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

Diff of /code/trunk/pcre_dfa_exec.c

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

revision 535 by ph10, Thu Jun 3 19:18:24 2010 UTC revision 881 by ph10, Sun Jan 15 18:07:05 2012 UTC
# Line 7  and semantics are as close as possible t Line 7  and semantics are as close as possible t
7  below for why this module is different).  below for why this module is different).
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2010 University of Cambridge             Copyright (c) 1997-2012 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 113  small value. Non-zero values in the tabl Line 113  small value. Non-zero values in the tabl
113  the character is to be found. ***NOTE*** If the start of this table is  the character is to be found. ***NOTE*** If the start of this table is
114  modified, the three tables that follow must also be modified. */  modified, the three tables that follow must also be modified. */
115    
116  static const uschar coptable[] = {  static const pcre_uint8 coptable[] = {
117    0,                             /* End                                    */    0,                             /* End                                    */
118    0, 0, 0, 0, 0,                 /* \A, \G, \K, \B, \b                     */    0, 0, 0, 0, 0,                 /* \A, \G, \K, \B, \b                     */
119    0, 0, 0, 0, 0, 0,              /* \D, \d, \S, \s, \W, \w                 */    0, 0, 0, 0, 0, 0,              /* \D, \d, \S, \s, \W, \w                 */
# Line 121  static const uschar coptable[] = { Line 121  static const uschar coptable[] = {
121    0, 0,                          /* \P, \p                                 */    0, 0,                          /* \P, \p                                 */
122    0, 0, 0, 0, 0,                 /* \R, \H, \h, \V, \v                     */    0, 0, 0, 0, 0,                 /* \R, \H, \h, \V, \v                     */
123    0,                             /* \X                                     */    0,                             /* \X                                     */
124    0, 0, 0, 0, 0,                 /* \Z, \z, Opt, ^, $                      */    0, 0, 0, 0, 0, 0,              /* \Z, \z, ^, ^M, $, $M                   */
125    1,                             /* Char                                   */    1,                             /* Char                                   */
126    1,                             /* Charnc                                 */    1,                             /* Chari                                  */
127    1,                             /* not                                    */    1,                             /* not                                    */
128      1,                             /* noti                                   */
129    /* Positive single-char repeats                                          */    /* Positive single-char repeats                                          */
130    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */
131    3, 3, 3,                       /* upto, minupto, exact                   */    1+IMM2_SIZE, 1+IMM2_SIZE,      /* upto, minupto                          */
132    1, 1, 1, 3,                    /* *+, ++, ?+, upto+                      */    1+IMM2_SIZE,                   /* exact                                  */
133      1, 1, 1, 1+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */
134      1, 1, 1, 1, 1, 1,              /* *I, *?I, +I, +?I, ?I, ??I              */
135      1+IMM2_SIZE, 1+IMM2_SIZE,      /* upto I, minupto I                      */
136      1+IMM2_SIZE,                   /* exact I                                */
137      1, 1, 1, 1+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */
138    /* Negative single-char repeats - only for chars < 256                   */    /* Negative single-char repeats - only for chars < 256                   */
139    1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */    1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */
140    3, 3, 3,                       /* NOT upto, minupto, exact               */    1+IMM2_SIZE, 1+IMM2_SIZE,      /* NOT upto, minupto                      */
141    1, 1, 1, 3,                    /* NOT *+, ++, ?+, updo+                  */    1+IMM2_SIZE,                   /* NOT exact                              */
142      1, 1, 1, 1+IMM2_SIZE,          /* NOT *+, ++, ?+, upto+                  */
143      1, 1, 1, 1, 1, 1,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */
144      1+IMM2_SIZE, 1+IMM2_SIZE,      /* NOT upto I, minupto I                  */
145      1+IMM2_SIZE,                   /* NOT exact I                            */
146      1, 1, 1, 1+IMM2_SIZE,          /* NOT *+I, ++I, ?+I, upto+I              */
147    /* Positive type repeats                                                 */    /* Positive type repeats                                                 */
148    1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */    1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */
149    3, 3, 3,                       /* Type upto, minupto, exact              */    1+IMM2_SIZE, 1+IMM2_SIZE,      /* Type upto, minupto                     */
150    1, 1, 1, 3,                    /* Type *+, ++, ?+, upto+                 */    1+IMM2_SIZE,                   /* Type exact                             */
151      1, 1, 1, 1+IMM2_SIZE,          /* Type *+, ++, ?+, upto+                 */
152    /* Character class & ref repeats                                         */    /* Character class & ref repeats                                         */
153    0, 0, 0, 0, 0, 0,              /* *, *?, +, +?, ?, ??                    */    0, 0, 0, 0, 0, 0,              /* *, *?, +, +?, ?, ??                    */
154    0, 0,                          /* CRRANGE, CRMINRANGE                    */    0, 0,                          /* CRRANGE, CRMINRANGE                    */
# Line 144  static const uschar coptable[] = { Line 156  static const uschar coptable[] = {
156    0,                             /* NCLASS                                 */    0,                             /* NCLASS                                 */
157    0,                             /* XCLASS - variable length               */    0,                             /* XCLASS - variable length               */
158    0,                             /* REF                                    */    0,                             /* REF                                    */
159      0,                             /* REFI                                   */
160    0,                             /* RECURSE                                */    0,                             /* RECURSE                                */
161    0,                             /* CALLOUT                                */    0,                             /* CALLOUT                                */
162    0,                             /* Alt                                    */    0,                             /* Alt                                    */
163    0,                             /* Ket                                    */    0,                             /* Ket                                    */
164    0,                             /* KetRmax                                */    0,                             /* KetRmax                                */
165    0,                             /* KetRmin                                */    0,                             /* KetRmin                                */
166      0,                             /* KetRpos                                */
167      0,                             /* Reverse                                */
168    0,                             /* Assert                                 */    0,                             /* Assert                                 */
169    0,                             /* Assert not                             */    0,                             /* Assert not                             */
170    0,                             /* Assert behind                          */    0,                             /* Assert behind                          */
171    0,                             /* Assert behind not                      */    0,                             /* Assert behind not                      */
172    0,                             /* Reverse                                */    0, 0,                          /* ONCE, ONCE_NC                          */
173    0, 0, 0, 0,                    /* ONCE, BRA, CBRA, COND                  */    0, 0, 0, 0, 0,                 /* BRA, BRAPOS, CBRA, CBRAPOS, COND       */
174    0, 0, 0,                       /* SBRA, SCBRA, SCOND                     */    0, 0, 0, 0, 0,                 /* SBRA, SBRAPOS, SCBRA, SCBRAPOS, SCOND  */
175    0, 0,                          /* CREF, NCREF                            */    0, 0,                          /* CREF, NCREF                            */
176    0, 0,                          /* RREF, NRREF                            */    0, 0,                          /* RREF, NRREF                            */
177    0,                             /* DEF                                    */    0,                             /* DEF                                    */
178    0, 0,                          /* BRAZERO, BRAMINZERO                    */    0, 0, 0,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */
179    0, 0, 0,                       /* MARK, PRUNE, PRUNE_ARG,                */    0, 0, 0,                       /* MARK, PRUNE, PRUNE_ARG                 */
180    0, 0, 0, 0,                    /* SKIP, SKIP_ARG, THEN, THEN_ARG,        */    0, 0, 0, 0,                    /* SKIP, SKIP_ARG, THEN, THEN_ARG         */
181    0, 0, 0, 0, 0                  /* COMMIT, FAIL, ACCEPT, CLOSE, SKIPZERO  */    0, 0, 0, 0,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */
182      0, 0                           /* CLOSE, SKIPZERO  */
183  };  };
184    
185  /* This table identifies those opcodes that inspect a character. It is used to  /* This table identifies those opcodes that inspect a character. It is used to
# Line 171  remember the fact that a character could Line 187  remember the fact that a character could
187  the subject is reached. ***NOTE*** If the start of this table is modified, the  the subject is reached. ***NOTE*** If the start of this table is modified, the
188  two tables that follow must also be modified. */  two tables that follow must also be modified. */
189    
190  static const uschar poptable[] = {  static const pcre_uint8 poptable[] = {
191    0,                             /* End                                    */    0,                             /* End                                    */
192    0, 0, 0, 1, 1,                 /* \A, \G, \K, \B, \b                     */    0, 0, 0, 1, 1,                 /* \A, \G, \K, \B, \b                     */
193    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */
# Line 179  static const uschar poptable[] = { Line 195  static const uschar poptable[] = {
195    1, 1,                          /* \P, \p                                 */    1, 1,                          /* \P, \p                                 */
196    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */
197    1,                             /* \X                                     */    1,                             /* \X                                     */
198    0, 0, 0, 0, 0,                 /* \Z, \z, Opt, ^, $                      */    0, 0, 0, 0, 0, 0,              /* \Z, \z, ^, ^M, $, $M                   */
199    1,                             /* Char                                   */    1,                             /* Char                                   */
200    1,                             /* Charnc                                 */    1,                             /* Chari                                  */
201    1,                             /* not                                    */    1,                             /* not                                    */
202      1,                             /* noti                                   */
203    /* Positive single-char repeats                                          */    /* Positive single-char repeats                                          */
204    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */
205    1, 1, 1,                       /* upto, minupto, exact                   */    1, 1, 1,                       /* upto, minupto, exact                   */
206    1, 1, 1, 1,                    /* *+, ++, ?+, upto+                      */    1, 1, 1, 1,                    /* *+, ++, ?+, upto+                      */
207      1, 1, 1, 1, 1, 1,              /* *I, *?I, +I, +?I, ?I, ??I              */
208      1, 1, 1,                       /* upto I, minupto I, exact I             */
209      1, 1, 1, 1,                    /* *+I, ++I, ?+I, upto+I                  */
210    /* Negative single-char repeats - only for chars < 256                   */    /* Negative single-char repeats - only for chars < 256                   */
211    1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */    1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */
212    1, 1, 1,                       /* NOT upto, minupto, exact               */    1, 1, 1,                       /* NOT upto, minupto, exact               */
213    1, 1, 1, 1,                    /* NOT *+, ++, ?+, upto+                  */    1, 1, 1, 1,                    /* NOT *+, ++, ?+, upto+                  */
214      1, 1, 1, 1, 1, 1,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */
215      1, 1, 1,                       /* NOT upto I, minupto I, exact I         */
216      1, 1, 1, 1,                    /* NOT *+I, ++I, ?+I, upto+I              */
217    /* Positive type repeats                                                 */    /* Positive type repeats                                                 */
218    1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */    1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */
219    1, 1, 1,                       /* Type upto, minupto, exact              */    1, 1, 1,                       /* Type upto, minupto, exact              */
# Line 202  static const uschar poptable[] = { Line 225  static const uschar poptable[] = {
225    1,                             /* NCLASS                                 */    1,                             /* NCLASS                                 */
226    1,                             /* XCLASS - variable length               */    1,                             /* XCLASS - variable length               */
227    0,                             /* REF                                    */    0,                             /* REF                                    */
228      0,                             /* REFI                                   */
229    0,                             /* RECURSE                                */    0,                             /* RECURSE                                */
230    0,                             /* CALLOUT                                */    0,                             /* CALLOUT                                */
231    0,                             /* Alt                                    */    0,                             /* Alt                                    */
232    0,                             /* Ket                                    */    0,                             /* Ket                                    */
233    0,                             /* KetRmax                                */    0,                             /* KetRmax                                */
234    0,                             /* KetRmin                                */    0,                             /* KetRmin                                */
235      0,                             /* KetRpos                                */
236      0,                             /* Reverse                                */
237    0,                             /* Assert                                 */    0,                             /* Assert                                 */
238    0,                             /* Assert not                             */    0,                             /* Assert not                             */
239    0,                             /* Assert behind                          */    0,                             /* Assert behind                          */
240    0,                             /* Assert behind not                      */    0,                             /* Assert behind not                      */
241    0,                             /* Reverse                                */    0, 0,                          /* ONCE, ONCE_NC                          */
242    0, 0, 0, 0,                    /* ONCE, BRA, CBRA, COND                  */    0, 0, 0, 0, 0,                 /* BRA, BRAPOS, CBRA, CBRAPOS, COND       */
243    0, 0, 0,                       /* SBRA, SCBRA, SCOND                     */    0, 0, 0, 0, 0,                 /* SBRA, SBRAPOS, SCBRA, SCBRAPOS, SCOND  */
244    0, 0,                          /* CREF, NCREF                            */    0, 0,                          /* CREF, NCREF                            */
245    0, 0,                          /* RREF, NRREF                            */    0, 0,                          /* RREF, NRREF                            */
246    0,                             /* DEF                                    */    0,                             /* DEF                                    */
247    0, 0,                          /* BRAZERO, BRAMINZERO                    */    0, 0, 0,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */
248    0, 0, 0,                       /* MARK, PRUNE, PRUNE_ARG,                */    0, 0, 0,                       /* MARK, PRUNE, PRUNE_ARG                 */
249    0, 0, 0, 0,                    /* SKIP, SKIP_ARG, THEN, THEN_ARG,        */    0, 0, 0, 0,                    /* SKIP, SKIP_ARG, THEN, THEN_ARG         */
250    0, 0, 0, 0, 0                  /* COMMIT, FAIL, ACCEPT, CLOSE, SKIPZERO  */    0, 0, 0, 0,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */
251      0, 0                           /* CLOSE, SKIPZERO                        */
252  };  };
253    
254  /* These 2 tables allow for compact code for testing for \D, \d, \S, \s, \W,  /* These 2 tables allow for compact code for testing for \D, \d, \S, \s, \W,
255  and \w */  and \w */
256    
257  static const uschar toptable1[] = {  static const pcre_uint8 toptable1[] = {
258    0, 0, 0, 0, 0, 0,    0, 0, 0, 0, 0, 0,
259    ctype_digit, ctype_digit,    ctype_digit, ctype_digit,
260    ctype_space, ctype_space,    ctype_space, ctype_space,
# Line 235  static const uschar toptable1[] = { Line 262  static const uschar toptable1[] = {
262    0, 0                            /* OP_ANY, OP_ALLANY */    0, 0                            /* OP_ANY, OP_ALLANY */
263  };  };
264    
265  static const uschar toptable2[] = {  static const pcre_uint8 toptable2[] = {
266    0, 0, 0, 0, 0, 0,    0, 0, 0, 0, 0, 0,
267    ctype_digit, 0,    ctype_digit, 0,
268    ctype_space, 0,    ctype_space, 0,
# Line 252  these structures in, is a vector of ints Line 279  these structures in, is a vector of ints
279  typedef struct stateblock {  typedef struct stateblock {
280    int offset;                     /* Offset to opcode */    int offset;                     /* Offset to opcode */
281    int count;                      /* Count for repeats */    int count;                      /* Count for repeats */
   int ims;                        /* ims flag bits */  
282    int data;                       /* Some use extra data */    int data;                       /* Some use extra data */
283  } stateblock;  } stateblock;
284    
# Line 275  Returns:       nothing Line 301  Returns:       nothing
301  */  */
302    
303  static void  static void
304  pchars(unsigned char *p, int length, FILE *f)  pchars(const pcre_uchar *p, int length, FILE *f)
305  {  {
306  int c;  int c;
307  while (length-- > 0)  while (length-- > 0)
# Line 308  Arguments: Line 334  Arguments:
334    offsetcount       size of same    offsetcount       size of same
335    workspace         vector of workspace    workspace         vector of workspace
336    wscount           size of same    wscount           size of same
   ims               the current ims flags  
337    rlevel            function call recursion level    rlevel            function call recursion level
   recursing         regex recursive call level  
338    
339  Returns:            > 0 => number of match offset pairs placed in offsets  Returns:            > 0 => number of match offset pairs placed in offsets
340                      = 0 => offsets overflowed; longest matches are present                      = 0 => offsets overflowed; longest matches are present
# Line 325  for the current character, one for the f Line 349  for the current character, one for the f
349      { \      { \
350      next_active_state->offset = (x); \      next_active_state->offset = (x); \
351      next_active_state->count  = (y); \      next_active_state->count  = (y); \
     next_active_state->ims    = ims; \  
352      next_active_state++; \      next_active_state++; \
353      DPRINTF(("%.*sADD_ACTIVE(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \      DPRINTF(("%.*sADD_ACTIVE(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \
354      } \      } \
# Line 336  for the current character, one for the f Line 359  for the current character, one for the f
359      { \      { \
360      next_active_state->offset = (x); \      next_active_state->offset = (x); \
361      next_active_state->count  = (y); \      next_active_state->count  = (y); \
     next_active_state->ims    = ims; \  
362      next_active_state->data   = (z); \      next_active_state->data   = (z); \
363      next_active_state++; \      next_active_state++; \
364      DPRINTF(("%.*sADD_ACTIVE_DATA(%d,%d,%d)\n", rlevel*2-2, SP, (x), (y), (z))); \      DPRINTF(("%.*sADD_ACTIVE_DATA(%d,%d,%d)\n", rlevel*2-2, SP, (x), (y), (z))); \
# Line 348  for the current character, one for the f Line 370  for the current character, one for the f
370      { \      { \
371      next_new_state->offset = (x); \      next_new_state->offset = (x); \
372      next_new_state->count  = (y); \      next_new_state->count  = (y); \
     next_new_state->ims    = ims; \  
373      next_new_state++; \      next_new_state++; \
374      DPRINTF(("%.*sADD_NEW(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \      DPRINTF(("%.*sADD_NEW(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \
375      } \      } \
# Line 359  for the current character, one for the f Line 380  for the current character, one for the f
380      { \      { \
381      next_new_state->offset = (x); \      next_new_state->offset = (x); \
382      next_new_state->count  = (y); \      next_new_state->count  = (y); \
     next_new_state->ims    = ims; \  
383      next_new_state->data   = (z); \      next_new_state->data   = (z); \
384      next_new_state++; \      next_new_state++; \
385      DPRINTF(("%.*sADD_NEW_DATA(%d,%d,%d)\n", rlevel*2-2, SP, (x), (y), (z))); \      DPRINTF(("%.*sADD_NEW_DATA(%d,%d,%d)\n", rlevel*2-2, SP, (x), (y), (z))); \
# Line 371  for the current character, one for the f Line 391  for the current character, one for the f
391  static int  static int
392  internal_dfa_exec(  internal_dfa_exec(
393    dfa_match_data *md,    dfa_match_data *md,
394    const uschar *this_start_code,    const pcre_uchar *this_start_code,
395    const uschar *current_subject,    const pcre_uchar *current_subject,
396    int start_offset,    int start_offset,
397    int *offsets,    int *offsets,
398    int offsetcount,    int offsetcount,
399    int *workspace,    int *workspace,
400    int wscount,    int wscount,
401    int ims,    int  rlevel)
   int  rlevel,  
   int  recursing)  
402  {  {
403  stateblock *active_states, *new_states, *temp_states;  stateblock *active_states, *new_states, *temp_states;
404  stateblock *next_active_state, *next_new_state;  stateblock *next_active_state, *next_new_state;
405    
406  const uschar *ctypes, *lcc, *fcc;  const pcre_uint8 *ctypes, *lcc, *fcc;
407  const uschar *ptr;  const pcre_uchar *ptr;
408  const uschar *end_code, *first_op;  const pcre_uchar *end_code, *first_op;
409    
410    dfa_recursion_info new_recursive;
411    
412  int active_count, new_count, match_count;  int active_count, new_count, match_count;
413    
414  /* Some fields in the md block are frequently referenced, so we load them into  /* Some fields in the md block are frequently referenced, so we load them into
415  independent variables in the hope that this will perform better. */  independent variables in the hope that this will perform better. */
416    
417  const uschar *start_subject = md->start_subject;  const pcre_uchar *start_subject = md->start_subject;
418  const uschar *end_subject = md->end_subject;  const pcre_uchar *end_subject = md->end_subject;
419  const uschar *start_code = md->start_code;  const pcre_uchar *start_code = md->start_code;
420    
421  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
422  BOOL utf8 = (md->poptions & PCRE_UTF8) != 0;  BOOL utf = (md->poptions & PCRE_UTF8) != 0;
423  #else  #else
424  BOOL utf8 = FALSE;  BOOL utf = FALSE;
425  #endif  #endif
426    
427  rlevel++;  rlevel++;
# Line 412  wscount = (wscount - (wscount % (INTS_PE Line 432  wscount = (wscount - (wscount % (INTS_PE
432            (2 * INTS_PER_STATEBLOCK);            (2 * INTS_PER_STATEBLOCK);
433    
434  DPRINTF(("\n%.*s---------------------\n"  DPRINTF(("\n%.*s---------------------\n"
435    "%.*sCall to internal_dfa_exec f=%d r=%d\n",    "%.*sCall to internal_dfa_exec f=%d\n",
436    rlevel*2-2, SP, rlevel*2-2, SP, rlevel, recursing));    rlevel*2-2, SP, rlevel*2-2, SP, rlevel));
437    
438  ctypes = md->tables + ctypes_offset;  ctypes = md->tables + ctypes_offset;
439  lcc = md->tables + lcc_offset;  lcc = md->tables + lcc_offset;
# Line 426  next_new_state = new_states = active_sta Line 446  next_new_state = new_states = active_sta
446  new_count = 0;  new_count = 0;
447    
448  first_op = this_start_code + 1 + LINK_SIZE +  first_op = this_start_code + 1 + LINK_SIZE +
449    ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA)? 2:0);    ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA ||
450        *this_start_code == OP_CBRAPOS || *this_start_code == OP_SCBRAPOS)
451        ? IMM2_SIZE:0);
452    
453  /* The first thing in any (sub) pattern is a bracket of some sort. Push all  /* The first thing in any (sub) pattern is a bracket of some sort. Push all
454  the alternative states onto the list, and find out where the end is. This  the alternative states onto the list, and find out where the end is. This
# Line 454  if (*first_op == OP_REVERSE) Line 476  if (*first_op == OP_REVERSE)
476    /* If we can't go back the amount required for the longest lookbehind    /* If we can't go back the amount required for the longest lookbehind
477    pattern, go back as far as we can; some alternatives may still be viable. */    pattern, go back as far as we can; some alternatives may still be viable. */
478    
479  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
480    /* In character mode we have to step back character by character */    /* In character mode we have to step back character by character */
481    
482    if (utf8)    if (utf)
483      {      {
484      for (gone_back = 0; gone_back < max_back; gone_back++)      for (gone_back = 0; gone_back < max_back; gone_back++)
485        {        {
486        if (current_subject <= start_subject) break;        if (current_subject <= start_subject) break;
487        current_subject--;        current_subject--;
488        while (current_subject > start_subject &&        ACROSSCHAR(current_subject > start_subject, *current_subject, current_subject--);
              (*current_subject & 0xc0) == 0x80)  
         current_subject--;  
489        }        }
490      }      }
491    else    else
# Line 525  else Line 545  else
545    else    else
546      {      {
547      int length = 1 + LINK_SIZE +      int length = 1 + LINK_SIZE +
548        ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA)? 2:0);        ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA ||
549            *this_start_code == OP_CBRAPOS || *this_start_code == OP_SCBRAPOS)
550            ? IMM2_SIZE:0);
551      do      do
552        {        {
553        ADD_NEW((int)(end_code - start_code + length), 0);        ADD_NEW((int)(end_code - start_code + length), 0);
# Line 538  else Line 560  else
560    
561  workspace[0] = 0;    /* Bit indicating which vector is current */  workspace[0] = 0;    /* Bit indicating which vector is current */
562    
563  DPRINTF(("%.*sEnd state = %d\n", rlevel*2-2, SP, end_code - start_code));  DPRINTF(("%.*sEnd state = %d\n", rlevel*2-2, SP, (int)(end_code - start_code)));
564    
565  /* Loop for scanning the subject */  /* Loop for scanning the subject */
566    
# Line 565  for (;;) Line 587  for (;;)
587    
588  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
589    printf("%.*sNext character: rest of subject = \"", rlevel*2-2, SP);    printf("%.*sNext character: rest of subject = \"", rlevel*2-2, SP);
590    pchars((uschar *)ptr, strlen((char *)ptr), stdout);    pchars(ptr, STRLEN_UC(ptr), stdout);
591    printf("\"\n");    printf("\"\n");
592    
593    printf("%.*sActive states: ", rlevel*2-2, SP);    printf("%.*sActive states: ", rlevel*2-2, SP);
# Line 586  for (;;) Line 608  for (;;)
608    if (ptr < end_subject)    if (ptr < end_subject)
609      {      {
610      clen = 1;        /* Number of bytes in the character */      clen = 1;        /* Number of bytes in the character */
611  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
612      if (utf8) { GETCHARLEN(c, ptr, clen); } else      if (utf) { GETCHARLEN(c, ptr, clen); } else
613  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
614      c = *ptr;      c = *ptr;
615      }      }
616    else    else
# Line 605  for (;;) Line 627  for (;;)
627    for (i = 0; i < active_count; i++)    for (i = 0; i < active_count; i++)
628      {      {
629      stateblock *current_state = active_states + i;      stateblock *current_state = active_states + i;
630      const uschar *code;      BOOL caseless = FALSE;
631        const pcre_uchar *code;
632      int state_offset = current_state->offset;      int state_offset = current_state->offset;
633      int count, codevalue, rrc;      int count, codevalue, rrc;
634    
# Line 616  for (;;) Line 639  for (;;)
639          else printf("0x%02x\n", c);          else printf("0x%02x\n", c);
640  #endif  #endif
641    
     /* This variable is referred to implicity in the ADD_xxx macros. */  
   
     ims = current_state->ims;  
   
642      /* A negative offset is a special case meaning "hold off going to this      /* A negative offset is a special case meaning "hold off going to this
643      (negated) state until the number of characters in the data field have      (negated) state until the number of characters in the data field have
644      been skipped". */      been skipped". */
# Line 678  for (;;) Line 697  for (;;)
697      if (coptable[codevalue] > 0)      if (coptable[codevalue] > 0)
698        {        {
699        dlen = 1;        dlen = 1;
700  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
701        if (utf8) { GETCHARLEN(d, (code + coptable[codevalue]), dlen); } else        if (utf) { GETCHARLEN(d, (code + coptable[codevalue]), dlen); } else
702  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
703        d = code[coptable[codevalue]];        d = code[coptable[codevalue]];
704        if (codevalue >= OP_TYPESTAR)        if (codevalue >= OP_TYPESTAR)
705          {          {
# Line 725  for (;;) Line 744  for (;;)
744    
745  /* ========================================================================== */  /* ========================================================================== */
746        /* Reached a closing bracket. If not at the end of the pattern, carry        /* Reached a closing bracket. If not at the end of the pattern, carry
747        on with the next opcode. Otherwise, unless we have an empty string and        on with the next opcode. For repeating opcodes, also add the repeat
748          state. Note that KETRPOS will always be encountered at the end of the
749          subpattern, because the possessive subpattern repeats are always handled
750          using recursive calls. Thus, it never adds any new states.
751    
752          At the end of the (sub)pattern, unless we have an empty string and
753        PCRE_NOTEMPTY is set, or PCRE_NOTEMPTY_ATSTART is set and we are at the        PCRE_NOTEMPTY is set, or PCRE_NOTEMPTY_ATSTART is set and we are at the
754        start of the subject, save the match data, shifting up all previous        start of the subject, save the match data, shifting up all previous
755        matches so we always have the longest first. */        matches so we always have the longest first. */
# Line 733  for (;;) Line 757  for (;;)
757        case OP_KET:        case OP_KET:
758        case OP_KETRMIN:        case OP_KETRMIN:
759        case OP_KETRMAX:        case OP_KETRMAX:
760          case OP_KETRPOS:
761        if (code != end_code)        if (code != end_code)
762          {          {
763          ADD_ACTIVE(state_offset + 1 + LINK_SIZE, 0);          ADD_ACTIVE(state_offset + 1 + LINK_SIZE, 0);
# Line 749  for (;;) Line 774  for (;;)
774                  current_subject > start_subject + md->start_offset)))                  current_subject > start_subject + md->start_offset)))
775            {            {
776            if (match_count < 0) match_count = (offsetcount >= 2)? 1 : 0;            if (match_count < 0) match_count = (offsetcount >= 2)? 1 : 0;
777              else if (match_count > 0 && ++match_count * 2 >= offsetcount)              else if (match_count > 0 && ++match_count * 2 > offsetcount)
778                match_count = 0;                match_count = 0;
779            count = ((match_count == 0)? offsetcount : match_count * 2) - 2;            count = ((match_count == 0)? offsetcount : match_count * 2) - 2;
780            if (count > 0) memmove(offsets + 2, offsets, count * sizeof(int));            if (count > 0) memmove(offsets + 2, offsets, count * sizeof(int));
# Line 795  for (;;) Line 820  for (;;)
820        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
821        case OP_CBRA:        case OP_CBRA:
822        case OP_SCBRA:        case OP_SCBRA:
823        ADD_ACTIVE((int)(code - start_code + 3 + LINK_SIZE),  0);        ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE + IMM2_SIZE),  0);
824        code += GET(code, 1);        code += GET(code, 1);
825        while (*code == OP_ALT)        while (*code == OP_ALT)
826          {          {
# Line 822  for (;;) Line 847  for (;;)
847    
848        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
849        case OP_CIRC:        case OP_CIRC:
850        if ((ptr == start_subject && (md->moptions & PCRE_NOTBOL) == 0) ||        if (ptr == start_subject && (md->moptions & PCRE_NOTBOL) == 0)
           ((ims & PCRE_MULTILINE) != 0 &&  
             ptr != end_subject &&  
             WAS_NEWLINE(ptr)))  
851          { ADD_ACTIVE(state_offset + 1, 0); }          { ADD_ACTIVE(state_offset + 1, 0); }
852        break;        break;
853    
854        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
855        case OP_EOD:        case OP_CIRCM:
856        if (ptr >= end_subject) { ADD_ACTIVE(state_offset + 1, 0); }        if ((ptr == start_subject && (md->moptions & PCRE_NOTBOL) == 0) ||
857              (ptr != end_subject && WAS_NEWLINE(ptr)))
858            { ADD_ACTIVE(state_offset + 1, 0); }
859        break;        break;
860    
861        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
862        case OP_OPT:        case OP_EOD:
863        ims = code[1];        if (ptr >= end_subject)
864        ADD_ACTIVE(state_offset + 2, 0);          {
865            if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
866              could_continue = TRUE;
867            else { ADD_ACTIVE(state_offset + 1, 0); }
868            }
869        break;        break;
870    
871        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
# Line 871  for (;;) Line 899  for (;;)
899    
900        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
901        case OP_EODN:        case OP_EODN:
902        if (clen == 0 || (IS_NEWLINE(ptr) && ptr == end_subject - md->nllen))        if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
903            could_continue = TRUE;
904          else if (clen == 0 || (IS_NEWLINE(ptr) && ptr == end_subject - md->nllen))
905          { ADD_ACTIVE(state_offset + 1, 0); }          { ADD_ACTIVE(state_offset + 1, 0); }
906        break;        break;
907    
# Line 879  for (;;) Line 909  for (;;)
909        case OP_DOLL:        case OP_DOLL:
910        if ((md->moptions & PCRE_NOTEOL) == 0)        if ((md->moptions & PCRE_NOTEOL) == 0)
911          {          {
912          if (clen == 0 ||          if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
913              could_continue = TRUE;
914            else if (clen == 0 ||
915              ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr) &&              ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr) &&
916                 ((ims & PCRE_MULTILINE) != 0 || ptr == end_subject - md->nllen)                 (ptr == end_subject - md->nllen)
917              ))              ))
918            { ADD_ACTIVE(state_offset + 1, 0); }            { ADD_ACTIVE(state_offset + 1, 0); }
919          }          }
920        else if ((ims & PCRE_MULTILINE) != 0 && IS_NEWLINE(ptr))        break;
921    
922          /*-----------------------------------------------------------------*/
923          case OP_DOLLM:
924          if ((md->moptions & PCRE_NOTEOL) == 0)
925            {
926            if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
927              could_continue = TRUE;
928            else if (clen == 0 ||
929                ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr)))
930              { ADD_ACTIVE(state_offset + 1, 0); }
931            }
932          else if (IS_NEWLINE(ptr))
933          { ADD_ACTIVE(state_offset + 1, 0); }          { ADD_ACTIVE(state_offset + 1, 0); }
934        break;        break;
935    
# Line 916  for (;;) Line 960  for (;;)
960    
961          if (ptr > start_subject)          if (ptr > start_subject)
962            {            {
963            const uschar *temp = ptr - 1;            const pcre_uchar *temp = ptr - 1;
964            if (temp < md->start_used_ptr) md->start_used_ptr = temp;            if (temp < md->start_used_ptr) md->start_used_ptr = temp;
965  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
966            if (utf8) BACKCHAR(temp);            if (utf) { BACKCHAR(temp); }
967  #endif  #endif
968            GETCHARTEST(d, temp);            GETCHARTEST(d, temp);
969  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
# Line 984  for (;;) Line 1028  for (;;)
1028            break;            break;
1029    
1030            case PT_GC:            case PT_GC:
1031            OK = _pcre_ucp_gentype[prop->chartype] == code[2];            OK = PRIV(ucp_gentype)[prop->chartype] == code[2];
1032            break;            break;
1033    
1034            case PT_PC:            case PT_PC:
# Line 998  for (;;) Line 1042  for (;;)
1042            /* These are specials for combination cases. */            /* These are specials for combination cases. */
1043    
1044            case PT_ALNUM:            case PT_ALNUM:
1045            OK = _pcre_ucp_gentype[prop->chartype] == ucp_L ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1046                 _pcre_ucp_gentype[prop->chartype] == ucp_N;                 PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1047            break;            break;
1048    
1049            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
1050            OK = _pcre_ucp_gentype[prop->chartype] == ucp_Z ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1051                 c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;                 c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1052            break;            break;
1053    
1054            case PT_PXSPACE:  /* POSIX space */            case PT_PXSPACE:  /* POSIX space */
1055            OK = _pcre_ucp_gentype[prop->chartype] == ucp_Z ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1056                 c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||                 c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1057                 c == CHAR_FF || c == CHAR_CR;                 c == CHAR_FF || c == CHAR_CR;
1058            break;            break;
1059    
1060            case PT_WORD:            case PT_WORD:
1061            OK = _pcre_ucp_gentype[prop->chartype] == ucp_L ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1062                 _pcre_ucp_gentype[prop->chartype] == ucp_N ||                 PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1063                 c == CHAR_UNDERSCORE;                 c == CHAR_UNDERSCORE;
1064            break;            break;
1065    
# Line 1117  for (;;) Line 1161  for (;;)
1161                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
1162            {            {
1163            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1164              { ADD_NEW(state_offset + 4, 0); }              { ADD_NEW(state_offset + 1 + IMM2_SIZE + 1, 0); }
1165            else            else
1166              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
1167            }            }
# Line 1128  for (;;) Line 1172  for (;;)
1172        case OP_TYPEUPTO:        case OP_TYPEUPTO:
1173        case OP_TYPEMINUPTO:        case OP_TYPEMINUPTO:
1174        case OP_TYPEPOSUPTO:        case OP_TYPEPOSUPTO:
1175        ADD_ACTIVE(state_offset + 4, 0);        ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0);
1176        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1177        if (clen > 0)        if (clen > 0)
1178          {          {
# Line 1143  for (;;) Line 1187  for (;;)
1187              next_active_state--;              next_active_state--;
1188              }              }
1189            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1190              { ADD_NEW(state_offset + 4, 0); }              { ADD_NEW(state_offset + 2 + IMM2_SIZE, 0); }
1191            else            else
1192              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
1193            }            }
# Line 1178  for (;;) Line 1222  for (;;)
1222            break;            break;
1223    
1224            case PT_GC:            case PT_GC:
1225            OK = _pcre_ucp_gentype[prop->chartype] == code[3];            OK = PRIV(ucp_gentype)[prop->chartype] == code[3];
1226            break;            break;
1227    
1228            case PT_PC:            case PT_PC:
# Line 1192  for (;;) Line 1236  for (;;)
1236            /* These are specials for combination cases. */            /* These are specials for combination cases. */
1237    
1238            case PT_ALNUM:            case PT_ALNUM:
1239            OK = _pcre_ucp_gentype[prop->chartype] == ucp_L ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1240                 _pcre_ucp_gentype[prop->chartype] == ucp_N;                 PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1241            break;            break;
1242    
1243            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
1244            OK = _pcre_ucp_gentype[prop->chartype] == ucp_Z ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1245                 c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;                 c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1246            break;            break;
1247    
1248            case PT_PXSPACE:  /* POSIX space */            case PT_PXSPACE:  /* POSIX space */
1249            OK = _pcre_ucp_gentype[prop->chartype] == ucp_Z ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1250                 c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||                 c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1251                 c == CHAR_FF || c == CHAR_CR;                 c == CHAR_FF || c == CHAR_CR;
1252            break;            break;
1253    
1254            case PT_WORD:            case PT_WORD:
1255            OK = _pcre_ucp_gentype[prop->chartype] == ucp_L ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1256                 _pcre_ucp_gentype[prop->chartype] == ucp_N ||                 PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1257                 c == CHAR_UNDERSCORE;                 c == CHAR_UNDERSCORE;
1258            break;            break;
1259    
# Line 1241  for (;;) Line 1285  for (;;)
1285        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
1286        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1287          {          {
1288          const uschar *nptr = ptr + clen;          const pcre_uchar *nptr = ptr + clen;
1289          int ncount = 0;          int ncount = 0;
1290          if (count > 0 && codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS)          if (count > 0 && codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS)
1291            {            {
# Line 1425  for (;;) Line 1469  for (;;)
1469            break;            break;
1470    
1471            case PT_GC:            case PT_GC:
1472            OK = _pcre_ucp_gentype[prop->chartype] == code[3];            OK = PRIV(ucp_gentype)[prop->chartype] == code[3];
1473            break;            break;
1474    
1475            case PT_PC:            case PT_PC:
# Line 1439  for (;;) Line 1483  for (;;)
1483            /* These are specials for combination cases. */            /* These are specials for combination cases. */
1484    
1485            case PT_ALNUM:            case PT_ALNUM:
1486            OK = _pcre_ucp_gentype[prop->chartype] == ucp_L ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1487                 _pcre_ucp_gentype[prop->chartype] == ucp_N;                 PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1488            break;            break;
1489    
1490            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
1491            OK = _pcre_ucp_gentype[prop->chartype] == ucp_Z ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1492                 c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;                 c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1493            break;            break;
1494    
1495            case PT_PXSPACE:  /* POSIX space */            case PT_PXSPACE:  /* POSIX space */
1496            OK = _pcre_ucp_gentype[prop->chartype] == ucp_Z ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1497                 c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||                 c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1498                 c == CHAR_FF || c == CHAR_CR;                 c == CHAR_FF || c == CHAR_CR;
1499            break;            break;
1500    
1501            case PT_WORD:            case PT_WORD:
1502            OK = _pcre_ucp_gentype[prop->chartype] == ucp_L ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1503                 _pcre_ucp_gentype[prop->chartype] == ucp_N ||                 PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1504                 c == CHAR_UNDERSCORE;                 c == CHAR_UNDERSCORE;
1505            break;            break;
1506    
# Line 1497  for (;;) Line 1541  for (;;)
1541        ADD_ACTIVE(state_offset + 2, 0);        ADD_ACTIVE(state_offset + 2, 0);
1542        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1543          {          {
1544          const uschar *nptr = ptr + clen;          const pcre_uchar *nptr = ptr + clen;
1545          int ncount = 0;          int ncount = 0;
1546          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSSTAR ||          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSSTAR ||
1547              codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSQUERY)              codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSQUERY)
# Line 1679  for (;;) Line 1723  for (;;)
1723        case OP_PROP_EXTRA + OP_TYPEMINUPTO:        case OP_PROP_EXTRA + OP_TYPEMINUPTO:
1724        case OP_PROP_EXTRA + OP_TYPEPOSUPTO:        case OP_PROP_EXTRA + OP_TYPEPOSUPTO:
1725        if (codevalue != OP_PROP_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_PROP_EXTRA + OP_TYPEEXACT)
1726          { ADD_ACTIVE(state_offset + 6, 0); }          { ADD_ACTIVE(state_offset + 1 + IMM2_SIZE + 3, 0); }
1727        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1728        if (clen > 0)        if (clen > 0)
1729          {          {
1730          BOOL OK;          BOOL OK;
1731          const ucd_record * prop = GET_UCD(c);          const ucd_record * prop = GET_UCD(c);
1732          switch(code[4])          switch(code[1 + IMM2_SIZE + 1])
1733            {            {
1734            case PT_ANY:            case PT_ANY:
1735            OK = TRUE;            OK = TRUE;
# Line 1697  for (;;) Line 1741  for (;;)
1741            break;            break;
1742    
1743            case PT_GC:            case PT_GC:
1744            OK = _pcre_ucp_gentype[prop->chartype] == code[5];            OK = PRIV(ucp_gentype)[prop->chartype] == code[1 + IMM2_SIZE + 2];
1745            break;            break;
1746    
1747            case PT_PC:            case PT_PC:
1748            OK = prop->chartype == code[5];            OK = prop->chartype == code[1 + IMM2_SIZE + 2];
1749            break;            break;
1750    
1751            case PT_SC:            case PT_SC:
1752            OK = prop->script == code[5];            OK = prop->script == code[1 + IMM2_SIZE + 2];
1753            break;            break;
1754    
1755            /* These are specials for combination cases. */            /* These are specials for combination cases. */
1756    
1757            case PT_ALNUM:            case PT_ALNUM:
1758            OK = _pcre_ucp_gentype[prop->chartype] == ucp_L ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1759                 _pcre_ucp_gentype[prop->chartype] == ucp_N;                 PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1760            break;            break;
1761    
1762            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
1763            OK = _pcre_ucp_gentype[prop->chartype] == ucp_Z ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1764                 c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;                 c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1765            break;            break;
1766    
1767            case PT_PXSPACE:  /* POSIX space */            case PT_PXSPACE:  /* POSIX space */
1768            OK = _pcre_ucp_gentype[prop->chartype] == ucp_Z ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1769                 c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||                 c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1770                 c == CHAR_FF || c == CHAR_CR;                 c == CHAR_FF || c == CHAR_CR;
1771            break;            break;
1772    
1773            case PT_WORD:            case PT_WORD:
1774            OK = _pcre_ucp_gentype[prop->chartype] == ucp_L ||            OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1775                 _pcre_ucp_gentype[prop->chartype] == ucp_N ||                 PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1776                 c == CHAR_UNDERSCORE;                 c == CHAR_UNDERSCORE;
1777            break;            break;
1778    
# Line 1747  for (;;) Line 1791  for (;;)
1791              next_active_state--;              next_active_state--;
1792              }              }
1793            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1794              { ADD_NEW(state_offset + 6, 0); }              { ADD_NEW(state_offset + 1 + IMM2_SIZE + 3, 0); }
1795            else            else
1796              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
1797            }            }
# Line 1760  for (;;) Line 1804  for (;;)
1804        case OP_EXTUNI_EXTRA + OP_TYPEMINUPTO:        case OP_EXTUNI_EXTRA + OP_TYPEMINUPTO:
1805        case OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO:        case OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO:
1806        if (codevalue != OP_EXTUNI_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_EXTUNI_EXTRA + OP_TYPEEXACT)
1807          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
1808        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1809        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1810          {          {
1811          const uschar *nptr = ptr + clen;          const pcre_uchar *nptr = ptr + clen;
1812          int ncount = 0;          int ncount = 0;
1813          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO)          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO)
1814            {            {
# Line 1781  for (;;) Line 1825  for (;;)
1825            nptr += ndlen;            nptr += ndlen;
1826            }            }
1827          if (++count >= GET2(code, 1))          if (++count >= GET2(code, 1))
1828            { ADD_NEW_DATA(-(state_offset + 4), 0, ncount); }            { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, ncount); }
1829          else          else
1830            { ADD_NEW_DATA(-state_offset, count, ncount); }            { ADD_NEW_DATA(-state_offset, count, ncount); }
1831          }          }
# Line 1794  for (;;) Line 1838  for (;;)
1838        case OP_ANYNL_EXTRA + OP_TYPEMINUPTO:        case OP_ANYNL_EXTRA + OP_TYPEMINUPTO:
1839        case OP_ANYNL_EXTRA + OP_TYPEPOSUPTO:        case OP_ANYNL_EXTRA + OP_TYPEPOSUPTO:
1840        if (codevalue != OP_ANYNL_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_ANYNL_EXTRA + OP_TYPEEXACT)
1841          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
1842        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1843        if (clen > 0)        if (clen > 0)
1844          {          {
# Line 1821  for (;;) Line 1865  for (;;)
1865              next_active_state--;              next_active_state--;
1866              }              }
1867            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1868              { ADD_NEW_DATA(-(state_offset + 4), 0, ncount); }              { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, ncount); }
1869            else            else
1870              { ADD_NEW_DATA(-state_offset, count, ncount); }              { ADD_NEW_DATA(-state_offset, count, ncount); }
1871            break;            break;
# Line 1838  for (;;) Line 1882  for (;;)
1882        case OP_VSPACE_EXTRA + OP_TYPEMINUPTO:        case OP_VSPACE_EXTRA + OP_TYPEMINUPTO:
1883        case OP_VSPACE_EXTRA + OP_TYPEPOSUPTO:        case OP_VSPACE_EXTRA + OP_TYPEPOSUPTO:
1884        if (codevalue != OP_VSPACE_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_VSPACE_EXTRA + OP_TYPEEXACT)
1885          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
1886        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1887        if (clen > 0)        if (clen > 0)
1888          {          {
# Line 1867  for (;;) Line 1911  for (;;)
1911              next_active_state--;              next_active_state--;
1912              }              }
1913            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1914              { ADD_NEW_DATA(-(state_offset + 4), 0, 0); }              { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, 0); }
1915            else            else
1916              { ADD_NEW_DATA(-state_offset, count, 0); }              { ADD_NEW_DATA(-state_offset, count, 0); }
1917            }            }
# Line 1880  for (;;) Line 1924  for (;;)
1924        case OP_HSPACE_EXTRA + OP_TYPEMINUPTO:        case OP_HSPACE_EXTRA + OP_TYPEMINUPTO:
1925        case OP_HSPACE_EXTRA + OP_TYPEPOSUPTO:        case OP_HSPACE_EXTRA + OP_TYPEPOSUPTO:
1926        if (codevalue != OP_HSPACE_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_HSPACE_EXTRA + OP_TYPEEXACT)
1927          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
1928        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1929        if (clen > 0)        if (clen > 0)
1930          {          {
# Line 1922  for (;;) Line 1966  for (;;)
1966              next_active_state--;              next_active_state--;
1967              }              }
1968            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1969              { ADD_NEW_DATA(-(state_offset + 4), 0, 0); }              { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, 0); }
1970            else            else
1971              { ADD_NEW_DATA(-state_offset, count, 0); }              { ADD_NEW_DATA(-state_offset, count, 0); }
1972            }            }
# Line 1941  for (;;) Line 1985  for (;;)
1985        break;        break;
1986    
1987        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
1988        case OP_CHARNC:        case OP_CHARI:
1989        if (clen == 0) break;        if (clen == 0) break;
1990    
1991  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
1992        if (utf8)        if (utf)
1993          {          {
1994          if (c == d) { ADD_NEW(state_offset + dlen + 1, 0); } else          if (c == d) { ADD_NEW(state_offset + dlen + 1, 0); } else
1995            {            {
1996            unsigned int othercase;            unsigned int othercase;
1997            if (c < 128) othercase = fcc[c]; else            if (c < 128)
1998                othercase = fcc[c];
1999            /* If we have Unicode property support, we can use it to test the            else
2000            other case of the character. */              /* If we have Unicode property support, we can use it to test the
2001                other case of the character. */
2002  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2003            othercase = UCD_OTHERCASE(c);              othercase = UCD_OTHERCASE(c);
2004  #else  #else
2005            othercase = NOTACHAR;              othercase = NOTACHAR;
2006  #endif  #endif
2007    
2008            if (d == othercase) { ADD_NEW(state_offset + dlen + 1, 0); }            if (d == othercase) { ADD_NEW(state_offset + dlen + 1, 0); }
2009            }            }
2010          }          }
2011        else        else
2012  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2013          /* Not UTF mode */
       /* Non-UTF-8 mode */  
2014          {          {
2015          if (lcc[c] == lcc[d]) { ADD_NEW(state_offset + 2, 0); }          if (TABLE_GET(c, lcc, c) == TABLE_GET(d, lcc, d))
2016              { ADD_NEW(state_offset + 2, 0); }
2017          }          }
2018        break;        break;
2019    
# Line 1983  for (;;) Line 2027  for (;;)
2027        case OP_EXTUNI:        case OP_EXTUNI:
2028        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
2029          {          {
2030          const uschar *nptr = ptr + clen;          const pcre_uchar *nptr = ptr + clen;
2031          int ncount = 0;          int ncount = 0;
2032          while (nptr < end_subject)          while (nptr < end_subject)
2033            {            {
# Line 2127  for (;;) Line 2171  for (;;)
2171        break;        break;
2172    
2173        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2174        /* Match a negated single character. This is only used for one-byte        /* Match a negated single character casefully. This is only used for
2175        characters, that is, we know that d < 256. The character we are        one-byte characters, that is, we know that d < 256. The character we are
2176        checking (c) can be multibyte. */        checking (c) can be multibyte. */
2177    
2178        case OP_NOT:        case OP_NOT:
2179        if (clen > 0)        if (clen > 0 && c != d) { ADD_NEW(state_offset + dlen + 1, 0); }
2180          {        break;
2181          unsigned int otherd = ((ims & PCRE_CASELESS) != 0)? fcc[d] : d;  
2182          if (c != d && c != otherd) { ADD_NEW(state_offset + dlen + 1, 0); }        /*-----------------------------------------------------------------*/
2183          }        /* Match a negated single character caselessly. This is only used for
2184          one-byte characters, that is, we know that d < 256. The character we are
2185          checking (c) can be multibyte. */
2186    
2187          case OP_NOTI:
2188          if (clen > 0 && c != d && c != fcc[d])
2189            { ADD_NEW(state_offset + dlen + 1, 0); }
2190        break;        break;
2191    
2192        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2193          case OP_PLUSI:
2194          case OP_MINPLUSI:
2195          case OP_POSPLUSI:
2196          case OP_NOTPLUSI:
2197          case OP_NOTMINPLUSI:
2198          case OP_NOTPOSPLUSI:
2199          caseless = TRUE;
2200          codevalue -= OP_STARI - OP_STAR;
2201    
2202          /* Fall through */
2203        case OP_PLUS:        case OP_PLUS:
2204        case OP_MINPLUS:        case OP_MINPLUS:
2205        case OP_POSPLUS:        case OP_POSPLUS:
# Line 2151  for (;;) Line 2211  for (;;)
2211        if (clen > 0)        if (clen > 0)
2212          {          {
2213          unsigned int otherd = NOTACHAR;          unsigned int otherd = NOTACHAR;
2214          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2215            {            {
2216  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2217            if (utf8 && d >= 128)            if (utf && d >= 128)
2218              {              {
2219  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2220              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2221  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2222              }              }
2223            else            else
2224  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2225            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2226            }            }
2227          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2228            {            {
# Line 2179  for (;;) Line 2239  for (;;)
2239        break;        break;
2240    
2241        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2242          case OP_QUERYI:
2243          case OP_MINQUERYI:
2244          case OP_POSQUERYI:
2245          case OP_NOTQUERYI:
2246          case OP_NOTMINQUERYI:
2247          case OP_NOTPOSQUERYI:
2248          caseless = TRUE;
2249          codevalue -= OP_STARI - OP_STAR;
2250          /* Fall through */
2251        case OP_QUERY:        case OP_QUERY:
2252        case OP_MINQUERY:        case OP_MINQUERY:
2253        case OP_POSQUERY:        case OP_POSQUERY:
# Line 2189  for (;;) Line 2258  for (;;)
2258        if (clen > 0)        if (clen > 0)
2259          {          {
2260          unsigned int otherd = NOTACHAR;          unsigned int otherd = NOTACHAR;
2261          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2262            {            {
2263  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2264            if (utf8 && d >= 128)            if (utf && d >= 128)
2265              {              {
2266  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2267              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2268  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2269              }              }
2270            else            else
2271  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2272            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2273            }            }
2274          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2275            {            {
# Line 2215  for (;;) Line 2284  for (;;)
2284        break;        break;
2285    
2286        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2287          case OP_STARI:
2288          case OP_MINSTARI:
2289          case OP_POSSTARI:
2290          case OP_NOTSTARI:
2291          case OP_NOTMINSTARI:
2292          case OP_NOTPOSSTARI:
2293          caseless = TRUE;
2294          codevalue -= OP_STARI - OP_STAR;
2295          /* Fall through */
2296        case OP_STAR:        case OP_STAR:
2297        case OP_MINSTAR:        case OP_MINSTAR:
2298        case OP_POSSTAR:        case OP_POSSTAR:
# Line 2225  for (;;) Line 2303  for (;;)
2303        if (clen > 0)        if (clen > 0)
2304          {          {
2305          unsigned int otherd = NOTACHAR;          unsigned int otherd = NOTACHAR;
2306          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2307            {            {
2308  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2309            if (utf8 && d >= 128)            if (utf && d >= 128)
2310              {              {
2311  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2312              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2313  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2314              }              }
2315            else            else
2316  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2317            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2318            }            }
2319          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2320            {            {
# Line 2251  for (;;) Line 2329  for (;;)
2329        break;        break;
2330    
2331        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2332          case OP_EXACTI:
2333          case OP_NOTEXACTI:
2334          caseless = TRUE;
2335          codevalue -= OP_STARI - OP_STAR;
2336          /* Fall through */
2337        case OP_EXACT:        case OP_EXACT:
2338        case OP_NOTEXACT:        case OP_NOTEXACT:
2339        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
2340        if (clen > 0)        if (clen > 0)
2341          {          {
2342          unsigned int otherd = NOTACHAR;          unsigned int otherd = NOTACHAR;
2343          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2344            {            {
2345  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2346            if (utf8 && d >= 128)            if (utf && d >= 128)
2347              {              {
2348  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2349              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2350  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2351              }              }
2352            else            else
2353  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2354            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2355            }            }
2356          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2357            {            {
2358            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
2359              { ADD_NEW(state_offset + dlen + 3, 0); }              { ADD_NEW(state_offset + dlen + 1 + IMM2_SIZE, 0); }
2360            else            else
2361              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
2362            }            }
# Line 2281  for (;;) Line 2364  for (;;)
2364        break;        break;
2365    
2366        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2367          case OP_UPTOI:
2368          case OP_MINUPTOI:
2369          case OP_POSUPTOI:
2370          case OP_NOTUPTOI:
2371          case OP_NOTMINUPTOI:
2372          case OP_NOTPOSUPTOI:
2373          caseless = TRUE;
2374          codevalue -= OP_STARI - OP_STAR;
2375          /* Fall through */
2376        case OP_UPTO:        case OP_UPTO:
2377        case OP_MINUPTO:        case OP_MINUPTO:
2378        case OP_POSUPTO:        case OP_POSUPTO:
2379        case OP_NOTUPTO:        case OP_NOTUPTO:
2380        case OP_NOTMINUPTO:        case OP_NOTMINUPTO:
2381        case OP_NOTPOSUPTO:        case OP_NOTPOSUPTO:
2382        ADD_ACTIVE(state_offset + dlen + 3, 0);        ADD_ACTIVE(state_offset + dlen + 1 + IMM2_SIZE, 0);
2383        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
2384        if (clen > 0)        if (clen > 0)
2385          {          {
2386          unsigned int otherd = NOTACHAR;          unsigned int otherd = NOTACHAR;
2387          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2388            {            {
2389  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2390            if (utf8 && d >= 128)            if (utf && d >= 128)
2391              {              {
2392  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2393              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2394  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2395              }              }
2396            else            else
2397  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2398            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2399            }            }
2400          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2401            {            {
# Line 2313  for (;;) Line 2405  for (;;)
2405              next_active_state--;              next_active_state--;
2406              }              }
2407            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
2408              { ADD_NEW(state_offset + dlen + 3, 0); }              { ADD_NEW(state_offset + dlen + 1 + IMM2_SIZE, 0); }
2409            else            else
2410              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
2411            }            }
# Line 2330  for (;;) Line 2422  for (;;)
2422          {          {
2423          BOOL isinclass = FALSE;          BOOL isinclass = FALSE;
2424          int next_state_offset;          int next_state_offset;
2425          const uschar *ecode;          const pcre_uchar *ecode;
2426    
2427          /* For a simple class, there is always just a 32-byte table, and we          /* For a simple class, there is always just a 32-byte table, and we
2428          can set isinclass from it. */          can set isinclass from it. */
2429    
2430          if (codevalue != OP_XCLASS)          if (codevalue != OP_XCLASS)
2431            {            {
2432            ecode = code + 33;            ecode = code + 1 + (32 / sizeof(pcre_uchar));
2433            if (clen > 0)            if (clen > 0)
2434              {              {
2435              isinclass = (c > 255)? (codevalue == OP_NCLASS) :              isinclass = (c > 255)? (codevalue == OP_NCLASS) :
2436                ((code[1 + c/8] & (1 << (c&7))) != 0);                ((((pcre_uint8 *)(code + 1))[c/8] & (1 << (c&7))) != 0);
2437              }              }
2438            }            }
2439    
# Line 2352  for (;;) Line 2444  for (;;)
2444          else          else
2445           {           {
2446           ecode = code + GET(code, 1);           ecode = code + GET(code, 1);
2447           if (clen > 0) isinclass = _pcre_xclass(c, code + 1 + LINK_SIZE);           if (clen > 0) isinclass = PRIV(xclass)(c, code + 1 + LINK_SIZE, utf);
2448           }           }
2449    
2450          /* At this point, isinclass is set for all kinds of class, and ecode          /* At this point, isinclass is set for all kinds of class, and ecode
# Line 2386  for (;;) Line 2478  for (;;)
2478            case OP_CRMINRANGE:            case OP_CRMINRANGE:
2479            count = current_state->count;  /* Already matched */            count = current_state->count;  /* Already matched */
2480            if (count >= GET2(ecode, 1))            if (count >= GET2(ecode, 1))
2481              { ADD_ACTIVE(next_state_offset + 5, 0); }              { ADD_ACTIVE(next_state_offset + 1 + 2 * IMM2_SIZE, 0); }
2482            if (isinclass)            if (isinclass)
2483              {              {
2484              int max = GET2(ecode, 3);              int max = GET2(ecode, 1 + IMM2_SIZE);
2485              if (++count >= max && max != 0)   /* Max 0 => no limit */              if (++count >= max && max != 0)   /* Max 0 => no limit */
2486                { ADD_NEW(next_state_offset + 5, 0); }                { ADD_NEW(next_state_offset + 1 + 2 * IMM2_SIZE, 0); }
2487              else              else
2488                { ADD_NEW(state_offset, count); }                { ADD_NEW(state_offset, count); }
2489              }              }
# Line 2422  for (;;) Line 2514  for (;;)
2514          int rc;          int rc;
2515          int local_offsets[2];          int local_offsets[2];
2516          int local_workspace[1000];          int local_workspace[1000];
2517          const uschar *endasscode = code + GET(code, 1);          const pcre_uchar *endasscode = code + GET(code, 1);
2518    
2519          while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);          while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);
2520    
# Line 2435  for (;;) Line 2527  for (;;)
2527            sizeof(local_offsets)/sizeof(int),    /* size of same */            sizeof(local_offsets)/sizeof(int),    /* size of same */
2528            local_workspace,                      /* workspace vector */            local_workspace,                      /* workspace vector */
2529            sizeof(local_workspace)/sizeof(int),  /* size of same */            sizeof(local_workspace)/sizeof(int),  /* size of same */
2530            ims,                                  /* the current ims flags */            rlevel);                              /* function recursion level */
           rlevel,                               /* function recursion level */  
           recursing);                           /* pass on regex recursion */  
2531    
2532          if (rc == PCRE_ERROR_DFA_UITEM) return rc;          if (rc == PCRE_ERROR_DFA_UITEM) return rc;
2533          if ((rc >= 0) == (codevalue == OP_ASSERT || codevalue == OP_ASSERTBACK))          if ((rc >= 0) == (codevalue == OP_ASSERT || codevalue == OP_ASSERTBACK))
# Line 2461  for (;;) Line 2551  for (;;)
2551          if (code[LINK_SIZE+1] == OP_CALLOUT)          if (code[LINK_SIZE+1] == OP_CALLOUT)
2552            {            {
2553            rrc = 0;            rrc = 0;
2554            if (pcre_callout != NULL)            if (PUBL(callout) != NULL)
2555              {              {
2556              pcre_callout_block cb;              PUBL(callout_block) cb;
2557              cb.version          = 1;   /* Version 1 of the callout block */              cb.version          = 1;   /* Version 1 of the callout block */
2558              cb.callout_number   = code[LINK_SIZE+2];              cb.callout_number   = code[LINK_SIZE+2];
2559              cb.offset_vector    = offsets;              cb.offset_vector    = offsets;
2560    #ifdef COMPILE_PCRE8
2561              cb.subject          = (PCRE_SPTR)start_subject;              cb.subject          = (PCRE_SPTR)start_subject;
2562    #else
2563                cb.subject          = (PCRE_SPTR16)start_subject;
2564    #endif
2565              cb.subject_length   = (int)(end_subject - start_subject);              cb.subject_length   = (int)(end_subject - start_subject);
2566              cb.start_match      = (int)(current_subject - start_subject);              cb.start_match      = (int)(current_subject - start_subject);
2567              cb.current_position = (int)(ptr - start_subject);              cb.current_position = (int)(ptr - start_subject);
# Line 2476  for (;;) Line 2570  for (;;)
2570              cb.capture_top      = 1;              cb.capture_top      = 1;
2571              cb.capture_last     = -1;              cb.capture_last     = -1;
2572              cb.callout_data     = md->callout_data;              cb.callout_data     = md->callout_data;
2573              if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc;   /* Abandon */              cb.mark             = NULL;   /* No (*MARK) support */
2574                if ((rrc = (*PUBL(callout))(&cb)) < 0) return rrc;   /* Abandon */
2575              }              }
2576            if (rrc > 0) break;                      /* Fail this thread */            if (rrc > 0) break;                      /* Fail this thread */
2577            code += _pcre_OP_lengths[OP_CALLOUT];    /* Skip callout data */            code += PRIV(OP_lengths)[OP_CALLOUT];    /* Skip callout data */
2578            }            }
2579    
2580          condcode = code[LINK_SIZE+1];          condcode = code[LINK_SIZE+1];
# Line 2500  for (;;) Line 2595  for (;;)
2595    
2596          else if (condcode == OP_RREF || condcode == OP_NRREF)          else if (condcode == OP_RREF || condcode == OP_NRREF)
2597            {            {
2598            int value = GET2(code, LINK_SIZE+2);            int value = GET2(code, LINK_SIZE + 2);
2599            if (value != RREF_ANY) return PCRE_ERROR_DFA_UCOND;            if (value != RREF_ANY) return PCRE_ERROR_DFA_UCOND;
2600            if (recursing > 0)            if (md->recursive != NULL)
2601              { ADD_ACTIVE(state_offset + LINK_SIZE + 4, 0); }              { ADD_ACTIVE(state_offset + LINK_SIZE + 2 + IMM2_SIZE, 0); }
2602            else { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }            else { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
2603            }            }
2604    
# Line 2512  for (;;) Line 2607  for (;;)
2607          else          else
2608            {            {
2609            int rc;            int rc;
2610            const uschar *asscode = code + LINK_SIZE + 1;            const pcre_uchar *asscode = code + LINK_SIZE + 1;
2611            const uschar *endasscode = asscode + GET(asscode, 1);            const pcre_uchar *endasscode = asscode + GET(asscode, 1);
2612    
2613            while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);            while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);
2614    
# Line 2526  for (;;) Line 2621  for (;;)
2621              sizeof(local_offsets)/sizeof(int),    /* size of same */              sizeof(local_offsets)/sizeof(int),    /* size of same */
2622              local_workspace,                      /* workspace vector */              local_workspace,                      /* workspace vector */
2623              sizeof(local_workspace)/sizeof(int),  /* size of same */              sizeof(local_workspace)/sizeof(int),  /* size of same */
2624              ims,                                  /* the current ims flags */              rlevel);                              /* function recursion level */
             rlevel,                               /* function recursion level */  
             recursing);                           /* pass on regex recursion */  
2625    
2626            if (rc == PCRE_ERROR_DFA_UITEM) return rc;            if (rc == PCRE_ERROR_DFA_UITEM) return rc;
2627            if ((rc >= 0) ==            if ((rc >= 0) ==
# Line 2543  for (;;) Line 2636  for (;;)
2636        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2637        case OP_RECURSE:        case OP_RECURSE:
2638          {          {
2639            dfa_recursion_info *ri;
2640          int local_offsets[1000];          int local_offsets[1000];
2641          int local_workspace[1000];          int local_workspace[1000];
2642            const pcre_uchar *callpat = start_code + GET(code, 1);
2643            int recno = (callpat == md->start_code)? 0 :
2644              GET2(callpat, 1 + LINK_SIZE);
2645          int rc;          int rc;
2646    
2647          DPRINTF(("%.*sStarting regex recursion %d\n", rlevel*2-2, SP,          DPRINTF(("%.*sStarting regex recursion\n", rlevel*2-2, SP));
2648            recursing + 1));  
2649            /* Check for repeating a recursion without advancing the subject
2650            pointer. This should catch convoluted mutual recursions. (Some simple
2651            cases are caught at compile time.) */
2652    
2653            for (ri = md->recursive; ri != NULL; ri = ri->prevrec)
2654              if (recno == ri->group_num && ptr == ri->subject_position)
2655                return PCRE_ERROR_RECURSELOOP;
2656    
2657            /* Remember this recursion and where we started it so as to
2658            catch infinite loops. */
2659    
2660            new_recursive.group_num = recno;
2661            new_recursive.subject_position = ptr;
2662            new_recursive.prevrec = md->recursive;
2663            md->recursive = &new_recursive;
2664    
2665          rc = internal_dfa_exec(          rc = internal_dfa_exec(
2666            md,                                   /* fixed match data */            md,                                   /* fixed match data */
2667            start_code + GET(code, 1),            /* this subexpression's code */            callpat,                              /* this subexpression's code */
2668            ptr,                                  /* where we currently are */            ptr,                                  /* where we currently are */
2669            (int)(ptr - start_subject),           /* start offset */            (int)(ptr - start_subject),           /* start offset */
2670            local_offsets,                        /* offset vector */            local_offsets,                        /* offset vector */
2671            sizeof(local_offsets)/sizeof(int),    /* size of same */            sizeof(local_offsets)/sizeof(int),    /* size of same */
2672            local_workspace,                      /* workspace vector */            local_workspace,                      /* workspace vector */
2673            sizeof(local_workspace)/sizeof(int),  /* size of same */            sizeof(local_workspace)/sizeof(int),  /* size of same */
2674            ims,                                  /* the current ims flags */            rlevel);                              /* function recursion level */
2675            rlevel,                               /* function recursion level */  
2676            recursing + 1);                       /* regex recurse level */          md->recursive = new_recursive.prevrec;  /* Done this recursion */
2677    
2678          DPRINTF(("%.*sReturn from regex recursion %d: rc=%d\n", rlevel*2-2, SP,          DPRINTF(("%.*sReturn from regex recursion: rc=%d\n", rlevel*2-2, SP,
2679            recursing + 1, rc));            rc));
2680    
2681          /* Ran out of internal offsets */          /* Ran out of internal offsets */
2682    
# Line 2578  for (;;) Line 2690  for (;;)
2690            {            {
2691            for (rc = rc*2 - 2; rc >= 0; rc -= 2)            for (rc = rc*2 - 2; rc >= 0; rc -= 2)
2692              {              {
2693              const uschar *p = start_subject + local_offsets[rc];              const pcre_uchar *p = start_subject + local_offsets[rc];
2694              const uschar *pp = start_subject + local_offsets[rc+1];              const pcre_uchar *pp = start_subject + local_offsets[rc+1];
2695              int charcount = local_offsets[rc+1] - local_offsets[rc];              int charcount = local_offsets[rc+1] - local_offsets[rc];
2696              while (p < pp) if ((*p++ & 0xc0) == 0x80) charcount--;  #ifdef SUPPORT_UTF
2697                while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
2698    #endif
2699              if (charcount > 0)              if (charcount > 0)
2700                {                {
2701                ADD_NEW_DATA(-(state_offset + LINK_SIZE + 1), 0, (charcount - 1));                ADD_NEW_DATA(-(state_offset + LINK_SIZE + 1), 0, (charcount - 1));
# Line 2597  for (;;) Line 2711  for (;;)
2711        break;        break;
2712    
2713        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2714          case OP_BRAPOS:
2715          case OP_SBRAPOS:
2716          case OP_CBRAPOS:
2717          case OP_SCBRAPOS:
2718          case OP_BRAPOSZERO:
2719            {
2720            int charcount, matched_count;
2721            const pcre_uchar *local_ptr = ptr;
2722            BOOL allow_zero;
2723    
2724            if (codevalue == OP_BRAPOSZERO)
2725              {
2726              allow_zero = TRUE;
2727              codevalue = *(++code);  /* Codevalue will be one of above BRAs */
2728              }
2729            else allow_zero = FALSE;
2730    
2731            /* Loop to match the subpattern as many times as possible as if it were
2732            a complete pattern. */
2733    
2734            for (matched_count = 0;; matched_count++)
2735              {
2736              int local_offsets[2];
2737              int local_workspace[1000];
2738    
2739              int rc = internal_dfa_exec(
2740                md,                                   /* fixed match data */
2741                code,                                 /* this subexpression's code */
2742                local_ptr,                            /* where we currently are */
2743                (int)(ptr - start_subject),           /* start offset */
2744                local_offsets,                        /* offset vector */
2745                sizeof(local_offsets)/sizeof(int),    /* size of same */
2746                local_workspace,                      /* workspace vector */
2747                sizeof(local_workspace)/sizeof(int),  /* size of same */
2748                rlevel);                              /* function recursion level */
2749    
2750              /* Failed to match */
2751    
2752              if (rc < 0)
2753                {
2754                if (rc != PCRE_ERROR_NOMATCH) return rc;
2755                break;
2756                }
2757    
2758              /* Matched: break the loop if zero characters matched. */
2759    
2760              charcount = local_offsets[1] - local_offsets[0];
2761              if (charcount == 0) break;
2762              local_ptr += charcount;    /* Advance temporary position ptr */
2763              }
2764    
2765            /* At this point we have matched the subpattern matched_count
2766            times, and local_ptr is pointing to the character after the end of the
2767            last match. */
2768    
2769            if (matched_count > 0 || allow_zero)
2770              {
2771              const pcre_uchar *end_subpattern = code;
2772              int next_state_offset;
2773    
2774              do { end_subpattern += GET(end_subpattern, 1); }
2775                while (*end_subpattern == OP_ALT);
2776              next_state_offset =
2777                (int)(end_subpattern - start_code + LINK_SIZE + 1);
2778    
2779              /* Optimization: if there are no more active states, and there
2780              are no new states yet set up, then skip over the subject string
2781              right here, to save looping. Otherwise, set up the new state to swing
2782              into action when the end of the matched substring is reached. */
2783    
2784              if (i + 1 >= active_count && new_count == 0)
2785                {
2786                ptr = local_ptr;
2787                clen = 0;
2788                ADD_NEW(next_state_offset, 0);
2789                }
2790              else
2791                {
2792                const pcre_uchar *p = ptr;
2793                const pcre_uchar *pp = local_ptr;
2794                charcount = (int)(pp - p);
2795    #ifdef SUPPORT_UTF
2796                while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
2797    #endif
2798                ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));
2799                }
2800              }
2801            }
2802          break;
2803    
2804          /*-----------------------------------------------------------------*/
2805        case OP_ONCE:        case OP_ONCE:
2806          case OP_ONCE_NC:
2807          {          {
2808          int local_offsets[2];          int local_offsets[2];
2809          int local_workspace[1000];          int local_workspace[1000];
# Line 2611  for (;;) Line 2817  for (;;)
2817            sizeof(local_offsets)/sizeof(int),    /* size of same */            sizeof(local_offsets)/sizeof(int),    /* size of same */
2818            local_workspace,                      /* workspace vector */            local_workspace,                      /* workspace vector */
2819            sizeof(local_workspace)/sizeof(int),  /* size of same */            sizeof(local_workspace)/sizeof(int),  /* size of same */
2820            ims,                                  /* the current ims flags */            rlevel);                              /* function recursion level */
           rlevel,                               /* function recursion level */  
           recursing);                           /* pass on regex recursion */  
2821    
2822          if (rc >= 0)          if (rc >= 0)
2823            {            {
2824            const uschar *end_subpattern = code;            const pcre_uchar *end_subpattern = code;
2825            int charcount = local_offsets[1] - local_offsets[0];            int charcount = local_offsets[1] - local_offsets[0];
2826            int next_state_offset, repeat_state_offset;            int next_state_offset, repeat_state_offset;
2827    
# Line 2647  for (;;) Line 2851  for (;;)
2851            /* Optimization: if there are no more active states, and there            /* Optimization: if there are no more active states, and there
2852            are no new states yet set up, then skip over the subject string            are no new states yet set up, then skip over the subject string
2853            right here, to save looping. Otherwise, set up the new state to swing            right here, to save looping. Otherwise, set up the new state to swing
2854            into action when the end of the substring is reached. */            into action when the end of the matched substring is reached. */
2855    
2856            else if (i + 1 >= active_count && new_count == 0)            else if (i + 1 >= active_count && new_count == 0)
2857              {              {
# Line 2670  for (;;) Line 2874  for (;;)
2874              }              }
2875            else            else
2876              {              {
2877              const uschar *p = start_subject + local_offsets[0];  #ifdef SUPPORT_UTF
2878              const uschar *pp = start_subject + local_offsets[1];              const pcre_uchar *p = start_subject + local_offsets[0];
2879              while (p < pp) if ((*p++ & 0xc0) == 0x80) charcount--;              const pcre_uchar *pp = start_subject + local_offsets[1];
2880                while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
2881    #endif
2882              ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));              ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));
2883              if (repeat_state_offset >= 0)              if (repeat_state_offset >= 0)
2884                { ADD_NEW_DATA(-repeat_state_offset, 0, (charcount - 1)); }                { ADD_NEW_DATA(-repeat_state_offset, 0, (charcount - 1)); }
2885              }              }
   
2886            }            }
2887          else if (rc != PCRE_ERROR_NOMATCH) return rc;          else if (rc != PCRE_ERROR_NOMATCH) return rc;
2888          }          }
# Line 2689  for (;;) Line 2894  for (;;)
2894    
2895        case OP_CALLOUT:        case OP_CALLOUT:
2896        rrc = 0;        rrc = 0;
2897        if (pcre_callout != NULL)        if (PUBL(callout) != NULL)
2898          {          {
2899          pcre_callout_block cb;          PUBL(callout_block) cb;
2900          cb.version          = 1;   /* Version 1 of the callout block */          cb.version          = 1;   /* Version 1 of the callout block */
2901          cb.callout_number   = code[1];          cb.callout_number   = code[1];
2902          cb.offset_vector    = offsets;          cb.offset_vector    = offsets;
2903    #ifdef COMPILE_PCRE8
2904          cb.subject          = (PCRE_SPTR)start_subject;          cb.subject          = (PCRE_SPTR)start_subject;
2905    #else
2906            cb.subject          = (PCRE_SPTR16)start_subject;
2907    #endif
2908          cb.subject_length   = (int)(end_subject - start_subject);          cb.subject_length   = (int)(end_subject - start_subject);
2909          cb.start_match      = (int)(current_subject - start_subject);          cb.start_match      = (int)(current_subject - start_subject);
2910          cb.current_position = (int)(ptr - start_subject);          cb.current_position = (int)(ptr - start_subject);
# Line 2704  for (;;) Line 2913  for (;;)
2913          cb.capture_top      = 1;          cb.capture_top      = 1;
2914          cb.capture_last     = -1;          cb.capture_last     = -1;
2915          cb.callout_data     = md->callout_data;          cb.callout_data     = md->callout_data;
2916          if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc;   /* Abandon */          cb.mark             = NULL;   /* No (*MARK) support */
2917            if ((rrc = (*PUBL(callout))(&cb)) < 0) return rrc;   /* Abandon */
2918          }          }
2919        if (rrc == 0)        if (rrc == 0)
2920          { ADD_ACTIVE(state_offset + _pcre_OP_lengths[OP_CALLOUT], 0); }          { ADD_ACTIVE(state_offset + PRIV(OP_lengths)[OP_CALLOUT], 0); }
2921        break;        break;
2922    
2923    
# Line 2744  for (;;) Line 2954  for (;;)
2954          ((md->moptions & PCRE_PARTIAL_SOFT) != 0 &&  /* Soft partial and */          ((md->moptions & PCRE_PARTIAL_SOFT) != 0 &&  /* Soft partial and */
2955           match_count < 0)                            /* no matches */           match_count < 0)                            /* no matches */
2956          ) &&                                         /* And... */          ) &&                                         /* And... */
2957          ptr >= end_subject &&                     /* Reached end of subject */          ptr >= end_subject &&                  /* Reached end of subject */
2958          ptr > current_subject)                    /* Matched non-empty string */          ptr > md->start_used_ptr)              /* Inspected non-empty string */
2959        {        {
2960        if (offsetcount >= 2)        if (offsetcount >= 2)
2961          {          {
# Line 2804  Returns:          > 0 => number of match Line 3014  Returns:          > 0 => number of match
3014                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
3015  */  */
3016    
3017    #ifdef COMPILE_PCRE8
3018  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
3019  pcre_dfa_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_dfa_exec(const pcre *argument_re, const pcre_extra *extra_data,
3020    const char *subject, int length, int start_offset, int options, int *offsets,    const char *subject, int length, int start_offset, int options, int *offsets,
3021    int offsetcount, int *workspace, int wscount)    int offsetcount, int *workspace, int wscount)
3022    #else
3023    PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
3024    pcre16_dfa_exec(const pcre16 *argument_re, const pcre16_extra *extra_data,
3025      PCRE_SPTR16 subject, int length, int start_offset, int options, int *offsets,
3026      int offsetcount, int *workspace, int wscount)
3027    #endif
3028  {  {
3029  real_pcre *re = (real_pcre *)argument_re;  REAL_PCRE *re = (REAL_PCRE *)argument_re;
3030  dfa_match_data match_block;  dfa_match_data match_block;
3031  dfa_match_data *md = &match_block;  dfa_match_data *md = &match_block;
3032  BOOL utf8, anchored, startline, firstline;  BOOL utf, anchored, startline, firstline;
3033  const uschar *current_subject, *end_subject, *lcc;  const pcre_uchar *current_subject, *end_subject;
   
 pcre_study_data internal_study;  
3034  const pcre_study_data *study = NULL;  const pcre_study_data *study = NULL;
 real_pcre internal_re;  
3035    
3036  const uschar *req_byte_ptr;  const pcre_uchar *req_char_ptr;
3037  const uschar *start_bits = NULL;  const pcre_uint8 *start_bits = NULL;
3038  BOOL first_byte_caseless = FALSE;  BOOL has_first_char = FALSE;
3039  BOOL req_byte_caseless = FALSE;  BOOL has_req_char = FALSE;
3040  int first_byte = -1;  pcre_uchar first_char = 0;
3041  int req_byte = -1;  pcre_uchar first_char2 = 0;
3042  int req_byte2 = -1;  pcre_uchar req_char = 0;
3043    pcre_uchar req_char2 = 0;
3044  int newline;  int newline;
3045    
3046  /* Plausibility checks */  /* Plausibility checks */
# Line 2835  if (re == NULL || subject == NULL || wor Line 3050  if (re == NULL || subject == NULL || wor
3050     (offsets == NULL && offsetcount > 0)) return PCRE_ERROR_NULL;     (offsets == NULL && offsetcount > 0)) return PCRE_ERROR_NULL;
3051  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;
3052  if (wscount < 20) return PCRE_ERROR_DFA_WSSIZE;  if (wscount < 20) return PCRE_ERROR_DFA_WSSIZE;
3053    if (start_offset < 0 || start_offset > length) return PCRE_ERROR_BADOFFSET;
3054    
3055  /* We need to find the pointer to any study data before we test for byte  /* We need to find the pointer to any study data before we test for byte
3056  flipping, so we scan the extra_data block first. This may set two fields in the  flipping, so we scan the extra_data block first. This may set two fields in the
# Line 2859  if (extra_data != NULL) Line 3075  if (extra_data != NULL)
3075    }    }
3076    
3077  /* Check that the first field in the block is the magic number. If it is not,  /* Check that the first field in the block is the magic number. If it is not,
3078  test for a regex that was compiled on a host of opposite endianness. If this is  return with PCRE_ERROR_BADMAGIC. However, if the magic number is equal to
3079  the case, flipped values are put in internal_re and internal_study if there was  REVERSED_MAGIC_NUMBER we return with PCRE_ERROR_BADENDIANNESS, which
3080  study data too. */  means that the pattern is likely compiled with different endianness. */
3081    
3082  if (re->magic_number != MAGIC_NUMBER)  if (re->magic_number != MAGIC_NUMBER)
3083    {    return re->magic_number == REVERSED_MAGIC_NUMBER?
3084    re = _pcre_try_flipped(re, &internal_re, study, &internal_study);      PCRE_ERROR_BADENDIANNESS:PCRE_ERROR_BADMAGIC;
3085    if (re == NULL) return PCRE_ERROR_BADMAGIC;  if ((re->flags & PCRE_MODE) == 0) return PCRE_ERROR_BADMODE;
   if (study != NULL) study = &internal_study;  
   }  
3086    
3087  /* Set some local values */  /* Set some local values */
3088    
3089  current_subject = (const unsigned char *)subject + start_offset;  current_subject = (const pcre_uchar *)subject + start_offset;
3090  end_subject = (const unsigned char *)subject + length;  end_subject = (const pcre_uchar *)subject + length;
3091  req_byte_ptr = current_subject - 1;  req_char_ptr = current_subject - 1;
3092    
3093  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3094  utf8 = (re->options & PCRE_UTF8) != 0;  /* PCRE_UTF16 has the same value as PCRE_UTF8. */
3095    utf = (re->options & PCRE_UTF8) != 0;
3096  #else  #else
3097  utf8 = FALSE;  utf = FALSE;
3098  #endif  #endif
3099    
3100  anchored = (options & (PCRE_ANCHORED|PCRE_DFA_RESTART)) != 0 ||  anchored = (options & (PCRE_ANCHORED|PCRE_DFA_RESTART)) != 0 ||
# Line 2887  anchored = (options & (PCRE_ANCHORED|PCR Line 3102  anchored = (options & (PCRE_ANCHORED|PCR
3102    
3103  /* The remaining fixed data for passing around. */  /* The remaining fixed data for passing around. */
3104    
3105  md->start_code = (const uschar *)argument_re +  md->start_code = (const pcre_uchar *)argument_re +
3106      re->name_table_offset + re->name_count * re->name_entry_size;      re->name_table_offset + re->name_count * re->name_entry_size;
3107  md->start_subject = (const unsigned char *)subject;  md->start_subject = (const pcre_uchar *)subject;
3108  md->end_subject = end_subject;  md->end_subject = end_subject;
3109  md->start_offset = start_offset;  md->start_offset = start_offset;
3110  md->moptions = options;  md->moptions = options;
# Line 2950  else Line 3165  else
3165  /* Check a UTF-8 string if required. Unfortunately there's no way of passing  /* Check a UTF-8 string if required. Unfortunately there's no way of passing
3166  back the character offset. */  back the character offset. */
3167    
3168  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3169  if (utf8 && (options & PCRE_NO_UTF8_CHECK) == 0)  if (utf && (options & PCRE_NO_UTF8_CHECK) == 0)
3170    {    {
3171    if (_pcre_valid_utf8((uschar *)subject, length) >= 0)    int erroroffset;
3172      return PCRE_ERROR_BADUTF8;    int errorcode = PRIV(valid_utf)((pcre_uchar *)subject, length, &erroroffset);
3173    if (start_offset > 0 && start_offset < length)    if (errorcode != 0)
3174      {      {
3175      int tb = ((uschar *)subject)[start_offset];      if (offsetcount >= 2)
     if (tb > 127)  
3176        {        {
3177        tb &= 0xc0;        offsets[0] = erroroffset;
3178        if (tb != 0 && tb != 0xc0) return PCRE_ERROR_BADUTF8_OFFSET;        offsets[1] = errorcode;
3179        }        }
3180        return (errorcode <= PCRE_UTF8_ERR5 && (options & PCRE_PARTIAL_HARD) != 0)?
3181          PCRE_ERROR_SHORTUTF8 : PCRE_ERROR_BADUTF8;
3182      }      }
3183      if (start_offset > 0 && start_offset < length &&
3184            NOT_FIRSTCHAR(((PCRE_PUCHAR)subject)[start_offset]))
3185        return PCRE_ERROR_BADUTF8_OFFSET;
3186    }    }
3187  #endif  #endif
3188    
# Line 2971  if (utf8 && (options & PCRE_NO_UTF8_CHEC Line 3190  if (utf8 && (options & PCRE_NO_UTF8_CHEC
3190  is a feature that makes it possible to save compiled regex and re-use them  is a feature that makes it possible to save compiled regex and re-use them
3191  in other programs later. */  in other programs later. */
3192    
3193  if (md->tables == NULL) md->tables = _pcre_default_tables;  if (md->tables == NULL) md->tables = PRIV(default_tables);
3194    
3195  /* The lower casing table and the "must be at the start of a line" flag are  /* The "must be at the start of a line" flags are used in a loop when finding
3196  used in a loop when finding where to start. */  where to start. */
3197    
 lcc = md->tables + lcc_offset;  
3198  startline = (re->flags & PCRE_STARTLINE) != 0;  startline = (re->flags & PCRE_STARTLINE) != 0;
3199  firstline = (re->options & PCRE_FIRSTLINE) != 0;  firstline = (re->options & PCRE_FIRSTLINE) != 0;
3200    
# Line 2990  if (!anchored) Line 3208  if (!anchored)
3208    {    {
3209    if ((re->flags & PCRE_FIRSTSET) != 0)    if ((re->flags & PCRE_FIRSTSET) != 0)
3210      {      {
3211      first_byte = re->first_byte & 255;      has_first_char = TRUE;
3212      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)      first_char = first_char2 = re->first_char;
3213        first_byte = lcc[first_byte];      if ((re->flags & PCRE_FCH_CASELESS) != 0)
3214          {
3215          first_char2 = TABLE_GET(first_char, md->tables + fcc_offset, first_char);
3216    #if defined SUPPORT_UCP && !(defined COMPILE_PCRE8)
3217          if (utf && first_char > 127)
3218            first_char2 = UCD_OTHERCASE(first_char);
3219    #endif
3220          }
3221      }      }
3222    else    else
3223      {      {
# Line 3007  character" set. */ Line 3232  character" set. */
3232    
3233  if ((re->flags & PCRE_REQCHSET) != 0)  if ((re->flags & PCRE_REQCHSET) != 0)
3234    {    {
3235    req_byte = re->req_byte & 255;    has_req_char = TRUE;
3236    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;    req_char = req_char2 = re->req_char;
3237    req_byte2 = (md->tables + fcc_offset)[req_byte];  /* case flipped */    if ((re->flags & PCRE_RCH_CASELESS) != 0)
3238        {
3239        req_char2 = TABLE_GET(req_char, md->tables + fcc_offset, req_char);
3240    #if defined SUPPORT_UCP && !(defined COMPILE_PCRE8)
3241        if (utf && req_char > 127)
3242          req_char2 = UCD_OTHERCASE(req_char);
3243    #endif
3244        }
3245    }    }
3246    
3247  /* Call the main matching function, looping for a non-anchored regex after a  /* Call the main matching function, looping for a non-anchored regex after a
# Line 3022  for (;;) Line 3254  for (;;)
3254    
3255    if ((options & PCRE_DFA_RESTART) == 0)    if ((options & PCRE_DFA_RESTART) == 0)
3256      {      {
3257      const uschar *save_end_subject = end_subject;      const pcre_uchar *save_end_subject = end_subject;
3258    
3259      /* If firstline is TRUE, the start of the match is constrained to the first      /* If firstline is TRUE, the start of the match is constrained to the first
3260      line of a multiline string. Implement this by temporarily adjusting      line of a multiline string. Implement this by temporarily adjusting
# Line 3031  for (;;) Line 3263  for (;;)
3263    
3264      if (firstline)      if (firstline)
3265        {        {
3266        USPTR t = current_subject;        PCRE_PUCHAR t = current_subject;
3267  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3268        if (utf8)        if (utf)
3269          {          {
3270          while (t < md->end_subject && !IS_NEWLINE(t))          while (t < md->end_subject && !IS_NEWLINE(t))
3271            {            {
3272            t++;            t++;
3273            while (t < end_subject && (*t & 0xc0) == 0x80) t++;            ACROSSCHAR(t < end_subject, *t, t++);
3274            }            }
3275          }          }
3276        else        else
# Line 3049  for (;;) Line 3281  for (;;)
3281    
3282      /* There are some optimizations that avoid running the match if a known      /* There are some optimizations that avoid running the match if a known
3283      starting point is not found. However, there is an option that disables      starting point is not found. However, there is an option that disables
3284      these, for testing and for ensuring that all callouts do actually occur. */      these, for testing and for ensuring that all callouts do actually occur.
3285        The option can be set in the regex by (*NO_START_OPT) or passed in
3286        match-time options. */
3287    
3288      if ((options & PCRE_NO_START_OPTIMIZE) == 0)      if (((options | re->options) & PCRE_NO_START_OPTIMIZE) == 0)
3289        {        {
3290        /* Advance to a known first byte. */        /* Advance to a known first char. */
3291    
3292        if (first_byte >= 0)        if (has_first_char)
3293          {          {
3294          if (first_byte_caseless)          if (first_char != first_char2)
3295            while (current_subject < end_subject &&            while (current_subject < end_subject &&
3296                   lcc[*current_subject] != first_byte)                *current_subject != first_char && *current_subject != first_char2)
3297              current_subject++;              current_subject++;
3298          else          else
3299            while (current_subject < end_subject &&            while (current_subject < end_subject &&
3300                   *current_subject != first_byte)                   *current_subject != first_char)
3301              current_subject++;              current_subject++;
3302          }          }
3303    
# Line 3073  for (;;) Line 3307  for (;;)
3307          {          {
3308          if (current_subject > md->start_subject + start_offset)          if (current_subject > md->start_subject + start_offset)
3309            {            {
3310  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3311            if (utf8)            if (utf)
3312              {              {
3313              while (current_subject < end_subject &&              while (current_subject < end_subject &&
3314                     !WAS_NEWLINE(current_subject))                     !WAS_NEWLINE(current_subject))
3315                {                {
3316                current_subject++;                current_subject++;
3317                while(current_subject < end_subject &&                ACROSSCHAR(current_subject < end_subject, *current_subject,
3318                      (*current_subject & 0xc0) == 0x80)                  current_subject++);
                 current_subject++;  
3319                }                }
3320              }              }
3321            else            else
# Line 3109  for (;;) Line 3342  for (;;)
3342          while (current_subject < end_subject)          while (current_subject < end_subject)
3343            {            {
3344            register unsigned int c = *current_subject;            register unsigned int c = *current_subject;
3345            if ((start_bits[c/8] & (1 << (c&7))) == 0) current_subject++;  #ifndef COMPILE_PCRE8
3346              else break;            if (c > 255) c = 255;
3347    #endif
3348              if ((start_bits[c/8] & (1 << (c&7))) == 0)
3349                {
3350                current_subject++;
3351    #if defined SUPPORT_UTF && defined COMPILE_PCRE8
3352                /* In non 8-bit mode, the iteration will stop for
3353                characters > 255 at the beginning or not stop at all. */
3354                if (utf)
3355                  ACROSSCHAR(current_subject < end_subject, *current_subject,
3356                    current_subject++);
3357    #endif
3358                }
3359              else break;
3360            }            }
3361          }          }
3362        }        }
# Line 3123  for (;;) Line 3369  for (;;)
3369      disabling is explicitly requested (and of course, by the test above, this      disabling is explicitly requested (and of course, by the test above, this
3370      code is not obeyed when restarting after a partial match). */      code is not obeyed when restarting after a partial match). */
3371    
3372      if ((options & PCRE_NO_START_OPTIMIZE) == 0 &&      if (((options | re->options) & PCRE_NO_START_OPTIMIZE) == 0 &&
3373          (options & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) == 0)          (options & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) == 0)
3374        {        {
3375        /* If the pattern was studied, a minimum subject length may be set. This        /* If the pattern was studied, a minimum subject length may be set. This
# Line 3135  for (;;) Line 3381  for (;;)
3381            (pcre_uint32)(end_subject - current_subject) < study->minlength)            (pcre_uint32)(end_subject - current_subject) < study->minlength)
3382          return PCRE_ERROR_NOMATCH;          return PCRE_ERROR_NOMATCH;
3383    
3384        /* If req_byte is set, we know that that character must appear in the        /* If req_char is set, we know that that character must appear in the
3385        subject for the match to succeed. If the first character is set, req_byte        subject for the match to succeed. If the first character is set, req_char
3386        must be later in the subject; otherwise the test starts at the match        must be later in the subject; otherwise the test starts at the match
3387        point. This optimization can save a huge amount of work in patterns with        point. This optimization can save a huge amount of work in patterns with
3388        nested unlimited repeats that aren't going to match. Writing separate        nested unlimited repeats that aren't going to match. Writing separate
# Line 3148  for (;;) Line 3394  for (;;)
3394        patterns. This showed up when somebody was matching /^C/ on a 32-megabyte        patterns. This showed up when somebody was matching /^C/ on a 32-megabyte
3395        string... so we don't do this when the string is sufficiently long. */        string... so we don't do this when the string is sufficiently long. */
3396    
3397        if (req_byte >= 0 && end_subject - current_subject < REQ_BYTE_MAX)        if (has_req_char && end_subject - current_subject < REQ_BYTE_MAX)
3398          {          {
3399          register const uschar *p = current_subject + ((first_byte >= 0)? 1 : 0);          register PCRE_PUCHAR p = current_subject + (has_first_char? 1:0);
3400    
3401          /* We don't need to repeat the search if we haven't yet reached the          /* We don't need to repeat the search if we haven't yet reached the
3402          place we found it at last time. */          place we found it at last time. */
3403    
3404          if (p > req_byte_ptr)          if (p > req_char_ptr)
3405            {            {
3406            if (req_byte_caseless)            if (req_char != req_char2)
3407              {              {
3408              while (p < end_subject)              while (p < end_subject)
3409                {                {
3410                register int pp = *p++;                register int pp = *p++;
3411                if (pp == req_byte || pp == req_byte2) { p--; break; }                if (pp == req_char || pp == req_char2) { p--; break; }
3412                }                }
3413              }              }
3414            else            else
3415              {              {
3416              while (p < end_subject)              while (p < end_subject)
3417                {                {
3418                if (*p++ == req_byte) { p--; break; }                if (*p++ == req_char) { p--; break; }
3419                }                }
3420              }              }
3421    
# Line 3182  for (;;) Line 3428  for (;;)
3428            found it, so that we don't search again next time round the loop if            found it, so that we don't search again next time round the loop if
3429            the start hasn't passed this character yet. */            the start hasn't passed this character yet. */
3430    
3431            req_byte_ptr = p;            req_char_ptr = p;
3432            }            }
3433          }          }
3434        }        }
# Line 3191  for (;;) Line 3437  for (;;)
3437    /* OK, now we can do the business */    /* OK, now we can do the business */
3438    
3439    md->start_used_ptr = current_subject;    md->start_used_ptr = current_subject;
3440      md->recursive = NULL;
3441    
3442    rc = internal_dfa_exec(    rc = internal_dfa_exec(
3443      md,                                /* fixed match data */      md,                                /* fixed match data */
# Line 3201  for (;;) Line 3448  for (;;)
3448      offsetcount,                       /* size of same */      offsetcount,                       /* size of same */
3449      workspace,                         /* workspace vector */      workspace,                         /* workspace vector */
3450      wscount,                           /* size of same */      wscount,                           /* size of same */
3451      re->options & (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL), /* ims flags */      0);                                /* function recurse level */
     0,                                 /* function recurse level */  
     0);                                /* regex recurse level */  
3452    
3453    /* Anything other than "no match" means we are done, always; otherwise, carry    /* Anything other than "no match" means we are done, always; otherwise, carry
3454    on only if not anchored. */    on only if not anchored. */
# Line 3215  for (;;) Line 3460  for (;;)
3460    
3461    if (firstline && IS_NEWLINE(current_subject)) break;    if (firstline && IS_NEWLINE(current_subject)) break;
3462    current_subject++;    current_subject++;
3463    if (utf8)  #ifdef SUPPORT_UTF
3464      if (utf)
3465      {      {
3466      while (current_subject < end_subject && (*current_subject & 0xc0) == 0x80)      ACROSSCHAR(current_subject < end_subject, *current_subject,
3467        current_subject++;        current_subject++);
3468      }      }
3469    #endif
3470    if (current_subject > end_subject) break;    if (current_subject > end_subject) break;
3471    
3472    /* If we have just passed a CR and we are now at a LF, and the pattern does    /* If we have just passed a CR and we are now at a LF, and the pattern does

Legend:
Removed from v.535  
changed lines
  Added in v.881

  ViewVC Help
Powered by ViewVC 1.1.5