
1283 Arguments:
1284 p pointer to first char after '{'
1285 minp pointer to int for min
1286 maxp pointer to int for max
1287 returned as -1 if no max
1288 errorcodeptr points to error code variable
1289
1290 Returns: pointer to '}' on success;
1291 current ptr on error, with errorcodeptr set non-zero
1292 */
1293
1294 static const pcre_uchar *
1295 read_repeat_counts(const pcre_uchar *p, int *minp, int *maxp, int *errorcodeptr)
1296 {
1297 int min = 0;
1298 int max = -1;
1299
1300 /* Read the minimum value and do a paranoid check: a negative value indicates
1301 an integer overflow. */
1302
1303 while (IS_DIGIT(*p)) min = min * 10 + *p++ - CHAR_0;
1304 if (min < 0 || min > 65535)
1305 {
1306 *errorcodeptr = ERR5;
1307 return p;
1308 }
1309
1310 /* Read the maximum value if there is one, and again do a paranoid on its size.
1311 Also, max must not be less than min. */
1312
1313 if (*p == CHAR_RIGHT_CURLY_BRACKET) max = min; else
1314 {
1315 if (*(++p) != CHAR_RIGHT_CURLY_BRACKET)
1316 {
1317 max = 0;
1318 while(IS_DIGIT(*p)) max = max * 10 + *p++ - CHAR_0;
1319 if (max < 0 || max > 65535)
1320 {
1321 *errorcodeptr = ERR5;
1322 return p;
1323 }
1324 if (max < min)
1325 {
1326 *errorcodeptr = ERR4;
1327 return p;
1328 }
1329 }
1330 }
1331
1332 /* Fill in the required variables, and pass back the pointer to the terminating
1333 '}'. */
1334
1335 *minp = min;
1336 *maxp = max;
1337 return p;
1338 }
1339
1340
1341
1342 /*************************************************
1343 * Subroutine for finding forward reference *
1344 *************************************************/
1345
1346 /* This recursive function is called only from find_parens() below. The
1347 top-level call starts at the beginning of the pattern. All other calls must
1348 start at a parenthesis. It scans along a pattern's text looking for capturing
1349 subpatterns, and counting them. If it finds a named pattern that matches the
1350 name it is given, it returns its number. Alternatively, if the name is NULL, it
1351 returns when it reaches a given numbered subpattern. Recursion is used to keep
1352 track of subpatterns that reset the capturing group numbers - the (?| feature.
1353
1354 This function was originally called only from the second pass, in which we know
1355 that if (?< or (?' or (?P< is encountered, the name will be correctly
1356 terminated because that is checked in the first pass. There is now one call to
1357 this function in the first pass, to check for a recursive back reference by
1358 name (so that we can make the whole group atomic). In this case, we need check
1359 only up to the current position in the pattern, and that is still OK because
1360 and previous occurrences will have been checked. To make this work, the test
1361 for "end of pattern" is a check against cd->end_pattern in the main loop,
1362 instead of looking for a binary zero. This means that the special first-pass
1363 call can adjust cd->end_pattern temporarily. (Checks for binary zero while
1364 processing items within the loop are OK, because afterwards the main loop will
1365 terminate.)
1366
1367 Arguments:
1368 ptrptr address of the current character pointer (updated)
1369 cd compile background data
1370 name name to seek, or NULL if seeking a numbered subpattern
1371 lorn name length, or subpattern number if name is NULL
1372 xmode TRUE if we are in /x mode
1373 utf TRUE if we are in UTF-8 / UTF-16 / UTF-32 mode
1374 count pointer to the current capturing subpattern number (updated)
1375
1376 Returns: the number of the named subpattern, or -1 if not found
1377 */
1378
1379 static int
1380 find_parens_sub(pcre_uchar **ptrptr, compile_data *cd, const pcre_uchar *name, int lorn,
1381 BOOL xmode, BOOL utf, int *count)
1382 {
1383 pcre_uchar *ptr = *ptrptr;
1384 int start_count = *count;
1385 int hwm_count = start_count;
1386 BOOL dup_parens = FALSE;
1387
1388 /* If the first character is a parenthesis, check on the type of group we are
1389 dealing with. The very first call may not start with a parenthesis. */
1390
1391 if (ptr[0] == CHAR_LEFT_PARENTHESIS)
1392 {
1393 /* Handle specials such as (*SKIP) or (*UTF8) etc. */
1394
1395 if (ptr[1] == CHAR_ASTERISK) ptr += 2;
1396
1397 /* Handle a normal, unnamed capturing parenthesis. */
1398
1399 else if (ptr[1] != CHAR_QUESTION_MARK)
1400 {
1401 *count += 1;
1402 if (name == NULL && *count == lorn) return *count;
1403 ptr++;
1404 }
1405
1406 /* All cases now have (? at the start. Remember when we are in a group
1407 where the parenthesis numbers are duplicated. */
1408
1409 else if (ptr[2] == CHAR_VERTICAL_LINE)
1410 {
1411 ptr += 3;
1412 dup_parens = TRUE;
1413 }
1414
1415 /* Handle comments; all characters are allowed until a ket is reached. */
1416
1417 else if (ptr[2] == CHAR_NUMBER_SIGN)
1418 {
1419 for (ptr += 3; *ptr != 0; ptr++) if (*ptr == CHAR_RIGHT_PARENTHESIS) break;
1420 goto FAIL_EXIT;
1421 }
1422
1423 /* Handle a condition. If it is an assertion, just carry on so that it
1424 is processed as normal. If not, skip to the closing parenthesis of the
1425 condition (there can't be any nested parens). */
1426
1427 else if (ptr[2] == CHAR_LEFT_PARENTHESIS)
1428 {
1429 ptr += 2;
1430 if (ptr[1] != CHAR_QUESTION_MARK)
1431 {
1432 while (*ptr != 0 && *ptr != CHAR_RIGHT_PARENTHESIS) ptr++;
1433 if (*ptr != 0) ptr++;
1434 }
1435 }
1436
1437 /* Start with (? but not a condition. */
1438
1439 else
1440 {
1441 ptr += 2;
1442 if (*ptr == CHAR_P) ptr++; /* Allow optional P */
1443
1444 /* We have to disambiguate (?<! and (?<= from (?<name> for named groups */
1445
1446 if ((*ptr == CHAR_LESS_THAN_SIGN && ptr[1] != CHAR_EXCLAMATION_MARK &&
1447 ptr[1] != CHAR_EQUALS_SIGN) || *ptr == CHAR_APOSTROPHE)
1448 {
1449 int term;
1450 const pcre_uchar *thisname;
1451 *count += 1;
1452 if (name == NULL && *count == lorn) return *count;
1453 term = *ptr++;
1454 if (term == CHAR_LESS_THAN_SIGN) term = CHAR_GREATER_THAN_SIGN;
1455 thisname = ptr;
1456 while (*ptr != term) ptr++;
1457 if (name != NULL && lorn == ptr - thisname &&
1458 STRNCMP_UC_UC(name, thisname, lorn) == 0)
1459 return *count;
1460 term++;
1461 }
1462 }
1463 }
1464
1465 /* Past any initial parenthesis handling, scan for parentheses or vertical
1466 bars. Stop if we get to cd->end_pattern. Note that this is important for the
1467 first-pass call when this value is temporarily adjusted to stop at the current
1468 position. So DO NOT change this to a test for binary zero. */
1469
1470 for (; ptr < cd->end_pattern; ptr++)
1471 {
1472 /* Skip over backslashed characters and also entire \Q...\E */
1473
1474 if (*ptr == CHAR_BACKSLASH)
1475 {
1476 if (*(++ptr) == 0) goto FAIL_EXIT;
1477 if (*ptr == CHAR_Q) for (;;)
1478 {
1479 while (*(++ptr) != 0 && *ptr != CHAR_BACKSLASH) {};
1480 if (*ptr == 0) goto FAIL_EXIT;
1481 if (*(++ptr) == CHAR_E) break;
1482 }
1483 continue;
1484 }
1485
1486 /* Skip over character classes; this logic must be similar to the way they
1487 are handled for real. If the first character is '^', skip it. Also, if the
1488 first few characters (either before or after ^) are \Q\E or \E we skip them
1489 too. This makes for compatibility with Perl. Note the use of STR macros to
1490 encode "Q\\E" so that it works in UTF-8 on EBCDIC platforms. */
1491
1492 if (*ptr == CHAR_LEFT_SQUARE_BRACKET)
1493 {
1494 BOOL negate_class = FALSE;
1495 for (;;)
1496 {
1497 if (ptr[1] == CHAR_BACKSLASH)
1498 {
1499 if (ptr[2] == CHAR_E)
1500 ptr+= 2;
1501 else if (STRNCMP_UC_C8(ptr + 2,
1502 STR_Q STR_BACKSLASH STR_E, 3) == 0)
1503 ptr += 4;
1504 else
1505 break;
1506 }
1507 else if (!negate_class && ptr[1] == CHAR_CIRCUMFLEX_ACCENT)
1508 {
1509 negate_class = TRUE;
1510 ptr++;
1511 }
1512 else break;
1513 }
1514
1515 /* If the next character is ']', it is a data character that must be
1516 skipped, except in JavaScript compatibility mode. */
1517
1518 if (ptr[1] == CHAR_RIGHT_SQUARE_BRACKET &&
1519 (cd->external_options & PCRE_JAVASCRIPT_COMPAT) == 0)
1520 ptr++;
1521
1522 while (*(++ptr) != CHAR_RIGHT_SQUARE_BRACKET)
1523 {
1524 if (*ptr == 0) return -1;
1525 if (*ptr == CHAR_BACKSLASH)
1526 {
1527 if (*(++ptr) == 0) goto FAIL_EXIT;
1528 if (*ptr == CHAR_Q) for (;;)
1529 {
1530 while (*(++ptr) != 0 && *ptr != CHAR_BACKSLASH) {};
1531 if (*ptr == 0) goto FAIL_EXIT;
1532 if (*(++ptr) == CHAR_E) break;
1533 }
1534 continue;
1535 }
1536 }
1537 continue;
1538 }
1539
1540 /* Skip comments in /x mode */
1541
1542 if (xmode && *ptr == CHAR_NUMBER_SIGN)
1543 {
1544 ptr++;
1545 while (*ptr != 0)
1546 {
1547 if (IS_NEWLINE(ptr)) { ptr += cd->nllen - 1; break; }
1548 ptr++;
1549 #ifdef SUPPORT_UTF
1550 if (utf) FORWARDCHAR(ptr);
1551 #endif
1552 }
1553 if (*ptr == 0) goto FAIL_EXIT;
1554 continue;
1555 }
1556
1557 /* Check for the special metacharacters */
1558
1559 if (*ptr == CHAR_LEFT_PARENTHESIS)
1560 {
1561 int rc = find_parens_sub(&ptr, cd, name, lorn, xmode, utf, count);
1562 if (rc > 0) return rc;
1563 if (*ptr == 0) goto FAIL_EXIT;
1564 }
1565
1566 else if (*ptr == CHAR_RIGHT_PARENTHESIS)
1567 {
1568 if (dup_parens && *count < hwm_count) *count = hwm_count;
1569 goto FAIL_EXIT;
1570 }
1571
1572 else if (*ptr == CHAR_VERTICAL_LINE && dup_parens)
1573 {
1574 if (*count > hwm_count) hwm_count = *count;
1575 *count = start_count;
1576 }
1577 }
1578
1579 FAIL_EXIT:
1580 *ptrptr = ptr;
1581 return -1;
1582 }
1583
1584
1585
1586
1587 /*************************************************
1588 * Find forward referenced subpattern *
1589 *************************************************/
1590
1591 /* This function scans along a pattern's text looking for capturing
1592 subpatterns, and counting them. If it finds a named pattern that matches the
1593 name it is given, it returns its number. Alternatively, if the name is NULL, it
1594 returns when it reaches a given numbered subpattern. This is used for forward
1595 references to subpatterns. We used to be able to start this scan from the
1596 current compiling point, using the current count value from cd->bracount, and
1597 do it all in a single loop, but the addition of the possibility of duplicate
1598 subpattern numbers means that we have to scan from the very start, in order to
1599 take account of such duplicates, and to use a recursive function to keep track
1600 of the different types of group.
1601
1602 Arguments:
1603 cd compile background data
1604 name name to seek, or NULL if seeking a numbered subpattern
1605 lorn name length, or subpattern number if name is NULL
1606 xmode TRUE if we are in /x mode
1607 utf TRUE if we are in UTF-8 / UTF-16 / UTF-32 mode
1608
1609 Returns: the number of the found subpattern, or -1 if not found
1610 */
1611
1612 static int
1613 find_parens(compile_data *cd, const pcre_uchar *name, int lorn, BOOL xmode,
1614 BOOL utf)
1615 {
1616 pcre_uchar *ptr = (pcre_uchar *)cd->start_pattern;
1617 int count = 0;
1618 int rc;
1619
1620 /* If the pattern does not start with an opening parenthesis, the first call
1621 to find_parens_sub() will scan right to the end (if necessary). However, if it
1622 does start with a parenthesis, find_parens_sub() will return when it hits the
1623 matching closing parens. That is why we have to have a loop. */
1624
1625 for (;;)
1626 {
1627 rc = find_parens_sub(&ptr, cd, name, lorn, xmode, utf, &count);
1628 if (rc > 0 || *ptr++ == 0) break;
1629 }
1630
1631 return rc;
1632 }
1633
1634
1635
1636
1637 /*************************************************
1638 * Find first significant op code *
1639 *************************************************/
1640
1641 /* This is called by several functions that scan a compiled expression looking
1642 for a fixed first character, or an anchoring op code etc. It skips over things
1643 that do not influence this. For some calls, it makes sense to skip negative
1644 forward and all backward assertions, and also the \b assertion; for others it
1645 does not.
1646
1647 Arguments:
1648 code pointer to the start of the group
1649 skipassert TRUE if certain assertions are to be skipped
1650
1651 Returns: pointer to the first significant opcode
1652 */
1653
1654 static const pcre_uchar*
1655 first_significant_code(const pcre_uchar *code, BOOL skipassert)
1656 {
1657 for (;;)
1658 {
1659 switch ((int)*code)
1660 {
1661 case OP_ASSERT_NOT:
1662 case OP_ASSERTBACK:
1663 case OP_ASSERTBACK_NOT:
1664 if (!skipassert) return code;
1665 do code += GET(code, 1); while (*code == OP_ALT);
1666 code += PRIV(OP_lengths)[*code];
1667 break;
1668
1669 case OP_WORD_BOUNDARY:
1670 case OP_NOT_WORD_BOUNDARY:
1671 if (!skipassert) return code;
1672 /* Fall through */
1673
1674 case OP_CALLOUT:
1675 case OP_CREF:
1676 case OP_NCREF:
1677 case OP_RREF:
1678 case OP_NRREF:
1679 case OP_DEF:
1680 code += PRIV(OP_lengths)[*code];
1681 break;
1682
1683 default:
1684 return code;
1685 }
1686 }
1687 /* Control never reaches here */
1688 }
1689
1690
1691
1692
1693 /*************************************************
1694 * Find the fixed length of a branch *
1695 *************************************************/
1696
1697 /* Scan a branch and compute the fixed length of subject that will match it,
1698 if the length is fixed. This is needed for dealing with backward assertions.
1699 In UTF8 mode, the result is in characters rather than bytes. The branch is
1700 temporarily terminated with OP_END when this function is called.
1701
1702 This function is called when a backward assertion is encountered, so that if it
1703 fails, the error message can point to the correct place in the pattern.
1704 However, we cannot do this when the assertion contains subroutine calls,
1705 because they can be forward references. We solve this by remembering this case
1706 and doing the check at the end; a flag specifies which mode we are running in.
1707
1708 Arguments:
1709 code points to the start of the pattern (the bracket)
1710 utf TRUE in UTF-8 / UTF-16 / UTF-32 mode
1711 atend TRUE if called when the pattern is complete
1712 cd the "compile data" structure
1713
1714 Returns: the fixed length,
1715 or -1 if there is no fixed length,
1716 or -2 if \C was encountered (in UTF-8 mode only)
1717 or -3 if an OP_RECURSE item was encountered and atend is FALSE
1718 or -4 if an unknown opcode was encountered (internal error)
1719 */
1720
1721 static int
1722 find_fixedlength(pcre_uchar *code, BOOL utf, BOOL atend, compile_data *cd)
1723 {
1724 int length = -1;
1725
1726 register int branchlength = 0;
1727 register pcre_uchar *cc = code + 1 + LINK_SIZE;
1728
1729 /* Scan along the opcodes for this branch. If we get to the end of the
1730 branch, check the length against that of the other branches. */
1731
1732 for (;;)
1733 {
1734 int d;
1735 pcre_uchar *ce, *cs;
1736 register int op = *cc;
1737
1738 switch (op)
1739 {
1740 /* We only need to continue for OP_CBRA (normal capturing bracket) and
1741 OP_BRA (normal non-capturing bracket) because the other variants of these
1742 opcodes are all concerned with unlimited repeated groups, which of course
1743 are not of fixed length. */
1744
1745 case OP_CBRA:
1746 case OP_BRA:
1747 case OP_ONCE:
1748 case OP_ONCE_NC:
1749 case OP_COND:
1750 d = find_fixedlength(cc + ((op == OP_CBRA)? IMM2_SIZE : 0), utf, atend, cd);
1751 if (d < 0) return d;
1752 branchlength += d;
1753 do cc += GET(cc, 1); while (*cc == OP_ALT);
1754 cc += 1 + LINK_SIZE;
1755 break;
1756
1757 /* Reached end of a branch; if it's a ket it is the end of a nested call.
1758 If it's ALT it is an alternation in a nested call. An ACCEPT is effectively
1759 an ALT. If it is END it's the end of the outer call. All can be handled by
1760 the same code. Note that we must not include the OP_KETRxxx opcodes here,
1761 because they all imply an unlimited repeat. */
1762
1763 case OP_ALT:
1764 case OP_KET:
1765 case OP_END:
1766 case OP_ACCEPT:
1767 case OP_ASSERT_ACCEPT:
1768 if (length < 0) length = branchlength;
1769 else if (length != branchlength) return -1;
1770 if (*cc != OP_ALT) return length;
1771 cc += 1 + LINK_SIZE;
1772 branchlength = 0;
1773 break;
1774
1775 /* A true recursion implies not fixed length, but a subroutine call may
1776 be OK. If the subroutine is a forward reference, we can't deal with
1777 it until the end of the pattern, so return -3. */
1778
1779 case OP_RECURSE:
1780 if (!atend) return -3;
1781 cs = ce = (pcre_uchar *)cd->start_code + GET(cc, 1); /* Start subpattern */
1782 do ce += GET(ce, 1); while (*ce == OP_ALT); /* End subpattern */
1783 if (cc > cs && cc < ce) return -1; /* Recursion */
1784 d = find_fixedlength(cs + IMM2_SIZE, utf, atend, cd);
1785 if (d < 0) return d;
1786 branchlength += d;
1787 cc += 1 + LINK_SIZE;
1788 break;
1789
1790 /* Skip over assertive subpatterns */
1791
1792 case OP_ASSERT:
1793 case OP_ASSERT_NOT:
1794 case OP_ASSERTBACK:
1795 case OP_ASSERTBACK_NOT:
1796 do cc += GET(cc, 1); while (*cc == OP_ALT);
1797 cc += PRIV(OP_lengths)[*cc];
1798 break;
1799
1800 /* Skip over things that don't match chars */
1801
1802 case OP_MARK:
1803 case OP_PRUNE_ARG:
1804 case OP_SKIP_ARG:
1805 case OP_THEN_ARG:
1806 cc += cc[1] + PRIV(OP_lengths)[*cc];
1807 break;
1808
1809 case OP_CALLOUT:
1810 case OP_CIRC:
1811 case OP_CIRCM:
1812 case OP_CLOSE:
1813 case OP_COMMIT:
1814 case OP_CREF:
1815 case OP_DEF:
1816 case OP_DOLL:
1817 case OP_DOLLM:
1818 case OP_EOD:
1819 case OP_EODN:
1820 case OP_FAIL:
1821 case OP_NCREF:
1822 case OP_NRREF:
1823 case OP_NOT_WORD_BOUNDARY:
1824 case OP_PRUNE:
1825 case OP_REVERSE:
1826 case OP_RREF:
1827 case OP_SET_SOM:
1828 case OP_SKIP:
1829 case OP_SOD:
1830 case OP_SOM:
1831 case OP_THEN:
1832 case OP_WORD_BOUNDARY:
1833 cc += PRIV(OP_lengths)[*cc];
1834 break;
1835
1836 /* Handle literal characters */
1837
1838 case OP_CHAR:
1839 case OP_CHARI:
1840 case OP_NOT:
1841 case OP_NOTI:
1842 branchlength++;
1843 cc += 2;
1844 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
1845 if (utf && HAS_EXTRALEN(cc[-1])) cc += GET_EXTRALEN(cc[-1]);
1846 #endif
1847 break;
1848
1849 /* Handle exact repetitions. The count is already in characters, but we
1850 need to skip over a multibyte character in UTF8 mode. */
1851
1852 case OP_EXACT:
1853 case OP_EXACTI:
1854 case OP_NOTEXACT:
1855 case OP_NOTEXACTI:
1856 branchlength += GET2(cc,1);
1857 cc += 2 + IMM2_SIZE;
1858 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
1859 if (utf && HAS_EXTRALEN(cc[-1])) cc += GET_EXTRALEN(cc[-1]);
1860 #endif
1861 break;
1862
1863 case OP_TYPEEXACT:
1864 branchlength += GET2(cc,1);
1865 if (cc[1 + IMM2_SIZE] == OP_PROP || cc[1 + IMM2_SIZE] == OP_NOTPROP)
1866 cc += 2;
1867 cc += 1 + IMM2_SIZE + 1;
1868 break;
1869
1870 /* Handle single-char matchers */
1871
1872 case OP_PROP:
1873 case OP_NOTPROP:
1874 cc += 2;
1875 /* Fall through */
1876
1877 case OP_HSPACE:
1878 case OP_VSPACE:
1879 case OP_NOT_HSPACE:
1880 case OP_NOT_VSPACE:
1881 case OP_NOT_DIGIT:
1882 case OP_DIGIT:
1883 case OP_NOT_WHITESPACE:
1884 case OP_WHITESPACE:
1885 case OP_NOT_WORDCHAR:
1886 case OP_WORDCHAR:
1887 case OP_ANY:
1888 case OP_ALLANY:
1889 branchlength++;
1890 cc++;
1891 break;
1892
1893 /* The single-byte matcher isn't allowed. This only happens in UTF-8 mode;
1894 otherwise \C is coded as OP_ALLANY. */
1895
1896 case OP_ANYBYTE:
1897 return -2;
1898
1899 /* Check a class for variable quantification */
1900
1901 #if defined SUPPORT_UTF || defined COMPILE_PCRE16 || defined COMPILE_PCRE32
1902 case OP_XCLASS:
1903 cc += GET(cc, 1) - PRIV(OP_lengths)[OP_CLASS];
1904 /* Fall through */
1905 #endif
1906
1907 case OP_CLASS:
1908 case OP_NCLASS:
1909 cc += PRIV(OP_lengths)[OP_CLASS];
1910
1911 switch (*cc)
1912 {
1913 case OP_CRPLUS:
1914 case OP_CRMINPLUS:
1915 case OP_CRSTAR:
1916 case OP_CRMINSTAR:
1917 case OP_CRQUERY:
1918 case OP_CRMINQUERY:
1919 return -1;
1920
1921 case OP_CRRANGE:
1922 case OP_CRMINRANGE:
1923 if (GET2(cc,1) != GET2(cc,1+IMM2_SIZE)) return -1;
1924 branchlength += GET2(cc,1);
1925 cc += 1 + 2 * IMM2_SIZE;
1926 break;
1927
1928 default:
1929 branchlength++;
1930 }
1931 break;
1932
1933 /* Anything else is variable length */
1934
1935 case OP_ANYNL:
1936 case OP_BRAMINZERO:
1937 case OP_BRAPOS:
1938 case OP_BRAPOSZERO:
1939 case OP_BRAZERO:
1940 case OP_CBRAPOS:
1941 case OP_EXTUNI:
1942 case OP_KETRMAX:
1943 case OP_KETRMIN:
1944 case OP_KETRPOS:
1945 case OP_MINPLUS:
1946 case OP_MINPLUSI:
1947 case OP_MINQUERY:
1948 case OP_MINQUERYI:
1949 case OP_MINSTAR:
1950 case OP_MINSTARI:
1951 case OP_MINUPTO:
1952 case OP_MINUPTOI:
1953 case OP_NOTMINPLUS:
1954 case OP_NOTMINPLUSI:
1955 case OP_NOTMINQUERY:
1956 case OP_NOTMINQUERYI:
1957 case OP_NOTMINSTAR:
1958 case OP_NOTMINSTARI:
1959 case OP_NOTMINUPTO:
1960 case OP_NOTMINUPTOI:
1961 case OP_NOTPLUS:
1962 case OP_NOTPLUSI:
1963 case OP_NOTPOSPLUS:
1964 case OP_NOTPOSPLUSI:
1965 case OP_NOTPOSQUERY:
1966 case OP_NOTPOSQUERYI:
1967 case OP_NOTPOSSTAR:
1968 case OP_NOTPOSSTARI:
1969 case OP_NOTPOSUPTO:
1970 case OP_NOTPOSUPTOI:
1971 case OP_NOTQUERY:
1972 case OP_NOTQUERYI:
1973 case OP_NOTSTAR:
1974 case OP_NOTSTARI:
1975 case OP_NOTUPTO:
1976 case OP_NOTUPTOI:
1977 case OP_PLUS:
1978 case OP_PLUSI:
1979 case OP_POSPLUS:
1980 case OP_POSPLUSI:
1981 case OP_POSQUERY:
1982 case OP_POSQUERYI:
1983 case OP_POSSTAR:
1984 case OP_POSSTARI:
1985 case OP_POSUPTO:
1986 case OP_POSUPTOI:
1987 case OP_QUERY:
1988 case OP_QUERYI:
1989 case OP_REF:
1990 case OP_REFI:
1991 case OP_SBRA:
1992 case OP_SBRAPOS:
1993 case OP_SCBRA:
1994 case OP_SCBRAPOS:
1995 case OP_SCOND:
1996 case OP_SKIPZERO:
1997 case OP_STAR:
1998 case OP_STARI:
1999 case OP_TYPEMINPLUS:
2000 case OP_TYPEMINQUERY:
2001 case OP_TYPEMINSTAR:
2002 case OP_TYPEMINUPTO:
2003 case OP_TYPEPLUS:
2004 case OP_TYPEPOSPLUS:
2005 case OP_TYPEPOSQUERY:
2006 case OP_TYPEPOSSTAR:
2007 case OP_TYPEPOSUPTO:
2008 case OP_TYPEQUERY:
2009 case OP_TYPESTAR:
2010 case OP_TYPEUPTO:
2011 case OP_UPTO:
2012 case OP_UPTOI:
2013 return -1;
2014
2015 /* Catch unrecognized opcodes so that when new ones are added they
2016 are not forgotten, as has happened in the past. */
2017
2018 default:
2019 return -4;
2020 }
2021 }
2022 /* Control never gets here */
2023 }
2024
2025
2026
2027
2028 /*************************************************
2029 * Scan compiled regex for specific bracket *
2030 *************************************************/
2031
2032 /* This little function scans through a compiled pattern until it finds a
2033 capturing bracket with the given number, or, if the number is negative, an
2034 instance of OP_REVERSE for a lookbehind. The function is global in the C sense
2035 so that it can be called from pcre_study() when finding the minimum matching
2036 length.
2037
2038 Arguments:
2039 code points to start of expression
2040 utf TRUE in UTF-8 / UTF-16 / UTF-32 mode
2041 number the required bracket number or negative to find a lookbehind
2042
2043 Returns: pointer to the opcode for the bracket, or NULL if not found
2044 */
2045
2046 const pcre_uchar *
2047 PRIV(find_bracket)(const pcre_uchar *code, BOOL utf, int number)
2048 {
2049 for (;;)
2050 {
2051 register int c = *code;
2052
2053 if (c == OP_END) return NULL;
2054
2055 /* XCLASS is used for classes that cannot be represented just by a bit
2056 map. This includes negated single high-valued characters. The length in
2057 the table is zero; the actual length is stored in the compiled code. */
2058
2059 if (c == OP_XCLASS) code += GET(code, 1);
2060
2061 /* Handle recursion */
2062
2063 else if (c == OP_REVERSE)
2064 {
2065 if (number < 0) return (pcre_uchar *)code;
2066 code += PRIV(OP_lengths)[c];
2067 }
2068
2069 /* Handle capturing bracket */
2070
2071 else if (c == OP_CBRA || c == OP_SCBRA ||
2072 c == OP_CBRAPOS || c == OP_SCBRAPOS)
2073 {
2074 int n = GET2(code, 1+LINK_SIZE);
2075 if (n == number) return (pcre_uchar *)code;
2076 code += PRIV(OP_lengths)[c];
2077 }
2078
2079 /* Otherwise, we can get the item's length from the table, except that for
2080 repeated character types, we have to test for \p and \P, which have an extra
2081 two bytes of parameters, and for MARK/PRUNE/SKIP/THEN with an argument, we
2082 must add in its length. */
2083
2084 else
2085 {
2086 switch(c)
2087 {
2088 case OP_TYPESTAR:
2089 case OP_TYPEMINSTAR:
2090 case OP_TYPEPLUS:
2091 case OP_TYPEMINPLUS:
2092 case OP_TYPEQUERY:
2093 case OP_TYPEMINQUERY:
2094 case OP_TYPEPOSSTAR:
2095 case OP_TYPEPOSPLUS:
2096 case OP_TYPEPOSQUERY:
2097 if (code[1] == OP_PROP || code[1] == OP_NOTPROP) code += 2;
2098 break;
2099
2100 case OP_TYPEUPTO:
2101 case OP_TYPEMINUPTO:
2102 case OP_TYPEEXACT:
2103 case OP_TYPEPOSUPTO:
2104 if (code[1 + IMM2_SIZE] == OP_PROP || code[1 + IMM2_SIZE] == OP_NOTPROP)
2105 code += 2;
2106 break;
2107
2108 case OP_MARK:
2109 case OP_PRUNE_ARG:
2110 case OP_SKIP_ARG:
2111 code += code[1];
2112 break;
2113
2114 case OP_THEN_ARG:
2115 code += code[1];
2116 break;
2117 }
2118
2119 /* Add in the fixed length from the table */
2120
2121 code += PRIV(OP_lengths)[c];
2122
2123 /* In UTF-8 mode, opcodes that are followed by a character may be followed by
2124 a multi-byte character. The length in the table is a minimum, so we have to
2125 arrange to skip the extra bytes. */
2126
2127 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
2128 if (utf) switch(c)
2129 {
2130 case OP_CHAR:
2131 case OP_CHARI:
2132 case OP_EXACT:
2133 case OP_EXACTI:
2134 case OP_UPTO:
2135 case OP_UPTOI:
2136 case OP_MINUPTO:
2137 case OP_MINUPTOI:
2138 case OP_POSUPTO:
2139 case OP_POSUPTOI:
2140 case OP_STAR:
2141 case OP_STARI:
2142 case OP_MINSTAR:
2143 case OP_MINSTARI:
2144 case OP_POSSTAR:
2145 case OP_POSSTARI:
2146 case OP_PLUS:
2147 case OP_PLUSI:
2148 case OP_MINPLUS:
2149 case OP_MINPLUSI:
2150 case OP_POSPLUS:
2151 case OP_POSPLUSI:
2152 case OP_QUERY:
2153 case OP_QUERYI:
2154 case OP_MINQUERY:
2155 case OP_MINQUERYI:
2156 case OP_POSQUERY:
2157 case OP_POSQUERYI:
2158 if (HAS_EXTRALEN(code[-1])) code += GET_EXTRALEN(code[-1]);
2159 break;
2160 }
2161 #else
2162 (void)(utf); /* Keep compiler happy by referencing function argument */
2163 #endif
2164 }
2165 }
2166 }
2167
2168
2169
2170 /*************************************************
2171 * Scan compiled regex for recursion reference *
2172 *************************************************/
2173
2174 /* This little function scans through a compiled pattern until it finds an
2175 instance of OP_RECURSE.
2176
2177 Arguments:
2178 code points to start of expression
2179 utf TRUE in UTF-8 / UTF-16 / UTF-32 mode
2180
2181 Returns: pointer to the opcode for OP_RECURSE, or NULL if not found
2182 */
2183
2184 static const pcre_uchar *
2185 find_recurse(const pcre_uchar *code, BOOL utf)
2186 {
2187 for (;;)
2188 {
2189 register int c = *code;
2190 if (c == OP_END) return NULL;
2191 if (c == OP_RECURSE) return code;
2192
2193 /* XCLASS is used for classes that cannot be represented just by a bit
2194 map. This includes negated single high-valued characters. The length in
2195 the table is zero; the actual length is stored in the compiled code. */
2196
2197 if (c == OP_XCLASS) code += GET(code, 1);
2198
2199 /* Otherwise, we can get the item's length from the table, except that for
2200 repeated character types, we have to test for \p and \P, which have an extra
2201 two bytes of parameters, and for MARK/PRUNE/SKIP/THEN with an argument, we
2202 must add in its length. */
2203
2204 else
2205 {
2206 switch(c)
2207 {
2208 case OP_TYPESTAR:
2209 case OP_TYPEMINSTAR:
2210 case OP_TYPEPLUS:
2211 case OP_TYPEMINPLUS:
2212 case OP_TYPEQUERY:
2213 case OP_TYPEMINQUERY:
2214 case OP_TYPEPOSSTAR:
2215 case OP_TYPEPOSPLUS:
2216 case OP_TYPEPOSQUERY:
2217 if (code[1] == OP_PROP || code[1] == OP_NOTPROP) code += 2;
2218 break;
2219
2220 case OP_TYPEPOSUPTO:
2221 case OP_TYPEUPTO:
2222 case OP_TYPEMINUPTO:
2223 case OP_TYPEEXACT:
2224 if (code[1 + IMM2_SIZE] == OP_PROP || code[1 + IMM2_SIZE] == OP_NOTPROP)
2225 code += 2;
2226 break;
2227
2228 case OP_MARK:
2229 case OP_PRUNE_ARG:
2230 case OP_SKIP_ARG:
2231 code += code[1];
2232 break;
2233
2234 case OP_THEN_ARG:
2235 code += code[1];
2236 break;
2237 }
2238
2239 /* Add in the fixed length from the table */
2240
2241 code += PRIV(OP_lengths)[c];
2242
2243 /* In UTF-8 mode, opcodes that are followed by a character may be followed
2244 by a multi-byte character. The length in the table is a minimum, so we have
2245 to arrange to skip the extra bytes. */
2246
2247 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
2248 if (utf) switch(c)
2249 {
2250 case OP_CHAR:
2251 case OP_CHARI:
2252 case OP_NOT:
2253 case OP_NOTI:
2254 case OP_EXACT:
2255 case OP_EXACTI:
2256 case OP_NOTEXACT:
2257 case OP_NOTEXACTI:
2258 case OP_UPTO:
2259 case OP_UPTOI:
2260 case OP_NOTUPTO:
2261 case OP_NOTUPTOI:
2262 case OP_MINUPTO:
2263 case OP_MINUPTOI:
2264 case OP_NOTMINUPTO:
2265 case OP_NOTMINUPTOI:
2266 case OP_POSUPTO:
2267 case OP_POSUPTOI:
2268 case OP_NOTPOSUPTO:
2269 case OP_NOTPOSUPTOI:
2270 case OP_STAR:
2271 case OP_STARI:
2272 case OP_NOTSTAR:
2273 case OP_NOTSTARI:
2274 case OP_MINSTAR:
2275 case OP_MINSTARI:
2276 case OP_NOTMINSTAR:
2277 case OP_NOTMINSTARI:
2278 case OP_POSSTAR:
2279 case OP_POSSTARI:
2280 case OP_NOTPOSSTAR:
2281 case OP_NOTPOSSTARI:
2282 case OP_PLUS:
2283 case OP_PLUSI:
2284 case OP_NOTPLUS:
2285 case OP_NOTPLUSI:
2286 case OP_MINPLUS:
2287 case OP_MINPLUSI:
2288 case OP_NOTMINPLUS:
2289 case OP_NOTMINPLUSI:
2290 case OP_POSPLUS:
2291 case OP_POSPLUSI:
2292 case OP_NOTPOSPLUS:
2293 case OP_NOTPOSPLUSI:
2294 case OP_QUERY:
2295 case OP_QUERYI:
2296 case OP_NOTQUERY:
2297 case OP_NOTQUERYI:
2298 case OP_MINQUERY:
2299 case OP_MINQUERYI:
2300 case OP_NOTMINQUERY:
2301 case OP_NOTMINQUERYI:
2302 case OP_POSQUERY:
2303 case OP_POSQUERYI:
2304 case OP_NOTPOSQUERY:
2305 case OP_NOTPOSQUERYI:
2306 if (HAS_EXTRALEN(code[-1])) code += GET_EXTRALEN(code[-1]);
2307 break;
2308 }
2309 #else
2310 (void)(utf); /* Keep compiler happy by referencing function argument */
2311 #endif
2312 }
2313 }
2314 }
2315
2316
2317
2318 /*************************************************
2319 * Scan compiled branch for non-emptiness *
2320 *************************************************/
2321
2322 /* This function scans through a branch of a compiled pattern to see whether it
2323 can match the empty string or not. It is called from could_be_empty()
2324 below and from compile_branch() when checking for an unlimited repeat of a
2325 group that can match nothing. Note that first_significant_code() skips over
2326 backward and negative forward assertions when its final argument is TRUE. If we
2327 hit an unclosed bracket, we return "empty" - this means we've struck an inner
2328 bracket whose current branch will already have been scanned.
2329
2330 Arguments:
2331 code points to start of search
2332 endcode points to where to stop
2333 utf TRUE if in UTF-8 / UTF-16 / UTF-32 mode
2334 cd contains pointers to tables etc.
2335
2336 Returns: TRUE if what is matched could be empty
2337 */
2338
2339 static BOOL
2340 could_be_empty_branch(const pcre_uchar *code, const pcre_uchar *endcode,
2341 BOOL utf, compile_data *cd)
2342 {
2343 register int c;
2344 for (code = first_significant_code(code + PRIV(OP_lengths)[*code], TRUE);
2345 code < endcode;
2346 code = first_significant_code(code + PRIV(OP_lengths)[c], TRUE))
2347 {
2348 const pcre_uchar *ccode;
2349
2350 c = *code;
2351
2352 /* Skip over forward assertions; the other assertions are skipped by
2353 first_significant_code() with a TRUE final argument. */
2354
2355 if (c == OP_ASSERT)
2356 {
2357 do code += GET(code, 1); while (*code == OP_ALT);
2358 c = *code;
2359 continue;
2360 }
2361
2362 /* For a recursion/subroutine call, if its end has been reached, which
2363 implies a backward reference subroutine call, we can scan it. If it's a
2364 forward reference subroutine call, we can't. To detect forward reference
2365 we have to scan up the list that is kept in the workspace. This function is
2366 called only when doing the real compile, not during the pre-compile that
2367 measures the size of the compiled pattern. */
2368
2369 if (c == OP_RECURSE)
2370 {
2371 const pcre_uchar *scode;
2372 BOOL empty_branch;
2373
2374 /* Test for forward reference */
2375
2376 for (scode = cd->start_workspace; scode < cd->hwm; scode += LINK_SIZE)
2377 if (GET(scode, 0) == code + 1 - cd->start_code) return TRUE;
2378
2379 /* Not a forward reference, test for completed backward reference */
2380
2381 empty_branch = FALSE;
2382 scode = cd->start_code + GET(code, 1);
2383 if (GET(scode, 1) == 0) return TRUE; /* Unclosed */
2384
2385 /* Completed backwards reference */
2386
2387 do
2388 {
2389 if (could_be_empty_branch(scode, endcode, utf, cd))
2390 {
2391 empty_branch = TRUE;
2392 break;
2393 }
2394 scode += GET(scode, 1);
2395 }
2396 while (*scode == OP_ALT);
2397
2398 if (!empty_branch) return FALSE; /* All branches are non-empty */
2399 continue;
2400 }
2401
2402 /* Groups with zero repeats can of course be empty; skip them. */
2403
2404 if (c == OP_BRAZERO || c == OP_BRAMINZERO || c == OP_SKIPZERO ||
2405 c == OP_BRAPOSZERO)
2406 {
2407 code += PRIV(OP_lengths)[c];
2408 do code += GET(code, 1); while (*code == OP_ALT);
2409 c = *code;
2410 continue;
2411 }
2412
2413 /* A nested group that is already marked as "could be empty" can just be
2414 skipped. */
2415
2416 if (c == OP_SBRA || c == OP_SBRAPOS ||
2417 c == OP_SCBRA || c == OP_SCBRAPOS)
2418 {
2419 do code += GET(code, 1); while (*code == OP_ALT);
2420 c = *code;
2421 continue;
2422 }
2423
2424 /* For other groups, scan the branches. */
2425
2426 if (c == OP_BRA || c == OP_BRAPOS ||
2427 c == OP_CBRA || c == OP_CBRAPOS ||
2428 c == OP_ONCE || c == OP_ONCE_NC ||
2429 c == OP_COND)
2430 {
2431 BOOL empty_branch;
2432 if (GET(code, 1) == 0) return TRUE; /* Hit unclosed bracket */
2433
2434 /* If a conditional group has only one branch, there is a second, implied,
2435 empty branch, so just skip over the conditional, because it could be empty.
2436 Otherwise, scan the individual branches of the group. */
2437
2438 if (c == OP_COND && code[GET(code, 1)] != OP_ALT)
2439 code += GET(code, 1);
2440 else
2441 {
2442 empty_branch = FALSE;
2443 do
2444 {
2445 if (!empty_branch && could_be_empty_branch(code, endcode, utf, cd))
2446 empty_branch = TRUE;
2447 code += GET(code, 1);
2448 }
2449 while (*code == OP_ALT);
2450 if (!empty_branch) return FALSE; /* All branches are non-empty */
2451 }
2452
2453 c = *code;
2454 continue;
2455 }
2456
2457 /* Handle the other opcodes */
2458
2459 switch (c)
2460 {
2461 /* Check for quantifiers after a class. XCLASS is used for classes that
2462 cannot be represented just by a bit map. This includes negated single
2463 high-valued characters. The length in PRIV(OP_lengths)[] is zero; the
2464 actual length is stored in the compiled code, so we must update "code"
2465 here. */
2466
2467 #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
2468 case OP_XCLASS:
2469 ccode = code += GET(code, 1);
2470 goto CHECK_CLASS_REPEAT;
2471 #endif
2472
2473 case OP_CLASS:
2474 case OP_NCLASS:
2475 ccode = code + PRIV(OP_lengths)[OP_CLASS];
2476
2477 #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
2478 CHECK_CLASS_REPEAT:
2479 #endif
2480
2481 switch (*ccode)
2482 {
2483 case OP_CRSTAR: /* These could be empty; continue */
2484 case OP_CRMINSTAR:
2485 case OP_CRQUERY:
2486 case OP_CRMINQUERY:
2487 break;
2488
2489 default: /* Non-repeat => class must match */
2490 case OP_CRPLUS: /* These repeats aren't empty */
2491 case OP_CRMINPLUS:
2492 return FALSE;
2493
2494 case OP_CRRANGE:
2495 case OP_CRMINRANGE:
2496 if (GET2(ccode, 1) > 0) return FALSE; /* Minimum > 0 */
2497 break;
2498 }
2499 break;
2500
2501 /* Opcodes that must match a character */
2502
2503 case OP_PROP:
2504 case OP_NOTPROP:
2505 case OP_EXTUNI:
2506 case OP_NOT_DIGIT:
2507 case OP_DIGIT:
2508 case OP_NOT_WHITESPACE:
2509 case OP_WHITESPACE:
2510 case OP_NOT_WORDCHAR:
2511 case OP_WORDCHAR:
2512 case OP_ANY:
2513 case OP_ALLANY:
2514 case OP_ANYBYTE:
2515 case OP_CHAR:
2516 case OP_CHARI:
2517 case OP_NOT:
2518 case OP_NOTI:
2519 case OP_PLUS:
2520 case OP_MINPLUS:
2521 case OP_POSPLUS:
2522 case OP_EXACT:
2523 case OP_NOTPLUS:
2524 case OP_NOTMINPLUS:
2525 case OP_NOTPOSPLUS:
2526 case OP_NOTEXACT:
2527 case OP_TYPEPLUS:
2528 case OP_TYPEMINPLUS:
2529 case OP_TYPEPOSPLUS:
2530 case OP_TYPEEXACT:
2531 return FALSE;
2532
2533 /* These are going to continue, as they may be empty, but we have to
2534 fudge the length for the \p and \P cases. */
2535
2536 case OP_TYPESTAR:
2537 case OP_TYPEMINSTAR:
2538 case OP_TYPEPOSSTAR:
2539 case OP_TYPEQUERY:
2540 case OP_TYPEMINQUERY:
2541 case OP_TYPEPOSQUERY:
2542 if (code[1] == OP_PROP || code[1] == OP_NOTPROP) code += 2;
2543 break;
2544
2545 /* Same for these */
2546
2547 case OP_TYPEUPTO:
2548 case OP_TYPEMINUPTO:
2549 case OP_TYPEPOSUPTO:
2550 if (code[1 + IMM2_SIZE] == OP_PROP || code[1 + IMM2_SIZE] == OP_NOTPROP)
2551 code += 2;
2552 break;
2553
2554 /* End of branch */
2555
2556 case OP_KET:
2557 case OP_KETRMAX:
2558 case OP_KETRMIN:
2559 case OP_KETRPOS:
2560 case OP_ALT:
2561 return TRUE;
2562
2563 /* In UTF-8 mode, STAR, MINSTAR, POSSTAR, QUERY, MINQUERY, POSQUERY, UPTO,
2564 MINUPTO, and POSUPTO may be followed by a multibyte character */
2565
2566 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
2567 case OP_STAR:
2568 case OP_STARI:
2569 case OP_MINSTAR:
2570 case OP_MINSTARI:
2571 case OP_POSSTAR:
2572 case OP_POSSTARI:
2573 case OP_QUERY:
2574 case OP_QUERYI:
2575 case OP_MINQUERY:
2576 case OP_MINQUERYI:
2577 case OP_POSQUERY:
2578 case OP_POSQUERYI:
2579 if (utf && HAS_EXTRALEN(code[1])) code += GET_EXTRALEN(code[1]);
2580 break;
2581
2582 case OP_UPTO:
2583 case OP_UPTOI:
2584 case OP_MINUPTO:
2585 case OP_MINUPTOI:
2586 case OP_POSUPTO:
2587 case OP_POSUPTOI:
2588 if (utf && HAS_EXTRALEN(code[1 + IMM2_SIZE])) code += GET_EXTRALEN(code[1 + IMM2_SIZE]);
2589 break;
2590 #endif
2591
2592 /* MARK, and PRUNE/SKIP/THEN with an argument must skip over the argument
2593 string. */
2594
2595 case OP_MARK:
2596 case OP_PRUNE_ARG:
2597 case OP_SKIP_ARG:
2598 code += code[1];
2599 break;
2600
2601 case OP_THEN_ARG:
2602 code += code[1];
2603 break;
2604
2605 /* None of the remaining opcodes are required to match a character. */
2606
2607 default:
2608 break;
2609 }
2610 }
2611
2612 return TRUE;
2613 }
2614
2615
2616
2617 /*************************************************
2618 * Scan compiled regex for non-emptiness *
2619 *************************************************/
2620
2621 /* This function is called to check for left recursive calls. We want to check
2622 the current branch of the current pattern to see if it could match the empty
2623 string. If it could, we must look outwards for branches at other levels,
2624 stopping when we pass beyond the bracket which is the subject of the recursion.
2625 This function is called only during the real compile, not during the
2626 pre-compile.
2627
2628 Arguments:
2629 code points to start of the recursion
2630 endcode points to where to stop (current RECURSE item)
2631 bcptr points to the chain of current (unclosed) branch starts
2632 utf TRUE if in UTF-8 / UTF-16 / UTF-32 mode
2633 cd pointers to tables etc
2634
2635 Returns: TRUE if what is matched could be empty
2636 */
2637
2638 static BOOL
2639 could_be_empty(const pcre_uchar *code, const pcre_uchar *endcode,
2640 branch_chain *bcptr, BOOL utf, compile_data *cd)
2641 {
2642 while (bcptr != NULL && bcptr->current_branch >= code)
2643 {
2644 if (!could_be_empty_branch(bcptr->current_branch, endcode, utf, cd))
2645 return FALSE;
2646 bcptr = bcptr->outer;
2647 }
2648 return TRUE;
2649 }
2650
2651
2652
2653 /*************************************************
2654 * Check for POSIX class syntax *
2655 *************************************************/
2656
2657 /* This function is called when the sequence "[:" or "[." or "[=" is
2658 encountered in a character class. It checks whether this is followed by a
2659 sequence of characters terminated by a matching ":]" or ".]" or "=]". If we
2660 reach an unescaped ']' without the special preceding character, return FALSE.
2661
2662 Originally, this function only recognized a sequence of letters between the
2663 terminators, but it seems that Perl recognizes any sequence of characters,
2664 though of course unknown POSIX names are subsequently rejected. Perl gives an
2665 "Unknown POSIX class" error for [:f\oo:] for example, where previously PCRE
2666 didn't consider this to be a POSIX class. Likewise for [:1234:].
2667
2668 The problem in trying to be exactly like Perl is in the handling of escapes. We
2669 have to be sure that [abc[:x\]pqr] is *not* treated as containing a POSIX
2670 class, but [abc[:x\]pqr:]] is (so that an error can be generated). The code
2671 below handles the special case of \], but does not try to do any other escape
2672 processing. This makes it different from Perl for cases such as [:l\ower:]
2673 where Perl recognizes it as the POSIX class "lower" but PCRE does not recognize
2674 "l\ower". This is a lesser evil that not diagnosing bad classes when Perl does,
2675 I think.
2676
2677 A user pointed out that PCRE was rejecting [:a[:digit:]] whereas Perl was not.
2678 It seems that the appearance of a nested POSIX class supersedes an apparent
2679 external class. For example, [:a[:digit:]b:] matches "a", "b", ":", or
2680 a digit.
2681
2682 In Perl, unescaped square brackets may also appear as part of class names. For
2683 example, [:a[:abc]b:] gives unknown POSIX class "[:abc]b:]". However, for
2684 [:a[:abc]b][b:] it gives unknown POSIX class "[:abc]b][b:]", which does not
2685 seem right at all. PCRE does not allow closing square brackets in POSIX class
2686 names.
2687
2688 Arguments:
2689 ptr pointer to the initial [
2690 endptr where to return the end pointer
2691
2692 Returns: TRUE or FALSE
2693 */
2694
2695 static BOOL
2696 check_posix_syntax(const pcre_uchar *ptr, const pcre_uchar **endptr)
2697 {
2698 int terminator; /* Don't combine these lines; the Solaris cc */
2699 terminator = *(++ptr); /* compiler warns about "non-constant" initializer. */
2700 for (++ptr; *ptr != 0; ptr++)
2701 {
2702 if (*ptr == CHAR_BACKSLASH && ptr[1] == CHAR_RIGHT_SQUARE_BRACKET)
2703 ptr++;
2704 else if (*ptr == CHAR_RIGHT_SQUARE_BRACKET) return FALSE;
2705 else
2706 {
2707 if (*ptr == terminator && ptr[1] == CHAR_RIGHT_SQUARE_BRACKET)
2708 {
2709 *endptr = ptr;
2710 return TRUE;
2711 }
2712 if (*ptr == CHAR_LEFT_SQUARE_BRACKET &&
2713 (ptr[1] == CHAR_COLON || ptr[1] == CHAR_DOT ||
2714 ptr[1] == CHAR_EQUALS_SIGN) &&
2715 check_posix_syntax(ptr, endptr))
2716 return FALSE;
2717 }
2718 }
2719 return FALSE;
2720 }
2721
2722
2723
2724
2725 /*************************************************
2726 * Check POSIX class name *
2727 *************************************************/
2728
2729 /* This function is called to check the name given in a POSIX-style class entry
2730 such as [:alnum:].
2731
2732 Arguments:
2733 ptr points to the first letter
2734 len the length of the name
2735
2736 Returns: a value representing the name, or -1 if unknown
2737 */
2738
2739 static int
2740 check_posix_name(const pcre_uchar *ptr, int len)
2741 {
2742 const char *pn = posix_names;
2743 register int yield = 0;
2744 while (posix_name_lengths[yield] != 0)
2745 {
2746 if (len == posix_name_lengths[yield] &&
2747 STRNCMP_UC_C8(ptr, pn, len) == 0) return yield;
2748 pn += posix_name_lengths[yield] + 1;
2749 yield++;
2750 }
2751 return -1;
2752 }
2753
2754
2755 /*************************************************
2756 * Adjust OP_RECURSE items in repeated group *
2757 *************************************************/
2758
2759 /* OP_RECURSE items contain an offset from the start of the regex to the group
2760 that is referenced. This means that groups can be replicated for fixed
2761 repetition simply by copying (because the recursion is allowed to refer to
2762 earlier groups that are outside the current group). However, when a group is
2763 optional (i.e. the minimum quantifier is zero), OP_BRAZERO or OP_SKIPZERO is
2764 inserted before it, after it has been compiled. This means that any OP_RECURSE
2765 items within it that refer to the group itself or any contained groups have to
2766 have their offsets adjusted. That one of the jobs of this function. Before it
2767 is called, the partially compiled regex must be temporarily terminated with
2768 OP_END.
2769
2770 This function has been extended with the possibility of forward references for
2771 recursions and subroutine calls. It must also check the list of such references
2772 for the group we are dealing with. If it finds that one of the recursions in
2773 the current group is on this list, it adjusts the offset in the list, not the
2774 value in the reference (which is a group number).
2775
2776 Arguments:
2777 group points to the start of the group
2778 adjust the amount by which the group is to be moved
2779 utf TRUE in UTF-8 / UTF-16 / UTF-32 mode
2780 cd contains pointers to tables etc.
2781 save_hwm the hwm forward reference pointer at the start of the group
2782
2783 Returns: nothing
2784 */
2785
2786 static void
2787 adjust_recurse(pcre_uchar *group, int adjust, BOOL utf, compile_data *cd,
2788 pcre_uchar *save_hwm)
2789 {
2790 pcre_uchar *ptr = group;
2791
2792 while ((ptr = (pcre_uchar *)find_recurse(ptr, utf)) != NULL)
2793 {
2794 int offset;
2795 pcre_uchar *hc;
2796
2797 /* See if this recursion is on the forward reference list. If so, adjust the
2798 reference. */
2799
2800 for (hc = save_hwm; hc < cd->hwm; hc += LINK_SIZE)
2801 {
2802 offset = GET(hc, 0);
2803 if (cd->start_code + offset == ptr + 1)
2804 {
2805 PUT(hc, 0, offset + adjust);
2806 break;
2807 }
2808 }
2809
2810 /* Otherwise, adjust the recursion offset if it's after the start of this
2811 group. */
2812
2813 if (hc >= cd->hwm)
2814 {
2815 offset = GET(ptr, 1);
2816 if (cd->start_code + offset >= group) PUT(ptr, 1, offset + adjust);
2817 }
2818
2819 ptr += 1 + LINK_SIZE;
2820 }
2821 }
2822
2823
2824
2825 /*************************************************
2826 * Insert an automatic callout point *
2827 *************************************************/
2828
2829 /* This function is called when the PCRE_AUTO_CALLOUT option is set, to insert
2830 callout points before each pattern item.
2831
2832 Arguments:
2833 code current code pointer
2834 ptr current pattern pointer
2835 cd pointers to tables etc
2836
2837 Returns: new code pointer
2838 */
2839
2840 static pcre_uchar *
2841 auto_callout(pcre_uchar *code, const pcre_uchar *ptr, compile_data *cd)
2842 {
2843 *code++ = OP_CALLOUT;
2844 *code++ = 255;
2845 PUT(code, 0, (int)(ptr - cd->start_pattern)); /* Pattern offset */
2846 PUT(code, LINK_SIZE, 0); /* Default length */
2847 return code + 2 * LINK_SIZE;
2848 }
2849
2850
2851
2852 /*************************************************
2853 * Complete a callout item *
2854 *************************************************/
2855
2856 /* A callout item contains the length of the next item in the pattern, which
2857 we can't fill in till after we have reached the relevant point. This is used
2858 for both automatic and manual callouts.
2859
2860 Arguments:
2861 previous_callout points to previous callout item
2862 ptr current pattern pointer
2863 cd pointers to tables etc
2864
2865 Returns: nothing
2866 */
2867
2868 static void
2869 complete_callout(pcre_uchar *previous_callout, const pcre_uchar *ptr, compile_data *cd)
2870 {
2871 int length = (int)(ptr - cd->start_pattern - GET(previous_callout, 2));
2872 PUT(previous_callout, 2 + LINK_SIZE, length);
2873 }
2874
2875
2876
2877 #ifdef SUPPORT_UCP
2878 /*************************************************
2879 * Get othercase range *
2880 *************************************************/
2881
2882 /* This function is passed the start and end of a class range, in UTF-8 mode
2883 with UCP support. It searches up the characters, looking for ranges of
2884 characters in the "other" case. Each call returns the next one, updating the
2885 start address. A character with multiple other cases is returned on its own
2886 with a special return value.
2887
2888 Arguments:
2889 cptr points to starting character value; updated
2890 d end value
2891 ocptr where to put start of othercase range
2892 odptr where to put end of othercase range
2893
2894 Yield: -1 when no more
2895 0 when a range is returned
2896 >0 the CASESET offset for char with multiple other cases
2897 in this case, ocptr contains the original
2898 */
2899
2900 static int
2901 get_othercase_range(unsigned int *cptr, unsigned int d, unsigned int *ocptr,
2902 unsigned int *odptr)
2903 {
2904 unsigned int c, othercase, next;
2905 int co;
2906
2907 /* Find the first character that has an other case. If it has multiple other
2908 cases, return its case offset value. */
2909
2910 for (c = *cptr; c <= d; c++)
2911 {
2912 if ((co = UCD_CASESET(c)) != 0)
2913 {
2914 *ocptr = c++; /* Character that has the set */
2915 *cptr = c; /* Rest of input range */
2916 return co;
2917 }
2918 if ((othercase = UCD_OTHERCASE(c)) != c) break;
2919 }
2920
2921 if (c > d) return -1; /* Reached end of range */
2922
2923 *ocptr = othercase;
2924 next = othercase + 1;
2925
2926 for (++c; c <= d; c++)
2927 {
2928 if (UCD_OTHERCASE(c) != next) break;
2929 next++;
2930 }
2931
2932 *odptr = next - 1; /* End of othercase range */
2933 *cptr = c; /* Rest of input range */
2934 return 0;
2935 }
2936
2937
2938
2939 /*************************************************
2940 * Check a character and a property *
2941 *************************************************/
2942
2943 /* This function is called by check_auto_possessive() when a property item
2944 is adjacent to a fixed character.
2945
2946 Arguments:
2947 c the character
2948 ptype the property type
2949 pdata the data for the type
2950 negated TRUE if it's a negated property (\P or \p{^)
2951
2952 Returns: TRUE if auto-possessifying is OK
2953 */
2954
2955 static BOOL
2956 check_char_prop(int c, int ptype, int pdata, BOOL negated)
2957 {
2958 #ifdef SUPPORT_UCP
2959 const pcre_uint32 *p;
2960 #endif
2961
2962 const ucd_record *prop = GET_UCD(c);
2963
2964 switch(ptype)
2965 {
2966 case PT_LAMP:
2967 return (prop->chartype == ucp_Lu ||
2968 prop->chartype == ucp_Ll ||
2969 prop->chartype == ucp_Lt) == negated;
2970
2971 case PT_GC:
2972 return (pdata == PRIV(ucp_gentype)[prop->chartype]) == negated;
2973
2974 case PT_PC:
2975 return (pdata == prop->chartype) == negated;
2976
2977 case PT_SC:
2978 return (pdata == prop->script) == negated;
2979
2980 /* These are specials */
2981
2982 case PT_ALNUM:
2983 return (PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2984 PRIV(ucp_gentype)[prop->chartype] == ucp_N) == negated;
2985
2986 case PT_SPACE: /* Perl space */
2987 return (PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2988 c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)
2989 == negated;
2990
2991 case PT_PXSPACE: /* POSIX space */
2992 return (PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2993 c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
2994 c == CHAR_FF || c == CHAR_CR)
2995 == negated;
2996
2997 case PT_WORD:
2998 return (PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2999 PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
3000 c == CHAR_UNDERSCORE) == negated;
3001
3002 #ifdef SUPPORT_UCP
3003 case PT_CLIST:
3004 p = PRIV(ucd_caseless_sets) + prop->caseset;
3005 for (;;)
3006 {
3007 if ((unsigned int)c < *p) return !negated;
3008 if ((unsigned int)c == *p++) return negated;
3009 }
3010 break; /* Control never reaches here */
3011 #endif
3012 }
3013
3014 return FALSE;
3015 }
3016 #endif /* SUPPORT_UCP */
3017
3018
3019
3020 /*************************************************
3021 * Check if auto-possessifying is possible *
3022 *************************************************/
3023
3024 /* This function is called for unlimited repeats of certain items, to see
3025 whether the next thing could possibly match the repeated item. If not, it makes
3026 sense to automatically possessify the repeated item.
3027
3028 Arguments:
3029 previous pointer to the repeated opcode
3030 utf TRUE in UTF-8 / UTF-16 / UTF-32 mode
3031 ptr next character in pattern
3032 options options bits
3033 cd contains pointers to tables etc.
3034
3035 Returns: TRUE if possessifying is wanted
3036 */
3037
3038 static BOOL
3039 check_auto_possessive(const pcre_uchar *previous, BOOL utf,
3040 const pcre_uchar *ptr, int options, compile_data *cd)
3041 {
3042 pcre_int32 c = NOTACHAR; // FIXMEchpe pcre_uint32
3043 pcre_int32 next;
3044 int escape;
3045 int op_code = *previous++;
3046
3047 /* Skip whitespace and comments in extended mode */
3048
3049 if ((options & PCRE_EXTENDED) != 0)
3050 {
3051 for (;;)
3052 {
3053 while (MAX_255(*ptr) && (cd->ctypes[*ptr] & ctype_space) != 0) ptr++;
3054 if (*ptr == CHAR_NUMBER_SIGN)
3055 {
3056 ptr++;
3057 while (*ptr != 0)
3058 {
3059 if (IS_NEWLINE(ptr)) { ptr += cd->nllen; break; }
3060 ptr++;
3061 #ifdef SUPPORT_UTF
3062 if (utf) FORWARDCHAR(ptr);
3063 #endif
3064 }
3065 }
3066 else break;
3067 }
3068 }
3069
3070 /* If the next item is one that we can handle, get its value. A non-negative
3071 value is a character, a negative value is an escape value. */
3072
3073 if (*ptr == CHAR_BACKSLASH)
3074 {
3075 int temperrorcode = 0;
3076 escape = check_escape(&ptr, &next, &temperrorcode, cd->bracount, options, FALSE);
3077 if (temperrorcode != 0) return FALSE;
3078 ptr++; /* Point after the escape sequence */
3079 }
3080 else if (!MAX_255(*ptr) || (cd->ctypes[*ptr] & ctype_meta) == 0)
3081 {
3082 escape = 0;
3083 #ifdef SUPPORT_UTF
3084 if (utf) { GETCHARINC(next, ptr); } else
3085 #endif
3086 next = *ptr++;
3087 }
3088 else return FALSE;
3089
3090 /* Skip whitespace and comments in extended mode */
3091
3092 if ((options & PCRE_EXTENDED) != 0)
3093 {
3094 for (;;)
3095 {
3096 while (MAX_255(*ptr) && (cd->ctypes[*ptr] & ctype_space) != 0) ptr++;
3097 if (*ptr == CHAR_NUMBER_SIGN)
3098 {
3099 ptr++;
3100 while (*ptr != 0)
3101 {
3102 if (IS_NEWLINE(ptr)) { ptr += cd->nllen; break; }
3103 ptr++;
3104 #ifdef SUPPORT_UTF
3105 if (utf) FORWARDCHAR(ptr);
3106 #endif
3107 }
3108 }
3109 else break;
3110 }
3111 }
3112
3113 /* If the next thing is itself optional, we have to give up. */
3114
3115 if (*ptr == CHAR_ASTERISK || *ptr == CHAR_QUESTION_MARK ||
3116 STRNCMP_UC_C8(ptr, STR_LEFT_CURLY_BRACKET STR_0 STR_COMMA, 3) == 0)
3117 return FALSE;
3118
3119 /* If the previous item is a character, get its value. */
3120
3121 if (op_code == OP_CHAR || op_code == OP_CHARI ||
3122 op_code == OP_NOT || op_code == OP_NOTI)
3123 //if (escape == 0) switch(op_code)
3124 {
3125 #ifdef SUPPORT_UTF
3126 GETCHARTEST(c, previous);
3127 #else
3128 c = *previous;
3129 #endif
3130 }
3131
3132 /* Now compare the next item with the previous opcode. First, handle cases when
3133 the next item is a character. */
3134
3135 if (escape == 0)
3136 {
3137 /* For a caseless UTF match, the next character may have more than one other
3138 case, which maps to the special PT_CLIST property. Check this first. */
3139
3140 #ifdef SUPPORT_UCP
3141 if (utf && (unsigned int)c != NOTACHAR && (options & PCRE_CASELESS) != 0)
3142 {
3143 int ocs = UCD_CASESET(next);
3144 if (ocs > 0) return check_char_prop(c, PT_CLIST, ocs, op_code >= OP_NOT);
3145 }
3146 #endif
3147
3148 switch(op_code)
3149 {
3150 case OP_CHAR:
3151 return c != next;
3152
3153 /* For CHARI (caseless character) we must check the other case. If we have
3154 Unicode property support, we can use it to test the other case of
3155 high-valued characters. We know that next can have only one other case,
3156 because multi-other-case characters are dealt with above. */
3157
3158 case OP_CHARI:
3159 if (c == next) return FALSE;
3160 #ifdef SUPPORT_UTF
3161 if (utf)
3162 {
3163 unsigned int othercase;
3164 if (next < 128) othercase = cd->fcc[next]; else
3165 #ifdef SUPPORT_UCP
3166 othercase = UCD_OTHERCASE((unsigned int)next);
3167 #else
3168 othercase = NOTACHAR;
3169 #endif
3170 return (unsigned int)c != othercase;
3171 }
3172 else
3173 #endif /* SUPPORT_UTF */
3174 return (c != TABLE_GET((unsigned int)next, cd->fcc, next)); /* Not UTF */
3175
3176 case OP_NOT:
3177 return c == next;
3178
3179 case OP_NOTI:
3180 if (c == next) return TRUE;
3181 #ifdef SUPPORT_UTF
3182 if (utf)
3183 {
3184 unsigned int othercase;
3185 if (next < 128) othercase = cd->fcc[next]; else
3186 #ifdef SUPPORT_UCP
3187 othercase = UCD_OTHERCASE((unsigned int)next);
3188 #else
3189 othercase = NOTACHAR;
3190 #endif
3191 return (unsigned int)c == othercase;
3192 }
3193 else
3194 #endif /* SUPPORT_UTF */
3195 return (c == TABLE_GET((unsigned int)next, cd->fcc, next)); /* Not UTF */
3196
3197 /* Note that OP_DIGIT etc. are generated only when PCRE_UCP is *not* set.
3198 When it is set, \d etc. are converted into OP_(NOT_)PROP codes. */
3199
3200 case OP_DIGIT:
3201 return next > 255 || (cd->ctypes[next] & ctype_digit) == 0;
3202
3203 case OP_NOT_DIGIT:
3204 return next <= 255 && (cd->ctypes[next] & ctype_digit) != 0;
3205
3206 case OP_WHITESPACE:
3207 return next > 255 || (cd->ctypes[next] & ctype_space) == 0;
3208
3209 case OP_NOT_WHITESPACE:
3210 return next <= 255 && (cd->ctypes[next] & ctype_space) != 0;
3211
3212 case OP_WORDCHAR:
3213 return next > 255 || (cd->ctypes[next] & ctype_word) == 0;
3214
3215 case OP_NOT_WORDCHAR:
3216 return next <= 255 && (cd->ctypes[next] & ctype_word) != 0;
3217
3218 case OP_HSPACE:
3219 case OP_NOT_HSPACE:
3220 switch(next)
3221 {
3222 HSPACE_CASES:
3223 return op_code == OP_NOT_HSPACE;
3224
3225 default:
3226 return op_code != OP_NOT_HSPACE;
3227 }
3228
3229 case OP_ANYNL:
3230 case OP_VSPACE:
3231 case OP_NOT_VSPACE:
3232 switch(next)
3233 {
3234 VSPACE_CASES:
3235 return op_code == OP_NOT_VSPACE;
3236
3237 default:
3238 return op_code != OP_NOT_VSPACE;
3239 }
3240
3241 #ifdef SUPPORT_UCP
3242 case OP_PROP:
3243 return check_char_prop(next, previous[0], previous[1], FALSE);
3244
3245 case OP_NOTPROP:
3246 return check_char_prop(next, previous[0], previous[1], TRUE);
3247 #endif
3248
3249 default:
3250 return FALSE;
3251 }
3252 }
3253
3254 /* Handle the case when the next item is \d, \s, etc. Note that when PCRE_UCP
3255 is set, \d turns into ESC_du rather than ESC_d, etc., so ESC_d etc. are
3256 generated only when PCRE_UCP is *not* set, that is, when only ASCII
3257 characteristics are recognized. Similarly, the opcodes OP_DIGIT etc. are
3258 replaced by OP_PROP codes when PCRE_UCP is set. */
3259
3260 switch(op_code)
3261 {
3262 case OP_CHAR:
3263 case OP_CHARI:
3264 switch(escape)
3265 {
3266 case ESC_d:
3267 return c > 255 || (cd->ctypes[c] & ctype_digit) == 0;
3268
3269 case ESC_D:
3270 return c <= 255 && (cd->ctypes[c] & ctype_digit) != 0;
3271
3272 case ESC_s:
3273 return c > 255 || (cd->ctypes[c] & ctype_space) == 0;
3274
3275 case ESC_S:
3276 return c <= 255 && (cd->ctypes[c] & ctype_space) != 0;
3277
3278 case ESC_w:
3279 return c > 255 || (cd->ctypes[c] & ctype_word) == 0;
3280
3281 case ESC_W:
3282 return c <= 255 && (cd->ctypes[c] & ctype_word) != 0;
3283
3284 case ESC_h:
3285 case ESC_H:
3286 switch(c)
3287 {
3288 HSPACE_CASES:
3289 return escape != ESC_h;
3290
3291 default:
3292 return escape == ESC_h;
3293 }
3294
3295 case ESC_v:
3296 case ESC_V:
3297 switch(c)
3298 {
3299 VSPACE_CASES:
3300 return escape != ESC_v;
3301
3302 default:
3303 return escape == ESC_v;
3304 }
3305
3306 /* When PCRE_UCP is set, these values get generated for \d etc. Find
3307 their substitutions and process them. The result will always be either
3308 ESC_p or ESC_P. Then fall through to process those values. */
3309
3310 #ifdef SUPPORT_UCP
3311 case ESC_du:
3312 case ESC_DU:
3313 case ESC_wu:
3314 case ESC_WU:
3315 case ESC_su:
3316 case ESC_SU:
3317 {
3318 int temperrorcode = 0;
3319 ptr = substitutes[escape - ESC_DU];
3320 escape = check_escape(&ptr, &next, &temperrorcode, 0, options, FALSE);
3321 if (temperrorcode != 0) return FALSE;
3322 ptr++; /* For compatibility */
3323 }
3324 /* Fall through */
3325
3326 case ESC_p:
3327 case ESC_P:
3328 {
3329 int ptype, pdata, errorcodeptr;
3330 BOOL negated;
3331
3332 ptr--; /* Make ptr point at the p or P */
3333 ptype = get_ucp(&ptr, &negated, &pdata, &errorcodeptr);
3334 if (ptype < 0) return FALSE;
3335 ptr++; /* Point past the final curly ket */
3336
3337 /* If the property item is optional, we have to give up. (When generated
3338 from \d etc by PCRE_UCP, this test will have been applied much earlier,
3339 to the original \d etc. At this point, ptr will point to a zero byte. */
3340
3341 if (*ptr == CHAR_ASTERISK || *ptr == CHAR_QUESTION_MARK ||
3342 STRNCMP_UC_C8(ptr, STR_LEFT_CURLY_BRACKET STR_0 STR_COMMA, 3) == 0)
3343 return FALSE;
3344
3345 /* Do the property check. */
3346
3347 return check_char_prop(c, ptype, pdata, (escape == ESC_P) != negated);
3348 }
3349 #endif
3350
3351 default:
3352 return FALSE;
3353 }
3354
3355 /* In principle, support for Unicode properties should be integrated here as
3356 well. It means re-organizing the above code so as to get hold of the property
3357 values before switching on the op-code. However, I wonder how many patterns
3358 combine ASCII \d etc with Unicode properties? (Note that if PCRE_UCP is set,
3359 these op-codes are never generated.) */
3360
3361 case OP_DIGIT:
3362 return escape == ESC_D || escape == ESC_s || escape == ESC_W ||
3363 escape == ESC_h || escape == ESC_v || escape == ESC_R;
3364
3365 case OP_NOT_DIGIT:
3366 return escape == ESC_d;
3367
3368 case OP_WHITESPACE:
3369 return escape == ESC_S || escape == ESC_d || escape == ESC_w;
3370
3371 case OP_NOT_WHITESPACE:
3372 return escape == ESC_s || escape == ESC_h || escape == ESC_v || escape == ESC_R;
3373
3374 case OP_HSPACE:
3375 return escape == ESC_S || escape == ESC_H || escape == ESC_d ||
3376 escape == ESC_w || escape == ESC_v || escape == ESC_R;
3377
3378 case OP_NOT_HSPACE:
3379 return escape == ESC_h;
3380
3381 /* Can't have \S in here because VT matches \S (Perl anomaly) */
3382 case OP_ANYNL:
3383 case OP_VSPACE:
3384 return escape == ESC_V || escape == ESC_d || escape == ESC_w;
3385
3386 case OP_NOT_VSPACE:
3387 return escape == ESC_v || escape == ESC_R;
3388
3389 case OP_WORDCHAR:
3390 return escape == ESC_W || escape == ESC_s || escape == ESC_h ||
3391 escape == ESC_v || escape == ESC_R;
3392
3393 case OP_NOT_WORDCHAR:
3394 return escape == ESC_w || escape == ESC_d;
3395
3396 default:
3397 return FALSE;
3398 }
3399
3400 /* Control does not reach here */
3401 }
3402
3403
3404
3405 /*************************************************
3406 * Add a character or range to a class *
3407 *************************************************/
3408
3409 /* This function packages up the logic of adding a character or range of
3410 characters to a class. The character values in the arguments will be within the
3411 valid values for the current mode (8-bit, 16-bit, UTF, etc). This function is
3412 mutually recursive with the function immediately below.
3413
3414 Arguments:
3415 classbits the bit map for characters < 256
3416 uchardptr points to the pointer for extra data
3417 options the options word
3418 cd contains pointers to tables etc.
3419 start start of range character
3420 end end of range character
3421
3422 Returns: the number of < 256 characters added
3423 the pointer to extra data is updated
3424 */
3425
3426 static int
3427 add_to_class(pcre_uint8 *classbits, pcre_uchar **uchardptr, int options,
3428 compile_data *cd, pcre_uint32 start, pcre_uint32 end)
3429 {
3430 pcre_uint32 c;
3431 int n8 = 0;
3432
3433 /* If caseless matching is required, scan the range and process alternate
3434 cases. In Unicode, there are 8-bit characters that have alternate cases that
3435 are greater than 255 and vice-versa. Sometimes we can just extend the original
3436 range. */
3437
3438 if ((options & PCRE_CASELESS) != 0)
3439 {
3440 #ifdef SUPPORT_UCP
3441 if ((options & PCRE_UTF8) != 0)
3442 {
3443 int rc;
3444 pcre_uint32 oc, od;
3445
3446 options &= ~PCRE_CASELESS; /* Remove for recursive calls */
3447 c = start;
3448
3449 while ((rc = get_othercase_range(&c, end, &oc, &od)) >= 0)
3450 {
3451 /* Handle a single character that has more than one other case. */
3452
3453 if (rc > 0) n8 += add_list_to_class(classbits, uchardptr, options, cd,
3454 PRIV(ucd_caseless_sets) + rc, oc);
3455
3456 /* Do nothing if the other case range is within the original range. */
3457
3458 else if (oc >= start && od <= end) continue;
3459
3460 /* Extend the original range if there is overlap, noting that if oc < c, we
3461 can't have od > end because a subrange is always shorter than the basic
3462 range. Otherwise, use a recursive call to add the additional range. */
3463
3464 else if (oc < start && od >= start - 1) start = oc; /* Extend downwards */
3465 else if (od > end && oc <= end + 1) end = od; /* Extend upwards */
3466 else n8 += add_to_class(classbits, uchardptr, options, cd, oc, od);
3467 }
3468 }
3469 else
3470 #endif /* SUPPORT_UCP */
3471
3472 /* Not UTF-mode, or no UCP */
3473
3474 for (c = start; c <= end && c < 256; c++)
3475 {
3476 SETBIT(classbits, cd->fcc[c]);
3477 n8++;
3478 }
3479 }
3480
3481 /* Now handle the original range. Adjust the final value according to the bit
3482 length - this means that the same lists of (e.g.) horizontal spaces can be used
3483 in all cases. */
3484
3485 #if defined COMPILE_PCRE8
3486 #ifdef SUPPORT_UTF
3487 if ((options & PCRE_UTF8) == 0)
3488 #endif
3489 if (end > 0xff) end = 0xff;
3490
3491 #elif defined COMPILE_PCRE16
3492 #ifdef SUPPORT_UTF
3493 if ((options & PCRE_UTF16) == 0)
3494 #endif
3495 if (end > 0xffff) end = 0xffff;
3496
3497 #endif /* COMPILE_PCRE[8|16] */
3498
3499 /* If all characters are less than 256, use the bit map. Otherwise use extra
3500 data. */
3501
3502 if (end < 0x100)
3503 {
3504 for (c = start; c <= end; c++)
3505 {
3506 n8++;
3507 SETBIT(classbits, c);
3508 }
3509 }
3510
3511 else
3512 {
3513 pcre_uchar *uchardata = *uchardptr;
3514
3515 #ifdef SUPPORT_UTF
3516 if ((options & PCRE_UTF8) != 0) /* All UTFs use the same flag bit */
3517 {
3518 if (start < end)
3519 {
3520 *uchardata++ = XCL_RANGE;
3521 uchardata += PRIV(ord2utf)(start, uchardata);
3522 uchardata += PRIV(ord2utf)(end, uchardata);
3523 }
3524 else if (start == end)
3525 {
3526 *uchardata++ = XCL_SINGLE;
3527 uchardata += PRIV(ord2utf)(start, uchardata);
3528 }
3529 }
3530 else
3531 #endif /* SUPPORT_UTF */
3532
3533 /* Without UTF support, character values are constrained by the bit length,
3534 and can only be > 256 for 16-bit and 32-bit libraries. */
3535
3536 #ifdef COMPILE_PCRE8
3537 {}
3538 #else
3539 if (start < end)
3540 {
3541 *uchardata++ = XCL_RANGE;
3542 *uchardata++ = start;
3543 *uchardata++ = end;
3544 }
3545 else if (start == end)
3546 {
3547 *uchardata++ = XCL_SINGLE;
3548 *uchardata++ = start;
3549 }
3550 #endif
3551
3552 *uchardptr = uchardata; /* Updata extra data pointer */
3553 }
3554
3555 return n8; /* Number of 8-bit characters */
3556 }
3557
3558
3559
3560
3561 /*************************************************
3562 * Add a list of characters to a class *
3563 *************************************************/
3564
3565 /* This function is used for adding a list of case-equivalent characters to a
3566 class, and also for adding a list of horizontal or vertical whitespace. If the
3567 list is in order (which it should be), ranges of characters are detected and
3568 handled appropriately. This function is mutually recursive with the function
3569 above.
3570
3571 Arguments:
3572 classbits the bit map for characters < 256
3573 uchardptr points to the pointer for extra data
3574 options the options word
3575 cd contains pointers to tables etc.
3576 p points to row of 32-bit values, terminated by NOTACHAR
3577 except character to omit; this is used when adding lists of
3578 case-equivalent characters to avoid including the one we
3579 already know about
3580
3581 Returns: the number of < 256 characters added
3582 the pointer to extra data is updated
3583 */
3584
3585 static int
3586 add_list_to_class(pcre_uint8 *classbits, pcre_uchar **uchardptr, int options,
3587 compile_data *cd, const pcre_uint32 *p, unsigned int except)
3588 {
3589 int n8 = 0;
3590 while (p[0] < NOTACHAR)
3591 {
3592 int n = 0;
3593 if (p[0] != except)
3594 {
3595 while(p[n+1] == p[0] + n + 1) n++;
3596 n8 += add_to_class(classbits, uchardptr, options, cd, p[0], p[n]);
3597 }
3598 p += n + 1;
3599 }
3600 return n8;
3601 }
3602
3603
3604
3605 /*************************************************
3606 * Add characters not in a list to a class *
3607 *************************************************/
3608
3609 /* This function is used for adding the complement of a list of horizontal or
3610 vertical whitespace to a class. The list must be in order.
3611
3612 Arguments:
3613 classbits the bit map for characters < 256
3614 uchardptr points to the pointer for extra data
3615 options the options word
3616 cd contains pointers to tables etc.
3617 p points to row of 32-bit values, terminated by NOTACHAR
3618
3619 Returns: the number of < 256 characters added
3620 the pointer to extra data is updated
3621 */
3622
3623 static int
3624 add_not_list_to_class(pcre_uint8 *classbits, pcre_uchar **uchardptr,
3625 int options, compile_data *cd, const pcre_uint32 *p)
3626 {
3627 BOOL utf = (options & PCRE_UTF8) != 0;
3628 int n8 = 0;
3629 if (p[0] > 0)
3630 n8 += add_to_class(classbits, uchardptr, options, cd, 0, p[0] - 1);
3631 while (p[0] < NOTACHAR)
3632 {
3633 while (p[1] == p[0] + 1) p++;
3634 n8 += add_to_class(classbits, uchardptr, options, cd, p[0] + 1,
3635 (p[1] == NOTACHAR) ? (utf ? 0x10ffffu : 0xffffffffu) : p[1] - 1);
3636 p++;
3637 }
3638 return n8;
3639 }
3640
3641
3642
3643 /*************************************************
3644 * Compile one branch *
3645 *************************************************/
3646
3647 /* Scan the pattern, compiling it into the a vector. If the options are
3648 changed during the branch, the pointer is used to change the external options
3649 bits. This function is used during the pre-compile phase when we are trying
3650 to find out the amount of memory needed, as well as during the real compile
3651 phase. The value of lengthptr distinguishes the two phases.
3652
3653 Arguments:
3654 optionsptr pointer to the option bits
3655 codeptr points to the pointer to the current code point
3656 ptrptr points to the current pattern pointer
3657 errorcodeptr points to error code variable
3658 firstcharptr set to initial literal character, or < 0 (REQ_UNSET, REQ_NONE)
3659 reqcharptr set to the last literal character required, else < 0
3660 bcptr points to current branch chain
3661 cond_depth conditional nesting depth
3662 cd contains pointers to tables etc.
3663 lengthptr NULL during the real compile phase
3664 points to length accumulator during pre-compile phase
3665
3666 Returns: TRUE on success
3667 FALSE, with *errorcodeptr set non-zero on error
3668 */
3669
3670 static BOOL
3671 compile_branch(int *optionsptr, pcre_uchar **codeptr,
3672 const pcre_uchar **ptrptr, int *errorcodeptr, pcre_int32 *firstcharptr,
3673 pcre_int32 *reqcharptr, branch_chain *bcptr, int cond_depth,
3674 compile_data *cd, int *lengthptr)
3675 {
3676 int repeat_type, op_type;
3677 int repeat_min = 0, repeat_max = 0; /* To please picky compilers */
3678 int bravalue = 0;
3679 int greedy_default, greedy_non_default;
3680 pcre_int32 firstchar, reqchar;
3681 pcre_int32 zeroreqchar, zerofirstchar;
3682 pcre_int32 req_caseopt, reqvary, tempreqvary;
3683 int options = *optionsptr; /* May change dynamically */
3684 int after_manual_callout = 0;
3685 int length_prevgroup = 0;
3686 register int c;
3687 int escape;
3688 register pcre_uchar *code = *codeptr;
3689 pcre_uchar *last_code = code;
3690 pcre_uchar *orig_code = code;
3691 pcre_uchar *tempcode;
3692 BOOL inescq = FALSE;
3693 BOOL groupsetfirstchar = FALSE;
3694 const pcre_uchar *ptr = *ptrptr;
3695 const pcre_uchar *tempptr;
3696 const pcre_uchar *nestptr = NULL;
3697 pcre_uchar *previous = NULL;
3698 pcre_uchar *previous_callout = NULL;
3699 pcre_uchar *save_hwm = NULL;
3700 pcre_uint8 classbits[32];
3701
3702 /* We can fish out the UTF-8 setting once and for all into a BOOL, but we
3703 must not do this for other options (e.g. PCRE_EXTENDED) because they may change
3704 dynamically as we process the pattern. */
3705
3706 #ifdef SUPPORT_UTF
3707 /* PCRE_UTF[16|32] have the same value as PCRE_UTF8. */
3708 BOOL utf = (options & PCRE_UTF8) != 0;
3709 pcre_uchar utf_chars[6];
3710 #else
3711 BOOL utf = FALSE;
3712 #endif
3713
3714 /* Helper variables for OP_XCLASS opcode (for characters > 255). We define
3715 class_uchardata always so that it can be passed to add_to_class() always,
3716 though it will not be used in non-UTF 8-bit cases. This avoids having to supply
3717 alternative calls for the different cases. */
3718
3719 pcre_uchar *class_uchardata;
3720 #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
3721 BOOL xclass;
3722 pcre_uchar *class_uchardata_base;
3723 #endif
3724
3725 #ifdef PCRE_DEBUG
3726 if (lengthptr != NULL) DPRINTF((">> start branch\n"));
3727 #endif
3728
3729 /* Set up the default and non-default settings for greediness */
3730
3731 greedy_default = ((options & PCRE_UNGREEDY) != 0);
3732 greedy_non_default = greedy_default ^ 1;
3733
3734 /* Initialize no first byte, no required byte. REQ_UNSET means "no char
3735 matching encountered yet". It gets changed to REQ_NONE if we hit something that
3736 matches a non-fixed char first char; reqchar just remains unset if we never
3737 find one.
3738
3739 When we hit a repeat whose minimum is zero, we may have to adjust these values
3740 to take the zero repeat into account. This is implemented by setting them to
3741 zerofirstbyte and zeroreqchar when such a repeat is encountered. The individual
3742 item types that can be repeated set these backoff variables appropriately. */
3743
3744 firstchar = reqchar = zerofirstchar = zeroreqchar = REQ_UNSET;
3745
3746 /* The variable req_caseopt contains either the REQ_CASELESS value
3747 or zero, according to the current setting of the caseless flag. The
3748 REQ_CASELESS leaves the lower 28 bit empty. It is added into the
3749 firstchar or reqchar variables to record the case status of the
3750 value. This is used only for ASCII characters. */
3751
3752 req_caseopt = ((options & PCRE_CASELESS) != 0)? REQ_CASELESS:0;
3753
3754 /* Switch on next character until the end of the branch */
3755
3756 for (;; ptr++)
3757 {
3758 BOOL negate_class;
3759 BOOL should_flip_negation;
3760 BOOL possessive_quantifier;
3761 BOOL is_quantifier;
3762 BOOL is_recurse;
3763 BOOL reset_bracount;
3764 int class_has_8bitchar;
3765 int class_one_char;
3766 int newoptions;
3767 int recno;
3768 int refsign;
3769 int skipbytes;
3770 int subreqchar;
3771 int subfirstchar;
3772 int terminator;
3773 int mclength;
3774 int tempbracount;
3775 int ec; // FIXMEchpe pcre_uint32
3776 pcre_uchar mcbuffer[8];
3777
3778 /* Get next character in the pattern */
3779
3780 c = *ptr;
3781
3782 /* If we are at the end of a nested substitution, revert to the outer level
3783 string. Nesting only happens one level deep. */
3784
3785 if (c == 0 && nestptr != NULL)
3786 {
3787 ptr = nestptr;
3788 nestptr = NULL;
3789 c = *ptr;
3790 }
3791
3792 /* If we are in the pre-compile phase, accumulate the length used for the
3793 previous cycle of this loop. */
3794
3795 if (lengthptr != NULL)
3796 {
3797 #ifdef PCRE_DEBUG
3798 if (code > cd->hwm) cd->hwm = code; /* High water info */
3799 #endif
3800 if (code > cd->start_workspace + cd->workspace_size -
3801 WORK_SIZE_SAFETY_MARGIN) /* Check for overrun */
3802 {
3803 *errorcodeptr = ERR52;
3804 goto FAILED;
3805 }
3806
3807 /* There is at least one situation where code goes backwards: this is the
3808 case of a zero quantifier after a class (e.g. [ab]{0}). At compile time,
3809 the class is simply eliminated. However, it is created first, so we have to
3810 allow memory for it. Therefore, don't ever reduce the length at this point.
3811 */
3812
3813 if (code < last_code) code = last_code;
3814
3815 /* Paranoid check for integer overflow */
3816
3817 if (OFLOW_MAX - *lengthptr < code - last_code)
3818 {
3819 *errorcodeptr = ERR20;
3820 goto FAILED;
3821 }
3822
3823 *lengthptr += (int)(code - last_code);
3824 DPRINTF(("length=%d added %d c=%c (0x%x)\n", *lengthptr,
3825 (int)(code - last_code), c, c));
3826
3827 /* If "previous" is set and it is not at the start of the work space, move
3828 it back to there, in order to avoid filling up the work space. Otherwise,
3829 if "previous" is NULL, reset the current code pointer to the start. */
3830
3831 if (previous != NULL)
3832 {
3833 if (previous > orig_code)
3834 {
3835 memmove(orig_code, previous, IN_UCHARS(code - previous));
3836 code -= previous - orig_code;
3837 previous = orig_code;
3838 }
3839 }
3840 else code = orig_code;
3841
3842 /* Remember where this code item starts so we can pick up the length
3843 next time round. */
3844
3845 last_code = code;
3846 }
3847
3848 /* In the real compile phase, just check the workspace used by the forward
3849 reference list. */
3850
3851 else if (cd->hwm > cd->start_workspace + cd->workspace_size -
3852 WORK_SIZE_SAFETY_MARGIN)
3853 {
3854 *errorcodeptr = ERR52;
3855 goto FAILED;
3856 }
3857
3858 /* If in \Q...\E, check for the end; if not, we have a literal */
3859
3860 if (inescq && c != 0)
3861 {
3862 if (c == CHAR_BACKSLASH && ptr[1] == CHAR_E)
3863 {
3864 inescq = FALSE;
3865 ptr++;
3866 continue;
3867 }
3868 else
3869 {
3870 if (previous_callout != NULL)
3871 {
3872 if (lengthptr == NULL) /* Don't attempt in pre-compile phase */
3873 complete_callout(previous_callout, ptr, cd);
3874 previous_callout = NULL;
3875 }
3876 if ((options & PCRE_AUTO_CALLOUT) != 0)
3877 {
3878 previous_callout = code;
3879 code = auto_callout(code, ptr, cd);
3880 }
3881 goto NORMAL_CHAR;
3882 }
3883 }
3884
3885 /* Fill in length of a previous callout, except when the next thing is
3886 a quantifier. */
3887
3888 is_quantifier =
3889 c == CHAR_ASTERISK || c == CHAR_PLUS || c == CHAR_QUESTION_MARK ||
3890 (c == CHAR_LEFT_CURLY_BRACKET && is_counted_repeat(ptr+1));
3891
3892 if (!is_quantifier && previous_callout != NULL &&
3893 after_manual_callout-- <= 0)
3894 {
3895 if (lengthptr == NULL) /* Don't attempt in pre-compile phase */
3896 complete_callout(previous_callout, ptr, cd);
3897 previous_callout = NULL;
3898 }
3899
3900 /* In extended mode, skip white space and comments. */
3901
3902 if ((options & PCRE_EXTENDED) != 0)
3903 {
3904 if (MAX_255(*ptr) && (cd->ctypes[c] & ctype_space) != 0) continue;
3905 if (c == CHAR_NUMBER_SIGN)
3906 {
3907 ptr++;
3908 while (*ptr != 0)
3909 {
3910 if (IS_NEWLINE(ptr)) { ptr += cd->nllen - 1; break; }
3911 ptr++;
3912 #ifdef SUPPORT_UTF
3913 if (utf) FORWARDCHAR(ptr);
3914 #endif
3915 }
3916 if (*ptr != 0) continue;
3917
3918 /* Else fall through to handle end of string */
3919 c = 0;
3920 }
3921 }
3922
3923 /* No auto callout for quantifiers. */
3924
3925 if ((options & PCRE_AUTO_CALLOUT) != 0 && !is_quantifier)
3926 {
3927 previous_callout = code;
3928 code = auto_callout(code, ptr, cd);
3929 }
3930
3931 switch(c)
3932 {
3933 /* ===================================================================*/
3934 case 0: /* The branch terminates at string end */
3935 case CHAR_VERTICAL_LINE: /* or | or ) */
3936 case CHAR_RIGHT_PARENTHESIS:
3937 *firstcharptr = firstchar;
3938 *reqcharptr = reqchar;
3939 *codeptr = code;
3940 *ptrptr = ptr;
3941 if (lengthptr != NULL)
3942 {
3943 if (OFLOW_MAX - *lengthptr < code - last_code)
3944 {
3945 *errorcodeptr = ERR20;
3946 goto FAILED;
3947 }
3948 *lengthptr += (int)(code - last_code); /* To include callout length */
3949 DPRINTF((">> end branch\n"));
3950 }
3951 return TRUE;
3952
3953
3954 /* ===================================================================*/
3955 /* Handle single-character metacharacters. In multiline mode, ^ disables
3956 the setting of any following char as a first character. */
3957
3958 case CHAR_CIRCUMFLEX_ACCENT:
3959 previous = NULL;
3960 if ((options & PCRE_MULTILINE) != 0)
3961 {
3962 if (firstchar == REQ_UNSET) firstchar = REQ_NONE;
3963 *code++ = OP_CIRCM;
3964 }
3965 else *code++ = OP_CIRC;
3966 break;
3967
3968 case CHAR_DOLLAR_SIGN:
3969 previous = NULL;
3970 *code++ = ((options & PCRE_MULTILINE) != 0)? OP_DOLLM : OP_DOLL;
3971 break;
3972
3973 /* There can never be a first char if '.' is first, whatever happens about
3974 repeats. The value of reqchar doesn't change either. */
3975
3976 case CHAR_DOT:
3977 if (firstchar == REQ_UNSET) firstchar = REQ_NONE;
3978 zerofirstchar = firstchar;
3979 zeroreqchar = reqchar;
3980 previous = code;
3981 *code++ = ((options & PCRE_DOTALL) != 0)? OP_ALLANY: OP_ANY;
3982 break;
3983
3984
3985 /* ===================================================================*/
3986 /* Character classes. If the included characters are all < 256, we build a
3987 32-byte bitmap of the permitted characters, except in the special case
3988 where there is only one such character. For negated classes, we build the
3989 map as usual, then invert it at the end. However, we use a different opcode
3990 so that data characters > 255 can be handled correctly.
3991
3992 If the class contains characters outside the 0-255 range, a different
3993 opcode is compiled. It may optionally have a bit map for characters < 256,
3994 but those above are are explicitly listed afterwards. A flag byte tells
3995 whether the bitmap is present, and whether this is a negated class or not.
3996
3997 In JavaScript compatibility mode, an isolated ']' causes an error. In
3998 default (Perl) mode, it is treated as a data character. */
3999
4000 case CHAR_RIGHT_SQUARE_BRACKET:
4001 if ((cd->external_options & PCRE_JAVASCRIPT_COMPAT) != 0)
4002 {
4003 *errorcodeptr = ERR64;
4004 goto FAILED;
4005 }
4006 goto NORMAL_CHAR;
4007
4008 case CHAR_LEFT_SQUARE_BRACKET:
4009 previous = code;
4010
4011 /* PCRE supports POSIX class stuff inside a class. Perl gives an error if
4012 they are encountered at the top level, so we'll do that too. */
4013
4014 if ((ptr[1] == CHAR_COLON || ptr[1] == CHAR_DOT ||
4015 ptr[1] == CHAR_EQUALS_SIGN) &&
4016 check_posix_syntax(ptr, &tempptr))
4017 {
4018 *errorcodeptr = (ptr[1] == CHAR_COLON)? ERR13 : ERR31;
4019 goto FAILED;
4020 }
4021
4022 /* If the first character is '^', set the negation flag and skip it. Also,
4023 if the first few characters (either before or after ^) are \Q\E or \E we
4024 skip them too. This makes for compatibility with Perl. */
4025
4026 negate_class = FALSE;
4027 for (;;)
4028 {
4029 c = *(++ptr);
4030 if (c == CHAR_BACKSLASH)
4031 {
4032 if (ptr[1] == CHAR_E)
4033 ptr++;
4034 else if (STRNCMP_UC_C8(ptr + 1, STR_Q STR_BACKSLASH STR_E, 3) == 0)
4035 ptr += 3;
4036 else
4037 break;
4038 }
4039 else if (!negate_class && c == CHAR_CIRCUMFLEX_ACCENT)
4040 negate_class = TRUE;
4041 else break;
4042 }
4043
4044 /* Empty classes are allowed in JavaScript compatibility mode. Otherwise,
4045 an initial ']' is taken as a data character -- the code below handles
4046 that. In JS mode, [] must always fail, so generate OP_FAIL, whereas
4047 [^] must match any character, so generate OP_ALLANY. */
4048
4049 if (c == CHAR_RIGHT_SQUARE_BRACKET &&
4050 (cd->external_options & PCRE_JAVASCRIPT_COMPAT) != 0)
4051 {
4052 *code++ = negate_class? OP_ALLANY : OP_FAIL;
4053 if (firstchar == REQ_UNSET) firstchar = REQ_NONE;
4054 zerofirstchar = firstchar;
4055 break;
4056 }
4057
4058 /* If a class contains a negative special such as \S, we need to flip the
4059 negation flag at the end, so that support for characters > 255 works
4060 correctly (they are all included in the class). */
4061
4062 should_flip_negation = FALSE;
4063
4064 /* For optimization purposes, we track some properties of the class:
4065 class_has_8bitchar will be non-zero if the class contains at least one <
4066 256 character; class_one_char will be 1 if the class contains just one
4067 character. */
4068
4069 class_has_8bitchar = 0;
4070 class_one_char = 0;
4071
4072 /* Initialize the 32-char bit map to all zeros. We build the map in a
4073 temporary bit of memory, in case the class contains fewer than two
4074 8-bit characters because in that case the compiled code doesn't use the bit
4075 map. */
4076
4077 memset(classbits, 0, 32 * sizeof(pcre_uint8));
4078
4079 #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
4080 xclass = FALSE;
4081 class_uchardata = code + LINK_SIZE + 2; /* For XCLASS items */
4082 class_uchardata_base = class_uchardata; /* Save the start */
4083 #endif
4084
4085 /* Process characters until ] is reached. By writing this as a "do" it
4086 means that an initial ] is taken as a data character. At the start of the
4087 loop, c contains the first byte of the character. */
4088
4089 if (c != 0) do
4090 {
4091 const pcre_uchar *oldptr;
4092
4093 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
4094 if (utf && HAS_EXTRALEN(c))
4095 { /* Braces are required because the */
4096 GETCHARLEN(c, ptr, ptr); /* macro generates multiple statements */
4097 }
4098 #endif
4099
4100 #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
4101 /* In the pre-compile phase, accumulate the length of any extra
4102 data and reset the pointer. This is so that very large classes that
4103 contain a zillion > 255 characters no longer overwrite the work space
4104 (which is on the stack). We have to remember that there was XCLASS data,
4105 however. */
4106
4107 if (lengthptr != NULL && class_uchardata > class_uchardata_base)
4108 {
4109 xclass = TRUE;
4110 *lengthptr += class_uchardata - class_uchardata_base;
4111 class_uchardata = class_uchardata_base;
4112 }
4113 #endif
4114
4115 /* Inside \Q...\E everything is literal except \E */
4116
4117 if (inescq)
4118 {
4119 if (c == CHAR_BACKSLASH && ptr[1] == CHAR_E) /* If we are at \E */
4120 {
4121 inescq = FALSE; /* Reset literal state */
4122 ptr++; /* Skip the 'E' */
4123 continue; /* Carry on with next */
4124 }
4125 goto CHECK_RANGE; /* Could be range if \E follows */
4126 }
4127
4128 /* Handle POSIX class names. Perl allows a negation extension of the
4129 form [:^name:]. A square bracket that doesn't match the syntax is
4130 treated as a literal. We also recognize the POSIX constructions
4131 [.ch.] and [=ch=] ("collating elements") and fault them, as Perl
4132 5.6 and 5.8 do. */
4133
4134 if (c == CHAR_LEFT_SQUARE_BRACKET &&
4135 (ptr[1] == CHAR_COLON || ptr[1] == CHAR_DOT ||
4136 ptr[1] == CHAR_EQUALS_SIGN) && check_posix_syntax(ptr, &tempptr))
4137 {
4138 BOOL local_negate = FALSE;
4139 int posix_class, taboffset, tabopt;
4140 register const pcre_uint8 *cbits = cd->cbits;
4141 pcre_uint8 pbits[32];
4142
4143 if (ptr[1] != CHAR_COLON)
4144 {
4145 *errorcodeptr = ERR31;
4146 goto FAILED;
4147 }
4148
4149 ptr += 2;
4150 if (*ptr == CHAR_CIRCUMFLEX_ACCENT)
4151 {
4152 local_negate = TRUE;
4153 should_flip_negation = TRUE; /* Note negative special */
4154 ptr++;
4155 }
4156
4157 posix_class = check_posix_name(ptr, (int)(tempptr - ptr));
4158 if (posix_class < 0)
4159 {
4160 *errorcodeptr = ERR30;
4161 goto FAILED;
4162 }
4163
4164 /* If matching is caseless, upper and lower are converted to
4165 alpha. This relies on the fact that the class table starts with
4166 alpha, lower, upper as the first 3 entries. */
4167
4168 if ((options & PCRE_CASELESS) != 0 && posix_class <= 2)
4169 posix_class = 0;
4170
4171 /* When PCRE_UCP is set, some of the POSIX classes are converted to
4172 different escape sequences that use Unicode properties. */
4173
4174 #ifdef SUPPORT_UCP
4175 if ((options & PCRE_UCP) != 0)
4176 {
4177 int pc = posix_class + ((local_negate)? POSIX_SUBSIZE/2 : 0);
4178 if (posix_substitutes[pc] != NULL)
4179 {
4180 nestptr = tempptr + 1;
4181 ptr = posix_substitutes[pc] - 1;
4182 continue;
4183 }
4184 }
4185 #endif
4186 /* In the non-UCP case, we build the bit map for the POSIX class in a
4187 chunk of local store because we may be adding and subtracting from it,
4188 and we don't want to subtract bits that may be in the main map already.
4189 At the end we or the result into the bit map that is being built. */
4190
4191 posix_class *= 3;
4192
4193 /* Copy in the first table (always present) */
4194
4195 memcpy(pbits, cbits + posix_class_maps[posix_class],
4196 32 * sizeof(pcre_uint8));
4197
4198 /* If there is a second table, add or remove it as required. */
4199
4200 taboffset = posix_class_maps[posix_class + 1];
4201 tabopt = posix_class_maps[posix_class + 2];
4202
4203 if (taboffset >= 0)
4204 {
4205 if (tabopt >= 0)
4206 for (c = 0; c < 32; c++) pbits[c] |= cbits[c + taboffset];
4207 else
4208 for (c = 0; c < 32; c++) pbits[c] &= ~cbits[c + taboffset];
4209 }
4210
4211 /* Now see if we need to remove any special characters. An option
4212 value of 1 removes vertical space and 2 removes underscore. */
4213
4214 if (tabopt < 0) tabopt = -tabopt;
4215 if (tabopt == 1) pbits[1] &= ~0x3c;
4216 else if (tabopt == 2) pbits[11] &= 0x7f;
4217
4218 /* Add the POSIX table or its complement into the main table that is
4219 being built and we are done. */
4220
4221 if (local_negate)
4222 for (c = 0; c < 32; c++) classbits[c] |= ~pbits[c];
4223 else
4224 for (c = 0; c < 32; c++) classbits[c] |= pbits[c];
4225
4226 ptr = tempptr + 1;
4227 /* Every class contains at least one < 256 character. */
4228 class_has_8bitchar = 1;
4229 /* Every class contains at least two characters. */
4230 class_one_char = 2;
4231 continue; /* End of POSIX syntax handling */
4232 }
4233
4234 /* Backslash may introduce a single character, or it may introduce one
4235 of the specials, which just set a flag. The sequence \b is a special
4236 case. Inside a class (and only there) it is treated as backspace. We
4237 assume that other escapes have more than one character in them, so
4238 speculatively set both class_has_8bitchar and class_one_char bigger
4239 than one. Unrecognized escapes fall through and are either treated
4240 as literal characters (by default), or are faulted if
4241 PCRE_EXTRA is set. */
4242
4243 if (c == CHAR_BACKSLASH)
4244 {
4245 escape = check_escape(&ptr, &ec, errorcodeptr, cd->bracount, options, TRUE);
4246
4247 if (*errorcodeptr != 0) goto FAILED;
4248
4249 if (escape == 0)
4250 c = ec;
4251 else if (escape == ESC_b) c = CHAR_BS; /* \b is backspace in a class */
4252 else if (escape == ESC_N) /* \N is not supported in a class */
4253 {
4254 *errorcodeptr = ERR71;
4255 goto FAILED;
4256 }
4257 else if (escape == ESC_Q) /* Handle start of quoted string */
4258 {
4259 if (ptr[1] == CHAR_BACKSLASH && ptr[2] == CHAR_E)
4260 {
4261 ptr += 2; /* avoid empty string */
4262 }
4263 else inescq = TRUE;
4264 continue;
4265 }
4266 else if (escape == ESC_E) continue; /* Ignore orphan \E */
4267
4268 else
4269 {
4270 register const pcre_uint8 *cbits = cd->cbits;
4271 /* Every class contains at least two < 256 characters. */
4272 class_has_8bitchar++;
4273 /* Every class contains at least two characters. */
4274 class_one_char += 2;
4275
4276 switch (escape)
4277 {
4278 #ifdef SUPPORT_UCP
4279 case ESC_du: /* These are the values given for \d etc */
4280 case ESC_DU: /* when PCRE_UCP is set. We replace the */
4281 case ESC_wu: /* escape sequence with an appropriate \p */
4282 case ESC_WU: /* or \P to test Unicode properties instead */
4283 case ESC_su: /* of the default ASCII testing. */
4284 case ESC_SU:
4285 nestptr = ptr;
4286 ptr = substitutes[escape - ESC_DU] - 1; /* Just before substitute */
4287 class_has_8bitchar--; /* Undo! */
4288 continue;
4289 #endif
4290 case ESC_d:
4291 for (c = 0; c < 32; c++) classbits[c] |= cbits[c+cbit_digit];
4292 continue;
4293
4294 case ESC_D:
4295 should_flip_negation = TRUE;
4296 for (c = 0; c < 32; c++) classbits[c] |= ~cbits[c+cbit_digit];
4297 continue;
4298
4299 case ESC_w:
4300 for (c = 0; c < 32; c++) classbits[c] |= cbits[c+cbit_word];
4301 continue;
4302
4303 case ESC_W:
4304 should_flip_negation = TRUE;
4305 for (c = 0; c < 32; c++) classbits[c] |= ~cbits[c+cbit_word];
4306 continue;
4307
4308 /* Perl 5.004 onwards omits VT from \s, but we must preserve it
4309 if it was previously set by something earlier in the character
4310 class. Luckily, the value of CHAR_VT is 0x0b in both ASCII and
4311 EBCDIC, so we lazily just adjust the appropriate bit. */
4312
4313 case ESC_s:
4314 classbits[0] |= cbits[cbit_space];
4315 classbits[1] |= cbits[cbit_space+1] & ~0x08;
4316 for (c = 2; c < 32; c++) classbits[c] |= cbits[c+cbit_space];
4317 continue;
4318
4319 case ESC_S:
4320 should_flip_negation = TRUE;
4321 for (c = 0; c < 32; c++) classbits[c] |= ~cbits[c+cbit_space];
4322 classbits[1] |= 0x08; /* Perl 5.004 onwards omits VT from \s */
4323 continue;
4324
4325 /* The rest apply in both UCP and non-UCP cases. */
4326
4327 case ESC_h:
4328 (void)add_list_to_class(classbits, &class_uchardata, options, cd,
4329 PRIV(hspace_list), NOTACHAR);
4330 continue;
4331
4332 case ESC_H:
4333 (void)add_not_list_to_class(classbits, &class_uchardata, options,
4334 cd, PRIV(hspace_list));
4335 continue;
4336
4337 case ESC_v:
4338 (void)add_list_to_class(classbits, &class_uchardata, options, cd,
4339 PRIV(vspace_list), NOTACHAR);
4340 continue;
4341
4342 case ESC_V:
4343 (void)add_not_list_to_class(classbits, &class_uchardata, options,
4344 cd, PRIV(vspace_list));
4345 continue;
4346
4347 #ifdef SUPPORT_UCP
4348 case ESC_p:
4349 case ESC_P:
4350 {
4351 BOOL negated;
4352 int pdata;
4353 int ptype = get_ucp(&ptr, &negated, &pdata, errorcodeptr);
4354 if (ptype < 0) goto FAILED;
4355 *class_uchardata++ = ((escape == ESC_p) != negated)?
4356 XCL_PROP : XCL_NOTPROP;
4357 *class_uchardata++ = ptype;
4358 *class_uchardata++ = pdata;
4359 class_has_8bitchar--; /* Undo! */
4360 continue;
4361 }
4362 #endif
4363 /* Unrecognized escapes are faulted if PCRE is running in its
4364 strict mode. By default, for compatibility with Perl, they are
4365 treated as literals. */
4366
4367 default:
4368 if ((options & PCRE_EXTRA) != 0)
4369 {
4370 *errorcodeptr = ERR7;
4371 goto FAILED;
4372 }
4373 class_has_8bitchar--; /* Undo the speculative increase. */
4374 class_one_char -= 2; /* Undo the speculative increase. */
4375 c = *ptr; /* Get the final character and fall through */
4376 break;
4377 }
4378 }
4379
4380 /* Fall through if the escape just defined a single character (c >= 0).
4381 This may be greater than 256. */
4382
4383 escape = 0;
4384
4385 } /* End of backslash handling */
4386
4387 /* A character may be followed by '-' to form a range. However, Perl does
4388 not permit ']' to be the end of the range. A '-' character at the end is
4389 treated as a literal. Perl ignores orphaned \E sequences entirely. The
4390 code for handling \Q and \E is messy. */
4391
4392 CHECK_RANGE:
4393 while (ptr[1] == CHAR_BACKSLASH && ptr[2] == CHAR_E)
4394 {
4395 inescq = FALSE;
4396 ptr += 2;
4397 }
4398 oldptr = ptr;
4399
4400 /* Remember if \r or \n were explicitly used */
4401
4402 if (c == CHAR_CR || c == CHAR_NL) cd->external_flags |= PCRE_HASCRORLF;
4403
4404 /* Check for range */
4405
4406 if (!inescq && ptr[1] == CHAR_MINUS)
4407 {
4408 int d;
4409 ptr += 2;
4410 while (*ptr == CHAR_BACKSLASH && ptr[1] == CHAR_E) ptr += 2;
4411
4412 /* If we hit \Q (not followed by \E) at this point, go into escaped
4413 mode. */
4414
4415 while (*ptr == CHAR_BACKSLASH && ptr[1] == CHAR_Q)
4416 {
4417 ptr += 2;
4418 if (*ptr == CHAR_BACKSLASH && ptr[1] == CHAR_E)
4419 { ptr += 2; continue; }
4420 inescq = TRUE;
4421 break;
4422 }
4423
4424 /* Minus (hyphen) at the end of a class is treated as a literal, so put
4425 back the pointer and jump to handle the character that preceded it. */
4426
4427 if (*ptr == 0 || (!inescq && *ptr == CHAR_RIGHT_SQUARE_BRACKET))
4428 {
4429 ptr = oldptr;
4430 goto CLASS_SINGLE_CHARACTER;
4431 }
4432
4433 /* Otherwise, we have a potential range; pick up the next character */
4434
4435 #ifdef SUPPORT_UTF
4436 if (utf)
4437 { /* Braces are required because the */
4438 GETCHARLEN(d, ptr, ptr); /* macro generates multiple statements */
4439 }
4440 else
4441 #endif
4442 d = *ptr; /* Not UTF-8 mode */
4443
4444 /* The second part of a range can be a single-character escape, but
4445 not any of the other escapes. Perl 5.6 treats a hyphen as a literal
4446 in such circumstances. */
4447
4448 if (!inescq && d == CHAR_BACKSLASH)
4449 {
4450 int descape;
4451 descape = check_escape(&ptr, &d, errorcodeptr, cd->bracount, options, TRUE);
4452 if (*errorcodeptr != 0) goto FAILED;
4453
4454 /* \b is backspace; any other special means the '-' was literal. */
4455
4456 if (descape > 0)
4457 {
4458 if (descape == ESC_b) d = CHAR_BS; else
4459 {
4460 ptr = oldptr;
4461 goto CLASS_SINGLE_CHARACTER; /* A few lines below */
4462 }
4463 }
4464 }
4465
4466 /* Check that the two values are in the correct order. Optimize
4467 one-character ranges. */
4468
4469 if (d < c)
4470 {
4471 *errorcodeptr = ERR8;
4472 goto FAILED;
4473 }
4474 if (d == c) goto CLASS_SINGLE_CHARACTER; /* A few lines below */
4475
4476 /* We have found a character range, so single character optimizations
4477 cannot be done anymore. Any value greater than 1 indicates that there
4478 is more than one character. */
4479
4480 class_one_char = 2;
4481
4482 /* Remember an explicit \r or \n, and add the range to the class. */
4483
4484 if (d == CHAR_CR || d == CHAR_NL) cd->external_flags |= PCRE_HASCRORLF;
4485
4486 class_has_8bitchar +=
4487 add_to_class(classbits, &class_uchardata, options, cd, c, d);
4488
4489 continue; /* Go get the next char in the class */
4490 }
4491
4492 /* Handle a single character - we can get here for a normal non-escape
4493 char, or after \ that introduces a single character or for an apparent
4494 range that isn't. Only the value 1 matters for class_one_char, so don't
4495 increase it if it is already 2 or more ... just in case there's a class
4496 with a zillion characters in it. */
4497
4498 CLASS_SINGLE_CHARACTER:
4499 if (class_one_char < 2) class_one_char++;
4500
4501 /* If class_one_char is 1, we have the first single character in the
4502 class, and there have been no prior ranges, or XCLASS items generated by
4503 escapes. If this is the final character in the class, we can optimize by
4504 turning the item into a 1-character OP_CHAR[I] if it's positive, or
4505 OP_NOT[I] if it's negative. In the positive case, it can cause firstchar
4506 to be set. Otherwise, there can be no first char if this item is first,
4507 whatever repeat count may follow. In the case of reqchar, save the
4508 previous value for reinstating. */
4509
4510 if (class_one_char == 1 && ptr[1] == CHAR_RIGHT_SQUARE_BRACKET)
4511 {
4512 ptr++;
4513 zeroreqchar = reqchar;
4514
4515 if (negate_class)
4516 {
4517 #ifdef SUPPORT_UCP
4518 // FIXMEchpe pcreuint32?
4519 int d;
4520 #endif
4521 if (firstchar == REQ_UNSET) firstchar = REQ_NONE;
4522 zerofirstchar = firstchar;
4523
4524 /* For caseless UTF-8 mode when UCP support is available, check
4525 whether this character has more than one other case. If so, generate
4526 a special OP_NOTPROP item instead of OP_NOTI. */
4527
4528 #ifdef SUPPORT_UCP
4529 if (utf && (options & PCRE_CASELESS) != 0 &&
4530 (d = UCD_CASESET(c)) != 0)
4531 {
4532 *code++ = OP_NOTPROP;
4533 *code++ = PT_CLIST;
4534 *code++ = d;
4535 }
4536 else
4537 #endif
4538 /* Char has only one other case, or UCP not available */
4539
4540 {
4541 *code++ = ((options & PCRE_CASELESS) != 0)? OP_NOTI: OP_NOT;
4542 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
4543 if (utf && c > MAX_VALUE_FOR_SINGLE_CHAR)
4544 code += PRIV(ord2utf)(c, code);
4545 else
4546 #endif
4547 *code++ = c;
4548 }
4549
4550 /* We are finished with this character class */
4551
4552 goto END_CLASS;
4553 }
4554
4555 /* For a single, positive character, get the value into mcbuffer, and
4556 then we can handle this with the normal one-character code. */
4557
4558 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
4559 if (utf && c > MAX_VALUE_FOR_SINGLE_CHAR)
4560 mclength = PRIV(ord2utf)(c, mcbuffer);
4561 else
4562 #endif
4563 {
4564 mcbuffer[0] = c;
4565 mclength = 1;
4566 }
4567 goto ONE_CHAR;
4568 } /* End of 1-char optimization */
4569
4570 /* There is more than one character in the class, or an XCLASS item
4571 has been generated. Add this character to the class. */
4572
4573 class_has_8bitchar +=
4574 add_to_class(classbits, &class_uchardata, options, cd, c, c);
4575 }
4576
4577 /* Loop until ']' reached. This "while" is the end of the "do" far above.
4578 If we are at the end of an internal nested string, revert to the outer
4579 string. */
4580
4581 while (((c = *(++ptr)) != 0 ||
4582 (nestptr != NULL &&
4583 (ptr = nestptr, nestptr = NULL, c = *(++ptr)) != 0)) &&
4584 (c != CHAR_RIGHT_SQUARE_BRACKET || inescq));
4585
4586 /* Check for missing terminating ']' */
4587
4588 if (c == 0)
4589 {
4590 *errorcodeptr = ERR6;
4591 goto FAILED;
4592 }
4593
4594 /* We will need an XCLASS if data has been placed in class_uchardata. In
4595 the second phase this is a sufficient test. However, in the pre-compile
4596 phase, class_uchardata gets emptied to prevent workspace overflow, so it
4597 only if the very last character in the class needs XCLASS will it contain
4598 anything at this point. For this reason, xclass gets set TRUE above when
4599 uchar_classdata is emptied, and that's why this code is the way it is here
4600 instead of just doing a test on class_uchardata below. */
4601
4602 #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
4603 if (class_uchardata > class_uchardata_base) xclass = TRUE;
4604 #endif
4605
4606 /* If this is the first thing in the branch, there can be no first char
4607 setting, whatever the repeat count. Any reqchar setting must remain
4608 unchanged after any kind of repeat. */
4609
4610 if (firstchar == REQ_UNSET) firstchar = REQ_NONE;
4611 zerofirstchar = firstchar;
4612 zeroreqchar = reqchar;
4613
4614 /* If there are characters with values > 255, we have to compile an
4615 extended class, with its own opcode, unless there was a negated special
4616 such as \S in the class, and PCRE_UCP is not set, because in that case all
4617 characters > 255 are in the class, so any that were explicitly given as
4618 well can be ignored. If (when there are explicit characters > 255 that must
4619 be listed) there are no characters < 256, we can omit the bitmap in the
4620 actual compiled code. */
4621
4622 #ifdef SUPPORT_UTF
4623 if (xclass && (!should_flip_negation || (options & PCRE_UCP) != 0))
4624 #elif !defined COMPILE_PCRE8
4625 if (xclass && !should_flip_negation)
4626 #endif
4627 #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
4628 {
4629 *class_uchardata++ = XCL_END; /* Marks the end of extra data */
4630 *code++ = OP_XCLASS;
4631 code += LINK_SIZE;
4632 *code = negate_class? XCL_NOT:0;
4633
4634 /* If the map is required, move up the extra data to make room for it;
4635 otherwise just move the code pointer to the end of the extra data. */
4636
4637 if (class_has_8bitchar > 0)
4638 {
4639 *code++ |= XCL_MAP;
4640 memmove(code + (32 / sizeof(pcre_uchar)), code,
4641 IN_UCHARS(class_uchardata - code));
4642 memcpy(code, classbits, 32);
4643 code = class_uchardata + (32 / sizeof(pcre_uchar));
4644 }
4645 else code = class_uchardata;
4646
4647 /* Now fill in the complete length of the item */
4648
4649 PUT(previous, 1, (int)(code - previous));
4650 break; /* End of class handling */
4651 }
4652 #endif
4653
4654 /* If there are no characters > 255, or they are all to be included or
4655 excluded, set the opcode to OP_CLASS or OP_NCLASS, depending on whether the
4656 whole class was negated and whether there were negative specials such as \S
4657 (non-UCP) in the class. Then copy the 32-byte map into the code vector,
4658 negating it if necessary. */
4659
4660 *code++ = (negate_class == should_flip_negation) ? OP_CLASS : OP_NCLASS;
4661 if (lengthptr == NULL) /* Save time in the pre-compile phase */
4662 {
4663 if (negate_class)
4664 for (c = 0; c < 32; c++) classbits[c] = ~classbits[c];
4665 memcpy(code, classbits, 32);
4666 }
4667 code += 32 / sizeof(pcre_uchar);
4668
4669 END_CLASS:
4670 break;
4671
4672
4673 /* ===================================================================*/
4674 /* Various kinds of repeat; '{' is not necessarily a quantifier, but this
4675 has been tested above. */
4676
4677 case CHAR_LEFT_CURLY_BRACKET:
4678 if (!is_quantifier) goto NORMAL_CHAR;
4679 ptr = read_repeat_counts(ptr+1, &repeat_min, &repeat_max, errorcodeptr);
4680 if (*errorcodeptr != 0) goto FAILED;
4681 goto REPEAT;
4682
4683 case CHAR_ASTERISK:
4684 repeat_min = 0;
4685 repeat_max = -1;
4686 goto REPEAT;
4687
4688 case CHAR_PLUS:
4689 repeat_min = 1;
4690 repeat_max = -1;
4691 goto REPEAT;
4692
4693 case CHAR_QUESTION_MARK:
4694 repeat_min = 0;
4695 repeat_max = 1;
4696
4697 REPEAT:
4698 if (previous == NULL)
4699 {
4700 *errorcodeptr = ERR9;
4701 goto FAILED;
4702 }
4703
4704 if (repeat_min == 0)
4705 {
4706 firstchar = zerofirstchar; /* Adjust for zero repeat */
4707 reqchar = zeroreqchar; /* Ditto */
4708 }
4709
4710 /* Remember whether this is a variable length repeat */
4711
4712 reqvary = (repeat_min == repeat_max)? 0 : REQ_VARY;
4713
4714 op_type = 0; /* Default single-char op codes */
4715 possessive_quantifier = FALSE; /* Default not possessive quantifier */
4716
4717 /* Save start of previous item, in case we have to move it up in order to
4718 insert something before it. */
4719
4720 tempcode = previous;
4721
4722 /* If the next character is '+', we have a possessive quantifier. This
4723 implies greediness, whatever the setting of the PCRE_UNGREEDY option.
4724 If the next character is '?' this is a minimizing repeat, by default,
4725 but if PCRE_UNGREEDY is set, it works the other way round. We change the
4726 repeat type to the non-default. */
4727
4728 if (ptr[1] == CHAR_PLUS)
4729 {
4730 repeat_type = 0; /* Force greedy */
4731 possessive_quantifier = TRUE;
4732 ptr++;
4733 }
4734 else if (ptr[1] == CHAR_QUESTION_MARK)
4735 {
4736 repeat_type = greedy_non_default;
4737 ptr++;
4738 }
4739 else repeat_type = greedy_default;
4740
4741 /* If previous was a recursion call, wrap it in atomic brackets so that
4742 previous becomes the atomic group. All recursions were so wrapped in the
4743 past, but it no longer happens for non-repeated recursions. In fact, the
4744 repeated ones could be re-implemented independently so as not to need this,
4745 but for the moment we rely on the code for repeating groups. */
4746
4747 if (*previous == OP_RECURSE)
4748 {
4749 memmove(previous + 1 + LINK_SIZE, previous, IN_UCHARS(1 + LINK_SIZE));
4750 *previous = OP_ONCE;
4751 PUT(previous, 1, 2 + 2*LINK_SIZE);
4752 previous[2 + 2*LINK_SIZE] = OP_KET;
4753 PUT(previous, 3 + 2*LINK_SIZE, 2 + 2*LINK_SIZE);
4754 code += 2 + 2 * LINK_SIZE;
4755 length_prevgroup = 3 + 3*LINK_SIZE;
4756
4757 /* When actually compiling, we need to check whether this was a forward
4758 reference, and if so, adjust the offset. */
4759
4760 if (lengthptr == NULL && cd->hwm >= cd->start_workspace + LINK_SIZE)
4761 {
4762 int offset = GET(cd->hwm, -LINK_SIZE);
4763 if (offset == previous + 1 - cd->start_code)
4764 PUT(cd->hwm, -LINK_SIZE, offset + 1 + LINK_SIZE);
4765 }
4766 }
4767
4768 /* Now handle repetition for the different types of item. */
4769
4770 /* If previous was a character or negated character match, abolish the item
4771 and generate a repeat item instead. If a char item has a minimum of more
4772 than one, ensure that it is set in reqchar - it might not be if a sequence
4773 such as x{3} is the first thing in a branch because the x will have gone
4774 into firstchar instead. */
4775
4776 if (*previous == OP_CHAR || *previous == OP_CHARI
4777 || *previous == OP_NOT || *previous == OP_NOTI)
4778 {
4779 switch (*previous)
4780 {
4781 default: /* Make compiler happy. */
4782 case OP_CHAR: op_type = OP_STAR - OP_STAR; break;
4783 case OP_CHARI: op_type = OP_STARI - OP_STAR; break;
4784 case OP_NOT: op_type = OP_NOTSTAR - OP_STAR; break;
4785 case OP_NOTI: op_type = OP_NOTSTARI - OP_STAR; break;
4786 }
4787
4788 /* Deal with UTF characters that take up more than one character. It's
4789 easier to write this out separately than try to macrify it. Use c to
4790 hold the length of the character in bytes, plus UTF_LENGTH to flag that
4791 it's a length rather than a small character. */
4792
4793 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
4794 if (utf && NOT_FIRSTCHAR(code[-1]))
4795 {
4796 pcre_uchar *lastchar = code - 1;
4797 BACKCHAR(lastchar);
4798 c = (int)(code - lastchar); /* Length of UTF-8 character */
4799 memcpy(utf_chars, lastchar, IN_UCHARS(c)); /* Save the char */
4800 c |= UTF_LENGTH; /* Flag c as a length */
4801 }
4802 else
4803 #endif /* SUPPORT_UTF */
4804
4805 /* Handle the case of a single charater - either with no UTF support, or
4806 with UTF disabled, or for a single character UTF character. */
4807 {
4808 c = code[-1];
4809 if (*previous <= OP_CHARI && repeat_min > 1)
4810 reqchar = c | req_caseopt | cd->req_varyopt;
4811 }
4812
4813 /* If the repetition is unlimited, it pays to see if the next thing on
4814 the line is something that cannot possibly match this character. If so,
4815 automatically possessifying this item gains some performance in the case
4816 where the match fails. */
4817
4818 if (!possessive_quantifier &&
4819 repeat_max < 0 &&
4820 check_auto_possessive(previous, utf, ptr + 1, options, cd))
4821 {
4822 repeat_type = 0; /* Force greedy */
4823 possessive_quantifier = TRUE;
4824 }
4825
4826 goto OUTPUT_SINGLE_REPEAT; /* Code shared with single character types */
4827 }
4828
4829 /* If previous was a character type match (\d or similar), abolish it and
4830 create a suitable repeat item. The code is shared with single-character
4831 repeats by setting op_type to add a suitable offset into repeat_type. Note
4832 the the Unicode property types will be present only when SUPPORT_UCP is
4833 defined, but we don't wrap the little bits of code here because it just
4834 makes it horribly messy. */
4835
4836 else if (*previous < OP_EODN)
4837 {
4838 pcre_uchar *oldcode;
4839 int prop_type, prop_value;
4840 op_type = OP_TYPESTAR - OP_STAR; /* Use type opcodes */
4841 c = *previous;
4842
4843 if (!possessive_quantifier &&
4844 repeat_max < 0 &&
4845 check_auto_possessive(previous, utf, ptr + 1, options, cd))
4846 {
4847 repeat_type = 0; /* Force greedy */
4848 possessive_quantifier = TRUE;
4849 }
4850
4851 OUTPUT_SINGLE_REPEAT:
4852 if (*previous == OP_PROP || *previous == OP_NOTPROP)
4853 {
4854 prop_type = previous[1];
4855 prop_value = previous[2];
4856 }
4857 else prop_type = prop_value = -1;
4858
4859 oldcode = code;
4860 code = previous; /* Usually overwrite previous item */
4861
4862 /* If the maximum is zero then the minimum must also be zero; Perl allows
4863 this case, so we do too - by simply omitting the item altogether. */
4864
4865 if (repeat_max == 0) goto END_REPEAT;
4866
4867 /*--------------------------------------------------------------------*/
4868 /* This code is obsolete from release 8.00; the restriction was finally
4869 removed: */
4870
4871 /* All real repeats make it impossible to handle partial matching (maybe
4872 one day we will be able to remove this restriction). */
4873
4874 /* if (repeat_max != 1) cd->external_flags |= PCRE_NOPARTIAL; */
4875 /*--------------------------------------------------------------------*/
4876
4877 /* Combine the op_type with the repeat_type */
4878
4879 repeat_type += op_type;
4880
4881 /* A minimum of zero is handled either as the special case * or ?, or as
4882 an UPTO, with the maximum given. */
4883
4884 if (repeat_min == 0)
4885 {
4886 if (repeat_max == -1) *code++ = OP_STAR + repeat_type;
4887 else if (repeat_max == 1) *code++ = OP_QUERY + repeat_type;
4888 else
4889 {
4890 *code++ = OP_UPTO + repeat_type;
4891 PUT2INC(code, 0, repeat_max);
4892 }
4893 }
4894
4895 /* A repeat minimum of 1 is optimized into some special cases. If the
4896 maximum is unlimited, we use OP_PLUS. Otherwise, the original item is
4897 left in place and, if the maximum is greater than 1, we use OP_UPTO with
4898 one less than the maximum. */
4899
4900 else if (repeat_min == 1)
4901 {
4902 if (repeat_max == -1)
4903 *code++ = OP_PLUS + repeat_type;
4904 else
4905 {
4906 code = oldcode; /* leave previous item in place */
4907 if (repeat_max == 1) goto END_REPEAT;
4908 *code++ = OP_UPTO + repeat_type;
4909 PUT2INC(code, 0, repeat_max - 1);
4910 }
4911 }
4912
4913 /* The case {n,n} is just an EXACT, while the general case {n,m} is
4914 handled as an EXACT followed by an UPTO. */
4915
4916 else
4917 {
4918 *code++ = OP_EXACT + op_type; /* NB EXACT doesn't have repeat_type */
4919 PUT2INC(code, 0, repeat_min);
4920
4921 /* If the maximum is unlimited, insert an OP_STAR. Before doing so,
4922 we have to insert the character for the previous code. For a repeated
4923 Unicode property match, there are two extra bytes that define the
4924 required property. In UTF-8 mode, long characters have their length in
4925 c, with the UTF_LENGTH bit as a flag. */
4926
4927 if (repeat_max < 0)
4928 {
4929 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
4930 if (utf && (c & UTF_LENGTH) != 0)
4931 {
4932 memcpy(code, utf_chars, IN_UCHARS(c & 7));
4933 code += c & 7;
4934 }
4935 else
4936 #endif
4937 {
4938 *code++ = c;
4939 if (prop_type >= 0)
4940 {
4941 *code++ = prop_type;
4942 *code++ = prop_value;
4943 }
4944 }
4945 *code++ = OP_STAR + repeat_type;
4946 }
4947
4948 /* Else insert an UPTO if the max is greater than the min, again
4949 preceded by the character, for the previously inserted code. If the
4950 UPTO is just for 1 instance, we can use QUERY instead. */
4951
4952 else if (repeat_max != repeat_min)
4953 {
4954 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
4955 if (utf && (c & UTF_LENGTH) != 0)
4956 {
4957 memcpy(code, utf_chars, IN_UCHARS(c & 7));
4958 code += c & 7;
4959 }
4960 else
4961 #endif
4962 *code++ = c;
4963 if (prop_type >= 0)
4964 {
4965 *code++ = prop_type;
4966 *code++ = prop_value;
4967 }
4968 repeat_max -= repeat_min;
4969
4970 if (repeat_max == 1)
4971 {
4972 *code++ = OP_QUERY + repeat_type;
4973 }
4974 else
4975 {
4976 *code++ = OP_UPTO + repeat_type;
4977 PUT2INC(code, 0, repeat_max);
4978 }
4979 }
4980 }
4981
4982 /* The character or character type itself comes last in all cases. */
4983
4984 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
4985 if (utf && (c & UTF_LENGTH) != 0)
4986 {
4987 memcpy(code, utf_chars, IN_UCHARS(c & 7));
4988 code += c & 7;
4989 }
4990 else
4991 #endif
4992 *code++ = c;
4993
4994 /* For a repeated Unicode property match, there are two extra bytes that
4995 define the required property. */
4996
4997 #ifdef SUPPORT_UCP
4998 if (prop_type >= 0)
4999 {
5000 *code++ = prop_type;
5001 *code++ = prop_value;
5002 }
5003 #endif
5004 }
5005
5006 /* If previous was a character class or a back reference, we put the repeat
5007 stuff after it, but just skip the item if the repeat was {0,0}. */
5008
5009 else if (*previous == OP_CLASS ||
5010 *previous == OP_NCLASS ||
5011 #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
5012 *previous == OP_XCLASS ||
5013 #endif
5014 *previous == OP_REF ||
5015 *previous == OP_REFI)
5016 {
5017 if (repeat_max == 0)
5018 {
5019 code = previous;
5020 goto END_REPEAT;
5021 }
5022
5023 /*--------------------------------------------------------------------*/
5024 /* This code is obsolete from release 8.00; the restriction was finally
5025 removed: */
5026
5027 /* All real repeats make it impossible to handle partial matching (maybe
5028 one day we will be able to remove this restriction). */
5029
5030 /* if (repeat_max != 1) cd->external_flags |= PCRE_NOPARTIAL; */
5031 /*--------------------------------------------------------------------*/
5032
5033 if (repeat_min == 0 && repeat_max == -1)
5034 *code++ = OP_CRSTAR + repeat_type;
5035 else if (repeat_min == 1 && repeat_max == -1)
5036 *code++ = OP_CRPLUS + repeat_type;
5037 else if (repeat_min == 0 && repeat_max == 1)
5038 *code++ = OP_CRQUERY + repeat_type;
5039 else
5040 {
5041 *code++ = OP_CRRANGE + repeat_type;
5042 PUT2INC(code, 0, repeat_min);
5043 if (repeat_max == -1) repeat_max = 0; /* 2-byte encoding for max */
5044 PUT2INC(code, 0, repeat_max);
5045 }
5046 }
5047
5048 /* If previous was a bracket group, we may have to replicate it in certain
5049 cases. Note that at this point we can encounter only the "basic" bracket
5050 opcodes such as BRA and CBRA, as this is the place where they get converted
5051 into the more special varieties such as BRAPOS and SBRA. A test for >=
5052 OP_ASSERT and <= OP_COND includes ASSERT, ASSERT_NOT, ASSERTBACK,
5053 ASSERTBACK_NOT, ONCE, BRA, CBRA, and COND. Originally, PCRE did not allow
5054 repetition of assertions, but now it does, for Perl compatibility. */
5055
5056 else if (*previous >= OP_ASSERT && *previous <= OP_COND)
5057 {
5058 register int i;
5059 int len = (int)(code - previous);
5060 pcre_uchar *bralink = NULL;
5061 pcre_uchar *brazeroptr = NULL;
5062
5063 /* Repeating a DEFINE group is pointless, but Perl allows the syntax, so
5064 we just ignore the repeat. */
5065
5066 if (*previous == OP_COND && previous[LINK_SIZE+1] == OP_DEF)
5067 goto END_REPEAT;
5068
5069 /* There is no sense in actually repeating assertions. The only potential
5070 use of repetition is in cases when the assertion is optional. Therefore,
5071 if the minimum is greater than zero, just ignore the repeat. If the
5072 maximum is not not zero or one, set it to 1. */
5073
5074 if (*previous < OP_ONCE) /* Assertion */
5075 {
5076 if (repeat_min > 0) goto END_REPEAT;
5077 if (repeat_max < 0 || repeat_max > 1) repeat_max = 1;
5078 }
5079
5080 /* The case of a zero minimum is special because of the need to stick
5081 OP_BRAZERO in front of it, and because the group appears once in the
5082 data, whereas in other cases it appears the minimum number of times. For
5083 this reason, it is simplest to treat this case separately, as otherwise
5084 the code gets far too messy. There are several special subcases when the
5085 minimum is zero. */
5086
5087 if (repeat_min == 0)
5088 {
5089 /* If the maximum is also zero, we used to just omit the group from the
5090 output altogether, like this:
5091
5092 ** if (repeat_max == 0)
5093 ** {
5094 ** code = previous;
5095 ** goto END_REPEAT;
5096 ** }
5097
5098 However, that fails when a group or a subgroup within it is referenced
5099 as a subroutine from elsewhere in the pattern, so now we stick in
5100 OP_SKIPZERO in front of it so that it is skipped on execution. As we
5101 don't have a list of which groups are referenced, we cannot do this
5102 selectively.
5103
5104 If the maximum is 1 or unlimited, we just have to stick in the BRAZERO
5105 and do no more at this point. However, we do need to adjust any
5106 OP_RECURSE calls inside the group that refer to the group itself or any
5107 internal or forward referenced group, because the offset is from the
5108 start of the whole regex. Temporarily terminate the pattern while doing
5109 this. */
5110
5111 if (repeat_max <= 1) /* Covers 0, 1, and unlimited */
5112 {
5113 *code = OP_END;
5114 adjust_recurse(previous, 1, utf, cd, save_hwm);
5115 memmove(previous + 1, previous, IN_UCHARS(len));
5116 code++;
5117 if (repeat_max == 0)
5118 {
5119 *previous++ = OP_SKIPZERO;
5120 goto END_REPEAT;
5121 }
5122 brazeroptr = previous; /* Save for possessive optimizing */
5123 *previous++ = OP_BRAZERO + repeat_type;
5124 }
5125
5126 /* If the maximum is greater than 1 and limited, we have to replicate
5127 in a nested fashion, sticking OP_BRAZERO before each set of brackets.
5128 The first one has to be handled carefully because it's the original
5129 copy, which has to be moved up. The remainder can be handled by code
5130 that is common with the non-zero minimum case below. We have to
5131 adjust the value or repeat_max, since one less copy is required. Once
5132 again, we may have to adjust any OP_RECURSE calls inside the group. */
5133
5134 else
5135 {
5136 int offset;
5137 *code = OP_END;
5138 adjust_recurse(previous, 2 + LINK_SIZE, utf, cd, save_hwm);
5139 memmove(previous + 2 + LINK_SIZE, previous, IN_UCHARS(len));
5140 code += 2 + LINK_SIZE;
5141 *previous++ = OP_BRAZERO + repeat_type;
5142 *previous++ = OP_BRA;
5143
5144 /* We chain together the bracket offset fields that have to be
5145 filled in later when the ends of the brackets are reached. */
5146
5147 offset = (bralink == NULL)? 0 : (int)(previous - bralink);
5148 bralink = previous;
5149 PUTINC(previous, 0, offset);
5150 }
5151
5152 repeat_max--;
5153 }
5154
5155 /* If the minimum is greater than zero, replicate the group as many
5156 times as necessary, and adjust the maximum to the number of subsequent
5157 copies that we need. If we set a first char from the group, and didn't
5158 set a required char, copy the latter from the former. If there are any
5159 forward reference subroutine calls in the group, there will be entries on
5160 the workspace list; replicate these with an appropriate increment. */
5161
5162 else
5163 {
5164 if (repeat_min > 1)
5165 {
5166 /* In the pre-compile phase, we don't actually do the replication. We
5167 just adjust the length as if we had. Do some paranoid checks for
5168 potential integer overflow. The INT64_OR_DOUBLE type is a 64-bit
5169 integer type when available, otherwise double. */
5170
5171 if (lengthptr != NULL)
5172 {
5173 int delta = (repeat_min - 1)*length_prevgroup;
5174 if ((INT64_OR_DOUBLE)(repeat_min - 1)*
5175 (INT64_OR_DOUBLE)length_prevgroup >
5176 (INT64_OR_DOUBLE)INT_MAX ||
5177 OFLOW_MAX - *lengthptr < delta)
5178 {
5179 *errorcodeptr = ERR20;
5180 goto FAILED;
5181 }
5182 *lengthptr += delta;
5183 }
5184
5185 /* This is compiling for real. If there is a set first byte for
5186 the group, and we have not yet set a "required byte", set it. Make
5187 sure there is enough workspace for copying forward references before
5188 doing the copy. */
5189
5190 else
5191 {
5192 if (groupsetfirstchar && reqchar < 0) reqchar = firstchar;
5193
5194 for (i = 1; i < repeat_min; i++)
5195 {
5196 pcre_uchar *hc;
5197 pcre_uchar *this_hwm = cd->hwm;
5198 memcpy(code, previous, IN_UCHARS(len));
5199
5200 while (cd->hwm > cd->start_workspace + cd->workspace_size -
5201 WORK_SIZE_SAFETY_MARGIN - (this_hwm - save_hwm))
5202 {
5203 int save_offset = save_hwm - cd->start_workspace;
5204 int this_offset = this_hwm - cd->start_workspace;
5205 *errorcodeptr = expand_workspace(cd);
5206 if (*errorcodeptr != 0) goto FAILED;
5207 save_hwm = (pcre_uchar *)cd->start_workspace + save_offset;
5208 this_hwm = (pcre_uchar *)cd->start_workspace + this_offset;
5209 }
5210
5211 for (hc = save_hwm; hc < this_hwm; hc += LINK_SIZE)
5212 {
5213 PUT(cd->hwm, 0, GET(hc, 0) + len);
5214 cd->hwm += LINK_SIZE;
5215 }
5216 save_hwm = this_hwm;
5217 code += len;
5218 }
5219 }
5220 }
5221
5222 if (repeat_max > 0) repeat_max -= repeat_min;
5223 }
5224
5225 /* This code is common to both the zero and non-zero minimum cases. If
5226 the maximum is limited, it replicates the group in a nested fashion,
5227 remembering the bracket starts on a stack. In the case of a zero minimum,
5228 the first one was set up above. In all cases the repeat_max now specifies
5229 the number of additional copies needed. Again, we must remember to
5230 replicate entries on the forward reference list. */
5231
5232 if (repeat_max >= 0)
5233 {
5234 /* In the pre-compile phase, we don't actually do the replication. We
5235 just adjust the length as if we had. For each repetition we must add 1
5236 to the length for BRAZERO and for all but the last repetition we must
5237 add 2 + 2*LINKSIZE to allow for the nesting that occurs. Do some
5238 paranoid checks to avoid integer overflow. The INT64_OR_DOUBLE type is
5239 a 64-bit integer type when available, otherwise double. */
5240
5241 if (lengthptr != NULL && repeat_max > 0)
5242 {
5243 int delta = repeat_max * (length_prevgroup + 1 + 2 + 2*LINK_SIZE) -
5244 2 - 2*LINK_SIZE; /* Last one doesn't nest */
5245 if ((INT64_OR_DOUBLE)repeat_max *
5246 (INT64_OR_DOUBLE)(length_prevgroup + 1 + 2 + 2*LINK_SIZE)
5247 > (INT64_OR_DOUBLE)INT_MAX ||
5248 OFLOW_MAX - *lengthptr < delta)
5249 {
5250 *errorcodeptr = ERR20;
5251 goto FAILED;
5252 }
5253 *lengthptr += delta;
5254 }
5255
5256 /* This is compiling for real */
5257
5258 else for (i = repeat_max - 1; i >= 0; i--)
5259 {
5260 pcre_uchar *hc;
5261 pcre_uchar *this_hwm = cd->hwm;
5262
5263 *code++ = OP_BRAZERO + repeat_type;
5264
5265 /* All but the final copy start a new nesting, maintaining the
5266 chain of brackets outstanding. */
5267
5268 if (i != 0)
5269 {
5270 int offset;
5271 *code++ = OP_BRA;
5272 offset = (bralink == NULL)? 0 : (int)(code - bralink);
5273 bralink = code;
5274 PUTINC(code, 0, offset);
5275 }
5276
5277 memcpy(code, previous, IN_UCHARS(len));
5278
5279 /* Ensure there is enough workspace for forward references before
5280 copying them. */
5281
5282 while (cd->hwm > cd->start_workspace + cd->workspace_size -
5283 WORK_SIZE_SAFETY_MARGIN - (this_hwm - save_hwm))
5284 {
5285 int save_offset = save_hwm - cd->start_workspace;
5286 int this_offset = this_hwm - cd->start_workspace;
5287 *errorcodeptr = expand_workspace(cd);
5288 if (*errorcodeptr != 0) goto FAILED;
5289 save_hwm = (pcre_uchar *)cd->start_workspace + save_offset;
5290 this_hwm = (pcre_uchar *)cd->start_workspace + this_offset;
5291 }
5292
5293 for (hc = save_hwm; hc < this_hwm; hc += LINK_SIZE)
5294 {
5295 PUT(cd->hwm, 0, GET(hc, 0) + len + ((i != 0)? 2+LINK_SIZE : 1));
5296 cd->hwm += LINK_SIZE;
5297 }
5298 save_hwm = this_hwm;
5299 code += len;
5300 }
5301
5302 /* Now chain through the pending brackets, and fill in their length
5303 fields (which are holding the chain links pro tem). */
5304
5305 while (bralink != NULL)
5306 {
5307 int oldlinkoffset;
5308 int offset = (int)(code - bralink + 1);
5309 pcre_uchar *bra = code - offset;
5310 oldlinkoffset = GET(bra, 1);
5311 bralink = (oldlinkoffset == 0)? NULL : bralink - oldlinkoffset;
5312 *code++ = OP_KET;
5313 PUTINC(code, 0, offset);
5314 PUT(bra, 1, offset);
5315 }
5316 }
5317
5318 /* If the maximum is unlimited, set a repeater in the final copy. For
5319 ONCE brackets, that's all we need to do. However, possessively repeated
5320 ONCE brackets can be converted into non-capturing brackets, as the
5321 behaviour of (?:xx)++ is the same as (?>xx)++ and this saves having to
5322 deal with possessive ONCEs specially.
5323
5324 Otherwise, when we are doing the actual compile phase, check to see
5325 whether this group is one that could match an empty string. If so,
5326 convert the initial operator to the S form (e.g. OP_BRA -> OP_SBRA) so
5327 that runtime checking can be done. [This check is also applied to ONCE
5328 groups at runtime, but in a different way.]
5329
5330 Then, if the quantifier was possessive and the bracket is not a
5331 conditional, we convert the BRA code to the POS form, and the KET code to
5332 KETRPOS. (It turns out to be convenient at runtime to detect this kind of
5333 subpattern at both the start and at the end.) The use of special opcodes
5334 makes it possible to reduce greatly the stack usage in pcre_exec(). If
5335 the group is preceded by OP_BRAZERO, convert this to OP_BRAPOSZERO.
5336
5337 Then, if the minimum number of matches is 1 or 0, cancel the possessive
5338 flag so that the default action below, of wrapping everything inside
5339 atomic brackets, does not happen. When the minimum is greater than 1,
5340 there will be earlier copies of the group, and so we still have to wrap
5341 the whole thing. */
5342
5343 else
5344 {
5345 pcre_uchar *ketcode = code - 1 - LINK_SIZE;
5346 pcre_uchar *bracode = ketcode - GET(ketcode, 1);
5347
5348 /* Convert possessive ONCE brackets to non-capturing */
5349
5350 if ((*bracode == OP_ONCE || *bracode == OP_ONCE_NC) &&
5351 possessive_quantifier) *bracode = OP_BRA;
5352
5353 /* For non-possessive ONCE brackets, all we need to do is to
5354 set the KET. */
5355
5356 if (*bracode == OP_ONCE || *bracode == OP_ONCE_NC)
5357 *ketcode = OP_KETRMAX + repeat_type;
5358
5359 /* Handle non-ONCE brackets and possessive ONCEs (which have been
5360 converted to non-capturing above). */
5361
5362 else
5363 {
5364 /* In the compile phase, check for empty string matching. */
5365
5366 if (lengthptr == NULL)
5367 {
5368 pcre_uchar *scode = bracode;
5369 do
5370 {
5371 if (could_be_empty_branch(scode, ketcode, utf, cd))
5372 {
5373 *bracode += OP_SBRA - OP_BRA;
5374 break;
5375 }
5376 scode += GET(scode, 1);
5377 }
5378 while (*scode == OP_ALT);
5379 }
5380
5381 /* Handle possessive quantifiers. */
5382
5383 if (possessive_quantifier)
5384 {
5385 /* For COND brackets, we wrap the whole thing in a possessively
5386 repeated non-capturing bracket, because we have not invented POS
5387 versions of the COND opcodes. Because we are moving code along, we
5388 must ensure that any pending recursive references are updated. */
5389
5390 if (*bracode == OP_COND || *bracode == OP_SCOND)
5391 {
5392 int nlen = (int)(code - bracode);
5393 *code = OP_END;
5394 adjust_recurse(bracode, 1 + LINK_SIZE, utf, cd, save_hwm);
5395 memmove(bracode + 1 + LINK_SIZE, bracode, IN_UCHARS(nlen));
5396 code += 1 + LINK_SIZE;
5397 nlen += 1 + LINK_SIZE;
5398 *bracode = OP_BRAPOS;
5399 *code++ = OP_KETRPOS;
5400 PUTINC(code, 0, nlen);
5401 PUT(bracode, 1, nlen);
5402 }
5403
5404 /* For non-COND brackets, we modify the BRA code and use KETRPOS. */
5405
5406 else
5407 {
5408 *bracode += 1; /* Switch to xxxPOS opcodes */
5409 *ketcode = OP_KETRPOS;
5410 }
5411
5412 /* If the minimum is zero, mark it as possessive, then unset the
5413 possessive flag when the minimum is 0 or 1. */
5414
5415 if (brazeroptr != NULL) *brazeroptr = OP_BRAPOSZERO;
5416 if (repeat_min < 2) possessive_quantifier = FALSE;
5417 }
5418
5419 /* Non-possessive quantifier */
5420
5421 else *ketcode = OP_KETRMAX + repeat_type;
5422 }
5423 }
5424 }
5425
5426 /* If previous is OP_FAIL, it was generated by an empty class [] in
5427 JavaScript mode. The other ways in which OP_FAIL can be generated, that is
5428 by (*FAIL) or (?!) set previous to NULL, which gives a "nothing to repeat"
5429 error above. We can just ignore the repeat in JS case. */
5430
5431 else if (*previous == OP_FAIL) goto END_REPEAT;
5432
5433 /* Else there's some kind of shambles */
5434
5435 else
5436 {
5437 *errorcodeptr = ERR11;
5438 goto FAILED;
5439 }
5440
5441 /* If the character following a repeat is '+', or if certain optimization
5442 tests above succeeded, possessive_quantifier is TRUE. For some opcodes,
5443 there are special alternative opcodes for this case. For anything else, we
5444 wrap the entire repeated item inside OP_ONCE brackets. Logically, the '+'
5445 notation is just syntactic sugar, taken from Sun's Java package, but the
5446 special opcodes can optimize it.
5447
5448 Some (but not all) possessively repeated subpatterns have already been
5449 completely handled in the code just above. For them, possessive_quantifier
5450 is always FALSE at this stage.
5451
5452 Note that the repeated item starts at tempcode, not at previous, which
5453 might be the first part of a string whose (former) last char we repeated.
5454
5455 Possessifying an 'exact' quantifier has no effect, so we can ignore it. But
5456 an 'upto' may follow. We skip over an 'exact' item, and then test the
5457 length of what remains before proceeding. */
5458
5459 if (possessive_quantifier)
5460 {
5461 int len;
5462
5463 if (*tempcode == OP_TYPEEXACT)
5464 tempcode += PRIV(OP_lengths)[*tempcode] +
5465 ((tempcode[1 + IMM2_SIZE] == OP_PROP
5466 || tempcode[1 + IMM2_SIZE] == OP_NOTPROP)? 2 : 0);
5467
5468 else if (*tempcode == OP_EXACT || *tempcode == OP_NOTEXACT)
5469 {
5470 tempcode += PRIV(OP_lengths)[*tempcode];
5471 #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
5472 if (utf && HAS_EXTRALEN(tempcode[-1]))
5473 tempcode += GET_EXTRALEN(tempcode[-1]);
5474 #endif
5475 }
5476
5477 len = (int)(code - tempcode);
5478 if (len > 0) switch (*tempcode)
5479 {
5480 case OP_STAR: *tempcode = OP_POSSTAR; break;
5481 case OP_PLUS: *tempcode = OP_POSPLUS; break;
5482 case OP_QUERY: *tempcode = OP_POSQUERY; break;
5483 case OP_UPTO: *tempcode = OP_POSUPTO; break;
5484
5485 case OP_STARI: *tempcode = OP_POSSTARI; break;
5486 case OP_PLUSI: *tempcode = OP_POSPLUSI; break;
5487 case OP_QUERYI: *tempcode = OP_POSQUERYI; break;
5488 case OP_UPTOI: *tempcode = OP_POSUPTOI; break;
5489
5490 case OP_NOTSTAR: *tempcode = OP_NOTPOSSTAR; break;
5491 case OP_NOTPLUS: *tempcode = OP_NOTPOSPLUS; break;
5492 case OP_NOTQUERY: *tempcode = OP_NOTPOSQUERY; break;
5493 case OP_NOTUPTO: *tempcode = OP_NOTPOSUPTO; break;
5494
5495 case OP_NOTSTARI: *tempcode = OP_NOTPOSSTARI; break;
5496 case OP_NOTPLUSI: *tempcode = OP_NOTPOSPLUSI; break;
5497 case OP_NOTQUERYI: *tempcode = OP_NOTPOSQUERYI; break;
5498 case OP_NOTUPTOI: *tempcode = OP_NOTPOSUPTOI; break;
5499
5500 case OP_TYPESTAR: *tempcode = OP_TYPEPOSSTAR; break;
5501 case OP_TYPEPLUS: *tempcode = OP_TYPEPOSPLUS; break;
5502 case OP_TYPEQUERY: *tempcode = OP_TYPEPOSQUERY; break;
5503 case OP_TYPEUPTO: *tempcode = OP_TYPEPOSUPTO; break;
5504
5505 /* Because we are moving code along, we must ensure that any
5506 pending recursive references are updated. */
5507
5508 default:
5509 *code = OP_END;
5510 adjust_recurse(tempcode, 1 + LINK_SIZE, utf, cd, save_hwm);
5511 memmove(tempcode + 1 + LINK_SIZE, tempcode, IN_UCHARS(len));
5512 code += 1 + LINK_SIZE;
5513 len += 1 + LINK_SIZE;
5514 tempcode[0] = OP_ONCE;
5515 *code++ = OP_KET;
5516 PUTINC(code, 0, len);
5517 PUT(tempcode, 1, len);
5518 break;
5519 }
5520 }
5521
5522 /* In all case we no longer have a previous item. We also set the
5523 "follows varying string" flag for subsequently encountered reqchars if
5524 it isn't already set and we have just passed a varying length item. */
5525
5526 END_REPEAT:
5527 previous = NULL;
5528 cd->req_varyopt |= reqvary;
5529 break;
5530
5531
5532 /* ===================================================================*/
5533 /* Start of nested parenthesized sub-expression, or comment or lookahead or
5534 lookbehind or option setting or condition or all the other extended
5535 parenthesis forms. */
5536
5537 case CHAR_LEFT_PARENTHESIS:
5538 newoptions = options;
5539 skipbytes = 0;
5540 bravalue = OP_CBRA;
5541 save_hwm = cd->hwm;
5542 reset_bracount = FALSE;
5543
5544 /* First deal with various "verbs" that can be introduced by '*'. */
5545
5546 ptr++;
5547 if (ptr[0] == CHAR_ASTERISK && (ptr[1] == ':'
5548 || (MAX_255(ptr[1]) && ((cd->ctypes[ptr[1]] & ctype_letter) != 0))))
5549 {
5550 int i, namelen;
5551 int arglen = 0;
5552 const char *vn = verbnames;
5553 const pcre_uchar *name = ptr + 1;
5554 const pcre_uchar *arg = NULL;
5555 previous = NULL;
5556 ptr++;
5557 while (MAX_255(*ptr) && (cd->ctypes[*ptr] & ctype_letter) != 0) ptr++;
5558 namelen = (int)(ptr - name);
5559
5560 /* It appears that Perl allows any characters whatsoever, other than
5561 a closing parenthesis, to appear in arguments, so we no longer insist on
5562 letters, digits, and underscores. */
5563
5564 if (*ptr == CHAR_COLON)
5565 {
5566 arg = ++ptr;
5567 while (*ptr != 0 && *ptr != CHAR_RIGHT_PARENTHESIS) ptr++;
5568 arglen = (int)(ptr - arg);
5569 if ((unsigned int)arglen > MAX_MARK)
5570 {
5571 *errorcodeptr = ERR75;
5572 goto FAILED;
5573 }
5574 }
5575
5576 if (*ptr != CHAR_RIGHT_PARENTHESIS)
5577 {
5578 *errorcodeptr = ERR60;
5579 goto FAILED;
5580 }
5581
5582 /* Scan the table of verb names */
5583
5584 for (i = 0; i < verbcount; i++)
5585 {
5586 if (namelen == verbs[i].len &&
5587 STRNCMP_UC_C8(name, vn, namelen) == 0)
5588 {
5589 int setverb;
5590
5591 /* Check for open captures before ACCEPT and convert it to
5592 ASSERT_ACCEPT if in an assertion. */
5593
5594 if (verbs[i].op == OP_ACCEPT)
5595 {
5596 open_capitem *oc;
5597 if (arglen != 0)
5598 {
5599 *errorcodeptr = ERR59;
5600 goto FAILED;
5601 }
5602 cd->had_accept = TRUE;
5603 for (oc = cd->open_caps; oc != NULL; oc = oc->next)
5604 {
5605 *code++ = OP_CLOSE;
5606 PUT2INC(code, 0, oc->number);
5607 }
5608 setverb = *code++ =
5609 (cd->assert_depth > 0)? OP_ASSERT_ACCEPT : OP_ACCEPT;
5610
5611 /* Do not set firstchar after *ACCEPT */
5612 if (firstchar == REQ_UNSET) firstchar = REQ_NONE;
5613 }
5614
5615 /* Handle other cases with/without an argument */
5616
5617 else if (arglen == 0)
5618 {
5619 if (verbs[i].op < 0) /* Argument is mandatory */
5620 {
5621 *errorcodeptr = ERR66;
5622 goto FAILED;
5623 }
5624 setverb = *code++ = verbs[i].op;
5625 }
5626
5627 else
5628 {
5629 if (verbs[i].op_arg < 0) /* Argument is forbidden */
5630 {
5631 *errorcodeptr = ERR59;
5632 goto FAILED;
5633 }
5634 setverb = *code++ = verbs[i].op_arg;
5635 *code++ = arglen;
5636 memcpy(code, arg, IN_UCHARS(arglen));
5637 code += arglen;
5638 *code++ = 0;
5639 }
5640
5641 switch (setverb)
5642 {
5643 case OP_THEN:
5644 case OP_THEN_ARG:
5645 cd->external_flags |= PCRE_HASTHEN;
5646 break;
5647
5648 case OP_PRUNE:
5649 case OP_PRUNE_ARG:
5650 case OP_SKIP:
5651 case OP_SKIP_ARG:
5652 cd->had_pruneorskip = TRUE;
5653 break;
5654 }
5655
5656 break; /* Found verb, exit loop */
5657 }
5658
5659 vn += verbs[i].len + 1;
5660 }
5661
5662 if (i < verbcount) continue; /* Successfully handled a verb */
5663 *errorcodeptr = ERR60; /* Verb not recognized */
5664 goto FAILED;
5665 }
5666
5667 /* Deal with the extended parentheses; all are introduced by '?', and the
5668 appearance of any of them means that this is not a capturing group. */
5669
5670 else if (*ptr == CHAR_QUESTION_MARK)
5671 {
5672 int i, set, unset, namelen;
5673 int *optset;
5674 const pcre_uchar *name;
5675 pcre_uchar *slot;
5676
5677 switch (*(++ptr))
5678 {
5679 case CHAR_NUMBER_SIGN: /* Comment; skip to ket */
5680 ptr++;
5681 while (*ptr != 0 && *ptr != CHAR_RIGHT_PARENTHESIS) ptr++;
5682 if (*ptr == 0)
5683 {
5684 *errorcodeptr = ERR18;
5685 goto FAILED;
5686 }
5687 continue;
5688
5689
5690 /* ------------------------------------------------------------ */
5691 case CHAR_VERTICAL_LINE: /* Reset capture count for each branch */
5692 reset_bracount = TRUE;
5693 /* Fall through */
5694
5695 /* ------------------------------------------------------------ */
5696 case CHAR_COLON: /* Non-capturing bracket */
5697 bravalue = OP_BRA;
5698 ptr++;
5699 break;
5700
5701
5702 /* ------------------------------------------------------------ */
5703 case CHAR_LEFT_PARENTHESIS:
5704 bravalue = OP_COND; /* Conditional group */
5705
5706 /* A condition can be an assertion, a number (referring to a numbered
5707 group), a name (referring to a named group), or 'R', referring to
5708 recursion. R<digits> and R&name are also permitted for recursion tests.
5709
5710 There are several syntaxes for testing a named group: (?(name)) is used
5711 by Python; Perl 5.10 onwards uses (?(<name>) or (?('name')).
5712
5713 There are two unfortunate ambiguities, caused by history. (a) 'R' can
5714 be the recursive thing or the name 'R' (and similarly for 'R' followed
5715 by digits), and (b) a number could be a name that consists of digits.
5716 In both cases, we look for a name first; if not found, we try the other
5717 cases. */
5718
5719 /* For conditions that are assertions, check the syntax, and then exit
5720 the switch. This will take control down to where bracketed groups,
5721 including assertions, are processed. */
5722
5723 if (ptr[1] == CHAR_QUESTION_MARK && (ptr[2] == CHAR_EQUALS_SIGN ||
5724 ptr[2] == CHAR_EXCLAMATION_MARK || ptr[2] == CHAR_LESS_THAN_SIGN))
5725 break;
5726
5727 /* Most other conditions use OP_CREF (a couple change to OP_RREF
5728 below), and all need to skip 1+IMM2_SIZE bytes at the start of the group. */
5729
5730 code[1+LINK_SIZE] = OP_CREF;
5731 skipbytes = 1+IMM2_SIZE;
5732 refsign = -1;
5733
5734 /* Check for a test for recursion in a named group. */
5735
5736 if (ptr[1] == CHAR_R && ptr[2] == CHAR_AMPERSAND)
5737 {
5738 terminator = -1;
5739 ptr += 2;
5740 code[1+LINK_SIZE] = OP_RREF; /* Change the type of test */
5741 }
5742
5743 /* Check for a test for a named group's having been set, using the Perl
5744 syntax (?(<name>) or (?('name') */
5745
5746 else if (ptr[1] == CHAR_LESS_THAN_SIGN)
5747 {
5748 terminator = CHAR_GREATER_THAN_SIGN;
5749 ptr++;
5750 }
5751 else if (ptr[1] == CHAR_APOSTROPHE)
5752 {
5753 terminator = CHAR_APOSTROPHE;
5754 ptr++;
5755 }
5756 else
5757 {
5758 terminator = 0;
5759 if (ptr[1] == CHAR_MINUS || ptr[1] == CHAR_PLUS) refsign = *(++ptr);
5760 }
5761
5762 /* We now expect to read a name; any thing else is an error */
5763
5764 if (!MAX_255(ptr[1]) || (cd->ctypes[ptr[1]] & ctype_word) == 0)
5765 {
5766 ptr += 1; /* To get the right offset */
5767 *errorcodeptr = ERR28;
5768 goto FAILED;
5769 }
5770
5771 /* Read the name, but also get it as a number if it's all digits */
5772
5773 recno = 0;
5774 name = ++ptr;
5775 while (MAX_255(*ptr) && (cd->ctypes[*ptr] & ctype_word) != 0)
5776 {
5777 if (recno >= 0)
5778 recno = (IS_DIGIT(*ptr))? recno * 10 + *ptr - CHAR_0 : -1;
5779 ptr++;
5780 }
5781 namelen = (int)(ptr - name);
5782
5783 if ((terminator > 0 && *ptr++ != terminator) ||
5784 *ptr++ != CHAR_RIGHT_PARENTHESIS)
5785 {
5786 ptr--; /* Error offset */
5787 *errorcodeptr = ERR26;
5788 goto FAILED;
5789 }
5790
5791 /* Do no further checking in the pre-compile phase. */
5792
5793 if (lengthptr != NULL) break;
5794
5795 /* In the real compile we do the work of looking for the actual
5796 reference. If the string started with "+" or "-" we require the rest to
5797 be digits, in which case recno will be set. */
5798
5799 if (refsign > 0)
5800 {
5801 if (recno <= 0)
5802 {
5803 *errorcodeptr = ERR58;
5804 goto FAILED;
5805 }
5806 recno = (refsign == CHAR_MINUS)?
5807 cd->bracount - recno + 1 : recno +cd->bracount;
5808 if (recno <= 0 || recno > cd->final_bracount)
5809 {
5810 *errorcodeptr = ERR15;
5811 goto FAILED;
5812 }
5813 PUT2(code, 2+LINK_SIZE, recno);
5814 break;
5815 }
5816
5817 /* Otherwise (did not start with "+" or "-"), start by looking for the
5818 name. If we find a name, add one to the opcode to change OP_CREF or
5819 OP_RREF into OP_NCREF or OP_NRREF. These behave exactly the same,
5820 except they record that the reference was originally to a name. The
5821 information is used to check duplicate names. */
5822
5823 slot = cd->name_table;
5824 for (i = 0; i < cd->names_found; i++)
5825 {
5826 if (STRNCMP_UC_UC(name, slot+IMM2_SIZE, namelen) == 0) break;
5827 slot += cd->name_entry_size;
5828 }
5829
5830 /* Found a previous named subpattern */
5831
5832 if (i < cd->names_found)
5833 {
5834 recno = GET2(slot, 0);
5835 PUT2(code, 2+LINK_SIZE, recno);
5836 code[1+LINK_SIZE]++;
5837 }
5838
5839 /* Search the pattern for a forward reference */
5840
5841 else if ((i = find_parens(cd, name, namelen,
5842 (options & PCRE_EXTENDED) != 0, utf)) > 0)
5843 {
5844 PUT2(code, 2+LINK_SIZE, i);
5845 code[1+LINK_SIZE]++;
5846 }
5847
5848 /* If terminator == 0 it means that the name followed directly after
5849 the opening parenthesis [e.g. (?(abc)...] and in this case there are
5850 some further alternatives to try. For the cases where terminator != 0
5851 [things like (?(<name>... or (?('name')... or (?(R&name)... ] we have
5852 now checked all the possibilities, so give an error. */
5853
5854 else if (terminator != 0)
5855 {
5856 *errorcodeptr = ERR15;
5857 goto FAILED;
5858 }
5859
5860 /* Check for (?(R) for recursion. Allow digits after R to specify a
5861 specific group number. */
5862
5863 else if (*name == CHAR_R)
5864 {
5865 recno = 0;
5866 for (i = 1; i < namelen; i++)
5867 {
5868 if (!IS_DIGIT(name[i]))
5869 {
5870 *errorcodeptr = ERR15;
5871 goto FAILED;
5872 }
5873 recno = recno * 10 + name[i] - CHAR_0;
5874 }
5875 if (recno == 0) recno = RREF_ANY;
5876 code[1+LINK_SIZE] = OP_RREF; /* Change test type */
5877 PUT2(code, 2+LINK_SIZE, recno);
5878 }
5879
5880 /* Similarly, check for the (?(DEFINE) "condition", which is always
5881 false. */
5882
5883 else if (namelen == 6 && STRNCMP_UC_C8(name, STRING_DEFINE, 6) == 0)
5884 {
5885 code[1+LINK_SIZE] = OP_DEF;
5886 skipbytes = 1;
5887 }
5888
5889 /* Check for the "name" actually being a subpattern number. We are
5890 in the second pass here, so final_bracount is set. */
5891
5892 else if (recno > 0 && recno <= cd->final_bracount)
5893 {
5894 PUT2(code, 2+LINK_SIZE, recno);
5895 }
5896
5897 /* Either an unidentified subpattern, or a reference to (?(0) */
5898
5899 else
5900 {
5901 *errorcodeptr = (recno == 0)? ERR35: ERR15;
5902 goto FAILED;
5903 }
5904 break;
5905
5906
5907 /* ------------------------------------------------------------ */
5908 case CHAR_EQUALS_SIGN: /* Positive lookahead */
5909 bravalue = OP_ASSERT;
5910 cd->assert_depth += 1;
5911 ptr++;
5912 break;
5913
5914
5915 /* ------------------------------------------------------------ */
5916 case CHAR_EXCLAMATION_MARK: /* Negative lookahead */
5917 ptr++;
5918 if (*ptr == CHAR_RIGHT_PARENTHESIS) /* Optimize (?!) */
5919 {
5920 *code++ = OP_FAIL;
5921 previous = NULL;
5922 continue;
5923 }
5924 bravalue = OP_ASSERT_NOT;
5925 cd->assert_depth += 1;
5926 break;
5927
5928
5929 /* ------------------------------------------------------------ */
5930 case CHAR_LESS_THAN_SIGN: /* Lookbehind or named define */
5931 switch (ptr[1])
5932 {
5933 case CHAR_EQUALS_SIGN: /* Positive lookbehind */
5934 bravalue = OP_ASSERTBACK;
5935 cd->assert_depth += 1;
5936 ptr += 2;
5937 break;
5938
5939 case CHAR_EXCLAMATION_MARK: /* Negative lookbehind */
5940 bravalue = OP_ASSERTBACK_NOT;
5941 cd->assert_depth += 1;
5942 ptr += 2;
5943 break;
5944
5945 default: /* Could be name define, else bad */
5946 if (MAX_255(ptr[1]) && (cd->ctypes[ptr[1]] & ctype_word) != 0)
5947 goto DEFINE_NAME;
5948 ptr++; /* Correct offset for error */
5949 *errorcodeptr = ERR24;
5950 goto FAILED;
5951 }
5952 break;
5953
5954
5955 /* ------------------------------------------------------------ */
5956 case CHAR_GREATER_THAN_SIGN: /* One-time brackets */
5957 bravalue = OP_ONCE;
5958 ptr++;
5959 break;
5960
5961
5962 /* ------------------------------------------------------------ */
5963 case CHAR_C: /* Callout - may be followed by digits; */
5964 previous_callout = code; /* Save for later completion */
5965 after_manual_callout = 1; /* Skip one item before completing */
5966 *code++ = OP_CALLOUT;
5967 {
5968 int n = 0;
5969 ptr++;
5970 while(IS_DIGIT(*ptr))
5971 n = n * 10 + *ptr++ - CHAR_0;
5972 if (*ptr != CHAR_RIGHT_PARENTHESIS)
5973 {
5974 *errorcodeptr = ERR39;
5975 goto FAILED;
5976 }
5977 if (n > 255)
5978 {
5979 *errorcodeptr = ERR38;
5980 goto FAILED;
5981 }
5982 *code++ = n;
5983 PUT(code, 0, (int)(ptr - cd->start_pattern + 1)); /* Pattern offset */
5984 PUT(code, LINK_SIZE, 0); /* Default length */
5985 code += 2 * LINK_SIZE;
5986 }
5987 previous = NULL;
5988 continue;
5989
5990
5991 /* ------------------------------------------------------------ */
5992 case CHAR_P: /* Python-style named subpattern handling */
5993 if (*(++ptr) == CHAR_EQUALS_SIGN ||
5994 *ptr == CHAR_GREATER_THAN_SIGN) /* Reference or recursion */
5995 {
5996 is_recurse = *ptr == CHAR_GREATER_THAN_SIGN;
5997 terminator = CHAR_RIGHT_PARENTHESIS;
5998 goto NAMED_REF_OR_RECURSE;
5999 }
6000 else if (*ptr != CHAR_LESS_THAN_SIGN) /* Test for Python-style defn */
6001 {
6002 *errorcodeptr = ERR41;
6003 goto FAILED;
6004 }
6005 /* Fall through to handle (?P< as (?< is handled */
6006
6007
6008 /* ------------------------------------------------------------ */
6009 DEFINE_NAME: /* Come here from (?< handling */
6010 case CHAR_APOSTROPHE:
6011 {
6012 terminator = (*ptr == CHAR_LESS_THAN_SIGN)?
6013 CHAR_GREATER_THAN_SIGN : CHAR_APOSTROPHE;
6014 name = ++ptr;
6015
6016 while (MAX_255(*ptr) && (cd->ctypes[*ptr] & ctype_word) != 0) ptr++;
6017 namelen = (int)(ptr - name);
6018
6019 /* In the pre-compile phase, just do a syntax check. */
6020
6021 if (lengthptr != NULL)
6022 {
6023 if (*ptr != terminator)
6024 {
6025 *errorcodeptr = ERR42;
6026 goto FAILED;
6027 }
6028 if (cd->names_found >= MAX_NAME_COUNT)
6029 {
6030 *errorcodeptr = ERR49;
6031 goto FAILED;
6032 }
6033 if (namelen + IMM2_SIZE + 1 > cd->name_entry_size)
6034 {
6035 cd->name_entry_size = namelen + IMM2_SIZE + 1;
6036 if (namelen > MAX_NAME_SIZE)
6037 {
6038 *errorcodeptr = ERR48;
6039 goto FAILED;
6040 }
6041 }
6042 }
6043
6044 /* In the real compile, create the entry in the table, maintaining
6045 alphabetical order. Duplicate names for different numbers are
6046 permitted only if PCRE_DUPNAMES is set. Duplicate names for the same
6047 number are always OK. (An existing number can be re-used if (?|
6048 appears in the pattern.) In either event, a duplicate name results in
6049 a duplicate entry in the table, even if the number is the same. This
6050 is because the number of names, and hence the table size, is computed
6051 in the pre-compile, and it affects various numbers and pointers which
6052 would all have to be modified, and the compiled code moved down, if
6053 duplicates with the same number were omitted from the table. This
6054 doesn't seem worth the hassle. However, *different* names for the
6055 same number are not permitted. */
6056
6057 else
6058 {
6059 BOOL dupname = FALSE;
6060 slot = cd->name_table;
6061
6062 for (i = 0; i < cd->names_found; i++)
6063 {
6064 int crc = memcmp(name, slot+IMM2_SIZE, IN_UCHARS(namelen));
6065 if (crc == 0)
6066 {
6067 if (slot[IMM2_SIZE+namelen] == 0)
6068 {
6069 if (GET2(slot, 0) != cd->bracount + 1 &&
6070 (options & PCRE_DUPNAMES) == 0)
6071 {
6072 *errorcodeptr = ERR43;
6073 goto FAILED;
6074 }
6075 else dupname = TRUE;
6076 }
6077 else crc = -1; /* Current name is a substring */
6078 }
6079
6080 /* Make space in the table and break the loop for an earlier
6081 name. For a duplicate or later name, carry on. We do this for
6082 duplicates so that in the simple case (when ?(| is not used) they
6083 are in order of their numbers. */
6084
6085 if (crc < 0)
6086 {
6087 memmove(slot + cd->name_entry_size, slot,
6088 IN_UCHARS((cd->names_found - i) * cd->name_entry_size));
6089 break;
6090 }
6091
6092 /* Continue the loop for a later or duplicate name */
6093
6094 slot += cd->name_entry_size;
6095 }
6096
6097 /* For non-duplicate names, check for a duplicate number before
6098 adding the new name. */
6099
6100 if (!dupname)
6101 {
6102 pcre_uchar *cslot = cd->name_table;
6103 for (i = 0; i < cd->names_found; i++)
6104 {
6105 if (cslot != slot)
6106 {
6107 if (GET2(cslot, 0) == cd->bracount + 1)
6108 {
6109 *errorcodeptr = ERR65;
6110 goto FAILED;
6111 }
6112 }
6113 else i--;
6114 cslot += cd->name_entry_size;
6115 }
6116 }
6117
6118 PUT2(slot, 0, cd->bracount + 1);
6119 memcpy(slot + IMM2_SIZE, name, IN_UCHARS(namelen));
6120 slot[IMM2_SIZE + namelen] = 0;
6121 }
6122 }
6123
6124 /* In both pre-compile and compile, count the number of names we've
6125 encountered. */
6126
6127 cd->names_found++;
6128 ptr++; /* Move past > or ' */
6129 goto NUMBERED_GROUP;
6130
6131
6132 /* ------------------------------------------------------------ */
6133 case CHAR_AMPERSAND: /* Perl recursion/subroutine syntax */
6134 terminator = CHAR_RIGHT_PARENTHESIS;
6135 is_recurse = TRUE;
6136 /* Fall through */
6137
6138 /* We come here from the Python syntax above that handles both
6139 references (?P=name) and recursion (?P>name), as well as falling
6140 through from the Perl recursion syntax (?&name). We also come here from
6141 the Perl \k<name> or \k'name' back reference syntax and the \k{name}
6142 .NET syntax, and the Oniguruma \g<...> and \g'...' subroutine syntax. */
6143
6144 NAMED_REF_OR_RECURSE:
6145 name = ++ptr;
6146 while (MAX_255(*ptr) && (cd->ctypes[*ptr] & ctype_word) != 0) ptr++;
6147 namelen = (int)(ptr - name);
6148
6149 /* In the pre-compile phase, do a syntax check. We used to just set
6150 a dummy reference number, because it was not used in the first pass.
6151 However, with the change of recursive back references to be atomic,
6152 we have to look for the number so that this state can be identified, as
6153 otherwise the incorrect length is computed. If it's not a backwards
6154 reference, the dummy number will do. */
6155
6156 if (lengthptr != NULL)
6157 {
6158 const pcre_uchar *temp;
6159
6160 if (namelen == 0)
6161 {
6162 *errorcodeptr = ERR62;
6163 goto FAILED;
6164 }
6165 if (*ptr != terminator)
6166 {
6167 *errorcodeptr = ERR42;
6168 goto FAILED;
6169 }
6170 if (namelen > MAX_NAME_SIZE)
6171 {
6172 *errorcodeptr = ERR48;
6173 goto FAILED;
6174 }
6175
6176 /* The name table does not exist in the first pass, so we cannot
6177 do a simple search as in the code below. Instead, we have to scan the
6178 pattern to find the number. It is important that we scan it only as
6179 far as we have got because the syntax of named subpatterns has not
6180 been checked for the rest of the pattern, and find_parens() assumes
6181 correct syntax. In any case, it's a waste of resources to scan
6182 further. We stop the scan at the current point by temporarily
6183 adjusting the value of cd->endpattern. */
6184
6185 temp = cd->end_pattern;
6186 cd->end_pattern = ptr;
6187 recno = find_parens(cd, name, namelen,
6188 (options & PCRE_EXTENDED) != 0, utf);
6189 cd->end_pattern = temp;
6190 if (recno < 0) recno = 0; /* Forward ref; set dummy number */
6191 }
6192
6193 /* In the real compile, seek the name in the table. We check the name
6194 first, and then check that we have reached the end of the name in the
6195 table. That way, if the name that is longer than any in the table,
6196 the comparison will fail without reading beyond the table entry. */
6197
6198 else
6199 {
6200 slot = cd->name_table;
6201 for (i = 0; i < cd->names_found; i++)
6202 {
6203 if (STRNCMP_UC_UC(name, slot+IMM2_SIZE, namelen) == 0 &&
6204 slot[IMM2_SIZE+namelen] == 0)
6205 break;
6206 slot += cd->name_entry_size;
6207 }
6208
6209 if (i < cd->names_found) /* Back reference */
6210 {
6211 recno = GET2(slot, 0);
6212 }
6213 else if ((recno = /* Forward back reference */
6214 find_parens(cd, name, namelen,
6215 (options & PCRE_EXTENDED) != 0, utf)) <= 0)
6216 {
6217 *errorcodeptr = ERR15;
6218 goto FAILED;
6219 }
6220 }
6221
6222 /* In both phases, we can now go to the code than handles numerical
6223 recursion or backreferences. */
6224
6225 if (is_recurse) goto HANDLE_RECURSION;
6226 else goto HANDLE_REFERENCE;
6227
6228
6229 /* ------------------------------------------------------------ */
6230 case CHAR_R: /* Recursion */
6231 ptr++; /* Same as (?0) */
6232 /* Fall through */
6233
6234
6235 /* ------------------------------------------------------------ */
6236 case CHAR_MINUS: case CHAR_PLUS: /* Recursion or subroutine */
6237 case CHAR_0: case CHAR_1: case CHAR_2: case CHAR_3: case CHAR_4:
6238 case CHAR_5: case CHAR_6: case CHAR_7: case CHAR_8: case CHAR_9:
6239 {
6240 const pcre_uchar *called;
6241 terminator = CHAR_RIGHT_PARENTHESIS;
6242
6243 /* Come here from the \g<...> and \g'...' code (Oniguruma
6244 compatibility). However, the syntax has been checked to ensure that
6245 the ... are a (signed) number, so that neither ERR63 nor ERR29 will
6246 be called on this path, nor with the jump to OTHER_CHAR_AFTER_QUERY
6247 ever be taken. */
6248
6249 HANDLE_NUMERICAL_RECURSION:
6250
6251 if ((refsign = *ptr) == CHAR_PLUS)
6252 {
6253 ptr++;
6254 if (!IS_DIGIT(*ptr))
6255 {
6256 *errorcodeptr = ERR63;
6257 goto FAILED;
6258 }
6259 }
6260 else if (refsign == CHAR_MINUS)
6261 {
6262 if (!IS_DIGIT(ptr[1]))
6263 goto OTHER_CHAR_AFTER_QUERY;
6264 ptr++;
6265 }
6266
6267 recno = 0;
6268 while(IS_DIGIT(*ptr))
6269 recno = recno * 10 + *ptr++ - CHAR_0;
6270
6271 if (*ptr != terminator)
6272 {
6273 *errorcodeptr = ERR29;
6274 goto FAILED;
6275 }
6276
6277 if (refsign == CHAR_MINUS)
6278 {
6279 if (recno == 0)
6280 {
6281 *errorcodeptr = ERR58;
6282 goto FAILED;
6283 }
6284 recno = cd->bracount - recno + 1;
6285 if (recno <= 0)
6286 {
6287 *errorcodeptr = ERR15;
6288 goto FAILED;
6289 }
6290 }
6291 else if (refsign == CHAR_PLUS)
6292 {
6293 if (recno == 0)
6294 {
6295 *errorcodeptr = ERR58;
6296 goto FAILED;
6297 }
6298 recno += cd->bracount;
6299 }
6300
6301 /* Come here from code above that handles a named recursion */
6302
6303 HANDLE_RECURSION:
6304
6305 previous = code;
6306 called = cd->start_code;
6307
6308 /* When we are actually compiling, find the bracket that is being
6309 referenced. Temporarily end the regex in case it doesn't exist before
6310 this point. If we end up with a forward reference, first check that
6311 the bracket does occur later so we can give the error (and position)
6312 now. Then remember this forward reference in the workspace so it can
6313 be filled in at the end. */
6314
6315 if (lengthptr == NULL)
6316 {
6317 *code = OP_END;
6318 if (recno != 0)
6319 called = PRIV(find_bracket)(cd->start_code, utf, recno);
6320
6321 /* Forward reference */
6322
6323 if (called == NULL)
6324 {
6325 if (find_parens(cd, NULL, recno,
6326 (options & PCRE_EXTENDED) != 0, utf) < 0)
6327 {
6328 *errorcodeptr = ERR15;
6329 goto FAILED;
6330 }
6331
6332 /* Fudge the value of "called" so that when it is inserted as an
6333 offset below, what it actually inserted is the reference number
6334 of the group. Then remember the forward reference. */
6335
6336 called = cd->start_code + recno;
6337 if (cd->hwm >= cd->start_workspace + cd->workspace_size -
6338 WORK_SIZE_SAFETY_MARGIN)
6339 {
6340 *errorcodeptr = expand_workspace(cd);
6341 if (*errorcodeptr != 0) goto FAILED;
6342 }
6343 PUTINC(cd->hwm, 0, (int)(code + 1 - cd->start_code));
6344 }
6345
6346 /* If not a forward reference, and the subpattern is still open,
6347 this is a recursive call. We check to see if this is a left
6348 recursion that could loop for ever, and diagnose that case. We
6349 must not, however, do this check if we are in a conditional
6350 subpattern because the condition might be testing for recursion in
6351 a pattern such as /(?(R)a+|(?R)b)/, which is perfectly valid.
6352 Forever loops are also detected at runtime, so those that occur in
6353 conditional subpatterns will be picked up then. */
6354
6355 else if (GET(called, 1) == 0 && cond_depth <= 0 &&
6356 could_be_empty(called, code, bcptr, utf, cd))
6357 {
6358 *errorcodeptr = ERR40;
6359 goto FAILED;
6360 }
6361 }
6362
6363 /* Insert the recursion/subroutine item. It does not have a set first
6364 character (relevant if it is repeated, because it will then be
6365 wrapped with ONCE brackets). */
6366
6367 *code = OP_RECURSE;
6368 PUT(code, 1, (int)(called - cd->start_code));
6369 code += 1 + LINK_SIZE;
6370 groupsetfirstchar = FALSE;
6371 }
6372
6373 /* Can't determine a first byte now */
6374
6375 if (firstchar == REQ_UNSET) firstchar = REQ_NONE;
6376 continue;
6377
6378
6379 /* ------------------------------------------------------------ */
6380 default: /* Other characters: check option setting */
6381 OTHER_CHAR_AFTER_QUERY:
6382 set = unset = 0;
6383 optset = &set;
6384
6385 while (*ptr != CHAR_RIGHT_PARENTHESIS && *ptr != CHAR_COLON)
6386 {
6387 switch (*ptr++)
6388 {
6389 case CHAR_MINUS: optset = &unset; break;
6390
6391 case CHAR_J: /* Record that it changed in the external options */
6392 *optset |= PCRE_DUPNAMES;
6393 cd->external_flags |= PCRE_JCHANGED;
6394 break;
6395
6396 case CHAR_i: *optset |= PCRE_CASELESS; break;
6397 case CHAR_m: *optset |= PCRE_MULTILINE; break;
6398 case CHAR_s: *optset |= PCRE_DOTALL; break;
6399 case CHAR_x: *optset |= PCRE_EXTENDED; break;
6400 case CHAR_U: *optset |= PCRE_UNGREEDY; break;
6401 case CHAR_X: *optset |= PCRE_EXTRA; break;
6402
6403 default: *errorcodeptr = ERR12;
6404 ptr--; /* Correct the offset */
6405 goto FAILED;
6406 }
6407 }
6408
6409 /* Set up the changed option bits, but don't change anything yet. */
6410
6411 newoptions = (options | set) & (~unset);
6412
6413 /* If the options ended with ')' this is not the start of a nested
6414 group with option changes, so the options change at this level. If this
6415 item is right at the start of the pattern, the options can be
6416 abstracted and made external in the pre-compile phase, and ignored in
6417 the compile phase. This can be helpful when matching -- for instance in
6418 caseless checking of required bytes.
6419
6420 If the code pointer is not (cd->start_code + 1 + LINK_SIZE), we are
6421 definitely *not* at the start of the pattern because something has been
6422 compiled. In the pre-compile phase, however, the code pointer can have
6423 that value after the start, because it gets reset as code is discarded
6424 during the pre-compile. However, this can happen only at top level - if
6425 we are within parentheses, the starting BRA will still be present. At
6426 any parenthesis level, the length value can be used to test if anything
6427 has been compiled at that level. Thus, a test for both these conditions
6428 is necessary to ensure we correctly detect the start of the pattern in
6429 both phases.
6430
6431 If we are not at the pattern start, reset the greedy defaults and the
6432 case value for firstchar and reqchar. */
6433
6434 if (*ptr == CHAR_RIGHT_PARENTHESIS)
6435 {
6436 if (code == cd->start_code + 1 + LINK_SIZE &&
6437 (lengthptr == NULL || *lengthptr == 2 + 2*LINK_SIZE))
6438 {
6439 cd->external_options = newoptions;
6440 }
6441 else
6442 {
6443 greedy_default = ((newoptions & PCRE_UNGREEDY) != 0);
6444 greedy_non_default = greedy_default ^ 1;
6445 req_caseopt = ((newoptions & PCRE_CASELESS) != 0)? REQ_CASELESS:0;
6446 }
6447
6448 /* Change options at this level, and pass them back for use
6449 in subsequent branches. */
6450
6451 *optionsptr = options = newoptions;
6452 previous = NULL; /* This item can't be repeated */
6453 continue; /* It is complete */
6454 }
6455
6456 /* If the options ended with ':' we are heading into a nested group
6457 with possible change of options. Such groups are non-capturing and are
6458 not assertions of any kind. All we need to do is skip over the ':';
6459 the newoptions value is handled below. */
6460
6461 bravalue = OP_BRA;
6462 ptr++;
6463 } /* End of switch for character following (? */
6464 } /* End of (? handling */
6465
6466 /* Opening parenthesis not followed by '*' or '?'. If PCRE_NO_AUTO_CAPTURE
6467 is set, all unadorned brackets become non-capturing and behave like (?:...)
6468 brackets. */
6469
6470 else if ((options & PCRE_NO_AUTO_CAPTURE) != 0)
6471 {
6472 bravalue = OP_BRA;
6473 }
6474
6475 /* Else we have a capturing group. */
6476
6477 else
6478 {
6479 NUMBERED_GROUP:
6480 cd->bracount += 1;
6481 PUT2(code, 1+LINK_SIZE, cd->bracount);
6482 skipbytes = IMM2_SIZE;
6483 }
6484
6485 /* Process nested bracketed regex. Assertions used not to be repeatable,
6486 but this was changed for Perl compatibility, so all kinds can now be
6487 repeated. We copy code into a non-register variable (tempcode) in order to
6488 be able to pass its address because some compilers complain otherwise. */
6489
6490 previous = code; /* For handling repetition */
6491 *code = bravalue;
6492 tempcode = code;
6493 tempreqvary = cd->req_varyopt; /* Save value before bracket */
6494 tempbracount = cd->bracount; /* Save value before bracket */
6495 length_prevgroup = 0; /* Initialize for pre-compile phase */
6496
6497 if (!compile_regex(
6498 newoptions, /* The complete new option state */
6499 &tempcode, /* Where to put code (updated) */
6500 &ptr, /* Input pointer (updated) */
6501 errorcodeptr, /* Where to put an error message */
6502 (bravalue == OP_ASSERTBACK ||
6503 bravalue == OP_ASSERTBACK_NOT), /* TRUE if back assert */
6504 reset_bracount, /* True if (?| group */
6505 skipbytes, /* Skip over bracket number */
6506 cond_depth +
6507 ((bravalue == OP_COND)?1:0), /* Depth of condition subpatterns */
6508 &subfirstchar, /* For possible first char */
6509 &subreqchar, /* For possible last char */
6510 bcptr, /* Current branch chain */
6511 cd, /* Tables block */
6512 (lengthptr == NULL)? NULL : /* Actual compile phase */
6513 &length_prevgroup /* Pre-compile phase */
6514 ))
6515 goto FAILED;
6516
6517 /* If this was an atomic group and there are no capturing groups within it,
6518 generate OP_ONCE_NC instead of OP_ONCE. */
6519
6520 if (bravalue == OP_ONCE && cd->bracount <= tempbracount)
6521 *code = OP_ONCE_NC;
6522
6523 if (bravalue >= OP_ASSERT && bravalue <= OP_ASSERTBACK_NOT)
6524 cd->assert_depth -= 1;
6525
6526 /* At the end of compiling, code is still pointing to the start of the
6527 group, while tempcode has been updated to point past the end of the group.
6528 The pattern pointer (ptr) is on the bracket.
6529
6530 If this is a conditional bracket, check that there are no more than
6531 two branches in the group, or just one if it's a DEFINE group. We do this
6532 in the real compile phase, not in the pre-pass, where the whole group may
6533 not be available. */
6534
6535 if (bravalue == OP_COND && lengthptr == NULL)
6536 {
6537 pcre_uchar *tc = code;
6538 int condcount = 0;
6539
6540 do {
6541 condcount++;
6542 tc += GET(tc,1);
6543 }
6544 while (*tc != OP_KET);
6545
6546 /* A DEFINE group is never obeyed inline (the "condition" is always
6547 false). It must have only one branch. */
6548
6549 if (code[LINK_SIZE+1] == OP_DEF)
6550 {
6551 if (condcount > 1)
6552 {
6553 *errorcodeptr = ERR54;
6554 goto FAILED;
6555 }
6556 bravalue = OP_DEF; /* Just a flag to suppress char handling below */
6557 }
6558
6559 /* A "normal" conditional group. If there is just one branch, we must not
6560 make use of its firstchar or reqchar, because this is equivalent to an
6561 empty second branch. */
6562
6563 else
6564 {
6565 if (condcount > 2)
6566 {
6567 *errorcodeptr = ERR27;
6568 goto FAILED;
6569 }
6570 if (condcount == 1) subfirstchar = subreqchar = REQ_NONE;
6571 }
6572 }
6573
6574 /* Error if hit end of pattern */
6575
6576 if (*ptr != CHAR_RIGHT_PARENTHESIS)
6577 {
6578 *errorcodeptr = ERR14;
6579 goto FAILED;
6580 }
6581
6582 /* In the pre-compile phase, update the length by the length of the group,
6583 less the brackets at either end. Then reduce the compiled code to just a
6584 set of non-capturing brackets so that it doesn't use much memory if it is
6585 duplicated by a quantifier.*/
6586
6587 if (lengthptr != NULL)
6588 {
6589 if (OFLOW_MAX - *lengthptr < length_prevgroup - 2 - 2*LINK_SIZE)
6590 {
6591 *errorcodeptr = ERR20;
6592 goto FAILED;
6593 }
6594 *lengthptr += length_prevgroup - 2 - 2*LINK_SIZE;
6595 code++; /* This already contains bravalue */
6596 PUTINC(code, 0, 1 + LINK_SIZE);
6597 *code++ = OP_KET;
6598 PUTINC(code, 0, 1 + LINK_SIZE);
6599 break; /* No need to waste time with special character handling */
6600 }
6601
6602 /* Otherwise update the main code pointer to the end of the group. */
6603
6604 code = tempcode;
6605
6606 /* For a DEFINE group, required and first character settings are not
6607 relevant. */
6608
6609 if (bravalue == OP_DEF) break;
6610
6611 /* Handle updating of the required and first characters for other types of
6612 group. Update for normal brackets of all kinds, and conditions with two
6613 branches (see code above). If the bracket is followed by a quantifier with
6614 zero repeat, we have to back off. Hence the definition of zeroreqchar and
6615 zerofirstchar outside the main loop so that they can be accessed for the
6616 back off. */
6617
6618 zeroreqchar = reqchar;
6619 zerofirstchar = firstchar;
6620 groupsetfirstchar = FALSE;
6621
6622 if (bravalue >= OP_ONCE)
6623 {
6624 /* If we have not yet set a firstchar in this branch, take it from the
6625 subpattern, remembering that it was set here so that a repeat of more
6626 than one can replicate it as reqchar if necessary. If the subpattern has
6627 no firstchar, set "none" for the whole branch. In both cases, a zero
6628 repeat forces firstchar to "none". */
6629
6630 if (firstchar == REQ_UNSET)
6631 {
6632 if (subfirstchar >= 0)
6633 {
6634 firstchar = subfirstchar;
6635 groupsetfirstchar = TRUE;
6636 }
6637 else firstchar = REQ_NONE;
6638 zerofirst