/[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 442 by ph10, Fri Sep 11 10:21:02 2009 UTC revision 1233 by ph10, Sat Dec 8 12:43:37 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) 1997-2009 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 38  POSSIBILITY OF SUCH DAMAGE. Line 38  POSSIBILITY OF SUCH DAMAGE.
38  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
39  */  */
40    
   
41  /* This module contains the external function pcre_dfa_exec(), which is an  /* This module contains the external function pcre_dfa_exec(), which is an
42  alternative matching function that uses a sort of DFA algorithm (not a true  alternative matching function that uses a sort of DFA algorithm (not a true
43  FSM). This is NOT Perl- compatible, but it has advantages in certain  FSM). This is NOT Perl-compatible, but it has advantages in certain
44  applications. */  applications. */
45    
46    
47  /* NOTE ABOUT PERFORMANCE: A user of this function sent some code that improved  /* NOTE ABOUT PERFORMANCE: A user of this function sent some code that improved
48  the performance of his patterns greatly. I could not use it as it stood, as it  the performance of his patterns greatly. I could not use it as it stood, as it
49  was not thread safe, and made assumptions about pattern sizes. Also, it caused  was not thread safe, and made assumptions about pattern sizes. Also, it caused
50  test 7 to loop, and test 9 to crash with a segfault.  test 7 to loop, and test 9 to crash with a segfault.
51    
52  The issue is the check for duplicate states, which is done by a simple linear  The issue is the check for duplicate states, which is done by a simple linear
# Line 68  was the extra time to initialize the ind Line 67  was the extra time to initialize the ind
67  of internal_dfa_exec(). (The supplied patch used a static vector, initialized  of internal_dfa_exec(). (The supplied patch used a static vector, initialized
68  only once - I suspect this was the cause of the problems with the tests.)  only once - I suspect this was the cause of the problems with the tests.)
69    
70  Overall, I concluded that the gains in some cases did not outweigh the losses  Overall, I concluded that the gains in some cases did not outweigh the losses
71  in others, so I abandoned this code. */  in others, so I abandoned this code. */
72    
73    
# Line 106  never stored, so we push them well clear Line 105  never stored, so we push them well clear
105    
106    
107  /* This table identifies those opcodes that are followed immediately by a  /* This table identifies those opcodes that are followed immediately by a
108  character that is to be tested in some way. This makes is possible to  character that is to be tested in some way. This makes it possible to
109  centralize the loading of these characters. In the case of Type * etc, the  centralize the loading of these characters. In the case of Type * etc, the
110  "character" is the opcode for \D, \d, \S, \s, \W, or \w, which will always be a  "character" is the opcode for \D, \d, \S, \s, \W, or \w, which will always be a
111  small value. ***NOTE*** If the start of this table is modified, the two tables  small value. Non-zero values in the table are the offsets from the opcode where
112  that follow must also be modified. */  the character is to be found. ***NOTE*** If the start of this table is
113    modified, the three tables that follow must also be modified. */
114    
115  static const uschar coptable[] = {  static const pcre_uint8 coptable[] = {
116    0,                             /* End                                    */    0,                             /* End                                    */
117    0, 0, 0, 0, 0,                 /* \A, \G, \K, \B, \b                     */    0, 0, 0, 0, 0,                 /* \A, \G, \K, \B, \b                     */
118    0, 0, 0, 0, 0, 0,              /* \D, \d, \S, \s, \W, \w                 */    0, 0, 0, 0, 0, 0,              /* \D, \d, \S, \s, \W, \w                 */
119    0, 0, 0,                       /* Any, AllAny, Anybyte                   */    0, 0, 0,                       /* Any, AllAny, Anybyte                   */
120    0, 0, 0,                       /* NOTPROP, PROP, EXTUNI                  */    0, 0,                          /* \P, \p                                 */
121    0, 0, 0, 0, 0,                 /* \R, \H, \h, \V, \v                     */    0, 0, 0, 0, 0,                 /* \R, \H, \h, \V, \v                     */
122    0, 0, 0, 0, 0,                 /* \Z, \z, Opt, ^, $                      */    0,                             /* \X                                     */
123      0, 0, 0, 0, 0, 0,              /* \Z, \z, ^, ^M, $, $M                   */
124    1,                             /* Char                                   */    1,                             /* Char                                   */
125    1,                             /* Charnc                                 */    1,                             /* Chari                                  */
126    1,                             /* not                                    */    1,                             /* not                                    */
127      1,                             /* noti                                   */
128    /* Positive single-char repeats                                          */    /* Positive single-char repeats                                          */
129    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */
130    3, 3, 3,                       /* upto, minupto, exact                   */    1+IMM2_SIZE, 1+IMM2_SIZE,      /* upto, minupto                          */
131    1, 1, 1, 3,                    /* *+, ++, ?+, upto+                      */    1+IMM2_SIZE,                   /* exact                                  */
132      1, 1, 1, 1+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */
133      1, 1, 1, 1, 1, 1,              /* *I, *?I, +I, +?I, ?I, ??I              */
134      1+IMM2_SIZE, 1+IMM2_SIZE,      /* upto I, minupto I                      */
135      1+IMM2_SIZE,                   /* exact I                                */
136      1, 1, 1, 1+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */
137    /* Negative single-char repeats - only for chars < 256                   */    /* Negative single-char repeats - only for chars < 256                   */
138    1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */    1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */
139    3, 3, 3,                       /* NOT upto, minupto, exact               */    1+IMM2_SIZE, 1+IMM2_SIZE,      /* NOT upto, minupto                      */
140    1, 1, 1, 3,                    /* NOT *+, ++, ?+, updo+                  */    1+IMM2_SIZE,                   /* NOT exact                              */
141      1, 1, 1, 1+IMM2_SIZE,          /* NOT *+, ++, ?+, upto+                  */
142      1, 1, 1, 1, 1, 1,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */
143      1+IMM2_SIZE, 1+IMM2_SIZE,      /* NOT upto I, minupto I                  */
144      1+IMM2_SIZE,                   /* NOT exact I                            */
145      1, 1, 1, 1+IMM2_SIZE,          /* NOT *+I, ++I, ?+I, upto+I              */
146    /* Positive type repeats                                                 */    /* Positive type repeats                                                 */
147    1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */    1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */
148    3, 3, 3,                       /* Type upto, minupto, exact              */    1+IMM2_SIZE, 1+IMM2_SIZE,      /* Type upto, minupto                     */
149    1, 1, 1, 3,                    /* Type *+, ++, ?+, upto+                 */    1+IMM2_SIZE,                   /* Type exact                             */
150      1, 1, 1, 1+IMM2_SIZE,          /* Type *+, ++, ?+, upto+                 */
151    /* Character class & ref repeats                                         */    /* Character class & ref repeats                                         */
152    0, 0, 0, 0, 0, 0,              /* *, *?, +, +?, ?, ??                    */    0, 0, 0, 0, 0, 0,              /* *, *?, +, +?, ?, ??                    */
153    0, 0,                          /* CRRANGE, CRMINRANGE                    */    0, 0,                          /* CRRANGE, CRMINRANGE                    */
# Line 142  static const uschar coptable[] = { Line 155  static const uschar coptable[] = {
155    0,                             /* NCLASS                                 */    0,                             /* NCLASS                                 */
156    0,                             /* XCLASS - variable length               */    0,                             /* XCLASS - variable length               */
157    0,                             /* REF                                    */    0,                             /* REF                                    */
158      0,                             /* REFI                                   */
159    0,                             /* RECURSE                                */    0,                             /* RECURSE                                */
160    0,                             /* CALLOUT                                */    0,                             /* CALLOUT                                */
161    0,                             /* Alt                                    */    0,                             /* Alt                                    */
162    0,                             /* Ket                                    */    0,                             /* Ket                                    */
163    0,                             /* KetRmax                                */    0,                             /* KetRmax                                */
164    0,                             /* KetRmin                                */    0,                             /* KetRmin                                */
165      0,                             /* KetRpos                                */
166      0,                             /* Reverse                                */
167    0,                             /* Assert                                 */    0,                             /* Assert                                 */
168    0,                             /* Assert not                             */    0,                             /* Assert not                             */
169    0,                             /* Assert behind                          */    0,                             /* Assert behind                          */
170    0,                             /* Assert behind not                      */    0,                             /* Assert behind not                      */
171      0, 0,                          /* ONCE, ONCE_NC                          */
172      0, 0, 0, 0, 0,                 /* BRA, BRAPOS, CBRA, CBRAPOS, COND       */
173      0, 0, 0, 0, 0,                 /* SBRA, SBRAPOS, SCBRA, SCBRAPOS, SCOND  */
174      0, 0,                          /* CREF, NCREF                            */
175      0, 0,                          /* RREF, NRREF                            */
176      0,                             /* DEF                                    */
177      0, 0, 0,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */
178      0, 0, 0,                       /* MARK, PRUNE, PRUNE_ARG                 */
179      0, 0, 0, 0,                    /* SKIP, SKIP_ARG, THEN, THEN_ARG         */
180      0, 0, 0, 0,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */
181      0, 0                           /* CLOSE, SKIPZERO  */
182    };
183    
184    /* This table identifies those opcodes that inspect a character. It is used to
185    remember the fact that a character could have been inspected when the end of
186    the subject is reached. ***NOTE*** If the start of this table is modified, the
187    two tables that follow must also be modified. */
188    
189    static const pcre_uint8 poptable[] = {
190      0,                             /* End                                    */
191      0, 0, 0, 1, 1,                 /* \A, \G, \K, \B, \b                     */
192      1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */
193      1, 1, 1,                       /* Any, AllAny, Anybyte                   */
194      1, 1,                          /* \P, \p                                 */
195      1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */
196      1,                             /* \X                                     */
197      0, 0, 0, 0, 0, 0,              /* \Z, \z, ^, ^M, $, $M                   */
198      1,                             /* Char                                   */
199      1,                             /* Chari                                  */
200      1,                             /* not                                    */
201      1,                             /* noti                                   */
202      /* Positive single-char repeats                                          */
203      1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */
204      1, 1, 1,                       /* upto, minupto, exact                   */
205      1, 1, 1, 1,                    /* *+, ++, ?+, upto+                      */
206      1, 1, 1, 1, 1, 1,              /* *I, *?I, +I, +?I, ?I, ??I              */
207      1, 1, 1,                       /* upto I, minupto I, exact I             */
208      1, 1, 1, 1,                    /* *+I, ++I, ?+I, upto+I                  */
209      /* Negative single-char repeats - only for chars < 256                   */
210      1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */
211      1, 1, 1,                       /* NOT upto, minupto, exact               */
212      1, 1, 1, 1,                    /* NOT *+, ++, ?+, upto+                  */
213      1, 1, 1, 1, 1, 1,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */
214      1, 1, 1,                       /* NOT upto I, minupto I, exact I         */
215      1, 1, 1, 1,                    /* NOT *+I, ++I, ?+I, upto+I              */
216      /* Positive type repeats                                                 */
217      1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */
218      1, 1, 1,                       /* Type upto, minupto, exact              */
219      1, 1, 1, 1,                    /* Type *+, ++, ?+, upto+                 */
220      /* Character class & ref repeats                                         */
221      1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */
222      1, 1,                          /* CRRANGE, CRMINRANGE                    */
223      1,                             /* CLASS                                  */
224      1,                             /* NCLASS                                 */
225      1,                             /* XCLASS - variable length               */
226      0,                             /* REF                                    */
227      0,                             /* REFI                                   */
228      0,                             /* RECURSE                                */
229      0,                             /* CALLOUT                                */
230      0,                             /* Alt                                    */
231      0,                             /* Ket                                    */
232      0,                             /* KetRmax                                */
233      0,                             /* KetRmin                                */
234      0,                             /* KetRpos                                */
235    0,                             /* Reverse                                */    0,                             /* Reverse                                */
236    0, 0, 0, 0,                    /* ONCE, BRA, CBRA, COND                  */    0,                             /* Assert                                 */
237    0, 0, 0,                       /* SBRA, SCBRA, SCOND                     */    0,                             /* Assert not                             */
238    0,                             /* CREF                                   */    0,                             /* Assert behind                          */
239    0,                             /* RREF                                   */    0,                             /* Assert behind not                      */
240      0, 0,                          /* ONCE, ONCE_NC                          */
241      0, 0, 0, 0, 0,                 /* BRA, BRAPOS, CBRA, CBRAPOS, COND       */
242      0, 0, 0, 0, 0,                 /* SBRA, SBRAPOS, SCBRA, SCBRAPOS, SCOND  */
243      0, 0,                          /* CREF, NCREF                            */
244      0, 0,                          /* RREF, NRREF                            */
245    0,                             /* DEF                                    */    0,                             /* DEF                                    */
246    0, 0,                          /* BRAZERO, BRAMINZERO                    */    0, 0, 0,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */
247    0, 0, 0, 0,                    /* PRUNE, SKIP, THEN, COMMIT              */    0, 0, 0,                       /* MARK, PRUNE, PRUNE_ARG                 */
248    0, 0, 0                        /* FAIL, ACCEPT, SKIPZERO                 */    0, 0, 0, 0,                    /* SKIP, SKIP_ARG, THEN, THEN_ARG         */
249      0, 0, 0, 0,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */
250      0, 0                           /* CLOSE, SKIPZERO                        */
251  };  };
252    
253  /* 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,
254  and \w */  and \w */
255    
256  static const uschar toptable1[] = {  static const pcre_uint8 toptable1[] = {
257    0, 0, 0, 0, 0, 0,    0, 0, 0, 0, 0, 0,
258    ctype_digit, ctype_digit,    ctype_digit, ctype_digit,
259    ctype_space, ctype_space,    ctype_space, ctype_space,
# Line 174  static const uschar toptable1[] = { Line 261  static const uschar toptable1[] = {
261    0, 0                            /* OP_ANY, OP_ALLANY */    0, 0                            /* OP_ANY, OP_ALLANY */
262  };  };
263    
264  static const uschar toptable2[] = {  static const pcre_uint8 toptable2[] = {
265    0, 0, 0, 0, 0, 0,    0, 0, 0, 0, 0, 0,
266    ctype_digit, 0,    ctype_digit, 0,
267    ctype_space, 0,    ctype_space, 0,
# Line 191  these structures in, is a vector of ints Line 278  these structures in, is a vector of ints
278  typedef struct stateblock {  typedef struct stateblock {
279    int offset;                     /* Offset to opcode */    int offset;                     /* Offset to opcode */
280    int count;                      /* Count for repeats */    int count;                      /* Count for repeats */
   int ims;                        /* ims flag bits */  
281    int data;                       /* Some use extra data */    int data;                       /* Some use extra data */
282  } stateblock;  } stateblock;
283    
284  #define INTS_PER_STATEBLOCK  (sizeof(stateblock)/sizeof(int))  #define INTS_PER_STATEBLOCK  (int)(sizeof(stateblock)/sizeof(int))
285    
286    
287  #ifdef DEBUG  #ifdef PCRE_DEBUG
288  /*************************************************  /*************************************************
289  *             Print character string             *  *             Print character string             *
290  *************************************************/  *************************************************/
# Line 214  Returns:       nothing Line 300  Returns:       nothing
300  */  */
301    
302  static void  static void
303  pchars(unsigned char *p, int length, FILE *f)  pchars(const pcre_uchar *p, int length, FILE *f)
304  {  {
305  int c;  pcre_uint32 c;
306  while (length-- > 0)  while (length-- > 0)
307    {    {
308    if (isprint(c = *(p++)))    if (isprint(c = *(p++)))
309      fprintf(f, "%c", c);      fprintf(f, "%c", c);
310    else    else
311      fprintf(f, "\\x%02x", c);      fprintf(f, "\\x{%02x}", c);
312    }    }
313  }  }
314  #endif  #endif
# Line 247  Arguments: Line 333  Arguments:
333    offsetcount       size of same    offsetcount       size of same
334    workspace         vector of workspace    workspace         vector of workspace
335    wscount           size of same    wscount           size of same
   ims               the current ims flags  
336    rlevel            function call recursion level    rlevel            function call recursion level
   recursing         regex recursive call level  
337    
338  Returns:            > 0 => number of match offset pairs placed in offsets  Returns:            > 0 => number of match offset pairs placed in offsets
339                      = 0 => offsets overflowed; longest matches are present                      = 0 => offsets overflowed; longest matches are present
# Line 264  for the current character, one for the f Line 348  for the current character, one for the f
348      { \      { \
349      next_active_state->offset = (x); \      next_active_state->offset = (x); \
350      next_active_state->count  = (y); \      next_active_state->count  = (y); \
     next_active_state->ims    = ims; \  
351      next_active_state++; \      next_active_state++; \
352      DPRINTF(("%.*sADD_ACTIVE(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \      DPRINTF(("%.*sADD_ACTIVE(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \
353      } \      } \
# Line 275  for the current character, one for the f Line 358  for the current character, one for the f
358      { \      { \
359      next_active_state->offset = (x); \      next_active_state->offset = (x); \
360      next_active_state->count  = (y); \      next_active_state->count  = (y); \
     next_active_state->ims    = ims; \  
361      next_active_state->data   = (z); \      next_active_state->data   = (z); \
362      next_active_state++; \      next_active_state++; \
363      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 287  for the current character, one for the f Line 369  for the current character, one for the f
369      { \      { \
370      next_new_state->offset = (x); \      next_new_state->offset = (x); \
371      next_new_state->count  = (y); \      next_new_state->count  = (y); \
     next_new_state->ims    = ims; \  
372      next_new_state++; \      next_new_state++; \
373      DPRINTF(("%.*sADD_NEW(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \      DPRINTF(("%.*sADD_NEW(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \
374      } \      } \
# Line 298  for the current character, one for the f Line 379  for the current character, one for the f
379      { \      { \
380      next_new_state->offset = (x); \      next_new_state->offset = (x); \
381      next_new_state->count  = (y); \      next_new_state->count  = (y); \
     next_new_state->ims    = ims; \  
382      next_new_state->data   = (z); \      next_new_state->data   = (z); \
383      next_new_state++; \      next_new_state++; \
384      DPRINTF(("%.*sADD_NEW_DATA(%d,%d,%d)\n", rlevel*2-2, SP, (x), (y), (z))); \      DPRINTF(("%.*sADD_NEW_DATA(%d,%d,%d) line %d\n", rlevel*2-2, SP, \
385          (x), (y), (z), __LINE__)); \
386      } \      } \
387    else return PCRE_ERROR_DFA_WSSIZE    else return PCRE_ERROR_DFA_WSSIZE
388    
# Line 310  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    BOOL reset_could_continue = FALSE;
428    
429  rlevel++;  rlevel++;
430  offsetcount &= (-2);  offsetcount &= (-2);
431    
# Line 351  wscount = (wscount - (wscount % (INTS_PE Line 434  wscount = (wscount - (wscount % (INTS_PE
434            (2 * INTS_PER_STATEBLOCK);            (2 * INTS_PER_STATEBLOCK);
435    
436  DPRINTF(("\n%.*s---------------------\n"  DPRINTF(("\n%.*s---------------------\n"
437    "%.*sCall to internal_dfa_exec f=%d r=%d\n",    "%.*sCall to internal_dfa_exec f=%d\n",
438    rlevel*2-2, SP, rlevel*2-2, SP, rlevel, recursing));    rlevel*2-2, SP, rlevel*2-2, SP, rlevel));
439    
440  ctypes = md->tables + ctypes_offset;  ctypes = md->tables + ctypes_offset;
441  lcc = md->tables + lcc_offset;  lcc = md->tables + lcc_offset;
# Line 365  next_new_state = new_states = active_sta Line 448  next_new_state = new_states = active_sta
448  new_count = 0;  new_count = 0;
449    
450  first_op = this_start_code + 1 + LINK_SIZE +  first_op = this_start_code + 1 + LINK_SIZE +
451    ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA)? 2:0);    ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA ||
452        *this_start_code == OP_CBRAPOS || *this_start_code == OP_SCBRAPOS)
453        ? IMM2_SIZE:0);
454    
455  /* 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
456  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 393  if (*first_op == OP_REVERSE) Line 478  if (*first_op == OP_REVERSE)
478    /* 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
479    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. */
480    
481  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
482    /* In character mode we have to step back character by character */    /* In character mode we have to step back character by character */
483    
484    if (utf8)    if (utf)
485      {      {
486      for (gone_back = 0; gone_back < max_back; gone_back++)      for (gone_back = 0; gone_back < max_back; gone_back++)
487        {        {
488        if (current_subject <= start_subject) break;        if (current_subject <= start_subject) break;
489        current_subject--;        current_subject--;
490        while (current_subject > start_subject &&        ACROSSCHAR(current_subject > start_subject, *current_subject, current_subject--);
              (*current_subject & 0xc0) == 0x80)  
         current_subject--;  
491        }        }
492      }      }
493    else    else
# Line 414  if (*first_op == OP_REVERSE) Line 497  if (*first_op == OP_REVERSE)
497    
498      {      {
499      gone_back = (current_subject - max_back < start_subject)?      gone_back = (current_subject - max_back < start_subject)?
500        current_subject - start_subject : max_back;        (int)(current_subject - start_subject) : max_back;
501      current_subject -= gone_back;      current_subject -= gone_back;
502      }      }
503    
504    /* Save the earliest consulted character */    /* Save the earliest consulted character */
505    
506    if (current_subject < md->start_used_ptr)    if (current_subject < md->start_used_ptr)
507      md->start_used_ptr = current_subject;      md->start_used_ptr = current_subject;
508    
509    /* Now we can process the individual branches. */    /* Now we can process the individual branches. */
510    
# Line 431  if (*first_op == OP_REVERSE) Line 514  if (*first_op == OP_REVERSE)
514      int back = GET(end_code, 2+LINK_SIZE);      int back = GET(end_code, 2+LINK_SIZE);
515      if (back <= gone_back)      if (back <= gone_back)
516        {        {
517        int bstate = end_code - start_code + 2 + 2*LINK_SIZE;        int bstate = (int)(end_code - start_code + 2 + 2*LINK_SIZE);
518        ADD_NEW_DATA(-bstate, 0, gone_back - back);        ADD_NEW_DATA(-bstate, 0, gone_back - back);
519        }        }
520      end_code += GET(end_code, 1);      end_code += GET(end_code, 1);
# Line 464  else Line 547  else
547    else    else
548      {      {
549      int length = 1 + LINK_SIZE +      int length = 1 + LINK_SIZE +
550        ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA)? 2:0);        ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA ||
551            *this_start_code == OP_CBRAPOS || *this_start_code == OP_SCBRAPOS)
552            ? IMM2_SIZE:0);
553      do      do
554        {        {
555        ADD_NEW(end_code - start_code + length, 0);        ADD_NEW((int)(end_code - start_code + length), 0);
556        end_code += GET(end_code, 1);        end_code += GET(end_code, 1);
557        length = 1 + LINK_SIZE;        length = 1 + LINK_SIZE;
558        }        }
# Line 477  else Line 562  else
562    
563  workspace[0] = 0;    /* Bit indicating which vector is current */  workspace[0] = 0;    /* Bit indicating which vector is current */
564    
565  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)));
566    
567  /* Loop for scanning the subject */  /* Loop for scanning the subject */
568    
# Line 486  for (;;) Line 571  for (;;)
571    {    {
572    int i, j;    int i, j;
573    int clen, dlen;    int clen, dlen;
574    unsigned int c, d;    pcre_uint32 c, d;
575    int forced_fail = 0;    int forced_fail = 0;
576    int reached_end = 0;    BOOL partial_newline = FALSE;
577      BOOL could_continue = reset_could_continue;
578      reset_could_continue = FALSE;
579    
580    /* Make the new state list into the active state list and empty the    /* Make the new state list into the active state list and empty the
581    new state list. */    new state list. */
# Line 502  for (;;) Line 589  for (;;)
589    workspace[0] ^= 1;              /* Remember for the restarting feature */    workspace[0] ^= 1;              /* Remember for the restarting feature */
590    workspace[1] = active_count;    workspace[1] = active_count;
591    
592  #ifdef DEBUG  #ifdef PCRE_DEBUG
593    printf("%.*sNext character: rest of subject = \"", rlevel*2-2, SP);    printf("%.*sNext character: rest of subject = \"", rlevel*2-2, SP);
594    pchars((uschar *)ptr, strlen((char *)ptr), stdout);    pchars(ptr, STRLEN_UC(ptr), stdout);
595    printf("\"\n");    printf("\"\n");
596    
597    printf("%.*sActive states: ", rlevel*2-2, SP);    printf("%.*sActive states: ", rlevel*2-2, SP);
# Line 524  for (;;) Line 611  for (;;)
611    
612    if (ptr < end_subject)    if (ptr < end_subject)
613      {      {
614      clen = 1;        /* Number of bytes in the character */      clen = 1;        /* Number of data items in the character */
615  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
616      if (utf8) { GETCHARLEN(c, ptr, clen); } else      GETCHARLENTEST(c, ptr, clen);
617  #endif  /* SUPPORT_UTF8 */  #else
618      c = *ptr;      c = *ptr;
619    #endif  /* SUPPORT_UTF */
620      }      }
621    else    else
622      {      {
# Line 544  for (;;) Line 632  for (;;)
632    for (i = 0; i < active_count; i++)    for (i = 0; i < active_count; i++)
633      {      {
634      stateblock *current_state = active_states + i;      stateblock *current_state = active_states + i;
635      const uschar *code;      BOOL caseless = FALSE;
636        const pcre_uchar *code;
637      int state_offset = current_state->offset;      int state_offset = current_state->offset;
638      int count, codevalue, rrc;      int codevalue, rrc;
639        unsigned int count;
640    
641  #ifdef DEBUG  #ifdef PCRE_DEBUG
642      printf ("%.*sProcessing state %d c=", rlevel*2-2, SP, state_offset);      printf ("%.*sProcessing state %d c=", rlevel*2-2, SP, state_offset);
643      if (clen == 0) printf("EOL\n");      if (clen == 0) printf("EOL\n");
644        else if (c > 32 && c < 127) printf("'%c'\n", c);        else if (c > 32 && c < 127) printf("'%c'\n", c);
645          else printf("0x%02x\n", c);          else printf("0x%02x\n", c);
646  #endif  #endif
647    
     /* This variable is referred to implicity in the ADD_xxx macros. */  
   
     ims = current_state->ims;  
   
648      /* 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
649      (negated) state until the number of characters in the data field have      (negated) state until the number of characters in the data field have
650      been skipped". */      been skipped". If the could_continue flag was passed over from a previous
651        state, arrange for it to passed on. */
652    
653      if (state_offset < 0)      if (state_offset < 0)
654        {        {
# Line 570  for (;;) Line 657  for (;;)
657          DPRINTF(("%.*sSkipping this character\n", rlevel*2-2, SP));          DPRINTF(("%.*sSkipping this character\n", rlevel*2-2, SP));
658          ADD_NEW_DATA(state_offset, current_state->count,          ADD_NEW_DATA(state_offset, current_state->count,
659            current_state->data - 1);            current_state->data - 1);
660            if (could_continue) reset_could_continue = TRUE;
661          continue;          continue;
662          }          }
663        else        else
# Line 578  for (;;) Line 666  for (;;)
666          }          }
667        }        }
668    
669      /* Check for a duplicate state with the same count, and skip if found.      /* Check for a duplicate state with the same count, and skip if found.
670      See the note at the head of this module about the possibility of improving      See the note at the head of this module about the possibility of improving
671      performance here. */      performance here. */
672    
# Line 597  for (;;) Line 685  for (;;)
685      code = start_code + state_offset;      code = start_code + state_offset;
686      codevalue = *code;      codevalue = *code;
687    
688        /* If this opcode inspects a character, but we are at the end of the
689        subject, remember the fact for use when testing for a partial match. */
690    
691        if (clen == 0 && poptable[codevalue] != 0)
692          could_continue = TRUE;
693    
694      /* If this opcode is followed by an inline character, load it. It is      /* If this opcode is followed by an inline character, load it. It is
695      tempting to test for the presence of a subject character here, but that      tempting to test for the presence of a subject character here, but that
696      is wrong, because sometimes zero repetitions of the subject are      is wrong, because sometimes zero repetitions of the subject are
697      permitted.      permitted.
698    
699      We also use this mechanism for opcodes such as OP_TYPEPLUS that take an      We also use this mechanism for opcodes such as OP_TYPEPLUS that take an
700      argument that is not a data character - but is always one byte long. We      argument that is not a data character - but is always one byte long because
701      have to take special action to deal with  \P, \p, \H, \h, \V, \v and \X in      the values are small. We have to take special action to deal with  \P, \p,
702      this case. To keep the other cases fast, convert these ones to new opcodes.      \H, \h, \V, \v and \X in this case. To keep the other cases fast, convert
703      */      these ones to new opcodes. */
704    
705      if (coptable[codevalue] > 0)      if (coptable[codevalue] > 0)
706        {        {
707        dlen = 1;        dlen = 1;
708  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
709        if (utf8) { GETCHARLEN(d, (code + coptable[codevalue]), dlen); } else        if (utf) { GETCHARLEN(d, (code + coptable[codevalue]), dlen); } else
710  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
711        d = code[coptable[codevalue]];        d = code[coptable[codevalue]];
712        if (codevalue >= OP_TYPESTAR)        if (codevalue >= OP_TYPESTAR)
713          {          {
# Line 643  for (;;) Line 737  for (;;)
737    
738      switch (codevalue)      switch (codevalue)
739        {        {
740    /* ========================================================================== */
741          /* These cases are never obeyed. This is a fudge that causes a compile-
742          time error if the vectors coptable or poptable, which are indexed by
743          opcode, are not the correct length. It seems to be the only way to do
744          such a check at compile time, as the sizeof() operator does not work
745          in the C preprocessor. */
746    
747          case OP_TABLE_LENGTH:
748          case OP_TABLE_LENGTH +
749            ((sizeof(coptable) == OP_TABLE_LENGTH) &&
750             (sizeof(poptable) == OP_TABLE_LENGTH)):
751          break;
752    
753  /* ========================================================================== */  /* ========================================================================== */
754        /* 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
755        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
756        PCRE_NOTEMPTY is set, or PCRE_NOTEMPTY_ATSTART is set and we are at the        state. Note that KETRPOS will always be encountered at the end of the
757          subpattern, because the possessive subpattern repeats are always handled
758          using recursive calls. Thus, it never adds any new states.
759    
760          At the end of the (sub)pattern, unless we have an empty string and
761          PCRE_NOTEMPTY is set, or PCRE_NOTEMPTY_ATSTART is set and we are at the
762        start of the subject, save the match data, shifting up all previous        start of the subject, save the match data, shifting up all previous
763        matches so we always have the longest first. */        matches so we always have the longest first. */
764    
765        case OP_KET:        case OP_KET:
766        case OP_KETRMIN:        case OP_KETRMIN:
767        case OP_KETRMAX:        case OP_KETRMAX:
768          case OP_KETRPOS:
769        if (code != end_code)        if (code != end_code)
770          {          {
771          ADD_ACTIVE(state_offset + 1 + LINK_SIZE, 0);          ADD_ACTIVE(state_offset + 1 + LINK_SIZE, 0);
# Line 662  for (;;) Line 774  for (;;)
774            ADD_ACTIVE(state_offset - GET(code, 1), 0);            ADD_ACTIVE(state_offset - GET(code, 1), 0);
775            }            }
776          }          }
777        else        else
778          {          {
779          reached_end++;    /* Count branches that reach the end */          if (ptr > current_subject ||
         if (ptr > current_subject ||  
780              ((md->moptions & PCRE_NOTEMPTY) == 0 &&              ((md->moptions & PCRE_NOTEMPTY) == 0 &&
781                ((md->moptions & PCRE_NOTEMPTY_ATSTART) == 0 ||                ((md->moptions & PCRE_NOTEMPTY_ATSTART) == 0 ||
782                  current_subject > start_subject + md->start_offset)))                  current_subject > start_subject + md->start_offset)))
783            {            {
784            if (match_count < 0) match_count = (offsetcount >= 2)? 1 : 0;            if (match_count < 0) match_count = (offsetcount >= 2)? 1 : 0;
785              else if (match_count > 0 && ++match_count * 2 >= offsetcount)              else if (match_count > 0 && ++match_count * 2 > offsetcount)
786                match_count = 0;                match_count = 0;
787            count = ((match_count == 0)? offsetcount : match_count * 2) - 2;            count = ((match_count == 0)? offsetcount : match_count * 2) - 2;
788            if (count > 0) memmove(offsets + 2, offsets, count * sizeof(int));            if (count > 0) memmove(offsets + 2, offsets, count * sizeof(int));
789            if (offsetcount >= 2)            if (offsetcount >= 2)
790              {              {
791              offsets[0] = current_subject - start_subject;              offsets[0] = (int)(current_subject - start_subject);
792              offsets[1] = ptr - start_subject;              offsets[1] = (int)(ptr - start_subject);
793              DPRINTF(("%.*sSet matched string = \"%.*s\"\n", rlevel*2-2, SP,              DPRINTF(("%.*sSet matched string = \"%.*s\"\n", rlevel*2-2, SP,
794                offsets[1] - offsets[0], current_subject));                offsets[1] - offsets[0], (char *)current_subject));
795              }              }
796            if ((md->moptions & PCRE_DFA_SHORTEST) != 0)            if ((md->moptions & PCRE_DFA_SHORTEST) != 0)
797              {              {
# Line 689  for (;;) Line 800  for (;;)
800                match_count, rlevel*2-2, SP));                match_count, rlevel*2-2, SP));
801              return match_count;              return match_count;
802              }              }
803            }            }
804          }          }
805        break;        break;
806    
# Line 700  for (;;) Line 811  for (;;)
811        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
812        case OP_ALT:        case OP_ALT:
813        do { code += GET(code, 1); } while (*code == OP_ALT);        do { code += GET(code, 1); } while (*code == OP_ALT);
814        ADD_ACTIVE(code - start_code, 0);        ADD_ACTIVE((int)(code - start_code), 0);
815        break;        break;
816    
817        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
# Line 708  for (;;) Line 819  for (;;)
819        case OP_SBRA:        case OP_SBRA:
820        do        do
821          {          {
822          ADD_ACTIVE(code - start_code + 1 + LINK_SIZE, 0);          ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE), 0);
823          code += GET(code, 1);          code += GET(code, 1);
824          }          }
825        while (*code == OP_ALT);        while (*code == OP_ALT);
# Line 717  for (;;) Line 828  for (;;)
828        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
829        case OP_CBRA:        case OP_CBRA:
830        case OP_SCBRA:        case OP_SCBRA:
831        ADD_ACTIVE(code - start_code + 3 + LINK_SIZE,  0);        ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE + IMM2_SIZE),  0);
832        code += GET(code, 1);        code += GET(code, 1);
833        while (*code == OP_ALT)        while (*code == OP_ALT)
834          {          {
835          ADD_ACTIVE(code - start_code + 1 + LINK_SIZE,  0);          ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE),  0);
836          code += GET(code, 1);          code += GET(code, 1);
837          }          }
838        break;        break;
# Line 732  for (;;) Line 843  for (;;)
843        ADD_ACTIVE(state_offset + 1, 0);        ADD_ACTIVE(state_offset + 1, 0);
844        code += 1 + GET(code, 2);        code += 1 + GET(code, 2);
845        while (*code == OP_ALT) code += GET(code, 1);        while (*code == OP_ALT) code += GET(code, 1);
846        ADD_ACTIVE(code - start_code + 1 + LINK_SIZE, 0);        ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE), 0);
847        break;        break;
848    
849        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
850        case OP_SKIPZERO:        case OP_SKIPZERO:
851        code += 1 + GET(code, 2);        code += 1 + GET(code, 2);
852        while (*code == OP_ALT) code += GET(code, 1);        while (*code == OP_ALT) code += GET(code, 1);
853        ADD_ACTIVE(code - start_code + 1 + LINK_SIZE, 0);        ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE), 0);
854        break;        break;
855    
856        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
857        case OP_CIRC:        case OP_CIRC:
858        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)))  
859          { ADD_ACTIVE(state_offset + 1, 0); }          { ADD_ACTIVE(state_offset + 1, 0); }
860        break;        break;
861    
862        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
863        case OP_EOD:        case OP_CIRCM:
864        if (ptr >= end_subject) { ADD_ACTIVE(state_offset + 1, 0); }        if ((ptr == start_subject && (md->moptions & PCRE_NOTBOL) == 0) ||
865              (ptr != end_subject && WAS_NEWLINE(ptr)))
866            { ADD_ACTIVE(state_offset + 1, 0); }
867        break;        break;
868    
869        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
870        case OP_OPT:        case OP_EOD:
871        ims = code[1];        if (ptr >= end_subject)
872        ADD_ACTIVE(state_offset + 2, 0);          {
873            if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
874              could_continue = TRUE;
875            else { ADD_ACTIVE(state_offset + 1, 0); }
876            }
877        break;        break;
878    
879        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
# Line 782  for (;;) Line 896  for (;;)
896        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
897        case OP_ANY:        case OP_ANY:
898        if (clen > 0 && !IS_NEWLINE(ptr))        if (clen > 0 && !IS_NEWLINE(ptr))
899          { ADD_NEW(state_offset + 1, 0); }          {
900            if (ptr + 1 >= md->end_subject &&
901                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
902                NLBLOCK->nltype == NLTYPE_FIXED &&
903                NLBLOCK->nllen == 2 &&
904                c == NLBLOCK->nl[0])
905              {
906              could_continue = partial_newline = TRUE;
907              }
908            else
909              {
910              ADD_NEW(state_offset + 1, 0);
911              }
912            }
913        break;        break;
914    
915        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
# Line 793  for (;;) Line 920  for (;;)
920    
921        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
922        case OP_EODN:        case OP_EODN:
923        if (clen == 0 || (IS_NEWLINE(ptr) && ptr == end_subject - md->nllen))        if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
924            could_continue = TRUE;
925          else if (clen == 0 || (IS_NEWLINE(ptr) && ptr == end_subject - md->nllen))
926          { ADD_ACTIVE(state_offset + 1, 0); }          { ADD_ACTIVE(state_offset + 1, 0); }
927        break;        break;
928    
# Line 801  for (;;) Line 930  for (;;)
930        case OP_DOLL:        case OP_DOLL:
931        if ((md->moptions & PCRE_NOTEOL) == 0)        if ((md->moptions & PCRE_NOTEOL) == 0)
932          {          {
933          if (clen == 0 ||          if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
934              could_continue = TRUE;
935            else if (clen == 0 ||
936              ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr) &&              ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr) &&
937                 ((ims & PCRE_MULTILINE) != 0 || ptr == end_subject - md->nllen)                 (ptr == end_subject - md->nllen)
938              ))              ))
939            { ADD_ACTIVE(state_offset + 1, 0); }            { ADD_ACTIVE(state_offset + 1, 0); }
940            else if (ptr + 1 >= md->end_subject &&
941                     (md->moptions & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) != 0 &&
942                     NLBLOCK->nltype == NLTYPE_FIXED &&
943                     NLBLOCK->nllen == 2 &&
944                     c == NLBLOCK->nl[0])
945              {
946              if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
947                {
948                reset_could_continue = TRUE;
949                ADD_NEW_DATA(-(state_offset + 1), 0, 1);
950                }
951              else could_continue = partial_newline = TRUE;
952              }
953          }          }
954        else if ((ims & PCRE_MULTILINE) != 0 && IS_NEWLINE(ptr))        break;
955    
956          /*-----------------------------------------------------------------*/
957          case OP_DOLLM:
958          if ((md->moptions & PCRE_NOTEOL) == 0)
959            {
960            if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
961              could_continue = TRUE;
962            else if (clen == 0 ||
963                ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr)))
964              { ADD_ACTIVE(state_offset + 1, 0); }
965            else if (ptr + 1 >= md->end_subject &&
966                     (md->moptions & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) != 0 &&
967                     NLBLOCK->nltype == NLTYPE_FIXED &&
968                     NLBLOCK->nllen == 2 &&
969                     c == NLBLOCK->nl[0])
970              {
971              if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
972                {
973                reset_could_continue = TRUE;
974                ADD_NEW_DATA(-(state_offset + 1), 0, 1);
975                }
976              else could_continue = partial_newline = TRUE;
977              }
978            }
979          else if (IS_NEWLINE(ptr))
980          { ADD_ACTIVE(state_offset + 1, 0); }          { ADD_ACTIVE(state_offset + 1, 0); }
981        break;        break;
982    
# Line 838  for (;;) Line 1007  for (;;)
1007    
1008          if (ptr > start_subject)          if (ptr > start_subject)
1009            {            {
1010            const uschar *temp = ptr - 1;            const pcre_uchar *temp = ptr - 1;
1011            if (temp < md->start_used_ptr) md->start_used_ptr = temp;            if (temp < md->start_used_ptr) md->start_used_ptr = temp;
1012  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
1013            if (utf8) BACKCHAR(temp);            if (utf) { BACKCHAR(temp); }
1014  #endif  #endif
1015            GETCHARTEST(d, temp);            GETCHARTEST(d, temp);
1016    #ifdef SUPPORT_UCP
1017              if ((md->poptions & PCRE_UCP) != 0)
1018                {
1019                if (d == '_') left_word = TRUE; else
1020                  {
1021                  int cat = UCD_CATEGORY(d);
1022                  left_word = (cat == ucp_L || cat == ucp_N);
1023                  }
1024                }
1025              else
1026    #endif
1027            left_word = d < 256 && (ctypes[d] & ctype_word) != 0;            left_word = d < 256 && (ctypes[d] & ctype_word) != 0;
1028            }            }
1029          else left_word = 0;          else left_word = FALSE;
1030    
1031          if (clen > 0)          if (clen > 0)
1032              {
1033    #ifdef SUPPORT_UCP
1034              if ((md->poptions & PCRE_UCP) != 0)
1035                {
1036                if (c == '_') right_word = TRUE; else
1037                  {
1038                  int cat = UCD_CATEGORY(c);
1039                  right_word = (cat == ucp_L || cat == ucp_N);
1040                  }
1041                }
1042              else
1043    #endif
1044            right_word = c < 256 && (ctypes[c] & ctype_word) != 0;            right_word = c < 256 && (ctypes[c] & ctype_word) != 0;
1045          else              /* This is a fudge to ensure that if this is the */            }
1046            {               /* last item in the pattern, we don't count it as */          else right_word = FALSE;
           reached_end--;  /* reached, thus disabling a partial match. */  
           right_word = 0;  
           }  
1047    
1048          if ((left_word == right_word) == (codevalue == OP_NOT_WORD_BOUNDARY))          if ((left_word == right_word) == (codevalue == OP_NOT_WORD_BOUNDARY))
1049            { ADD_ACTIVE(state_offset + 1, 0); }            { ADD_ACTIVE(state_offset + 1, 0); }
# Line 873  for (;;) Line 1062  for (;;)
1062        if (clen > 0)        if (clen > 0)
1063          {          {
1064          BOOL OK;          BOOL OK;
1065            const pcre_uint32 *cp;
1066          const ucd_record * prop = GET_UCD(c);          const ucd_record * prop = GET_UCD(c);
1067          switch(code[1])          switch(code[1])
1068            {            {
# Line 881  for (;;) Line 1071  for (;;)
1071            break;            break;
1072    
1073            case PT_LAMP:            case PT_LAMP:
1074            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
1075                   prop->chartype == ucp_Lt;
1076            break;            break;
1077    
1078            case PT_GC:            case PT_GC:
1079            OK = _pcre_ucp_gentype[prop->chartype] == code[2];            OK = PRIV(ucp_gentype)[prop->chartype] == code[2];
1080            break;            break;
1081    
1082            case PT_PC:            case PT_PC:
# Line 896  for (;;) Line 1087  for (;;)
1087            OK = prop->script == code[2];            OK = prop->script == code[2];
1088            break;            break;
1089    
1090              /* These are specials for combination cases. */
1091    
1092              case PT_ALNUM:
1093              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1094                   PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1095              break;
1096    
1097              case PT_SPACE:    /* Perl space */
1098              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1099                   c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1100              break;
1101    
1102              case PT_PXSPACE:  /* POSIX space */
1103              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1104                   c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1105                   c == CHAR_FF || c == CHAR_CR;
1106              break;
1107    
1108              case PT_WORD:
1109              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1110                   PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1111                   c == CHAR_UNDERSCORE;
1112              break;
1113    
1114              case PT_CLIST:
1115              cp = PRIV(ucd_caseless_sets) + code[2];
1116              for (;;)
1117                {
1118                if (c < *cp) { OK = FALSE; break; }
1119                if (c == *cp++) { OK = TRUE; break; }
1120                }
1121              break;
1122    
1123            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
1124    
1125            default:            default:
# Line 923  for (;;) Line 1147  for (;;)
1147        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
1148        if (clen > 0)        if (clen > 0)
1149          {          {
1150          if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||          if (d == OP_ANY && ptr + 1 >= md->end_subject &&
1151                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
1152                NLBLOCK->nltype == NLTYPE_FIXED &&
1153                NLBLOCK->nllen == 2 &&
1154                c == NLBLOCK->nl[0])
1155              {
1156              could_continue = partial_newline = TRUE;
1157              }
1158            else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1159              (c < 256 &&              (c < 256 &&
1160                (d != OP_ANY || !IS_NEWLINE(ptr)) &&                (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1161                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
# Line 946  for (;;) Line 1178  for (;;)
1178        ADD_ACTIVE(state_offset + 2, 0);        ADD_ACTIVE(state_offset + 2, 0);
1179        if (clen > 0)        if (clen > 0)
1180          {          {
1181          if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||          if (d == OP_ANY && ptr + 1 >= md->end_subject &&
1182                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
1183                NLBLOCK->nltype == NLTYPE_FIXED &&
1184                NLBLOCK->nllen == 2 &&
1185                c == NLBLOCK->nl[0])
1186              {
1187              could_continue = partial_newline = TRUE;
1188              }
1189            else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1190              (c < 256 &&              (c < 256 &&
1191                (d != OP_ANY || !IS_NEWLINE(ptr)) &&                (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1192                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
# Line 968  for (;;) Line 1208  for (;;)
1208        ADD_ACTIVE(state_offset + 2, 0);        ADD_ACTIVE(state_offset + 2, 0);
1209        if (clen > 0)        if (clen > 0)
1210          {          {
1211          if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||          if (d == OP_ANY && ptr + 1 >= md->end_subject &&
1212                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
1213                NLBLOCK->nltype == NLTYPE_FIXED &&
1214                NLBLOCK->nllen == 2 &&
1215                c == NLBLOCK->nl[0])
1216              {
1217              could_continue = partial_newline = TRUE;
1218              }
1219            else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1220              (c < 256 &&              (c < 256 &&
1221                (d != OP_ANY || !IS_NEWLINE(ptr)) &&                (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1222                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
# Line 988  for (;;) Line 1236  for (;;)
1236        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1237        if (clen > 0)        if (clen > 0)
1238          {          {
1239          if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||          if (d == OP_ANY && ptr + 1 >= md->end_subject &&
1240                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
1241                NLBLOCK->nltype == NLTYPE_FIXED &&
1242                NLBLOCK->nllen == 2 &&
1243                c == NLBLOCK->nl[0])
1244              {
1245              could_continue = partial_newline = TRUE;
1246              }
1247            else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1248              (c < 256 &&              (c < 256 &&
1249                (d != OP_ANY || !IS_NEWLINE(ptr)) &&                (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1250                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
1251            {            {
1252            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1253              { ADD_NEW(state_offset + 4, 0); }              { ADD_NEW(state_offset + 1 + IMM2_SIZE + 1, 0); }
1254            else            else
1255              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
1256            }            }
# Line 1005  for (;;) Line 1261  for (;;)
1261        case OP_TYPEUPTO:        case OP_TYPEUPTO:
1262        case OP_TYPEMINUPTO:        case OP_TYPEMINUPTO:
1263        case OP_TYPEPOSUPTO:        case OP_TYPEPOSUPTO:
1264        ADD_ACTIVE(state_offset + 4, 0);        ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0);
1265        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1266        if (clen > 0)        if (clen > 0)
1267          {          {
1268          if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||          if (d == OP_ANY && ptr + 1 >= md->end_subject &&
1269                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
1270                NLBLOCK->nltype == NLTYPE_FIXED &&
1271                NLBLOCK->nllen == 2 &&
1272                c == NLBLOCK->nl[0])
1273              {
1274              could_continue = partial_newline = TRUE;
1275              }
1276            else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1277              (c < 256 &&              (c < 256 &&
1278                (d != OP_ANY || !IS_NEWLINE(ptr)) &&                (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1279                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
# Line 1020  for (;;) Line 1284  for (;;)
1284              next_active_state--;              next_active_state--;
1285              }              }
1286            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1287              { ADD_NEW(state_offset + 4, 0); }              { ADD_NEW(state_offset + 2 + IMM2_SIZE, 0); }
1288            else            else
1289              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
1290            }            }
# Line 1042  for (;;) Line 1306  for (;;)
1306        if (clen > 0)        if (clen > 0)
1307          {          {
1308          BOOL OK;          BOOL OK;
1309            const pcre_uint32 *cp;
1310          const ucd_record * prop = GET_UCD(c);          const ucd_record * prop = GET_UCD(c);
1311          switch(code[2])          switch(code[2])
1312            {            {
# Line 1050  for (;;) Line 1315  for (;;)
1315            break;            break;
1316    
1317            case PT_LAMP:            case PT_LAMP:
1318            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
1319                prop->chartype == ucp_Lt;
1320            break;            break;
1321    
1322            case PT_GC:            case PT_GC:
1323            OK = _pcre_ucp_gentype[prop->chartype] == code[3];            OK = PRIV(ucp_gentype)[prop->chartype] == code[3];
1324            break;            break;
1325    
1326            case PT_PC:            case PT_PC:
# Line 1065  for (;;) Line 1331  for (;;)
1331            OK = prop->script == code[3];            OK = prop->script == code[3];
1332            break;            break;
1333    
1334              /* These are specials for combination cases. */
1335    
1336              case PT_ALNUM:
1337              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1338                   PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1339              break;
1340    
1341              case PT_SPACE:    /* Perl space */
1342              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1343                   c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1344              break;
1345    
1346              case PT_PXSPACE:  /* POSIX space */
1347              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1348                   c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1349                   c == CHAR_FF || c == CHAR_CR;
1350              break;
1351    
1352              case PT_WORD:
1353              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1354                   PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1355                   c == CHAR_UNDERSCORE;
1356              break;
1357    
1358              case PT_CLIST:
1359              cp = PRIV(ucd_caseless_sets) + code[3];
1360              for (;;)
1361                {
1362                if (c < *cp) { OK = FALSE; break; }
1363                if (c == *cp++) { OK = TRUE; break; }
1364                }
1365              break;
1366    
1367            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
1368    
1369            default:            default:
# Line 1091  for (;;) Line 1390  for (;;)
1390        case OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS:        case OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS:
1391        count = current_state->count;  /* Already matched */        count = current_state->count;  /* Already matched */
1392        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
1393        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0)
1394          {          {
1395          const uschar *nptr = ptr + clen;          int lgb, rgb;
1396            const pcre_uchar *nptr = ptr + clen;
1397          int ncount = 0;          int ncount = 0;
1398          if (count > 0 && codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS)          if (count > 0 && codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS)
1399            {            {
1400            active_count--;           /* Remove non-match possibility */            active_count--;           /* Remove non-match possibility */
1401            next_active_state--;            next_active_state--;
1402            }            }
1403            lgb = UCD_GRAPHBREAK(c);
1404          while (nptr < end_subject)          while (nptr < end_subject)
1405            {            {
1406            int nd;            dlen = 1;
1407            int ndlen = 1;            if (!utf) d = *nptr; else { GETCHARLEN(d, nptr, dlen); }
1408            GETCHARLEN(nd, nptr, ndlen);            rgb = UCD_GRAPHBREAK(d);
1409            if (UCD_CATEGORY(nd) != ucp_M) break;            if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
1410            ncount++;            ncount++;
1411            nptr += ndlen;            lgb = rgb;
1412              nptr += dlen;
1413            }            }
1414          count++;          count++;
1415          ADD_NEW_DATA(-state_offset, count, ncount);          ADD_NEW_DATA(-state_offset, count, ncount);
# Line 1126  for (;;) Line 1428  for (;;)
1428          int ncount = 0;          int ncount = 0;
1429          switch (c)          switch (c)
1430            {            {
1431            case 0x000b:            case CHAR_VT:
1432            case 0x000c:            case CHAR_FF:
1433            case 0x0085:            case CHAR_NEL:
1434    #ifndef EBCDIC
1435            case 0x2028:            case 0x2028:
1436            case 0x2029:            case 0x2029:
1437    #endif  /* Not EBCDIC */
1438            if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;            if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
1439            goto ANYNL01;            goto ANYNL01;
1440    
1441            case 0x000d:            case CHAR_CR:
1442            if (ptr + 1 < end_subject && ptr[1] == 0x0a) ncount = 1;            if (ptr + 1 < end_subject && RAWUCHARTEST(ptr + 1) == CHAR_LF) ncount = 1;
1443            /* Fall through */            /* Fall through */
1444    
1445            ANYNL01:            ANYNL01:
1446            case 0x000a:            case CHAR_LF:
1447            if (count > 0 && codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSPLUS)            if (count > 0 && codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSPLUS)
1448              {              {
1449              active_count--;           /* Remove non-match possibility */              active_count--;           /* Remove non-match possibility */
# Line 1166  for (;;) Line 1470  for (;;)
1470          BOOL OK;          BOOL OK;
1471          switch (c)          switch (c)
1472            {            {
1473            case 0x000a:            VSPACE_CASES:
           case 0x000b:  
           case 0x000c:  
           case 0x000d:  
           case 0x0085:  
           case 0x2028:  
           case 0x2029:  
1474            OK = TRUE;            OK = TRUE;
1475            break;            break;
1476    
# Line 1205  for (;;) Line 1503  for (;;)
1503          BOOL OK;          BOOL OK;
1504          switch (c)          switch (c)
1505            {            {
1506            case 0x09:      /* HT */            HSPACE_CASES:
           case 0x20:      /* SPACE */  
           case 0xa0:      /* NBSP */  
           case 0x1680:    /* OGHAM SPACE MARK */  
           case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
           case 0x2000:    /* EN QUAD */  
           case 0x2001:    /* EM QUAD */  
           case 0x2002:    /* EN SPACE */  
           case 0x2003:    /* EM SPACE */  
           case 0x2004:    /* THREE-PER-EM SPACE */  
           case 0x2005:    /* FOUR-PER-EM SPACE */  
           case 0x2006:    /* SIX-PER-EM SPACE */  
           case 0x2007:    /* FIGURE SPACE */  
           case 0x2008:    /* PUNCTUATION SPACE */  
           case 0x2009:    /* THIN SPACE */  
           case 0x200A:    /* HAIR SPACE */  
           case 0x202f:    /* NARROW NO-BREAK SPACE */  
           case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
           case 0x3000:    /* IDEOGRAPHIC SPACE */  
1507            OK = TRUE;            OK = TRUE;
1508            break;            break;
1509    
# Line 1264  for (;;) Line 1544  for (;;)
1544        if (clen > 0)        if (clen > 0)
1545          {          {
1546          BOOL OK;          BOOL OK;
1547            const pcre_uint32 *cp;
1548          const ucd_record * prop = GET_UCD(c);          const ucd_record * prop = GET_UCD(c);
1549          switch(code[2])          switch(code[2])
1550            {            {
# Line 1272  for (;;) Line 1553  for (;;)
1553            break;            break;
1554    
1555            case PT_LAMP:            case PT_LAMP:
1556            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
1557                prop->chartype == ucp_Lt;
1558            break;            break;
1559    
1560            case PT_GC:            case PT_GC:
1561            OK = _pcre_ucp_gentype[prop->chartype] == code[3];            OK = PRIV(ucp_gentype)[prop->chartype] == code[3];
1562            break;            break;
1563    
1564            case PT_PC:            case PT_PC:
# Line 1287  for (;;) Line 1569  for (;;)
1569            OK = prop->script == code[3];            OK = prop->script == code[3];
1570            break;            break;
1571    
1572              /* These are specials for combination cases. */
1573    
1574              case PT_ALNUM:
1575              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1576                   PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1577              break;
1578    
1579              case PT_SPACE:    /* Perl space */
1580              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1581                   c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1582              break;
1583    
1584              case PT_PXSPACE:  /* POSIX space */
1585              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1586                   c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1587                   c == CHAR_FF || c == CHAR_CR;
1588              break;
1589    
1590              case PT_WORD:
1591              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1592                   PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1593                   c == CHAR_UNDERSCORE;
1594              break;
1595    
1596              case PT_CLIST:
1597              cp = PRIV(ucd_caseless_sets) + code[3];
1598              for (;;)
1599                {
1600                if (c < *cp) { OK = FALSE; break; }
1601                if (c == *cp++) { OK = TRUE; break; }
1602                }
1603              break;
1604    
1605            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
1606    
1607            default:            default:
# Line 1322  for (;;) Line 1637  for (;;)
1637        QS2:        QS2:
1638    
1639        ADD_ACTIVE(state_offset + 2, 0);        ADD_ACTIVE(state_offset + 2, 0);
1640        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0)
1641          {          {
1642          const uschar *nptr = ptr + clen;          int lgb, rgb;
1643            const pcre_uchar *nptr = ptr + clen;
1644          int ncount = 0;          int ncount = 0;
1645          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSSTAR ||          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSSTAR ||
1646              codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSQUERY)              codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSQUERY)
# Line 1332  for (;;) Line 1648  for (;;)
1648            active_count--;           /* Remove non-match possibility */            active_count--;           /* Remove non-match possibility */
1649            next_active_state--;            next_active_state--;
1650            }            }
1651            lgb = UCD_GRAPHBREAK(c);
1652          while (nptr < end_subject)          while (nptr < end_subject)
1653            {            {
1654            int nd;            dlen = 1;
1655            int ndlen = 1;            if (!utf) d = *nptr; else { GETCHARLEN(d, nptr, dlen); }
1656            GETCHARLEN(nd, nptr, ndlen);            rgb = UCD_GRAPHBREAK(d);
1657            if (UCD_CATEGORY(nd) != ucp_M) break;            if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
1658            ncount++;            ncount++;
1659            nptr += ndlen;            lgb = rgb;
1660              nptr += dlen;
1661            }            }
1662          ADD_NEW_DATA(-(state_offset + count), 0, ncount);          ADD_NEW_DATA(-(state_offset + count), 0, ncount);
1663          }          }
# Line 1365  for (;;) Line 1683  for (;;)
1683          int ncount = 0;          int ncount = 0;
1684          switch (c)          switch (c)
1685            {            {
1686            case 0x000b:            case CHAR_VT:
1687            case 0x000c:            case CHAR_FF:
1688            case 0x0085:            case CHAR_NEL:
1689    #ifndef EBCDIC
1690            case 0x2028:            case 0x2028:
1691            case 0x2029:            case 0x2029:
1692    #endif  /* Not EBCDIC */
1693            if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;            if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
1694            goto ANYNL02;            goto ANYNL02;
1695    
1696            case 0x000d:            case CHAR_CR:
1697            if (ptr + 1 < end_subject && ptr[1] == 0x0a) ncount = 1;            if (ptr + 1 < end_subject && RAWUCHARTEST(ptr + 1) == CHAR_LF) ncount = 1;
1698            /* Fall through */            /* Fall through */
1699    
1700            ANYNL02:            ANYNL02:
1701            case 0x000a:            case CHAR_LF:
1702            if (codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSSTAR ||            if (codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSSTAR ||
1703                codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSQUERY)                codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSQUERY)
1704              {              {
1705              active_count--;           /* Remove non-match possibility */              active_count--;           /* Remove non-match possibility */
1706              next_active_state--;              next_active_state--;
1707              }              }
1708            ADD_NEW_DATA(-(state_offset + count), 0, ncount);            ADD_NEW_DATA(-(state_offset + (int)count), 0, ncount);
1709            break;            break;
1710    
1711            default:            default:
# Line 1413  for (;;) Line 1733  for (;;)
1733          BOOL OK;          BOOL OK;
1734          switch (c)          switch (c)
1735            {            {
1736            case 0x000a:            VSPACE_CASES:
           case 0x000b:  
           case 0x000c:  
           case 0x000d:  
           case 0x0085:  
           case 0x2028:  
           case 0x2029:  
1737            OK = TRUE;            OK = TRUE;
1738            break;            break;
1739    
# Line 1435  for (;;) Line 1749  for (;;)
1749              active_count--;           /* Remove non-match possibility */              active_count--;           /* Remove non-match possibility */
1750              next_active_state--;              next_active_state--;
1751              }              }
1752            ADD_NEW_DATA(-(state_offset + count), 0, 0);            ADD_NEW_DATA(-(state_offset + (int)count), 0, 0);
1753            }            }
1754          }          }
1755        break;        break;
# Line 1459  for (;;) Line 1773  for (;;)
1773          BOOL OK;          BOOL OK;
1774          switch (c)          switch (c)
1775            {            {
1776            case 0x09:      /* HT */            HSPACE_CASES:
           case 0x20:      /* SPACE */  
           case 0xa0:      /* NBSP */  
           case 0x1680:    /* OGHAM SPACE MARK */  
           case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
           case 0x2000:    /* EN QUAD */  
           case 0x2001:    /* EM QUAD */  
           case 0x2002:    /* EN SPACE */  
           case 0x2003:    /* EM SPACE */  
           case 0x2004:    /* THREE-PER-EM SPACE */  
           case 0x2005:    /* FOUR-PER-EM SPACE */  
           case 0x2006:    /* SIX-PER-EM SPACE */  
           case 0x2007:    /* FIGURE SPACE */  
           case 0x2008:    /* PUNCTUATION SPACE */  
           case 0x2009:    /* THIN SPACE */  
           case 0x200A:    /* HAIR SPACE */  
           case 0x202f:    /* NARROW NO-BREAK SPACE */  
           case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
           case 0x3000:    /* IDEOGRAPHIC SPACE */  
1777            OK = TRUE;            OK = TRUE;
1778            break;            break;
1779    
# Line 1494  for (;;) Line 1790  for (;;)
1790              active_count--;           /* Remove non-match possibility */              active_count--;           /* Remove non-match possibility */
1791              next_active_state--;              next_active_state--;
1792              }              }
1793            ADD_NEW_DATA(-(state_offset + count), 0, 0);            ADD_NEW_DATA(-(state_offset + (int)count), 0, 0);
1794            }            }
1795          }          }
1796        break;        break;
# Line 1506  for (;;) Line 1802  for (;;)
1802        case OP_PROP_EXTRA + OP_TYPEMINUPTO:        case OP_PROP_EXTRA + OP_TYPEMINUPTO:
1803        case OP_PROP_EXTRA + OP_TYPEPOSUPTO:        case OP_PROP_EXTRA + OP_TYPEPOSUPTO:
1804        if (codevalue != OP_PROP_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_PROP_EXTRA + OP_TYPEEXACT)
1805          { ADD_ACTIVE(state_offset + 6, 0); }          { ADD_ACTIVE(state_offset + 1 + IMM2_SIZE + 3, 0); }
1806        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1807        if (clen > 0)        if (clen > 0)
1808          {          {
1809          BOOL OK;          BOOL OK;
1810            const pcre_uint32 *cp;
1811          const ucd_record * prop = GET_UCD(c);          const ucd_record * prop = GET_UCD(c);
1812          switch(code[4])          switch(code[1 + IMM2_SIZE + 1])
1813            {            {
1814            case PT_ANY:            case PT_ANY:
1815            OK = TRUE;            OK = TRUE;
1816            break;            break;
1817    
1818            case PT_LAMP:            case PT_LAMP:
1819            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
1820                prop->chartype == ucp_Lt;
1821            break;            break;
1822    
1823            case PT_GC:            case PT_GC:
1824            OK = _pcre_ucp_gentype[prop->chartype] == code[5];            OK = PRIV(ucp_gentype)[prop->chartype] == code[1 + IMM2_SIZE + 2];
1825            break;            break;
1826    
1827            case PT_PC:            case PT_PC:
1828            OK = prop->chartype == code[5];            OK = prop->chartype == code[1 + IMM2_SIZE + 2];
1829            break;            break;
1830    
1831            case PT_SC:            case PT_SC:
1832            OK = prop->script == code[5];            OK = prop->script == code[1 + IMM2_SIZE + 2];
1833              break;
1834    
1835              /* These are specials for combination cases. */
1836    
1837              case PT_ALNUM:
1838              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1839                   PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1840              break;
1841    
1842              case PT_SPACE:    /* Perl space */
1843              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1844                   c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1845              break;
1846    
1847              case PT_PXSPACE:  /* POSIX space */
1848              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1849                   c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1850                   c == CHAR_FF || c == CHAR_CR;
1851              break;
1852    
1853              case PT_WORD:
1854              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1855                   PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1856                   c == CHAR_UNDERSCORE;
1857              break;
1858    
1859              case PT_CLIST:
1860              cp = PRIV(ucd_caseless_sets) + code[1 + IMM2_SIZE + 2];
1861              for (;;)
1862                {
1863                if (c < *cp) { OK = FALSE; break; }
1864                if (c == *cp++) { OK = TRUE; break; }
1865                }
1866            break;            break;
1867    
1868            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
# Line 1549  for (;;) Line 1880  for (;;)
1880              next_active_state--;              next_active_state--;
1881              }              }
1882            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1883              { ADD_NEW(state_offset + 6, 0); }              { ADD_NEW(state_offset + 1 + IMM2_SIZE + 3, 0); }
1884            else            else
1885              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
1886            }            }
# Line 1562  for (;;) Line 1893  for (;;)
1893        case OP_EXTUNI_EXTRA + OP_TYPEMINUPTO:        case OP_EXTUNI_EXTRA + OP_TYPEMINUPTO:
1894        case OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO:        case OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO:
1895        if (codevalue != OP_EXTUNI_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_EXTUNI_EXTRA + OP_TYPEEXACT)
1896          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
1897        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1898        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0)
1899          {          {
1900          const uschar *nptr = ptr + clen;          int lgb, rgb;
1901            const pcre_uchar *nptr = ptr + clen;
1902          int ncount = 0;          int ncount = 0;
1903          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO)          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO)
1904            {            {
1905            active_count--;           /* Remove non-match possibility */            active_count--;           /* Remove non-match possibility */
1906            next_active_state--;            next_active_state--;
1907            }            }
1908            lgb = UCD_GRAPHBREAK(c);
1909          while (nptr < end_subject)          while (nptr < end_subject)
1910            {            {
1911            int nd;            dlen = 1;
1912            int ndlen = 1;            if (!utf) d = *nptr; else { GETCHARLEN(d, nptr, dlen); }
1913            GETCHARLEN(nd, nptr, ndlen);            rgb = UCD_GRAPHBREAK(d);
1914            if (UCD_CATEGORY(nd) != ucp_M) break;            if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
1915            ncount++;            ncount++;
1916            nptr += ndlen;            lgb = rgb;
1917              nptr += dlen;
1918            }            }
1919            if (nptr >= end_subject && (md->moptions & PCRE_PARTIAL_HARD) != 0)
1920                reset_could_continue = TRUE;
1921          if (++count >= GET2(code, 1))          if (++count >= GET2(code, 1))
1922            { ADD_NEW_DATA(-(state_offset + 4), 0, ncount); }            { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, ncount); }
1923          else          else
1924            { ADD_NEW_DATA(-state_offset, count, ncount); }            { ADD_NEW_DATA(-state_offset, count, ncount); }
1925          }          }
# Line 1596  for (;;) Line 1932  for (;;)
1932        case OP_ANYNL_EXTRA + OP_TYPEMINUPTO:        case OP_ANYNL_EXTRA + OP_TYPEMINUPTO:
1933        case OP_ANYNL_EXTRA + OP_TYPEPOSUPTO:        case OP_ANYNL_EXTRA + OP_TYPEPOSUPTO:
1934        if (codevalue != OP_ANYNL_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_ANYNL_EXTRA + OP_TYPEEXACT)
1935          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
1936        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1937        if (clen > 0)        if (clen > 0)
1938          {          {
1939          int ncount = 0;          int ncount = 0;
1940          switch (c)          switch (c)
1941            {            {
1942            case 0x000b:            case CHAR_VT:
1943            case 0x000c:            case CHAR_FF:
1944            case 0x0085:            case CHAR_NEL:
1945    #ifndef EBCDIC
1946            case 0x2028:            case 0x2028:
1947            case 0x2029:            case 0x2029:
1948    #endif  /* Not EBCDIC */
1949            if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;            if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
1950            goto ANYNL03;            goto ANYNL03;
1951    
1952            case 0x000d:            case CHAR_CR:
1953            if (ptr + 1 < end_subject && ptr[1] == 0x0a) ncount = 1;            if (ptr + 1 < end_subject && RAWUCHARTEST(ptr + 1) == CHAR_LF) ncount = 1;
1954            /* Fall through */            /* Fall through */
1955    
1956            ANYNL03:            ANYNL03:
1957            case 0x000a:            case CHAR_LF:
1958            if (codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSUPTO)            if (codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSUPTO)
1959              {              {
1960              active_count--;           /* Remove non-match possibility */              active_count--;           /* Remove non-match possibility */
1961              next_active_state--;              next_active_state--;
1962              }              }
1963            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1964              { ADD_NEW_DATA(-(state_offset + 4), 0, ncount); }              { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, ncount); }
1965            else            else
1966              { ADD_NEW_DATA(-state_offset, count, ncount); }              { ADD_NEW_DATA(-state_offset, count, ncount); }
1967            break;            break;
# Line 1640  for (;;) Line 1978  for (;;)
1978        case OP_VSPACE_EXTRA + OP_TYPEMINUPTO:        case OP_VSPACE_EXTRA + OP_TYPEMINUPTO:
1979        case OP_VSPACE_EXTRA + OP_TYPEPOSUPTO:        case OP_VSPACE_EXTRA + OP_TYPEPOSUPTO:
1980        if (codevalue != OP_VSPACE_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_VSPACE_EXTRA + OP_TYPEEXACT)
1981          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
1982        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1983        if (clen > 0)        if (clen > 0)
1984          {          {
1985          BOOL OK;          BOOL OK;
1986          switch (c)          switch (c)
1987            {            {
1988            case 0x000a:            VSPACE_CASES:
           case 0x000b:  
           case 0x000c:  
           case 0x000d:  
           case 0x0085:  
           case 0x2028:  
           case 0x2029:  
1989            OK = TRUE;            OK = TRUE;
1990            break;            break;
1991    
# Line 1669  for (;;) Line 2001  for (;;)
2001              next_active_state--;              next_active_state--;
2002              }              }
2003            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
2004              { ADD_NEW_DATA(-(state_offset + 4), 0, 0); }              { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, 0); }
2005            else            else
2006              { ADD_NEW_DATA(-state_offset, count, 0); }              { ADD_NEW_DATA(-state_offset, count, 0); }
2007            }            }
# Line 1682  for (;;) Line 2014  for (;;)
2014        case OP_HSPACE_EXTRA + OP_TYPEMINUPTO:        case OP_HSPACE_EXTRA + OP_TYPEMINUPTO:
2015        case OP_HSPACE_EXTRA + OP_TYPEPOSUPTO:        case OP_HSPACE_EXTRA + OP_TYPEPOSUPTO:
2016        if (codevalue != OP_HSPACE_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_HSPACE_EXTRA + OP_TYPEEXACT)
2017          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
2018        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
2019        if (clen > 0)        if (clen > 0)
2020          {          {
2021          BOOL OK;          BOOL OK;
2022          switch (c)          switch (c)
2023            {            {
2024            case 0x09:      /* HT */            HSPACE_CASES:
           case 0x20:      /* SPACE */  
           case 0xa0:      /* NBSP */  
           case 0x1680:    /* OGHAM SPACE MARK */  
           case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
           case 0x2000:    /* EN QUAD */  
           case 0x2001:    /* EM QUAD */  
           case 0x2002:    /* EN SPACE */  
           case 0x2003:    /* EM SPACE */  
           case 0x2004:    /* THREE-PER-EM SPACE */  
           case 0x2005:    /* FOUR-PER-EM SPACE */  
           case 0x2006:    /* SIX-PER-EM SPACE */  
           case 0x2007:    /* FIGURE SPACE */  
           case 0x2008:    /* PUNCTUATION SPACE */  
           case 0x2009:    /* THIN SPACE */  
           case 0x200A:    /* HAIR SPACE */  
           case 0x202f:    /* NARROW NO-BREAK SPACE */  
           case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
           case 0x3000:    /* IDEOGRAPHIC SPACE */  
2025            OK = TRUE;            OK = TRUE;
2026            break;            break;
2027    
# Line 1724  for (;;) Line 2038  for (;;)
2038              next_active_state--;              next_active_state--;
2039              }              }
2040            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
2041              { ADD_NEW_DATA(-(state_offset + 4), 0, 0); }              { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, 0); }
2042            else            else
2043              { ADD_NEW_DATA(-state_offset, count, 0); }              { ADD_NEW_DATA(-state_offset, count, 0); }
2044            }            }
# Line 1743  for (;;) Line 2057  for (;;)
2057        break;        break;
2058    
2059        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2060        case OP_CHARNC:        case OP_CHARI:
2061        if (clen == 0) break;        if (clen == 0) break;
2062    
2063  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2064        if (utf8)        if (utf)
2065          {          {
2066          if (c == d) { ADD_NEW(state_offset + dlen + 1, 0); } else          if (c == d) { ADD_NEW(state_offset + dlen + 1, 0); } else
2067            {            {
2068            unsigned int othercase;            unsigned int othercase;
2069            if (c < 128) othercase = fcc[c]; else            if (c < 128)
2070                othercase = fcc[c];
2071            /* If we have Unicode property support, we can use it to test the            else
2072            other case of the character. */              /* If we have Unicode property support, we can use it to test the
2073                other case of the character. */
2074  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2075            othercase = UCD_OTHERCASE(c);              othercase = UCD_OTHERCASE(c);
2076  #else  #else
2077            othercase = NOTACHAR;              othercase = NOTACHAR;
2078  #endif  #endif
2079    
2080            if (d == othercase) { ADD_NEW(state_offset + dlen + 1, 0); }            if (d == othercase) { ADD_NEW(state_offset + dlen + 1, 0); }
2081            }            }
2082          }          }
2083        else        else
2084  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2085          /* Not UTF mode */
       /* Non-UTF-8 mode */  
2086          {          {
2087          if (lcc[c] == lcc[d]) { ADD_NEW(state_offset + 2, 0); }          if (TABLE_GET(c, lcc, c) == TABLE_GET(d, lcc, d))
2088              { ADD_NEW(state_offset + 2, 0); }
2089          }          }
2090        break;        break;
2091    
# Line 1783  for (;;) Line 2097  for (;;)
2097        to wait for them to pass before continuing. */        to wait for them to pass before continuing. */
2098    
2099        case OP_EXTUNI:        case OP_EXTUNI:
2100        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0)
2101          {          {
2102          const uschar *nptr = ptr + clen;          int lgb, rgb;
2103            const pcre_uchar *nptr = ptr + clen;
2104          int ncount = 0;          int ncount = 0;
2105            lgb = UCD_GRAPHBREAK(c);
2106          while (nptr < end_subject)          while (nptr < end_subject)
2107            {            {
2108            int nclen = 1;            dlen = 1;
2109            GETCHARLEN(c, nptr, nclen);            if (!utf) d = *nptr; else { GETCHARLEN(d, nptr, dlen); }
2110            if (UCD_CATEGORY(c) != ucp_M) break;            rgb = UCD_GRAPHBREAK(d);
2111              if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
2112            ncount++;            ncount++;
2113            nptr += nclen;            lgb = rgb;
2114              nptr += dlen;
2115            }            }
2116            if (nptr >= end_subject && (md->moptions & PCRE_PARTIAL_HARD) != 0)
2117                reset_could_continue = TRUE;
2118          ADD_NEW_DATA(-(state_offset + 1), 0, ncount);          ADD_NEW_DATA(-(state_offset + 1), 0, ncount);
2119          }          }
2120        break;        break;
# Line 1808  for (;;) Line 2128  for (;;)
2128        case OP_ANYNL:        case OP_ANYNL:
2129        if (clen > 0) switch(c)        if (clen > 0) switch(c)
2130          {          {
2131          case 0x000b:          case CHAR_VT:
2132          case 0x000c:          case CHAR_FF:
2133          case 0x0085:          case CHAR_NEL:
2134    #ifndef EBCDIC
2135          case 0x2028:          case 0x2028:
2136          case 0x2029:          case 0x2029:
2137    #endif  /* Not EBCDIC */
2138          if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;          if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
2139    
2140          case 0x000a:          case CHAR_LF:
2141          ADD_NEW(state_offset + 1, 0);          ADD_NEW(state_offset + 1, 0);
2142          break;          break;
2143    
2144          case 0x000d:          case CHAR_CR:
2145          if (ptr + 1 < end_subject && ptr[1] == 0x0a)          if (ptr + 1 >= end_subject)
2146              {
2147              ADD_NEW(state_offset + 1, 0);
2148              if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
2149                reset_could_continue = TRUE;
2150              }
2151            else if (RAWUCHARTEST(ptr + 1) == CHAR_LF)
2152            {            {
2153            ADD_NEW_DATA(-(state_offset + 1), 0, 1);            ADD_NEW_DATA(-(state_offset + 1), 0, 1);
2154            }            }
# Line 1836  for (;;) Line 2164  for (;;)
2164        case OP_NOT_VSPACE:        case OP_NOT_VSPACE:
2165        if (clen > 0) switch(c)        if (clen > 0) switch(c)
2166          {          {
2167          case 0x000a:          VSPACE_CASES:
         case 0x000b:  
         case 0x000c:  
         case 0x000d:  
         case 0x0085:  
         case 0x2028:  
         case 0x2029:  
2168          break;          break;
2169    
2170          default:          default:
# Line 1855  for (;;) Line 2177  for (;;)
2177        case OP_VSPACE:        case OP_VSPACE:
2178        if (clen > 0) switch(c)        if (clen > 0) switch(c)
2179          {          {
2180          case 0x000a:          VSPACE_CASES:
         case 0x000b:  
         case 0x000c:  
         case 0x000d:  
         case 0x0085:  
         case 0x2028:  
         case 0x2029:  
2181          ADD_NEW(state_offset + 1, 0);          ADD_NEW(state_offset + 1, 0);
2182          break;          break;
2183    
2184          default: break;          default:
2185            break;
2186          }          }
2187        break;        break;
2188    
# Line 1873  for (;;) Line 2190  for (;;)
2190        case OP_NOT_HSPACE:        case OP_NOT_HSPACE:
2191        if (clen > 0) switch(c)        if (clen > 0) switch(c)
2192          {          {
2193          case 0x09:      /* HT */          HSPACE_CASES:
         case 0x20:      /* SPACE */  
         case 0xa0:      /* NBSP */  
         case 0x1680:    /* OGHAM SPACE MARK */  
         case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
         case 0x2000:    /* EN QUAD */  
         case 0x2001:    /* EM QUAD */  
         case 0x2002:    /* EN SPACE */  
         case 0x2003:    /* EM SPACE */  
         case 0x2004:    /* THREE-PER-EM SPACE */  
         case 0x2005:    /* FOUR-PER-EM SPACE */  
         case 0x2006:    /* SIX-PER-EM SPACE */  
         case 0x2007:    /* FIGURE SPACE */  
         case 0x2008:    /* PUNCTUATION SPACE */  
         case 0x2009:    /* THIN SPACE */  
         case 0x200A:    /* HAIR SPACE */  
         case 0x202f:    /* NARROW NO-BREAK SPACE */  
         case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
         case 0x3000:    /* IDEOGRAPHIC SPACE */  
2194          break;          break;
2195    
2196          default:          default:
# Line 1904  for (;;) Line 2203  for (;;)
2203        case OP_HSPACE:        case OP_HSPACE:
2204        if (clen > 0) switch(c)        if (clen > 0) switch(c)
2205          {          {
2206          case 0x09:      /* HT */          HSPACE_CASES:
         case 0x20:      /* SPACE */  
         case 0xa0:      /* NBSP */  
         case 0x1680:    /* OGHAM SPACE MARK */  
         case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
         case 0x2000:    /* EN QUAD */  
         case 0x2001:    /* EM QUAD */  
         case 0x2002:    /* EN SPACE */  
         case 0x2003:    /* EM SPACE */  
         case 0x2004:    /* THREE-PER-EM SPACE */  
         case 0x2005:    /* FOUR-PER-EM SPACE */  
         case 0x2006:    /* SIX-PER-EM SPACE */  
         case 0x2007:    /* FIGURE SPACE */  
         case 0x2008:    /* PUNCTUATION SPACE */  
         case 0x2009:    /* THIN SPACE */  
         case 0x200A:    /* HAIR SPACE */  
         case 0x202f:    /* NARROW NO-BREAK SPACE */  
         case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
         case 0x3000:    /* IDEOGRAPHIC SPACE */  
2207          ADD_NEW(state_offset + 1, 0);          ADD_NEW(state_offset + 1, 0);
2208          break;          break;
2209    
2210            default:
2211            break;
2212          }          }
2213        break;        break;
2214    
2215        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2216        /* Match a negated single character. This is only used for one-byte        /* Match a negated single character casefully. */
       characters, that is, we know that d < 256. The character we are  
       checking (c) can be multibyte. */  
2217    
2218        case OP_NOT:        case OP_NOT:
2219          if (clen > 0 && c != d) { ADD_NEW(state_offset + dlen + 1, 0); }
2220          break;
2221    
2222          /*-----------------------------------------------------------------*/
2223          /* Match a negated single character caselessly. */
2224    
2225          case OP_NOTI:
2226        if (clen > 0)        if (clen > 0)
2227          {          {
2228          unsigned int otherd = ((ims & PCRE_CASELESS) != 0)? fcc[d] : d;          unsigned int otherd;
2229          if (c != d && c != otherd) { ADD_NEW(state_offset + dlen + 1, 0); }  #ifdef SUPPORT_UTF
2230            if (utf && d >= 128)
2231              {
2232    #ifdef SUPPORT_UCP
2233              otherd = UCD_OTHERCASE(d);
2234    #endif  /* SUPPORT_UCP */
2235              }
2236            else
2237    #endif  /* SUPPORT_UTF */
2238            otherd = TABLE_GET(d, fcc, d);
2239            if (c != d && c != otherd)
2240              { ADD_NEW(state_offset + dlen + 1, 0); }
2241          }          }
2242        break;        break;
2243    
2244        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2245          case OP_PLUSI:
2246          case OP_MINPLUSI:
2247          case OP_POSPLUSI:
2248          case OP_NOTPLUSI:
2249          case OP_NOTMINPLUSI:
2250          case OP_NOTPOSPLUSI:
2251          caseless = TRUE;
2252          codevalue -= OP_STARI - OP_STAR;
2253    
2254          /* Fall through */
2255        case OP_PLUS:        case OP_PLUS:
2256        case OP_MINPLUS:        case OP_MINPLUS:
2257        case OP_POSPLUS:        case OP_POSPLUS:
# Line 1952  for (;;) Line 2262  for (;;)
2262        if (count > 0) { ADD_ACTIVE(state_offset + dlen + 1, 0); }        if (count > 0) { ADD_ACTIVE(state_offset + dlen + 1, 0); }
2263        if (clen > 0)        if (clen > 0)
2264          {          {
2265          unsigned int otherd = NOTACHAR;          pcre_uint32 otherd = NOTACHAR;
2266          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2267            {            {
2268  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2269            if (utf8 && d >= 128)            if (utf && d >= 128)
2270              {              {
2271  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2272              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2273  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2274              }              }
2275            else            else
2276  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2277            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2278            }            }
2279          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2280            {            {
# Line 1981  for (;;) Line 2291  for (;;)
2291        break;        break;
2292    
2293        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2294          case OP_QUERYI:
2295          case OP_MINQUERYI:
2296          case OP_POSQUERYI:
2297          case OP_NOTQUERYI:
2298          case OP_NOTMINQUERYI:
2299          case OP_NOTPOSQUERYI:
2300          caseless = TRUE;
2301          codevalue -= OP_STARI - OP_STAR;
2302          /* Fall through */
2303        case OP_QUERY:        case OP_QUERY:
2304        case OP_MINQUERY:        case OP_MINQUERY:
2305        case OP_POSQUERY:        case OP_POSQUERY:
# Line 1990  for (;;) Line 2309  for (;;)
2309        ADD_ACTIVE(state_offset + dlen + 1, 0);        ADD_ACTIVE(state_offset + dlen + 1, 0);
2310        if (clen > 0)        if (clen > 0)
2311          {          {
2312          unsigned int otherd = NOTACHAR;          pcre_uint32 otherd = NOTACHAR;
2313          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2314            {            {
2315  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2316            if (utf8 && d >= 128)            if (utf && d >= 128)
2317              {              {
2318  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2319              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2320  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2321              }              }
2322            else            else
2323  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2324            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2325            }            }
2326          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2327            {            {
# Line 2017  for (;;) Line 2336  for (;;)
2336        break;        break;
2337    
2338        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2339          case OP_STARI:
2340          case OP_MINSTARI:
2341          case OP_POSSTARI:
2342          case OP_NOTSTARI:
2343          case OP_NOTMINSTARI:
2344          case OP_NOTPOSSTARI:
2345          caseless = TRUE;
2346          codevalue -= OP_STARI - OP_STAR;
2347          /* Fall through */
2348        case OP_STAR:        case OP_STAR:
2349        case OP_MINSTAR:        case OP_MINSTAR:
2350        case OP_POSSTAR:        case OP_POSSTAR:
# Line 2026  for (;;) Line 2354  for (;;)
2354        ADD_ACTIVE(state_offset + dlen + 1, 0);        ADD_ACTIVE(state_offset + dlen + 1, 0);
2355        if (clen > 0)        if (clen > 0)
2356          {          {
2357          unsigned int otherd = NOTACHAR;          pcre_uint32 otherd = NOTACHAR;
2358          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2359            {            {
2360  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2361            if (utf8 && d >= 128)            if (utf && d >= 128)
2362              {              {
2363  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2364              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2365  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2366              }              }
2367            else            else
2368  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2369            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2370            }            }
2371          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2372            {            {
# Line 2053  for (;;) Line 2381  for (;;)
2381        break;        break;
2382    
2383        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2384          case OP_EXACTI:
2385          case OP_NOTEXACTI:
2386          caseless = TRUE;
2387          codevalue -= OP_STARI - OP_STAR;
2388          /* Fall through */
2389        case OP_EXACT:        case OP_EXACT:
2390        case OP_NOTEXACT:        case OP_NOTEXACT:
2391        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
2392        if (clen > 0)        if (clen > 0)
2393          {          {
2394          unsigned int otherd = NOTACHAR;          pcre_uint32 otherd = NOTACHAR;
2395          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2396            {            {
2397  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2398            if (utf8 && d >= 128)            if (utf && d >= 128)
2399              {              {
2400  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2401              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2402  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2403              }              }
2404            else            else
2405  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2406            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2407            }            }
2408          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2409            {            {
2410            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
2411              { ADD_NEW(state_offset + dlen + 3, 0); }              { ADD_NEW(state_offset + dlen + 1 + IMM2_SIZE, 0); }
2412            else            else
2413              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
2414            }            }
# Line 2083  for (;;) Line 2416  for (;;)
2416        break;        break;
2417    
2418        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2419          case OP_UPTOI:
2420          case OP_MINUPTOI:
2421          case OP_POSUPTOI:
2422          case OP_NOTUPTOI:
2423          case OP_NOTMINUPTOI:
2424          case OP_NOTPOSUPTOI:
2425          caseless = TRUE;
2426          codevalue -= OP_STARI - OP_STAR;
2427          /* Fall through */
2428        case OP_UPTO:        case OP_UPTO:
2429        case OP_MINUPTO:        case OP_MINUPTO:
2430        case OP_POSUPTO:        case OP_POSUPTO:
2431        case OP_NOTUPTO:        case OP_NOTUPTO:
2432        case OP_NOTMINUPTO:        case OP_NOTMINUPTO:
2433        case OP_NOTPOSUPTO:        case OP_NOTPOSUPTO:
2434        ADD_ACTIVE(state_offset + dlen + 3, 0);        ADD_ACTIVE(state_offset + dlen + 1 + IMM2_SIZE, 0);
2435        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
2436        if (clen > 0)        if (clen > 0)
2437          {          {
2438          unsigned int otherd = NOTACHAR;          pcre_uint32 otherd = NOTACHAR;
2439          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2440            {            {
2441  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2442            if (utf8 && d >= 128)            if (utf && d >= 128)
2443              {              {
2444  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2445              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2446  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2447              }              }
2448            else            else
2449  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2450            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2451            }            }
2452          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2453            {            {
# Line 2115  for (;;) Line 2457  for (;;)
2457              next_active_state--;              next_active_state--;
2458              }              }
2459            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
2460              { ADD_NEW(state_offset + dlen + 3, 0); }              { ADD_NEW(state_offset + dlen + 1 + IMM2_SIZE, 0); }
2461            else            else
2462              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
2463            }            }
# Line 2132  for (;;) Line 2474  for (;;)
2474          {          {
2475          BOOL isinclass = FALSE;          BOOL isinclass = FALSE;
2476          int next_state_offset;          int next_state_offset;
2477          const uschar *ecode;          const pcre_uchar *ecode;
2478    
2479          /* 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
2480          can set isinclass from it. */          can set isinclass from it. */
2481    
2482          if (codevalue != OP_XCLASS)          if (codevalue != OP_XCLASS)
2483            {            {
2484            ecode = code + 33;            ecode = code + 1 + (32 / sizeof(pcre_uchar));
2485            if (clen > 0)            if (clen > 0)
2486              {              {
2487              isinclass = (c > 255)? (codevalue == OP_NCLASS) :              isinclass = (c > 255)? (codevalue == OP_NCLASS) :
2488                ((code[1 + c/8] & (1 << (c&7))) != 0);                ((((pcre_uint8 *)(code + 1))[c/8] & (1 << (c&7))) != 0);
2489              }              }
2490            }            }
2491    
# Line 2154  for (;;) Line 2496  for (;;)
2496          else          else
2497           {           {
2498           ecode = code + GET(code, 1);           ecode = code + GET(code, 1);
2499           if (clen > 0) isinclass = _pcre_xclass(c, code + 1 + LINK_SIZE);           if (clen > 0) isinclass = PRIV(xclass)(c, code + 1 + LINK_SIZE, utf);
2500           }           }
2501    
2502          /* 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
2503          points to the byte after the end of the class. If there is a          points to the byte after the end of the class. If there is a
2504          quantifier, this is where it will be. */          quantifier, this is where it will be. */
2505    
2506          next_state_offset = ecode - start_code;          next_state_offset = (int)(ecode - start_code);
2507    
2508          switch (*ecode)          switch (*ecode)
2509            {            {
# Line 2188  for (;;) Line 2530  for (;;)
2530            case OP_CRMINRANGE:            case OP_CRMINRANGE:
2531            count = current_state->count;  /* Already matched */            count = current_state->count;  /* Already matched */
2532            if (count >= GET2(ecode, 1))            if (count >= GET2(ecode, 1))
2533              { ADD_ACTIVE(next_state_offset + 5, 0); }              { ADD_ACTIVE(next_state_offset + 1 + 2 * IMM2_SIZE, 0); }
2534            if (isinclass)            if (isinclass)
2535              {              {
2536              int max = GET2(ecode, 3);              unsigned int max = GET2(ecode, 1 + IMM2_SIZE);
2537              if (++count >= max && max != 0)   /* Max 0 => no limit */              if (++count >= max && max != 0)   /* Max 0 => no limit */
2538                { ADD_NEW(next_state_offset + 5, 0); }                { ADD_NEW(next_state_offset + 1 + 2 * IMM2_SIZE, 0); }
2539              else              else
2540                { ADD_NEW(state_offset, count); }                { ADD_NEW(state_offset, count); }
2541              }              }
# Line 2224  for (;;) Line 2566  for (;;)
2566          int rc;          int rc;
2567          int local_offsets[2];          int local_offsets[2];
2568          int local_workspace[1000];          int local_workspace[1000];
2569          const uschar *endasscode = code + GET(code, 1);          const pcre_uchar *endasscode = code + GET(code, 1);
2570    
2571          while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);          while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);
2572    
# Line 2232  for (;;) Line 2574  for (;;)
2574            md,                                   /* static match data */            md,                                   /* static match data */
2575            code,                                 /* this subexpression's code */            code,                                 /* this subexpression's code */
2576            ptr,                                  /* where we currently are */            ptr,                                  /* where we currently are */
2577            ptr - start_subject,                  /* start offset */            (int)(ptr - start_subject),           /* start offset */
2578            local_offsets,                        /* offset vector */            local_offsets,                        /* offset vector */
2579            sizeof(local_offsets)/sizeof(int),    /* size of same */            sizeof(local_offsets)/sizeof(int),    /* size of same */
2580            local_workspace,                      /* workspace vector */            local_workspace,                      /* workspace vector */
2581            sizeof(local_workspace)/sizeof(int),  /* size of same */            sizeof(local_workspace)/sizeof(int),  /* size of same */
2582            ims,                                  /* the current ims flags */            rlevel);                              /* function recursion level */
           rlevel,                               /* function recursion level */  
           recursing);                           /* pass on regex recursion */  
2583    
2584            if (rc == PCRE_ERROR_DFA_UITEM) return rc;
2585          if ((rc >= 0) == (codevalue == OP_ASSERT || codevalue == OP_ASSERTBACK))          if ((rc >= 0) == (codevalue == OP_ASSERT || codevalue == OP_ASSERTBACK))
2586              { ADD_ACTIVE(endasscode + LINK_SIZE + 1 - start_code, 0); }              { ADD_ACTIVE((int)(endasscode + LINK_SIZE + 1 - start_code), 0); }
2587          }          }
2588        break;        break;
2589    
# Line 2262  for (;;) Line 2603  for (;;)
2603          if (code[LINK_SIZE+1] == OP_CALLOUT)          if (code[LINK_SIZE+1] == OP_CALLOUT)
2604            {            {
2605            rrc = 0;            rrc = 0;
2606            if (pcre_callout != NULL)            if (PUBL(callout) != NULL)
2607              {              {
2608              pcre_callout_block cb;              PUBL(callout_block) cb;
2609              cb.version          = 1;   /* Version 1 of the callout block */              cb.version          = 1;   /* Version 1 of the callout block */
2610              cb.callout_number   = code[LINK_SIZE+2];              cb.callout_number   = code[LINK_SIZE+2];
2611              cb.offset_vector    = offsets;              cb.offset_vector    = offsets;
2612    #if defined COMPILE_PCRE8
2613              cb.subject          = (PCRE_SPTR)start_subject;              cb.subject          = (PCRE_SPTR)start_subject;
2614              cb.subject_length   = end_subject - start_subject;  #elif defined COMPILE_PCRE16
2615              cb.start_match      = current_subject - start_subject;              cb.subject          = (PCRE_SPTR16)start_subject;
2616              cb.current_position = ptr - start_subject;  #elif defined COMPILE_PCRE32
2617                cb.subject          = (PCRE_SPTR32)start_subject;
2618    #endif
2619                cb.subject_length   = (int)(end_subject - start_subject);
2620                cb.start_match      = (int)(current_subject - start_subject);
2621                cb.current_position = (int)(ptr - start_subject);
2622              cb.pattern_position = GET(code, LINK_SIZE + 3);              cb.pattern_position = GET(code, LINK_SIZE + 3);
2623              cb.next_item_length = GET(code, 3 + 2*LINK_SIZE);              cb.next_item_length = GET(code, 3 + 2*LINK_SIZE);
2624              cb.capture_top      = 1;              cb.capture_top      = 1;
2625              cb.capture_last     = -1;              cb.capture_last     = -1;
2626              cb.callout_data     = md->callout_data;              cb.callout_data     = md->callout_data;
2627              if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc;   /* Abandon */              cb.mark             = NULL;   /* No (*MARK) support */
2628                if ((rrc = (*PUBL(callout))(&cb)) < 0) return rrc;   /* Abandon */
2629              }              }
2630            if (rrc > 0) break;                      /* Fail this thread */            if (rrc > 0) break;                      /* Fail this thread */
2631            code += _pcre_OP_lengths[OP_CALLOUT];    /* Skip callout data */            code += PRIV(OP_lengths)[OP_CALLOUT];    /* Skip callout data */
2632            }            }
2633    
2634          condcode = code[LINK_SIZE+1];          condcode = code[LINK_SIZE+1];
2635    
2636          /* Back reference conditions are not supported */          /* Back reference conditions are not supported */
2637    
2638          if (condcode == OP_CREF) return PCRE_ERROR_DFA_UCOND;          if (condcode == OP_CREF || condcode == OP_NCREF)
2639              return PCRE_ERROR_DFA_UCOND;
2640    
2641          /* The DEFINE condition is always false */          /* The DEFINE condition is always false */
2642    
# Line 2298  for (;;) Line 2647  for (;;)
2647          which means "test if in any recursion". We can't test for specifically          which means "test if in any recursion". We can't test for specifically
2648          recursed groups. */          recursed groups. */
2649    
2650          else if (condcode == OP_RREF)          else if (condcode == OP_RREF || condcode == OP_NRREF)
2651            {            {
2652            int value = GET2(code, LINK_SIZE+2);            int value = GET2(code, LINK_SIZE + 2);
2653            if (value != RREF_ANY) return PCRE_ERROR_DFA_UCOND;            if (value != RREF_ANY) return PCRE_ERROR_DFA_UCOND;
2654            if (recursing > 0)            if (md->recursive != NULL)
2655              { ADD_ACTIVE(state_offset + LINK_SIZE + 4, 0); }              { ADD_ACTIVE(state_offset + LINK_SIZE + 2 + IMM2_SIZE, 0); }
2656            else { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }            else { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
2657            }            }
2658    
# Line 2312  for (;;) Line 2661  for (;;)
2661          else          else
2662            {            {
2663            int rc;            int rc;
2664            const uschar *asscode = code + LINK_SIZE + 1;            const pcre_uchar *asscode = code + LINK_SIZE + 1;
2665            const uschar *endasscode = asscode + GET(asscode, 1);            const pcre_uchar *endasscode = asscode + GET(asscode, 1);
2666    
2667            while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);            while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);
2668    
# Line 2321  for (;;) Line 2670  for (;;)
2670              md,                                   /* fixed match data */              md,                                   /* fixed match data */
2671              asscode,                              /* this subexpression's code */              asscode,                              /* this subexpression's code */
2672              ptr,                                  /* where we currently are */              ptr,                                  /* where we currently are */
2673              ptr - start_subject,                  /* start offset */              (int)(ptr - start_subject),           /* start offset */
2674              local_offsets,                        /* offset vector */              local_offsets,                        /* offset vector */
2675              sizeof(local_offsets)/sizeof(int),    /* size of same */              sizeof(local_offsets)/sizeof(int),    /* size of same */
2676              local_workspace,                      /* workspace vector */              local_workspace,                      /* workspace vector */
2677              sizeof(local_workspace)/sizeof(int),  /* size of same */              sizeof(local_workspace)/sizeof(int),  /* size of same */
2678              ims,                                  /* the current ims flags */              rlevel);                              /* function recursion level */
             rlevel,                               /* function recursion level */  
             recursing);                           /* pass on regex recursion */  
2679    
2680              if (rc == PCRE_ERROR_DFA_UITEM) return rc;
2681            if ((rc >= 0) ==            if ((rc >= 0) ==
2682                  (condcode == OP_ASSERT || condcode == OP_ASSERTBACK))                  (condcode == OP_ASSERT || condcode == OP_ASSERTBACK))
2683              { ADD_ACTIVE(endasscode + LINK_SIZE + 1 - start_code, 0); }              { ADD_ACTIVE((int)(endasscode + LINK_SIZE + 1 - start_code), 0); }
2684            else            else
2685              { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }              { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
2686            }            }
# Line 2342  for (;;) Line 2690  for (;;)
2690        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2691        case OP_RECURSE:        case OP_RECURSE:
2692          {          {
2693            dfa_recursion_info *ri;
2694          int local_offsets[1000];          int local_offsets[1000];
2695          int local_workspace[1000];          int local_workspace[1000];
2696            const pcre_uchar *callpat = start_code + GET(code, 1);
2697            int recno = (callpat == md->start_code)? 0 :
2698              GET2(callpat, 1 + LINK_SIZE);
2699          int rc;          int rc;
2700    
2701          DPRINTF(("%.*sStarting regex recursion %d\n", rlevel*2-2, SP,          DPRINTF(("%.*sStarting regex recursion\n", rlevel*2-2, SP));
2702            recursing + 1));  
2703            /* Check for repeating a recursion without advancing the subject
2704            pointer. This should catch convoluted mutual recursions. (Some simple
2705            cases are caught at compile time.) */
2706    
2707            for (ri = md->recursive; ri != NULL; ri = ri->prevrec)
2708              if (recno == ri->group_num && ptr == ri->subject_position)
2709                return PCRE_ERROR_RECURSELOOP;
2710    
2711            /* Remember this recursion and where we started it so as to
2712            catch infinite loops. */
2713    
2714            new_recursive.group_num = recno;
2715            new_recursive.subject_position = ptr;
2716            new_recursive.prevrec = md->recursive;
2717            md->recursive = &new_recursive;
2718    
2719          rc = internal_dfa_exec(          rc = internal_dfa_exec(
2720            md,                                   /* fixed match data */            md,                                   /* fixed match data */
2721            start_code + GET(code, 1),            /* this subexpression's code */            callpat,                              /* this subexpression's code */
2722            ptr,                                  /* where we currently are */            ptr,                                  /* where we currently are */
2723            ptr - start_subject,                  /* start offset */            (int)(ptr - start_subject),           /* start offset */
2724            local_offsets,                        /* offset vector */            local_offsets,                        /* offset vector */
2725            sizeof(local_offsets)/sizeof(int),    /* size of same */            sizeof(local_offsets)/sizeof(int),    /* size of same */
2726            local_workspace,                      /* workspace vector */            local_workspace,                      /* workspace vector */
2727            sizeof(local_workspace)/sizeof(int),  /* size of same */            sizeof(local_workspace)/sizeof(int),  /* size of same */
2728            ims,                                  /* the current ims flags */            rlevel);                              /* function recursion level */
           rlevel,                               /* function recursion level */  
           recursing + 1);                       /* regex recurse level */  
2729    
2730          DPRINTF(("%.*sReturn from regex recursion %d: rc=%d\n", rlevel*2-2, SP,          md->recursive = new_recursive.prevrec;  /* Done this recursion */
2731            recursing + 1, rc));  
2732            DPRINTF(("%.*sReturn from regex recursion: rc=%d\n", rlevel*2-2, SP,
2733              rc));
2734    
2735          /* Ran out of internal offsets */          /* Ran out of internal offsets */
2736    
# Line 2377  for (;;) Line 2744  for (;;)
2744            {            {
2745            for (rc = rc*2 - 2; rc >= 0; rc -= 2)            for (rc = rc*2 - 2; rc >= 0; rc -= 2)
2746              {              {
             const uschar *p = start_subject + local_offsets[rc];  
             const uschar *pp = start_subject + local_offsets[rc+1];  
2747              int charcount = local_offsets[rc+1] - local_offsets[rc];              int charcount = local_offsets[rc+1] - local_offsets[rc];
2748              while (p < pp) if ((*p++ & 0xc0) == 0x80) charcount--;  #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
2749                if (utf)
2750                  {
2751                  const pcre_uchar *p = start_subject + local_offsets[rc];
2752                  const pcre_uchar *pp = start_subject + local_offsets[rc+1];
2753                  while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
2754                  }
2755    #endif
2756              if (charcount > 0)              if (charcount > 0)
2757                {                {
2758                ADD_NEW_DATA(-(state_offset + LINK_SIZE + 1), 0, (charcount - 1));                ADD_NEW_DATA(-(state_offset + LINK_SIZE + 1), 0, (charcount - 1));
# Line 2396  for (;;) Line 2768  for (;;)
2768        break;        break;
2769    
2770        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2771          case OP_BRAPOS:
2772          case OP_SBRAPOS:
2773          case OP_CBRAPOS:
2774          case OP_SCBRAPOS:
2775          case OP_BRAPOSZERO:
2776            {
2777            int charcount, matched_count;
2778            const pcre_uchar *local_ptr = ptr;
2779            BOOL allow_zero;
2780    
2781            if (codevalue == OP_BRAPOSZERO)
2782              {
2783              allow_zero = TRUE;
2784              codevalue = *(++code);  /* Codevalue will be one of above BRAs */
2785              }
2786            else allow_zero = FALSE;
2787    
2788            /* Loop to match the subpattern as many times as possible as if it were
2789            a complete pattern. */
2790    
2791            for (matched_count = 0;; matched_count++)
2792              {
2793              int local_offsets[2];
2794              int local_workspace[1000];
2795    
2796              int rc = internal_dfa_exec(
2797                md,                                   /* fixed match data */
2798                code,                                 /* this subexpression's code */
2799                local_ptr,                            /* where we currently are */
2800                (int)(ptr - start_subject),           /* start offset */
2801                local_offsets,                        /* offset vector */
2802                sizeof(local_offsets)/sizeof(int),    /* size of same */
2803                local_workspace,                      /* workspace vector */
2804                sizeof(local_workspace)/sizeof(int),  /* size of same */
2805                rlevel);                              /* function recursion level */
2806    
2807              /* Failed to match */
2808    
2809              if (rc < 0)
2810                {
2811                if (rc != PCRE_ERROR_NOMATCH) return rc;
2812                break;
2813                }
2814    
2815              /* Matched: break the loop if zero characters matched. */
2816    
2817              charcount = local_offsets[1] - local_offsets[0];
2818              if (charcount == 0) break;
2819              local_ptr += charcount;    /* Advance temporary position ptr */
2820              }
2821    
2822            /* At this point we have matched the subpattern matched_count
2823            times, and local_ptr is pointing to the character after the end of the
2824            last match. */
2825    
2826            if (matched_count > 0 || allow_zero)
2827              {
2828              const pcre_uchar *end_subpattern = code;
2829              int next_state_offset;
2830    
2831              do { end_subpattern += GET(end_subpattern, 1); }
2832                while (*end_subpattern == OP_ALT);
2833              next_state_offset =
2834                (int)(end_subpattern - start_code + LINK_SIZE + 1);
2835    
2836              /* Optimization: if there are no more active states, and there
2837              are no new states yet set up, then skip over the subject string
2838              right here, to save looping. Otherwise, set up the new state to swing
2839              into action when the end of the matched substring is reached. */
2840    
2841              if (i + 1 >= active_count && new_count == 0)
2842                {
2843                ptr = local_ptr;
2844                clen = 0;
2845                ADD_NEW(next_state_offset, 0);
2846                }
2847              else
2848                {
2849                const pcre_uchar *p = ptr;
2850                const pcre_uchar *pp = local_ptr;
2851                charcount = (int)(pp - p);
2852    #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
2853                if (utf) while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
2854    #endif
2855                ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));
2856                }
2857              }
2858            }
2859          break;
2860    
2861          /*-----------------------------------------------------------------*/
2862        case OP_ONCE:        case OP_ONCE:
2863          case OP_ONCE_NC:
2864          {          {
2865          int local_offsets[2];          int local_offsets[2];
2866          int local_workspace[1000];          int local_workspace[1000];
# Line 2405  for (;;) Line 2869  for (;;)
2869            md,                                   /* fixed match data */            md,                                   /* fixed match data */
2870            code,                                 /* this subexpression's code */            code,                                 /* this subexpression's code */
2871            ptr,                                  /* where we currently are */            ptr,                                  /* where we currently are */
2872            ptr - start_subject,                  /* start offset */            (int)(ptr - start_subject),           /* start offset */
2873            local_offsets,                        /* offset vector */            local_offsets,                        /* offset vector */
2874            sizeof(local_offsets)/sizeof(int),    /* size of same */            sizeof(local_offsets)/sizeof(int),    /* size of same */
2875            local_workspace,                      /* workspace vector */            local_workspace,                      /* workspace vector */
2876            sizeof(local_workspace)/sizeof(int),  /* size of same */            sizeof(local_workspace)/sizeof(int),  /* size of same */
2877            ims,                                  /* the current ims flags */            rlevel);                              /* function recursion level */
           rlevel,                               /* function recursion level */  
           recursing);                           /* pass on regex recursion */  
2878    
2879          if (rc >= 0)          if (rc >= 0)
2880            {            {
2881            const uschar *end_subpattern = code;            const pcre_uchar *end_subpattern = code;
2882            int charcount = local_offsets[1] - local_offsets[0];            int charcount = local_offsets[1] - local_offsets[0];
2883            int next_state_offset, repeat_state_offset;            int next_state_offset, repeat_state_offset;
2884    
2885            do { end_subpattern += GET(end_subpattern, 1); }            do { end_subpattern += GET(end_subpattern, 1); }
2886              while (*end_subpattern == OP_ALT);              while (*end_subpattern == OP_ALT);
2887            next_state_offset = end_subpattern - start_code + LINK_SIZE + 1;            next_state_offset =
2888                (int)(end_subpattern - start_code + LINK_SIZE + 1);
2889    
2890            /* If the end of this subpattern is KETRMAX or KETRMIN, we must            /* If the end of this subpattern is KETRMAX or KETRMIN, we must
2891            arrange for the repeat state also to be added to the relevant list.            arrange for the repeat state also to be added to the relevant list.
# Line 2430  for (;;) Line 2893  for (;;)
2893    
2894            repeat_state_offset = (*end_subpattern == OP_KETRMAX ||            repeat_state_offset = (*end_subpattern == OP_KETRMAX ||
2895                                   *end_subpattern == OP_KETRMIN)?                                   *end_subpattern == OP_KETRMIN)?
2896              end_subpattern - start_code - GET(end_subpattern, 1) : -1;              (int)(end_subpattern - start_code - GET(end_subpattern, 1)) : -1;
2897    
2898            /* If we have matched an empty string, add the next state at the            /* If we have matched an empty string, add the next state at the
2899            current character pointer. This is important so that the duplicate            current character pointer. This is important so that the duplicate
# Line 2445  for (;;) Line 2908  for (;;)
2908            /* Optimization: if there are no more active states, and there            /* Optimization: if there are no more active states, and there
2909            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
2910            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
2911            into action when the end of the substring is reached. */            into action when the end of the matched substring is reached. */
2912    
2913            else if (i + 1 >= active_count && new_count == 0)            else if (i + 1 >= active_count && new_count == 0)
2914              {              {
# Line 2468  for (;;) Line 2931  for (;;)
2931              }              }
2932            else            else
2933              {              {
2934              const uschar *p = start_subject + local_offsets[0];  #if defined SUPPORT_UTF && !defined COMPILE_PCRE32
2935              const uschar *pp = start_subject + local_offsets[1];              if (utf)
2936              while (p < pp) if ((*p++ & 0xc0) == 0x80) charcount--;                {
2937                  const pcre_uchar *p = start_subject + local_offsets[0];
2938                  const pcre_uchar *pp = start_subject + local_offsets[1];
2939                  while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
2940                  }
2941    #endif
2942              ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));              ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));
2943              if (repeat_state_offset >= 0)              if (repeat_state_offset >= 0)
2944                { ADD_NEW_DATA(-repeat_state_offset, 0, (charcount - 1)); }                { ADD_NEW_DATA(-repeat_state_offset, 0, (charcount - 1)); }
2945              }              }
   
2946            }            }
2947          else if (rc != PCRE_ERROR_NOMATCH) return rc;          else if (rc != PCRE_ERROR_NOMATCH) return rc;
2948          }          }
# Line 2487  for (;;) Line 2954  for (;;)
2954    
2955        case OP_CALLOUT:        case OP_CALLOUT:
2956        rrc = 0;        rrc = 0;
2957        if (pcre_callout != NULL)        if (PUBL(callout) != NULL)
2958          {          {
2959          pcre_callout_block cb;          PUBL(callout_block) cb;
2960          cb.version          = 1;   /* Version 1 of the callout block */          cb.version          = 1;   /* Version 1 of the callout block */
2961          cb.callout_number   = code[1];          cb.callout_number   = code[1];
2962          cb.offset_vector    = offsets;          cb.offset_vector    = offsets;
2963    #if defined COMPILE_PCRE8
2964          cb.subject          = (PCRE_SPTR)start_subject;          cb.subject          = (PCRE_SPTR)start_subject;
2965          cb.subject_length   = end_subject - start_subject;  #elif defined COMPILE_PCRE16
2966          cb.start_match      = current_subject - start_subject;          cb.subject          = (PCRE_SPTR16)start_subject;
2967          cb.current_position = ptr - start_subject;  #elif defined COMPILE_PCRE32
2968            cb.subject          = (PCRE_SPTR32)start_subject;
2969    #endif
2970            cb.subject_length   = (int)(end_subject - start_subject);
2971            cb.start_match      = (int)(current_subject - start_subject);
2972            cb.current_position = (int)(ptr - start_subject);
2973          cb.pattern_position = GET(code, 2);          cb.pattern_position = GET(code, 2);
2974          cb.next_item_length = GET(code, 2 + LINK_SIZE);          cb.next_item_length = GET(code, 2 + LINK_SIZE);
2975          cb.capture_top      = 1;          cb.capture_top      = 1;
2976          cb.capture_last     = -1;          cb.capture_last     = -1;
2977          cb.callout_data     = md->callout_data;          cb.callout_data     = md->callout_data;
2978          if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc;   /* Abandon */          cb.mark             = NULL;   /* No (*MARK) support */
2979            if ((rrc = (*PUBL(callout))(&cb)) < 0) return rrc;   /* Abandon */
2980          }          }
2981        if (rrc == 0)        if (rrc == 0)
2982          { ADD_ACTIVE(state_offset + _pcre_OP_lengths[OP_CALLOUT], 0); }          { ADD_ACTIVE(state_offset + PRIV(OP_lengths)[OP_CALLOUT], 0); }
2983        break;        break;
2984    
2985    
# Line 2521  for (;;) Line 2995  for (;;)
2995    /* We have finished the processing at the current subject character. If no    /* We have finished the processing at the current subject character. If no
2996    new states have been set for the next character, we have found all the    new states have been set for the next character, we have found all the
2997    matches that we are going to find. If we are at the top level and partial    matches that we are going to find. If we are at the top level and partial
2998    matching has been requested, check for appropriate conditions. The "forced_    matching has been requested, check for appropriate conditions.
2999    fail" variable counts the number of (*F) encountered for the character. If it  
3000    is equal to the original active_count (saved in workspace[1]) it means that    The "forced_ fail" variable counts the number of (*F) encountered for the
3001    (*F) was found on every active state. In this case we don't want to give a    character. If it is equal to the original active_count (saved in
3002    partial match. */    workspace[1]) it means that (*F) was found on every active state. In this
3003      case we don't want to give a partial match.
3004    
3005      The "could_continue" variable is true if a state could have continued but
3006      for the fact that the end of the subject was reached. */
3007    
3008    if (new_count <= 0)    if (new_count <= 0)
3009      {      {
3010      if (rlevel == 1 &&                               /* Top level, and */      if (rlevel == 1 &&                               /* Top level, and */
3011          reached_end != workspace[1] &&               /* Not all reached end */          could_continue &&                            /* Some could go on, and */
3012          forced_fail != workspace[1] &&               /* Not all forced fail & */          forced_fail != workspace[1] &&               /* Not all forced fail & */
3013          (                                            /* either... */          (                                            /* either... */
3014          (md->moptions & PCRE_PARTIAL_HARD) != 0      /* Hard partial */          (md->moptions & PCRE_PARTIAL_HARD) != 0      /* Hard partial */
# Line 2538  for (;;) Line 3016  for (;;)
3016          ((md->moptions & PCRE_PARTIAL_SOFT) != 0 &&  /* Soft partial and */          ((md->moptions & PCRE_PARTIAL_SOFT) != 0 &&  /* Soft partial and */
3017           match_count < 0)                            /* no matches */           match_count < 0)                            /* no matches */
3018          ) &&                                         /* And... */          ) &&                                         /* And... */
3019          ptr >= end_subject &&                     /* Reached end of subject */          (
3020          ptr > current_subject)                    /* Matched non-empty string */          partial_newline ||                           /* Either partial NL */
3021              (                                          /* or ... */
3022              ptr >= end_subject &&                /* End of subject and */
3023              ptr > md->start_used_ptr)            /* Inspected non-empty string */
3024              )
3025            )
3026        {        {
3027        if (offsetcount >= 2)        if (offsetcount >= 2)
3028          {          {
3029          offsets[0] = md->start_used_ptr - start_subject;          offsets[0] = (int)(md->start_used_ptr - start_subject);
3030          offsets[1] = end_subject - start_subject;          offsets[1] = (int)(end_subject - start_subject);
3031          }          }
3032        match_count = PCRE_ERROR_PARTIAL;        match_count = PCRE_ERROR_PARTIAL;
3033        }        }
# Line 2598  Returns:          > 0 => number of match Line 3081  Returns:          > 0 => number of match
3081                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
3082  */  */
3083    
3084    #if defined COMPILE_PCRE8
3085  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
3086  pcre_dfa_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_dfa_exec(const pcre *argument_re, const pcre_extra *extra_data,
3087    const char *subject, int length, int start_offset, int options, int *offsets,    const char *subject, int length, int start_offset, int options, int *offsets,
3088    int offsetcount, int *workspace, int wscount)    int offsetcount, int *workspace, int wscount)
3089    #elif defined COMPILE_PCRE16
3090    PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
3091    pcre16_dfa_exec(const pcre16 *argument_re, const pcre16_extra *extra_data,
3092      PCRE_SPTR16 subject, int length, int start_offset, int options, int *offsets,
3093      int offsetcount, int *workspace, int wscount)
3094    #elif defined COMPILE_PCRE32
3095    PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
3096    pcre32_dfa_exec(const pcre32 *argument_re, const pcre32_extra *extra_data,
3097      PCRE_SPTR32 subject, int length, int start_offset, int options, int *offsets,
3098      int offsetcount, int *workspace, int wscount)
3099    #endif
3100  {  {
3101  real_pcre *re = (real_pcre *)argument_re;  REAL_PCRE *re = (REAL_PCRE *)argument_re;
3102  dfa_match_data match_block;  dfa_match_data match_block;
3103  dfa_match_data *md = &match_block;  dfa_match_data *md = &match_block;
3104  BOOL utf8, anchored, startline, firstline;  BOOL utf, anchored, startline, firstline;
3105  const uschar *current_subject, *end_subject, *lcc;  const pcre_uchar *current_subject, *end_subject;
   
 pcre_study_data internal_study;  
3106  const pcre_study_data *study = NULL;  const pcre_study_data *study = NULL;
 real_pcre internal_re;  
3107    
3108  const uschar *req_byte_ptr;  const pcre_uchar *req_char_ptr;
3109  const uschar *start_bits = NULL;  const pcre_uint8 *start_bits = NULL;
3110  BOOL first_byte_caseless = FALSE;  BOOL has_first_char = FALSE;
3111  BOOL req_byte_caseless = FALSE;  BOOL has_req_char = FALSE;
3112  int first_byte = -1;  pcre_uchar first_char = 0;
3113  int req_byte = -1;  pcre_uchar first_char2 = 0;
3114  int req_byte2 = -1;  pcre_uchar req_char = 0;
3115    pcre_uchar req_char2 = 0;
3116  int newline;  int newline;
3117    
3118  /* Plausibility checks */  /* Plausibility checks */
# Line 2629  if (re == NULL || subject == NULL || wor Line 3122  if (re == NULL || subject == NULL || wor
3122     (offsets == NULL && offsetcount > 0)) return PCRE_ERROR_NULL;     (offsets == NULL && offsetcount > 0)) return PCRE_ERROR_NULL;
3123  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;
3124  if (wscount < 20) return PCRE_ERROR_DFA_WSSIZE;  if (wscount < 20) return PCRE_ERROR_DFA_WSSIZE;
3125    if (length < 0) return PCRE_ERROR_BADLENGTH;
3126    if (start_offset < 0 || start_offset > length) return PCRE_ERROR_BADOFFSET;
3127    
3128    /* Check that the first field in the block is the magic number. If it is not,
3129    return with PCRE_ERROR_BADMAGIC. However, if the magic number is equal to
3130    REVERSED_MAGIC_NUMBER we return with PCRE_ERROR_BADENDIANNESS, which
3131    means that the pattern is likely compiled with different endianness. */
3132    
3133    if (re->magic_number != MAGIC_NUMBER)
3134      return re->magic_number == REVERSED_MAGIC_NUMBER?
3135        PCRE_ERROR_BADENDIANNESS:PCRE_ERROR_BADMAGIC;
3136    if ((re->flags & PCRE_MODE) == 0) return PCRE_ERROR_BADMODE;
3137    
3138  /* We need to find the pointer to any study data before we test for byte  /* If restarting after a partial match, do some sanity checks on the contents
3139  flipping, so we scan the extra_data block first. This may set two fields in the  of the workspace. */
3140  match block, so we must initialize them beforehand. However, the other fields  
3141  in the match block must not be set until after the byte flipping. */  if ((options & PCRE_DFA_RESTART) != 0)
3142      {
3143      if ((workspace[0] & (-2)) != 0 || workspace[1] < 1 ||
3144        workspace[1] > (wscount - 2)/INTS_PER_STATEBLOCK)
3145          return PCRE_ERROR_DFA_BADRESTART;
3146      }
3147    
3148    /* Set up study, callout, and table data */
3149    
3150  md->tables = re->tables;  md->tables = re->tables;
3151  md->callout_data = NULL;  md->callout_data = NULL;
# Line 2652  if (extra_data != NULL) Line 3164  if (extra_data != NULL)
3164      md->tables = extra_data->tables;      md->tables = extra_data->tables;
3165    }    }
3166    
 /* Check that the first field in the block is the magic number. If it is not,  
 test for a regex that was compiled on a host of opposite endianness. If this is  
 the case, flipped values are put in internal_re and internal_study if there was  
 study data too. */  
   
 if (re->magic_number != MAGIC_NUMBER)  
   {  
   re = _pcre_try_flipped(re, &internal_re, study, &internal_study);  
   if (re == NULL) return PCRE_ERROR_BADMAGIC;  
   if (study != NULL) study = &internal_study;  
   }  
   
3167  /* Set some local values */  /* Set some local values */
3168    
3169  current_subject = (const unsigned char *)subject + start_offset;  current_subject = (const pcre_uchar *)subject + start_offset;
3170  end_subject = (const unsigned char *)subject + length;  end_subject = (const pcre_uchar *)subject + length;
3171  req_byte_ptr = current_subject - 1;  req_char_ptr = current_subject - 1;
3172    
3173  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3174  utf8 = (re->options & PCRE_UTF8) != 0;  /* PCRE_UTF(16|32) have the same value as PCRE_UTF8. */
3175    utf = (re->options & PCRE_UTF8) != 0;
3176  #else  #else
3177  utf8 = FALSE;  utf = FALSE;
3178  #endif  #endif
3179    
3180  anchored = (options & (PCRE_ANCHORED|PCRE_DFA_RESTART)) != 0 ||  anchored = (options & (PCRE_ANCHORED|PCRE_DFA_RESTART)) != 0 ||
# Line 2681  anchored = (options & (PCRE_ANCHORED|PCR Line 3182  anchored = (options & (PCRE_ANCHORED|PCR
3182    
3183  /* The remaining fixed data for passing around. */  /* The remaining fixed data for passing around. */
3184    
3185  md->start_code = (const uschar *)argument_re +  md->start_code = (const pcre_uchar *)argument_re +
3186      re->name_table_offset + re->name_count * re->name_entry_size;      re->name_table_offset + re->name_count * re->name_entry_size;
3187  md->start_subject = (const unsigned char *)subject;  md->start_subject = (const pcre_uchar *)subject;
3188  md->end_subject = end_subject;  md->end_subject = end_subject;
3189  md->start_offset = start_offset;  md->start_offset = start_offset;
3190  md->moptions = options;  md->moptions = options;
# Line 2744  else Line 3245  else
3245  /* 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
3246  back the character offset. */  back the character offset. */
3247    
3248  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3249  if (utf8 && (options & PCRE_NO_UTF8_CHECK) == 0)  if (utf && (options & PCRE_NO_UTF8_CHECK) == 0)
3250    {    {
3251    if (_pcre_valid_utf8((uschar *)subject, length) >= 0)    int erroroffset;
3252      return PCRE_ERROR_BADUTF8;    int errorcode = PRIV(valid_utf)((pcre_uchar *)subject, length, &erroroffset);
3253    if (start_offset > 0 && start_offset < length)    if (errorcode != 0)
3254      {      {
3255      int tb = ((uschar *)subject)[start_offset];      if (offsetcount >= 2)
     if (tb > 127)  
3256        {        {
3257        tb &= 0xc0;        offsets[0] = erroroffset;
3258        if (tb != 0 && tb != 0xc0) return PCRE_ERROR_BADUTF8_OFFSET;        offsets[1] = errorcode;
3259        }        }
3260    #if defined COMPILE_PCRE8
3261        return (errorcode <= PCRE_UTF8_ERR5 && (options & PCRE_PARTIAL_HARD) != 0) ?
3262          PCRE_ERROR_SHORTUTF8 : PCRE_ERROR_BADUTF8;
3263    #elif defined COMPILE_PCRE16
3264        return (errorcode <= PCRE_UTF16_ERR1 && (options & PCRE_PARTIAL_HARD) != 0) ?
3265          PCRE_ERROR_SHORTUTF16 : PCRE_ERROR_BADUTF16;
3266    #elif defined COMPILE_PCRE32
3267        return PCRE_ERROR_BADUTF32;
3268    #endif
3269      }      }
3270    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
3271      if (start_offset > 0 && start_offset < length &&
3272            NOT_FIRSTCHAR(((PCRE_PUCHAR)subject)[start_offset]))
3273        return PCRE_ERROR_BADUTF8_OFFSET;
3274    #endif
3275    }    }
3276  #endif  #endif
3277    
# Line 2765  if (utf8 && (options & PCRE_NO_UTF8_CHEC Line 3279  if (utf8 && (options & PCRE_NO_UTF8_CHEC
3279  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
3280  in other programs later. */  in other programs later. */
3281    
3282  if (md->tables == NULL) md->tables = _pcre_default_tables;  if (md->tables == NULL) md->tables = PRIV(default_tables);
3283    
3284  /* 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
3285  used in a loop when finding where to start. */  where to start. */
3286    
 lcc = md->tables + lcc_offset;  
3287  startline = (re->flags & PCRE_STARTLINE) != 0;  startline = (re->flags & PCRE_STARTLINE) != 0;
3288  firstline = (re->options & PCRE_FIRSTLINE) != 0;  firstline = (re->options & PCRE_FIRSTLINE) != 0;
3289    
# Line 2784  if (!anchored) Line 3297  if (!anchored)
3297    {    {
3298    if ((re->flags & PCRE_FIRSTSET) != 0)    if ((re->flags & PCRE_FIRSTSET) != 0)
3299      {      {
3300      first_byte = re->first_byte & 255;      has_first_char = TRUE;
3301      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)      first_char = first_char2 = (pcre_uchar)(re->first_char);
3302        first_byte = lcc[first_byte];      if ((re->flags & PCRE_FCH_CASELESS) != 0)
3303          {
3304          first_char2 = TABLE_GET(first_char, md->tables + fcc_offset, first_char);
3305    #if defined SUPPORT_UCP && !(defined COMPILE_PCRE8)
3306          if (utf && first_char > 127)
3307            first_char2 = UCD_OTHERCASE(first_char);
3308    #endif
3309          }
3310      }      }
3311    else    else
3312      {      {
3313      if (startline && study != NULL &&      if (!startline && study != NULL &&
3314           (study->options & PCRE_STUDY_MAPPED) != 0)           (study->flags & PCRE_STUDY_MAPPED) != 0)
3315        start_bits = study->start_bits;        start_bits = study->start_bits;
3316      }      }
3317    }    }
# Line 2801  character" set. */ Line 3321  character" set. */
3321    
3322  if ((re->flags & PCRE_REQCHSET) != 0)  if ((re->flags & PCRE_REQCHSET) != 0)
3323    {    {
3324    req_byte = re->req_byte & 255;    has_req_char = TRUE;
3325    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;    req_char = req_char2 = (pcre_uchar)(re->req_char);
3326    req_byte2 = (md->tables + fcc_offset)[req_byte];  /* case flipped */    if ((re->flags & PCRE_RCH_CASELESS) != 0)
3327        {
3328        req_char2 = TABLE_GET(req_char, md->tables + fcc_offset, req_char);
3329    #if defined SUPPORT_UCP && !(defined COMPILE_PCRE8)
3330        if (utf && req_char > 127)
3331          req_char2 = UCD_OTHERCASE(req_char);
3332    #endif
3333        }
3334    }    }
3335    
3336  /* 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 2816  for (;;) Line 3343  for (;;)
3343    
3344    if ((options & PCRE_DFA_RESTART) == 0)    if ((options & PCRE_DFA_RESTART) == 0)
3345      {      {
3346      const uschar *save_end_subject = end_subject;      const pcre_uchar *save_end_subject = end_subject;
3347    
3348      /* 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
3349      line of a multiline string. Implement this by temporarily adjusting      line of a multiline string. Implement this by temporarily adjusting
# Line 2825  for (;;) Line 3352  for (;;)
3352    
3353      if (firstline)      if (firstline)
3354        {        {
3355        USPTR t = current_subject;        PCRE_PUCHAR t = current_subject;
3356  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3357        if (utf8)        if (utf)
3358          {          {
3359          while (t < md->end_subject && !IS_NEWLINE(t))          while (t < md->end_subject && !IS_NEWLINE(t))
3360            {            {
3361            t++;            t++;
3362            while (t < end_subject && (*t & 0xc0) == 0x80) t++;            ACROSSCHAR(t < end_subject, *t, t++);
3363            }            }
3364          }          }
3365        else        else
# Line 2842  for (;;) Line 3369  for (;;)
3369        }        }
3370    
3371      /* There are some optimizations that avoid running the match if a known      /* There are some optimizations that avoid running the match if a known
3372      starting point is not found, or if a known later character is not present.      starting point is not found. However, there is an option that disables
3373      However, there is an option that disables these, for testing and for      these, for testing and for ensuring that all callouts do actually occur.
3374      ensuring that all callouts do actually occur. */      The option can be set in the regex by (*NO_START_OPT) or passed in
3375        match-time options. */
3376    
3377      if ((options & PCRE_NO_START_OPTIMIZE) == 0)      if (((options | re->options) & PCRE_NO_START_OPTIMIZE) == 0)
3378        {        {
3379          /* Advance to a known first char. */
3380    
3381        /* Advance to a known first byte. */        if (has_first_char)
   
       if (first_byte >= 0)  
3382          {          {
3383          if (first_byte_caseless)          if (first_char != first_char2)
3384              {
3385              pcre_uchar csc;
3386            while (current_subject < end_subject &&            while (current_subject < end_subject &&
3387                   lcc[*current_subject] != first_byte)                   (csc = RAWUCHARTEST(current_subject)) != first_char && csc != first_char2)
3388              current_subject++;              current_subject++;
3389              }
3390          else          else
3391            while (current_subject < end_subject &&            while (current_subject < end_subject &&
3392                   *current_subject != first_byte)                   RAWUCHARTEST(current_subject) != first_char)
3393              current_subject++;              current_subject++;
3394          }          }
3395    
# Line 2869  for (;;) Line 3399  for (;;)
3399          {          {
3400          if (current_subject > md->start_subject + start_offset)          if (current_subject > md->start_subject + start_offset)
3401            {            {
3402  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3403            if (utf8)            if (utf)
3404              {              {
3405              while (current_subject < end_subject &&              while (current_subject < end_subject &&
3406                     !WAS_NEWLINE(current_subject))                     !WAS_NEWLINE(current_subject))
3407                {                {
3408                current_subject++;                current_subject++;
3409                while(current_subject < end_subject &&                ACROSSCHAR(current_subject < end_subject, *current_subject,
3410                      (*current_subject & 0xc0) == 0x80)                  current_subject++);
                 current_subject++;  
3411                }                }
3412              }              }
3413            else            else
# Line 2890  for (;;) Line 3419  for (;;)
3419            ANYCRLF, and we are now at a LF, advance the match position by one            ANYCRLF, and we are now at a LF, advance the match position by one
3420            more character. */            more character. */
3421    
3422            if (current_subject[-1] == CHAR_CR &&            if (RAWUCHARTEST(current_subject - 1) == CHAR_CR &&
3423                 (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&                 (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&
3424                 current_subject < end_subject &&                 current_subject < end_subject &&
3425                 *current_subject == CHAR_NL)                 RAWUCHARTEST(current_subject) == CHAR_NL)
3426              current_subject++;              current_subject++;
3427            }            }
3428          }          }
# Line 2904  for (;;) Line 3433  for (;;)
3433          {          {
3434          while (current_subject < end_subject)          while (current_subject < end_subject)
3435            {            {
3436            register unsigned int c = *current_subject;            register pcre_uint32 c = RAWUCHARTEST(current_subject);
3437            if ((start_bits[c/8] & (1 << (c&7))) == 0) current_subject++;  #ifndef COMPILE_PCRE8
3438              else break;            if (c > 255) c = 255;
3439    #endif
3440              if ((start_bits[c/8] & (1 << (c&7))) == 0)
3441                {
3442                current_subject++;
3443    #if defined SUPPORT_UTF && defined COMPILE_PCRE8
3444                /* In non 8-bit mode, the iteration will stop for
3445                characters > 255 at the beginning or not stop at all. */
3446                if (utf)
3447                  ACROSSCHAR(current_subject < end_subject, *current_subject,
3448                    current_subject++);
3449    #endif
3450                }
3451              else break;
3452            }            }
3453          }          }
3454        }        }
# Line 2914  for (;;) Line 3456  for (;;)
3456      /* Restore fudged end_subject */      /* Restore fudged end_subject */
3457    
3458      end_subject = save_end_subject;      end_subject = save_end_subject;
     }  
3459    
3460    /* If req_byte is set, we know that that character must appear in the subject      /* The following two optimizations are disabled for partial matching or if
3461    for the match to succeed. If the first character is set, req_byte must be      disabling is explicitly requested (and of course, by the test above, this
3462    later in the subject; otherwise the test starts at the match point. This      code is not obeyed when restarting after a partial match). */
   optimization can save a huge amount of work in patterns with nested unlimited  
   repeats that aren't going to match. Writing separate code for cased/caseless  
   versions makes it go faster, as does using an autoincrement and backing off  
   on a match.  
   
   HOWEVER: when the subject string is very, very long, searching to its end can  
   take a long time, and give bad performance on quite ordinary patterns. This  
   showed up when somebody was matching /^C/ on a 32-megabyte string... so we  
   don't do this when the string is sufficiently long.  
   
   ALSO: this processing is disabled when partial matching is requested, and can  
   also be explicitly deactivated. Furthermore, we have to disable when  
   restarting after a partial match, because the required character may have  
   already been matched. */  
   
   if ((options & PCRE_NO_START_OPTIMIZE) == 0 &&  
       req_byte >= 0 &&  
       end_subject - current_subject < REQ_BYTE_MAX &&  
       (options & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_RESTART)) == 0)  
     {  
     register const uschar *p = current_subject + ((first_byte >= 0)? 1 : 0);  
   
     /* We don't need to repeat the search if we haven't yet reached the  
     place we found it at last time. */  
3463    
3464      if (p > req_byte_ptr)      if (((options | re->options) & PCRE_NO_START_OPTIMIZE) == 0 &&
3465            (options & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) == 0)
3466        {        {
3467        if (req_byte_caseless)        /* If the pattern was studied, a minimum subject length may be set. This
3468          {        is a lower bound; no actual string of that length may actually match the
3469          while (p < end_subject)        pattern. Although the value is, strictly, in characters, we treat it as
3470            {        bytes to avoid spending too much time in this optimization. */
3471            register int pp = *p++;  
3472            if (pp == req_byte || pp == req_byte2) { p--; break; }        if (study != NULL && (study->flags & PCRE_STUDY_MINLEN) != 0 &&
3473            }            (pcre_uint32)(end_subject - current_subject) < study->minlength)
3474          }          return PCRE_ERROR_NOMATCH;
3475        else  
3476          /* If req_char is set, we know that that character must appear in the
3477          subject for the match to succeed. If the first character is set, req_char
3478          must be later in the subject; otherwise the test starts at the match
3479          point. This optimization can save a huge amount of work in patterns with
3480          nested unlimited repeats that aren't going to match. Writing separate
3481          code for cased/caseless versions makes it go faster, as does using an
3482          autoincrement and backing off on a match.
3483    
3484          HOWEVER: when the subject string is very, very long, searching to its end
3485          can take a long time, and give bad performance on quite ordinary
3486          patterns. This showed up when somebody was matching /^C/ on a 32-megabyte
3487          string... so we don't do this when the string is sufficiently long. */
3488    
3489          if (has_req_char && end_subject - current_subject < REQ_BYTE_MAX)
3490          {          {
3491          while (p < end_subject)          register PCRE_PUCHAR p = current_subject + (has_first_char? 1:0);
3492    
3493            /* We don't need to repeat the search if we haven't yet reached the
3494            place we found it at last time. */
3495    
3496            if (p > req_char_ptr)
3497            {            {
3498            if (*p++ == req_byte) { p--; break; }            if (req_char != req_char2)
3499            }              {
3500          }              while (p < end_subject)
3501                  {
3502                  register pcre_uint32 pp = RAWUCHARINCTEST(p);
3503                  if (pp == req_char || pp == req_char2) { p--; break; }
3504                  }
3505                }
3506              else
3507                {
3508                while (p < end_subject)
3509                  {
3510                  if (RAWUCHARINCTEST(p) == req_char) { p--; break; }
3511                  }
3512                }
3513    
3514        /* If we can't find the required character, break the matching loop,            /* If we can't find the required character, break the matching loop,
3515        which will cause a return or PCRE_ERROR_NOMATCH. */            which will cause a return or PCRE_ERROR_NOMATCH. */
3516    
3517        if (p >= end_subject) break;            if (p >= end_subject) break;
3518    
3519        /* If we have found the required character, save the point where we            /* If we have found the required character, save the point where we
3520        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
3521        the start hasn't passed this character yet. */            the start hasn't passed this character yet. */
3522    
3523        req_byte_ptr = p;            req_char_ptr = p;
3524              }
3525            }
3526        }        }
3527      }      }   /* End of optimizations that are done when not restarting */
3528    
3529    /* OK, now we can do the business */    /* OK, now we can do the business */
3530    
3531    md->start_used_ptr = current_subject;    md->start_used_ptr = current_subject;
3532      md->recursive = NULL;
3533    
3534    rc = internal_dfa_exec(    rc = internal_dfa_exec(
3535      md,                                /* fixed match data */      md,                                /* fixed match data */
3536      md->start_code,                    /* this subexpression's code */      md->start_code,                    /* this subexpression's code */
# Line 2988  for (;;) Line 3540  for (;;)
3540      offsetcount,                       /* size of same */      offsetcount,                       /* size of same */
3541      workspace,                         /* workspace vector */      workspace,                         /* workspace vector */
3542      wscount,                           /* size of same */      wscount,                           /* size of same */
3543      re->options & (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL), /* ims flags */      0);                                /* function recurse level */
     0,                                 /* function recurse level */  
     0);                                /* regex recurse level */  
3544    
3545    /* Anything other than "no match" means we are done, always; otherwise, carry    /* Anything other than "no match" means we are done, always; otherwise, carry
3546    on only if not anchored. */    on only if not anchored. */
# Line 3002  for (;;) Line 3552  for (;;)
3552    
3553    if (firstline && IS_NEWLINE(current_subject)) break;    if (firstline && IS_NEWLINE(current_subject)) break;
3554    current_subject++;    current_subject++;
3555    if (utf8)  #ifdef SUPPORT_UTF
3556      if (utf)
3557      {      {
3558      while (current_subject < end_subject && (*current_subject & 0xc0) == 0x80)      ACROSSCHAR(current_subject < end_subject, *current_subject,
3559        current_subject++;        current_subject++);
3560      }      }
3561    #endif
3562    if (current_subject > end_subject) break;    if (current_subject > end_subject) break;
3563    
3564    /* 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
3565    not contain any explicit matches for \r or \n, and the newline option is CRLF    not contain any explicit matches for \r or \n, and the newline option is CRLF
3566    or ANY or ANYCRLF, advance the match position by one more character. */    or ANY or ANYCRLF, advance the match position by one more character. */
3567    
3568    if (current_subject[-1] == CHAR_CR &&    if (RAWUCHARTEST(current_subject - 1) == CHAR_CR &&
3569        current_subject < end_subject &&        current_subject < end_subject &&
3570        *current_subject == CHAR_NL &&        RAWUCHARTEST(current_subject) == CHAR_NL &&
3571        (re->flags & PCRE_HASCRORLF) == 0 &&        (re->flags & PCRE_HASCRORLF) == 0 &&
3572          (md->nltype == NLTYPE_ANY ||          (md->nltype == NLTYPE_ANY ||
3573           md->nltype == NLTYPE_ANYCRLF ||           md->nltype == NLTYPE_ANYCRLF ||

Legend:
Removed from v.442  
changed lines
  Added in v.1233

  ViewVC Help
Powered by ViewVC 1.1.5