/[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 498 by ph10, Wed Mar 3 19:29:38 2010 UTC revision 982 by ph10, Wed Jun 20 15:15:27 2012 UTC
# Line 7  and semantics are as close as possible t Line 7  and semantics are as close as possible t
7  below for why this module is different).  below for why this module is different).
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2010 University of Cambridge             Copyright (c) 1997-2012 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 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    
# 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. Non-zero values in the table are the offsets from the opcode where  small value. Non-zero values in the table are the offsets from the opcode where
112  the character is to be found. ***NOTE*** If the start of this table is  the character is to be found. ***NOTE*** If the start of this table is
113  modified, the three tables that follow must also be modified. */  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                 */
# Line 121  static const uschar coptable[] = { Line 120  static const uschar coptable[] = {
120    0, 0,                          /* \P, \p                                 */    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,                             /* \X                                     */    0,                             /* \X                                     */
123    0, 0, 0, 0, 0,                 /* \Z, \z, Opt, ^, $                      */    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 144  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,                             /* Reverse                                */    0, 0,                          /* ONCE, ONCE_NC                          */
172    0, 0, 0, 0,                    /* ONCE, BRA, CBRA, COND                  */    0, 0, 0, 0, 0,                 /* BRA, BRAPOS, CBRA, CBRAPOS, COND       */
173    0, 0, 0,                       /* SBRA, SCBRA, SCOND                     */    0, 0, 0, 0, 0,                 /* SBRA, SBRAPOS, SCBRA, SCBRAPOS, SCOND  */
174    0, 0,                          /* CREF, NCREF                            */    0, 0,                          /* CREF, NCREF                            */
175    0, 0,                          /* RREF, NRREF                            */    0, 0,                          /* RREF, NRREF                            */
176    0,                             /* DEF                                    */    0,                             /* DEF                                    */
177    0, 0,                          /* BRAZERO, BRAMINZERO                    */    0, 0, 0,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */
178    0, 0, 0, 0,                    /* PRUNE, SKIP, THEN, COMMIT              */    0, 0, 0,                       /* MARK, PRUNE, PRUNE_ARG                 */
179    0, 0, 0, 0                     /* FAIL, ACCEPT, CLOSE, SKIPZERO          */    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  /* This table identifies those opcodes that inspect a character. It is used to
# Line 170  remember the fact that a character could Line 186  remember the fact that a character could
186  the subject is reached. ***NOTE*** If the start of this table is modified, the  the subject is reached. ***NOTE*** If the start of this table is modified, the
187  two tables that follow must also be modified. */  two tables that follow must also be modified. */
188    
189  static const uschar poptable[] = {  static const pcre_uint8 poptable[] = {
190    0,                             /* End                                    */    0,                             /* End                                    */
191    0, 0, 0, 1, 1,                 /* \A, \G, \K, \B, \b                     */    0, 0, 0, 1, 1,                 /* \A, \G, \K, \B, \b                     */
192    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */
# Line 178  static const uschar poptable[] = { Line 194  static const uschar poptable[] = {
194    1, 1,                          /* \P, \p                                 */    1, 1,                          /* \P, \p                                 */
195    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */
196    1,                             /* \X                                     */    1,                             /* \X                                     */
197    0, 0, 0, 0, 0,                 /* \Z, \z, Opt, ^, $                      */    0, 0, 0, 0, 0, 0,              /* \Z, \z, ^, ^M, $, $M                   */
198    1,                             /* Char                                   */    1,                             /* Char                                   */
199    1,                             /* Charnc                                 */    1,                             /* Chari                                  */
200    1,                             /* not                                    */    1,                             /* not                                    */
201      1,                             /* noti                                   */
202    /* Positive single-char repeats                                          */    /* Positive single-char repeats                                          */
203    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */
204    1, 1, 1,                       /* upto, minupto, exact                   */    1, 1, 1,                       /* upto, minupto, exact                   */
205    1, 1, 1, 1,                    /* *+, ++, ?+, upto+                      */    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                   */    /* Negative single-char repeats - only for chars < 256                   */
210    1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */    1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */
211    1, 1, 1,                       /* NOT upto, minupto, exact               */    1, 1, 1,                       /* NOT upto, minupto, exact               */
212    1, 1, 1, 1,                    /* NOT *+, ++, ?+, upto+                  */    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                                                 */    /* Positive type repeats                                                 */
217    1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */    1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */
218    1, 1, 1,                       /* Type upto, minupto, exact              */    1, 1, 1,                       /* Type upto, minupto, exact              */
# Line 201  static const uschar poptable[] = { Line 224  static const uschar poptable[] = {
224    1,                             /* NCLASS                                 */    1,                             /* NCLASS                                 */
225    1,                             /* XCLASS - variable length               */    1,                             /* XCLASS - variable length               */
226    0,                             /* REF                                    */    0,                             /* REF                                    */
227      0,                             /* REFI                                   */
228    0,                             /* RECURSE                                */    0,                             /* RECURSE                                */
229    0,                             /* CALLOUT                                */    0,                             /* CALLOUT                                */
230    0,                             /* Alt                                    */    0,                             /* Alt                                    */
231    0,                             /* Ket                                    */    0,                             /* Ket                                    */
232    0,                             /* KetRmax                                */    0,                             /* KetRmax                                */
233    0,                             /* KetRmin                                */    0,                             /* KetRmin                                */
234      0,                             /* KetRpos                                */
235      0,                             /* Reverse                                */
236    0,                             /* Assert                                 */    0,                             /* Assert                                 */
237    0,                             /* Assert not                             */    0,                             /* Assert not                             */
238    0,                             /* Assert behind                          */    0,                             /* Assert behind                          */
239    0,                             /* Assert behind not                      */    0,                             /* Assert behind not                      */
240    0,                             /* Reverse                                */    0, 0,                          /* ONCE, ONCE_NC                          */
241    0, 0, 0, 0,                    /* ONCE, BRA, CBRA, COND                  */    0, 0, 0, 0, 0,                 /* BRA, BRAPOS, CBRA, CBRAPOS, COND       */
242    0, 0, 0,                       /* SBRA, SCBRA, SCOND                     */    0, 0, 0, 0, 0,                 /* SBRA, SBRAPOS, SCBRA, SCBRAPOS, SCOND  */
243    0, 0,                          /* CREF, NCREF                            */    0, 0,                          /* CREF, NCREF                            */
244    0, 0,                          /* RREF, NRREF                            */    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, 0                     /* FAIL, ACCEPT, CLOSE, 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 233  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 250  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 PCRE_DEBUG  #ifdef PCRE_DEBUG
# Line 273  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;  int c;
306  while (length-- > 0)  while (length-- > 0)
# Line 306  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 323  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 334  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 346  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 357  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 369  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 410  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 424  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 452  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 473  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    
# Line 490  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 523  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 536  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 547  for (;;) Line 573  for (;;)
573    int clen, dlen;    int clen, dlen;
574    unsigned int c, d;    unsigned int c, d;
575    int forced_fail = 0;    int forced_fail = 0;
576    BOOL could_continue = FALSE;    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 563  for (;;) Line 591  for (;;)
591    
592  #ifdef PCRE_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 583  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      if (utf) { GETCHARLEN(c, ptr, clen); } else
617  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
618      c = *ptr;      c = *ptr;
619      }      }
620    else    else
# Line 603  for (;;) Line 631  for (;;)
631    for (i = 0; i < active_count; i++)    for (i = 0; i < active_count; i++)
632      {      {
633      stateblock *current_state = active_states + i;      stateblock *current_state = active_states + i;
634      const uschar *code;      BOOL caseless = FALSE;
635        const pcre_uchar *code;
636      int state_offset = current_state->offset;      int state_offset = current_state->offset;
637      int count, codevalue, rrc;      int count, codevalue, rrc;
638    
# Line 614  for (;;) Line 643  for (;;)
643          else printf("0x%02x\n", c);          else printf("0x%02x\n", c);
644  #endif  #endif
645    
     /* This variable is referred to implicity in the ADD_xxx macros. */  
   
     ims = current_state->ims;  
   
646      /* 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
647      (negated) state until the number of characters in the data field have      (negated) state until the number of characters in the data field have
648      been skipped". */      been skipped". If the could_continue flag was passed over from a previous
649        state, arrange for it to passed on. */
650    
651      if (state_offset < 0)      if (state_offset < 0)
652        {        {
# Line 629  for (;;) Line 655  for (;;)
655          DPRINTF(("%.*sSkipping this character\n", rlevel*2-2, SP));          DPRINTF(("%.*sSkipping this character\n", rlevel*2-2, SP));
656          ADD_NEW_DATA(state_offset, current_state->count,          ADD_NEW_DATA(state_offset, current_state->count,
657            current_state->data - 1);            current_state->data - 1);
658            if (could_continue) reset_could_continue = TRUE;
659          continue;          continue;
660          }          }
661        else        else
# Line 668  for (;;) Line 695  for (;;)
695      permitted.      permitted.
696    
697      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
698      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
699      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,
700      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
701      */      these ones to new opcodes. */
702    
703      if (coptable[codevalue] > 0)      if (coptable[codevalue] > 0)
704        {        {
705        dlen = 1;        dlen = 1;
706  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
707        if (utf8) { GETCHARLEN(d, (code + coptable[codevalue]), dlen); } else        if (utf) { GETCHARLEN(d, (code + coptable[codevalue]), dlen); } else
708  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
709        d = code[coptable[codevalue]];        d = code[coptable[codevalue]];
710        if (codevalue >= OP_TYPESTAR)        if (codevalue >= OP_TYPESTAR)
711          {          {
# Line 714  for (;;) Line 741  for (;;)
741        opcode, are not the correct length. It seems to be the only way to do        opcode, are not the correct length. It seems to be the only way to do
742        such a check at compile time, as the sizeof() operator does not work        such a check at compile time, as the sizeof() operator does not work
743        in the C preprocessor. */        in the C preprocessor. */
744    
745        case OP_TABLE_LENGTH:        case OP_TABLE_LENGTH:
746        case OP_TABLE_LENGTH +        case OP_TABLE_LENGTH +
747          ((sizeof(coptable) == OP_TABLE_LENGTH) &&          ((sizeof(coptable) == OP_TABLE_LENGTH) &&
748           (sizeof(poptable) == OP_TABLE_LENGTH)):           (sizeof(poptable) == OP_TABLE_LENGTH)):
749        break;        break;
750    
751  /* ========================================================================== */  /* ========================================================================== */
752        /* 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
753        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
754          state. Note that KETRPOS will always be encountered at the end of the
755          subpattern, because the possessive subpattern repeats are always handled
756          using recursive calls. Thus, it never adds any new states.
757    
758          At the end of the (sub)pattern, unless we have an empty string and
759        PCRE_NOTEMPTY is set, or PCRE_NOTEMPTY_ATSTART is set and we are at the        PCRE_NOTEMPTY is set, or PCRE_NOTEMPTY_ATSTART is set and we are at the
760        start of the subject, save the match data, shifting up all previous        start of the subject, save the match data, shifting up all previous
761        matches so we always have the longest first. */        matches so we always have the longest first. */
# Line 731  for (;;) Line 763  for (;;)
763        case OP_KET:        case OP_KET:
764        case OP_KETRMIN:        case OP_KETRMIN:
765        case OP_KETRMAX:        case OP_KETRMAX:
766          case OP_KETRPOS:
767        if (code != end_code)        if (code != end_code)
768          {          {
769          ADD_ACTIVE(state_offset + 1 + LINK_SIZE, 0);          ADD_ACTIVE(state_offset + 1 + LINK_SIZE, 0);
# Line 747  for (;;) Line 780  for (;;)
780                  current_subject > start_subject + md->start_offset)))                  current_subject > start_subject + md->start_offset)))
781            {            {
782            if (match_count < 0) match_count = (offsetcount >= 2)? 1 : 0;            if (match_count < 0) match_count = (offsetcount >= 2)? 1 : 0;
783              else if (match_count > 0 && ++match_count * 2 >= offsetcount)              else if (match_count > 0 && ++match_count * 2 > offsetcount)
784                match_count = 0;                match_count = 0;
785            count = ((match_count == 0)? offsetcount : match_count * 2) - 2;            count = ((match_count == 0)? offsetcount : match_count * 2) - 2;
786            if (count > 0) memmove(offsets + 2, offsets, count * sizeof(int));            if (count > 0) memmove(offsets + 2, offsets, count * sizeof(int));
787            if (offsetcount >= 2)            if (offsetcount >= 2)
788              {              {
789              offsets[0] = current_subject - start_subject;              offsets[0] = (int)(current_subject - start_subject);
790              offsets[1] = ptr - start_subject;              offsets[1] = (int)(ptr - start_subject);
791              DPRINTF(("%.*sSet matched string = \"%.*s\"\n", rlevel*2-2, SP,              DPRINTF(("%.*sSet matched string = \"%.*s\"\n", rlevel*2-2, SP,
792                offsets[1] - offsets[0], current_subject));                offsets[1] - offsets[0], (char *)current_subject));
793              }              }
794            if ((md->moptions & PCRE_DFA_SHORTEST) != 0)            if ((md->moptions & PCRE_DFA_SHORTEST) != 0)
795              {              {
# Line 776  for (;;) Line 809  for (;;)
809        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
810        case OP_ALT:        case OP_ALT:
811        do { code += GET(code, 1); } while (*code == OP_ALT);        do { code += GET(code, 1); } while (*code == OP_ALT);
812        ADD_ACTIVE(code - start_code, 0);        ADD_ACTIVE((int)(code - start_code), 0);
813        break;        break;
814    
815        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
# Line 784  for (;;) Line 817  for (;;)
817        case OP_SBRA:        case OP_SBRA:
818        do        do
819          {          {
820          ADD_ACTIVE(code - start_code + 1 + LINK_SIZE, 0);          ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE), 0);
821          code += GET(code, 1);          code += GET(code, 1);
822          }          }
823        while (*code == OP_ALT);        while (*code == OP_ALT);
# Line 793  for (;;) Line 826  for (;;)
826        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
827        case OP_CBRA:        case OP_CBRA:
828        case OP_SCBRA:        case OP_SCBRA:
829        ADD_ACTIVE(code - start_code + 3 + LINK_SIZE,  0);        ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE + IMM2_SIZE),  0);
830        code += GET(code, 1);        code += GET(code, 1);
831        while (*code == OP_ALT)        while (*code == OP_ALT)
832          {          {
833          ADD_ACTIVE(code - start_code + 1 + LINK_SIZE,  0);          ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE),  0);
834          code += GET(code, 1);          code += GET(code, 1);
835          }          }
836        break;        break;
# Line 808  for (;;) Line 841  for (;;)
841        ADD_ACTIVE(state_offset + 1, 0);        ADD_ACTIVE(state_offset + 1, 0);
842        code += 1 + GET(code, 2);        code += 1 + GET(code, 2);
843        while (*code == OP_ALT) code += GET(code, 1);        while (*code == OP_ALT) code += GET(code, 1);
844        ADD_ACTIVE(code - start_code + 1 + LINK_SIZE, 0);        ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE), 0);
845        break;        break;
846    
847        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
848        case OP_SKIPZERO:        case OP_SKIPZERO:
849        code += 1 + GET(code, 2);        code += 1 + GET(code, 2);
850        while (*code == OP_ALT) code += GET(code, 1);        while (*code == OP_ALT) code += GET(code, 1);
851        ADD_ACTIVE(code - start_code + 1 + LINK_SIZE, 0);        ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE), 0);
852        break;        break;
853    
854        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
855        case OP_CIRC:        case OP_CIRC:
856        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)))  
857          { ADD_ACTIVE(state_offset + 1, 0); }          { ADD_ACTIVE(state_offset + 1, 0); }
858        break;        break;
859    
860        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
861        case OP_EOD:        case OP_CIRCM:
862        if (ptr >= end_subject) { ADD_ACTIVE(state_offset + 1, 0); }        if ((ptr == start_subject && (md->moptions & PCRE_NOTBOL) == 0) ||
863              (ptr != end_subject && WAS_NEWLINE(ptr)))
864            { ADD_ACTIVE(state_offset + 1, 0); }
865        break;        break;
866    
867        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
868        case OP_OPT:        case OP_EOD:
869        ims = code[1];        if (ptr >= end_subject)
870        ADD_ACTIVE(state_offset + 2, 0);          {
871            if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
872              could_continue = TRUE;
873            else { ADD_ACTIVE(state_offset + 1, 0); }
874            }
875        break;        break;
876    
877        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
# Line 858  for (;;) Line 894  for (;;)
894        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
895        case OP_ANY:        case OP_ANY:
896        if (clen > 0 && !IS_NEWLINE(ptr))        if (clen > 0 && !IS_NEWLINE(ptr))
897          { ADD_NEW(state_offset + 1, 0); }          {
898            if (ptr + 1 >= md->end_subject &&
899                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
900                NLBLOCK->nltype == NLTYPE_FIXED &&
901                NLBLOCK->nllen == 2 &&
902                c == NLBLOCK->nl[0])
903              {
904              could_continue = partial_newline = TRUE;
905              }
906            else
907              {
908              ADD_NEW(state_offset + 1, 0);
909              }
910            }
911        break;        break;
912    
913        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
# Line 869  for (;;) Line 918  for (;;)
918    
919        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
920        case OP_EODN:        case OP_EODN:
921        if (clen == 0 || (IS_NEWLINE(ptr) && ptr == end_subject - md->nllen))        if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
922            could_continue = TRUE;
923          else if (clen == 0 || (IS_NEWLINE(ptr) && ptr == end_subject - md->nllen))
924          { ADD_ACTIVE(state_offset + 1, 0); }          { ADD_ACTIVE(state_offset + 1, 0); }
925        break;        break;
926    
# Line 877  for (;;) Line 928  for (;;)
928        case OP_DOLL:        case OP_DOLL:
929        if ((md->moptions & PCRE_NOTEOL) == 0)        if ((md->moptions & PCRE_NOTEOL) == 0)
930          {          {
931          if (clen == 0 ||          if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
932              could_continue = TRUE;
933            else if (clen == 0 ||
934              ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr) &&              ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr) &&
935                 ((ims & PCRE_MULTILINE) != 0 || ptr == end_subject - md->nllen)                 (ptr == end_subject - md->nllen)
936              ))              ))
937            { ADD_ACTIVE(state_offset + 1, 0); }            { ADD_ACTIVE(state_offset + 1, 0); }
938            else if (ptr + 1 >= md->end_subject &&
939                     (md->moptions & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) != 0 &&
940                     NLBLOCK->nltype == NLTYPE_FIXED &&
941                     NLBLOCK->nllen == 2 &&
942                     c == NLBLOCK->nl[0])
943              {
944              if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
945                {
946                reset_could_continue = TRUE;
947                ADD_NEW_DATA(-(state_offset + 1), 0, 1);
948                }
949              else could_continue = partial_newline = TRUE;
950              }
951          }          }
952        else if ((ims & PCRE_MULTILINE) != 0 && IS_NEWLINE(ptr))        break;
953    
954          /*-----------------------------------------------------------------*/
955          case OP_DOLLM:
956          if ((md->moptions & PCRE_NOTEOL) == 0)
957            {
958            if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
959              could_continue = TRUE;
960            else if (clen == 0 ||
961                ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr)))
962              { ADD_ACTIVE(state_offset + 1, 0); }
963            else if (ptr + 1 >= md->end_subject &&
964                     (md->moptions & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) != 0 &&
965                     NLBLOCK->nltype == NLTYPE_FIXED &&
966                     NLBLOCK->nllen == 2 &&
967                     c == NLBLOCK->nl[0])
968              {
969              if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
970                {
971                reset_could_continue = TRUE;
972                ADD_NEW_DATA(-(state_offset + 1), 0, 1);
973                }
974              else could_continue = partial_newline = TRUE;
975              }
976            }
977          else if (IS_NEWLINE(ptr))
978          { ADD_ACTIVE(state_offset + 1, 0); }          { ADD_ACTIVE(state_offset + 1, 0); }
979        break;        break;
980    
# Line 914  for (;;) Line 1005  for (;;)
1005    
1006          if (ptr > start_subject)          if (ptr > start_subject)
1007            {            {
1008            const uschar *temp = ptr - 1;            const pcre_uchar *temp = ptr - 1;
1009            if (temp < md->start_used_ptr) md->start_used_ptr = temp;            if (temp < md->start_used_ptr) md->start_used_ptr = temp;
1010  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
1011            if (utf8) BACKCHAR(temp);            if (utf) { BACKCHAR(temp); }
1012  #endif  #endif
1013            GETCHARTEST(d, temp);            GETCHARTEST(d, temp);
1014    #ifdef SUPPORT_UCP
1015              if ((md->poptions & PCRE_UCP) != 0)
1016                {
1017                if (d == '_') left_word = TRUE; else
1018                  {
1019                  int cat = UCD_CATEGORY(d);
1020                  left_word = (cat == ucp_L || cat == ucp_N);
1021                  }
1022                }
1023              else
1024    #endif
1025            left_word = d < 256 && (ctypes[d] & ctype_word) != 0;            left_word = d < 256 && (ctypes[d] & ctype_word) != 0;
1026            }            }
1027          else left_word = 0;          else left_word = FALSE;
1028    
1029          if (clen > 0)          if (clen > 0)
1030              {
1031    #ifdef SUPPORT_UCP
1032              if ((md->poptions & PCRE_UCP) != 0)
1033                {
1034                if (c == '_') right_word = TRUE; else
1035                  {
1036                  int cat = UCD_CATEGORY(c);
1037                  right_word = (cat == ucp_L || cat == ucp_N);
1038                  }
1039                }
1040              else
1041    #endif
1042            right_word = c < 256 && (ctypes[c] & ctype_word) != 0;            right_word = c < 256 && (ctypes[c] & ctype_word) != 0;
1043          else right_word = 0;            }
1044            else right_word = FALSE;
1045    
1046          if ((left_word == right_word) == (codevalue == OP_NOT_WORD_BOUNDARY))          if ((left_word == right_word) == (codevalue == OP_NOT_WORD_BOUNDARY))
1047            { ADD_ACTIVE(state_offset + 1, 0); }            { ADD_ACTIVE(state_offset + 1, 0); }
# Line 953  for (;;) Line 1068  for (;;)
1068            break;            break;
1069    
1070            case PT_LAMP:            case PT_LAMP:
1071            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
1072                   prop->chartype == ucp_Lt;
1073            break;            break;
1074    
1075            case PT_GC:            case PT_GC:
1076            OK = _pcre_ucp_gentype[prop->chartype] == code[2];            OK = PRIV(ucp_gentype)[prop->chartype] == code[2];
1077            break;            break;
1078    
1079            case PT_PC:            case PT_PC:
# Line 968  for (;;) Line 1084  for (;;)
1084            OK = prop->script == code[2];            OK = prop->script == code[2];
1085            break;            break;
1086    
1087              /* These are specials for combination cases. */
1088    
1089              case PT_ALNUM:
1090              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1091                   PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1092              break;
1093    
1094              case PT_SPACE:    /* Perl space */
1095              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1096                   c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1097              break;
1098    
1099              case PT_PXSPACE:  /* POSIX space */
1100              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1101                   c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1102                   c == CHAR_FF || c == CHAR_CR;
1103              break;
1104    
1105              case PT_WORD:
1106              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1107                   PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1108                   c == CHAR_UNDERSCORE;
1109              break;
1110    
1111            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
1112    
1113            default:            default:
# Line 995  for (;;) Line 1135  for (;;)
1135        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
1136        if (clen > 0)        if (clen > 0)
1137          {          {
1138          if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||          if (d == OP_ANY && ptr + 1 >= md->end_subject &&
1139                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
1140                NLBLOCK->nltype == NLTYPE_FIXED &&
1141                NLBLOCK->nllen == 2 &&
1142                c == NLBLOCK->nl[0])
1143              {
1144              could_continue = partial_newline = TRUE;
1145              }
1146            else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1147              (c < 256 &&              (c < 256 &&
1148                (d != OP_ANY || !IS_NEWLINE(ptr)) &&                (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1149                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
# Line 1018  for (;;) Line 1166  for (;;)
1166        ADD_ACTIVE(state_offset + 2, 0);        ADD_ACTIVE(state_offset + 2, 0);
1167        if (clen > 0)        if (clen > 0)
1168          {          {
1169          if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||          if (d == OP_ANY && ptr + 1 >= md->end_subject &&
1170                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
1171                NLBLOCK->nltype == NLTYPE_FIXED &&
1172                NLBLOCK->nllen == 2 &&
1173                c == NLBLOCK->nl[0])
1174              {
1175              could_continue = partial_newline = TRUE;
1176              }
1177            else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1178              (c < 256 &&              (c < 256 &&
1179                (d != OP_ANY || !IS_NEWLINE(ptr)) &&                (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1180                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
# Line 1040  for (;;) Line 1196  for (;;)
1196        ADD_ACTIVE(state_offset + 2, 0);        ADD_ACTIVE(state_offset + 2, 0);
1197        if (clen > 0)        if (clen > 0)
1198          {          {
1199          if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||          if (d == OP_ANY && ptr + 1 >= md->end_subject &&
1200                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
1201                NLBLOCK->nltype == NLTYPE_FIXED &&
1202                NLBLOCK->nllen == 2 &&
1203                c == NLBLOCK->nl[0])
1204              {
1205              could_continue = partial_newline = TRUE;
1206              }
1207            else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1208              (c < 256 &&              (c < 256 &&
1209                (d != OP_ANY || !IS_NEWLINE(ptr)) &&                (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1210                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
# Line 1060  for (;;) Line 1224  for (;;)
1224        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1225        if (clen > 0)        if (clen > 0)
1226          {          {
1227          if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||          if (d == OP_ANY && ptr + 1 >= md->end_subject &&
1228                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
1229                NLBLOCK->nltype == NLTYPE_FIXED &&
1230                NLBLOCK->nllen == 2 &&
1231                c == NLBLOCK->nl[0])
1232              {
1233              could_continue = partial_newline = TRUE;
1234              }
1235            else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1236              (c < 256 &&              (c < 256 &&
1237                (d != OP_ANY || !IS_NEWLINE(ptr)) &&                (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1238                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
1239            {            {
1240            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1241              { ADD_NEW(state_offset + 4, 0); }              { ADD_NEW(state_offset + 1 + IMM2_SIZE + 1, 0); }
1242            else            else
1243              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
1244            }            }
# Line 1077  for (;;) Line 1249  for (;;)
1249        case OP_TYPEUPTO:        case OP_TYPEUPTO:
1250        case OP_TYPEMINUPTO:        case OP_TYPEMINUPTO:
1251        case OP_TYPEPOSUPTO:        case OP_TYPEPOSUPTO:
1252        ADD_ACTIVE(state_offset + 4, 0);        ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0);
1253        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1254        if (clen > 0)        if (clen > 0)
1255          {          {
1256          if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||          if (d == OP_ANY && ptr + 1 >= md->end_subject &&
1257                (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
1258                NLBLOCK->nltype == NLTYPE_FIXED &&
1259                NLBLOCK->nllen == 2 &&
1260                c == NLBLOCK->nl[0])
1261              {
1262              could_continue = partial_newline = TRUE;
1263              }
1264            else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
1265              (c < 256 &&              (c < 256 &&
1266                (d != OP_ANY || !IS_NEWLINE(ptr)) &&                (d != OP_ANY || !IS_NEWLINE(ptr)) &&
1267                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))                ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
# Line 1092  for (;;) Line 1272  for (;;)
1272              next_active_state--;              next_active_state--;
1273              }              }
1274            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1275              { ADD_NEW(state_offset + 4, 0); }              { ADD_NEW(state_offset + 2 + IMM2_SIZE, 0); }
1276            else            else
1277              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
1278            }            }
# Line 1122  for (;;) Line 1302  for (;;)
1302            break;            break;
1303    
1304            case PT_LAMP:            case PT_LAMP:
1305            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
1306                prop->chartype == ucp_Lt;
1307            break;            break;
1308    
1309            case PT_GC:            case PT_GC:
1310            OK = _pcre_ucp_gentype[prop->chartype] == code[3];            OK = PRIV(ucp_gentype)[prop->chartype] == code[3];
1311            break;            break;
1312    
1313            case PT_PC:            case PT_PC:
# Line 1137  for (;;) Line 1318  for (;;)
1318            OK = prop->script == code[3];            OK = prop->script == code[3];
1319            break;            break;
1320    
1321              /* These are specials for combination cases. */
1322    
1323              case PT_ALNUM:
1324              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1325                   PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1326              break;
1327    
1328              case PT_SPACE:    /* Perl space */
1329              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1330                   c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1331              break;
1332    
1333              case PT_PXSPACE:  /* POSIX space */
1334              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1335                   c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1336                   c == CHAR_FF || c == CHAR_CR;
1337              break;
1338    
1339              case PT_WORD:
1340              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1341                   PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1342                   c == CHAR_UNDERSCORE;
1343              break;
1344    
1345            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
1346    
1347            default:            default:
# Line 1165  for (;;) Line 1370  for (;;)
1370        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
1371        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1372          {          {
1373          const uschar *nptr = ptr + clen;          const pcre_uchar *nptr = ptr + clen;
1374          int ncount = 0;          int ncount = 0;
1375          if (count > 0 && codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS)          if (count > 0 && codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS)
1376            {            {
# Line 1344  for (;;) Line 1549  for (;;)
1549            break;            break;
1550    
1551            case PT_LAMP:            case PT_LAMP:
1552            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
1553                prop->chartype == ucp_Lt;
1554            break;            break;
1555    
1556            case PT_GC:            case PT_GC:
1557            OK = _pcre_ucp_gentype[prop->chartype] == code[3];            OK = PRIV(ucp_gentype)[prop->chartype] == code[3];
1558            break;            break;
1559    
1560            case PT_PC:            case PT_PC:
# Line 1359  for (;;) Line 1565  for (;;)
1565            OK = prop->script == code[3];            OK = prop->script == code[3];
1566            break;            break;
1567    
1568              /* These are specials for combination cases. */
1569    
1570              case PT_ALNUM:
1571              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1572                   PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1573              break;
1574    
1575              case PT_SPACE:    /* Perl space */
1576              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1577                   c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1578              break;
1579    
1580              case PT_PXSPACE:  /* POSIX space */
1581              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1582                   c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1583                   c == CHAR_FF || c == CHAR_CR;
1584              break;
1585    
1586              case PT_WORD:
1587              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1588                   PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1589                   c == CHAR_UNDERSCORE;
1590              break;
1591    
1592            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
1593    
1594            default:            default:
# Line 1396  for (;;) Line 1626  for (;;)
1626        ADD_ACTIVE(state_offset + 2, 0);        ADD_ACTIVE(state_offset + 2, 0);
1627        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1628          {          {
1629          const uschar *nptr = ptr + clen;          const pcre_uchar *nptr = ptr + clen;
1630          int ncount = 0;          int ncount = 0;
1631          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSSTAR ||          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSSTAR ||
1632              codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSQUERY)              codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSQUERY)
# Line 1578  for (;;) Line 1808  for (;;)
1808        case OP_PROP_EXTRA + OP_TYPEMINUPTO:        case OP_PROP_EXTRA + OP_TYPEMINUPTO:
1809        case OP_PROP_EXTRA + OP_TYPEPOSUPTO:        case OP_PROP_EXTRA + OP_TYPEPOSUPTO:
1810        if (codevalue != OP_PROP_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_PROP_EXTRA + OP_TYPEEXACT)
1811          { ADD_ACTIVE(state_offset + 6, 0); }          { ADD_ACTIVE(state_offset + 1 + IMM2_SIZE + 3, 0); }
1812        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1813        if (clen > 0)        if (clen > 0)
1814          {          {
1815          BOOL OK;          BOOL OK;
1816          const ucd_record * prop = GET_UCD(c);          const ucd_record * prop = GET_UCD(c);
1817          switch(code[4])          switch(code[1 + IMM2_SIZE + 1])
1818            {            {
1819            case PT_ANY:            case PT_ANY:
1820            OK = TRUE;            OK = TRUE;
1821            break;            break;
1822    
1823            case PT_LAMP:            case PT_LAMP:
1824            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
1825                prop->chartype == ucp_Lt;
1826            break;            break;
1827    
1828            case PT_GC:            case PT_GC:
1829            OK = _pcre_ucp_gentype[prop->chartype] == code[5];            OK = PRIV(ucp_gentype)[prop->chartype] == code[1 + IMM2_SIZE + 2];
1830            break;            break;
1831    
1832            case PT_PC:            case PT_PC:
1833            OK = prop->chartype == code[5];            OK = prop->chartype == code[1 + IMM2_SIZE + 2];
1834            break;            break;
1835    
1836            case PT_SC:            case PT_SC:
1837            OK = prop->script == code[5];            OK = prop->script == code[1 + IMM2_SIZE + 2];
1838              break;
1839    
1840              /* These are specials for combination cases. */
1841    
1842              case PT_ALNUM:
1843              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1844                   PRIV(ucp_gentype)[prop->chartype] == ucp_N;
1845              break;
1846    
1847              case PT_SPACE:    /* Perl space */
1848              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1849                   c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
1850              break;
1851    
1852              case PT_PXSPACE:  /* POSIX space */
1853              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
1854                   c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
1855                   c == CHAR_FF || c == CHAR_CR;
1856              break;
1857    
1858              case PT_WORD:
1859              OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
1860                   PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
1861                   c == CHAR_UNDERSCORE;
1862            break;            break;
1863    
1864            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
# Line 1621  for (;;) Line 1876  for (;;)
1876              next_active_state--;              next_active_state--;
1877              }              }
1878            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1879              { ADD_NEW(state_offset + 6, 0); }              { ADD_NEW(state_offset + 1 + IMM2_SIZE + 3, 0); }
1880            else            else
1881              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
1882            }            }
# Line 1634  for (;;) Line 1889  for (;;)
1889        case OP_EXTUNI_EXTRA + OP_TYPEMINUPTO:        case OP_EXTUNI_EXTRA + OP_TYPEMINUPTO:
1890        case OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO:        case OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO:
1891        if (codevalue != OP_EXTUNI_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_EXTUNI_EXTRA + OP_TYPEEXACT)
1892          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
1893        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1894        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1895          {          {
1896          const uschar *nptr = ptr + clen;          const pcre_uchar *nptr = ptr + clen;
1897          int ncount = 0;          int ncount = 0;
1898          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO)          if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO)
1899            {            {
# Line 1654  for (;;) Line 1909  for (;;)
1909            ncount++;            ncount++;
1910            nptr += ndlen;            nptr += ndlen;
1911            }            }
1912            if (nptr >= end_subject && (md->moptions & PCRE_PARTIAL_HARD) != 0)
1913                reset_could_continue = TRUE;
1914          if (++count >= GET2(code, 1))          if (++count >= GET2(code, 1))
1915            { ADD_NEW_DATA(-(state_offset + 4), 0, ncount); }            { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, ncount); }
1916          else          else
1917            { ADD_NEW_DATA(-state_offset, count, ncount); }            { ADD_NEW_DATA(-state_offset, count, ncount); }
1918          }          }
# Line 1668  for (;;) Line 1925  for (;;)
1925        case OP_ANYNL_EXTRA + OP_TYPEMINUPTO:        case OP_ANYNL_EXTRA + OP_TYPEMINUPTO:
1926        case OP_ANYNL_EXTRA + OP_TYPEPOSUPTO:        case OP_ANYNL_EXTRA + OP_TYPEPOSUPTO:
1927        if (codevalue != OP_ANYNL_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_ANYNL_EXTRA + OP_TYPEEXACT)
1928          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
1929        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1930        if (clen > 0)        if (clen > 0)
1931          {          {
# Line 1695  for (;;) Line 1952  for (;;)
1952              next_active_state--;              next_active_state--;
1953              }              }
1954            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
1955              { ADD_NEW_DATA(-(state_offset + 4), 0, ncount); }              { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, ncount); }
1956            else            else
1957              { ADD_NEW_DATA(-state_offset, count, ncount); }              { ADD_NEW_DATA(-state_offset, count, ncount); }
1958            break;            break;
# Line 1712  for (;;) Line 1969  for (;;)
1969        case OP_VSPACE_EXTRA + OP_TYPEMINUPTO:        case OP_VSPACE_EXTRA + OP_TYPEMINUPTO:
1970        case OP_VSPACE_EXTRA + OP_TYPEPOSUPTO:        case OP_VSPACE_EXTRA + OP_TYPEPOSUPTO:
1971        if (codevalue != OP_VSPACE_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_VSPACE_EXTRA + OP_TYPEEXACT)
1972          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
1973        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1974        if (clen > 0)        if (clen > 0)
1975          {          {
# Line 1741  for (;;) Line 1998  for (;;)
1998              next_active_state--;              next_active_state--;
1999              }              }
2000            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
2001              { ADD_NEW_DATA(-(state_offset + 4), 0, 0); }              { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, 0); }
2002            else            else
2003              { ADD_NEW_DATA(-state_offset, count, 0); }              { ADD_NEW_DATA(-state_offset, count, 0); }
2004            }            }
# Line 1754  for (;;) Line 2011  for (;;)
2011        case OP_HSPACE_EXTRA + OP_TYPEMINUPTO:        case OP_HSPACE_EXTRA + OP_TYPEMINUPTO:
2012        case OP_HSPACE_EXTRA + OP_TYPEPOSUPTO:        case OP_HSPACE_EXTRA + OP_TYPEPOSUPTO:
2013        if (codevalue != OP_HSPACE_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_HSPACE_EXTRA + OP_TYPEEXACT)
2014          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
2015        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
2016        if (clen > 0)        if (clen > 0)
2017          {          {
# Line 1796  for (;;) Line 2053  for (;;)
2053              next_active_state--;              next_active_state--;
2054              }              }
2055            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
2056              { ADD_NEW_DATA(-(state_offset + 4), 0, 0); }              { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, 0); }
2057            else            else
2058              { ADD_NEW_DATA(-state_offset, count, 0); }              { ADD_NEW_DATA(-state_offset, count, 0); }
2059            }            }
# Line 1815  for (;;) Line 2072  for (;;)
2072        break;        break;
2073    
2074        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2075        case OP_CHARNC:        case OP_CHARI:
2076        if (clen == 0) break;        if (clen == 0) break;
2077    
2078  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2079        if (utf8)        if (utf)
2080          {          {
2081          if (c == d) { ADD_NEW(state_offset + dlen + 1, 0); } else          if (c == d) { ADD_NEW(state_offset + dlen + 1, 0); } else
2082            {            {
2083            unsigned int othercase;            unsigned int othercase;
2084            if (c < 128) othercase = fcc[c]; else            if (c < 128)
2085                othercase = fcc[c];
2086            /* If we have Unicode property support, we can use it to test the            else
2087            other case of the character. */              /* If we have Unicode property support, we can use it to test the
2088                other case of the character. */
2089  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2090            othercase = UCD_OTHERCASE(c);              othercase = UCD_OTHERCASE(c);
2091  #else  #else
2092            othercase = NOTACHAR;              othercase = NOTACHAR;
2093  #endif  #endif
2094    
2095            if (d == othercase) { ADD_NEW(state_offset + dlen + 1, 0); }            if (d == othercase) { ADD_NEW(state_offset + dlen + 1, 0); }
2096            }            }
2097          }          }
2098        else        else
2099  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2100          /* Not UTF mode */
       /* Non-UTF-8 mode */  
2101          {          {
2102          if (lcc[c] == lcc[d]) { ADD_NEW(state_offset + 2, 0); }          if (TABLE_GET(c, lcc, c) == TABLE_GET(d, lcc, d))
2103              { ADD_NEW(state_offset + 2, 0); }
2104          }          }
2105        break;        break;
2106    
# Line 1857  for (;;) Line 2114  for (;;)
2114        case OP_EXTUNI:        case OP_EXTUNI:
2115        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
2116          {          {
2117          const uschar *nptr = ptr + clen;          const pcre_uchar *nptr = ptr + clen;
2118          int ncount = 0;          int ncount = 0;
2119          while (nptr < end_subject)          while (nptr < end_subject)
2120            {            {
# Line 1867  for (;;) Line 2124  for (;;)
2124            ncount++;            ncount++;
2125            nptr += nclen;            nptr += nclen;
2126            }            }
2127            if (nptr >= end_subject && (md->moptions & PCRE_PARTIAL_HARD) != 0)
2128                reset_could_continue = TRUE;
2129          ADD_NEW_DATA(-(state_offset + 1), 0, ncount);          ADD_NEW_DATA(-(state_offset + 1), 0, ncount);
2130          }          }
2131        break;        break;
# Line 1892  for (;;) Line 2151  for (;;)
2151          break;          break;
2152    
2153          case 0x000d:          case 0x000d:
2154          if (ptr + 1 < end_subject && ptr[1] == 0x0a)          if (ptr + 1 >= end_subject)
2155              {
2156              ADD_NEW(state_offset + 1, 0);
2157              if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
2158                reset_could_continue = TRUE;
2159              }
2160            else if (ptr[1] == 0x0a)
2161            {            {
2162            ADD_NEW_DATA(-(state_offset + 1), 0, 1);            ADD_NEW_DATA(-(state_offset + 1), 0, 1);
2163            }            }
# Line 2001  for (;;) Line 2266  for (;;)
2266        break;        break;
2267    
2268        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2269        /* 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. */  
2270    
2271        case OP_NOT:        case OP_NOT:
2272          if (clen > 0 && c != d) { ADD_NEW(state_offset + dlen + 1, 0); }
2273          break;
2274    
2275          /*-----------------------------------------------------------------*/
2276          /* Match a negated single character caselessly. */
2277    
2278          case OP_NOTI:
2279        if (clen > 0)        if (clen > 0)
2280          {          {
2281          unsigned int otherd = ((ims & PCRE_CASELESS) != 0)? fcc[d] : d;          unsigned int otherd;
2282          if (c != d && c != otherd) { ADD_NEW(state_offset + dlen + 1, 0); }  #ifdef SUPPORT_UTF
2283            if (utf && d >= 128)
2284              {
2285    #ifdef SUPPORT_UCP
2286              otherd = UCD_OTHERCASE(d);
2287    #endif  /* SUPPORT_UCP */
2288              }
2289            else
2290    #endif  /* SUPPORT_UTF */
2291            otherd = TABLE_GET(d, fcc, d);
2292            if (c != d && c != otherd)
2293              { ADD_NEW(state_offset + dlen + 1, 0); }
2294          }          }
2295        break;        break;
2296    
2297        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2298          case OP_PLUSI:
2299          case OP_MINPLUSI:
2300          case OP_POSPLUSI:
2301          case OP_NOTPLUSI:
2302          case OP_NOTMINPLUSI:
2303          case OP_NOTPOSPLUSI:
2304          caseless = TRUE;
2305          codevalue -= OP_STARI - OP_STAR;
2306    
2307          /* Fall through */
2308        case OP_PLUS:        case OP_PLUS:
2309        case OP_MINPLUS:        case OP_MINPLUS:
2310        case OP_POSPLUS:        case OP_POSPLUS:
# Line 2025  for (;;) Line 2316  for (;;)
2316        if (clen > 0)        if (clen > 0)
2317          {          {
2318          unsigned int otherd = NOTACHAR;          unsigned int otherd = NOTACHAR;
2319          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2320            {            {
2321  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2322            if (utf8 && d >= 128)            if (utf && d >= 128)
2323              {              {
2324  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2325              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2326  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2327              }              }
2328            else            else
2329  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2330            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2331            }            }
2332          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2333            {            {
# Line 2053  for (;;) Line 2344  for (;;)
2344        break;        break;
2345    
2346        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2347          case OP_QUERYI:
2348          case OP_MINQUERYI:
2349          case OP_POSQUERYI:
2350          case OP_NOTQUERYI:
2351          case OP_NOTMINQUERYI:
2352          case OP_NOTPOSQUERYI:
2353          caseless = TRUE;
2354          codevalue -= OP_STARI - OP_STAR;
2355          /* Fall through */
2356        case OP_QUERY:        case OP_QUERY:
2357        case OP_MINQUERY:        case OP_MINQUERY:
2358        case OP_POSQUERY:        case OP_POSQUERY:
# Line 2063  for (;;) Line 2363  for (;;)
2363        if (clen > 0)        if (clen > 0)
2364          {          {
2365          unsigned int otherd = NOTACHAR;          unsigned int otherd = NOTACHAR;
2366          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2367            {            {
2368  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2369            if (utf8 && d >= 128)            if (utf && d >= 128)
2370              {              {
2371  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2372              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2373  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2374              }              }
2375            else            else
2376  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2377            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2378            }            }
2379          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2380            {            {
# Line 2089  for (;;) Line 2389  for (;;)
2389        break;        break;
2390    
2391        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2392          case OP_STARI:
2393          case OP_MINSTARI:
2394          case OP_POSSTARI:
2395          case OP_NOTSTARI:
2396          case OP_NOTMINSTARI:
2397          case OP_NOTPOSSTARI:
2398          caseless = TRUE;
2399          codevalue -= OP_STARI - OP_STAR;
2400          /* Fall through */
2401        case OP_STAR:        case OP_STAR:
2402        case OP_MINSTAR:        case OP_MINSTAR:
2403        case OP_POSSTAR:        case OP_POSSTAR:
# Line 2099  for (;;) Line 2408  for (;;)
2408        if (clen > 0)        if (clen > 0)
2409          {          {
2410          unsigned int otherd = NOTACHAR;          unsigned int otherd = NOTACHAR;
2411          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2412            {            {
2413  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2414            if (utf8 && d >= 128)            if (utf && d >= 128)
2415              {              {
2416  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2417              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2418  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2419              }              }
2420            else            else
2421  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2422            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2423            }            }
2424          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2425            {            {
# Line 2125  for (;;) Line 2434  for (;;)
2434        break;        break;
2435    
2436        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2437          case OP_EXACTI:
2438          case OP_NOTEXACTI:
2439          caseless = TRUE;
2440          codevalue -= OP_STARI - OP_STAR;
2441          /* Fall through */
2442        case OP_EXACT:        case OP_EXACT:
2443        case OP_NOTEXACT:        case OP_NOTEXACT:
2444        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
2445        if (clen > 0)        if (clen > 0)
2446          {          {
2447          unsigned int otherd = NOTACHAR;          unsigned int otherd = NOTACHAR;
2448          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2449            {            {
2450  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2451            if (utf8 && d >= 128)            if (utf && d >= 128)
2452              {              {
2453  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2454              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2455  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2456              }              }
2457            else            else
2458  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2459            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2460            }            }
2461          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2462            {            {
2463            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
2464              { ADD_NEW(state_offset + dlen + 3, 0); }              { ADD_NEW(state_offset + dlen + 1 + IMM2_SIZE, 0); }
2465            else            else
2466              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
2467            }            }
# Line 2155  for (;;) Line 2469  for (;;)
2469        break;        break;
2470    
2471        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2472          case OP_UPTOI:
2473          case OP_MINUPTOI:
2474          case OP_POSUPTOI:
2475          case OP_NOTUPTOI:
2476          case OP_NOTMINUPTOI:
2477          case OP_NOTPOSUPTOI:
2478          caseless = TRUE;
2479          codevalue -= OP_STARI - OP_STAR;
2480          /* Fall through */
2481        case OP_UPTO:        case OP_UPTO:
2482        case OP_MINUPTO:        case OP_MINUPTO:
2483        case OP_POSUPTO:        case OP_POSUPTO:
2484        case OP_NOTUPTO:        case OP_NOTUPTO:
2485        case OP_NOTMINUPTO:        case OP_NOTMINUPTO:
2486        case OP_NOTPOSUPTO:        case OP_NOTPOSUPTO:
2487        ADD_ACTIVE(state_offset + dlen + 3, 0);        ADD_ACTIVE(state_offset + dlen + 1 + IMM2_SIZE, 0);
2488        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
2489        if (clen > 0)        if (clen > 0)
2490          {          {
2491          unsigned int otherd = NOTACHAR;          unsigned int otherd = NOTACHAR;
2492          if ((ims & PCRE_CASELESS) != 0)          if (caseless)
2493            {            {
2494  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2495            if (utf8 && d >= 128)            if (utf && d >= 128)
2496              {              {
2497  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2498              otherd = UCD_OTHERCASE(d);              otherd = UCD_OTHERCASE(d);
2499  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2500              }              }
2501            else            else
2502  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
2503            otherd = fcc[d];            otherd = TABLE_GET(d, fcc, d);
2504            }            }
2505          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))          if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
2506            {            {
# Line 2187  for (;;) Line 2510  for (;;)
2510              next_active_state--;              next_active_state--;
2511              }              }
2512            if (++count >= GET2(code, 1))            if (++count >= GET2(code, 1))
2513              { ADD_NEW(state_offset + dlen + 3, 0); }              { ADD_NEW(state_offset + dlen + 1 + IMM2_SIZE, 0); }
2514            else            else
2515              { ADD_NEW(state_offset, count); }              { ADD_NEW(state_offset, count); }
2516            }            }
# Line 2204  for (;;) Line 2527  for (;;)
2527          {          {
2528          BOOL isinclass = FALSE;          BOOL isinclass = FALSE;
2529          int next_state_offset;          int next_state_offset;
2530          const uschar *ecode;          const pcre_uchar *ecode;
2531    
2532          /* 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
2533          can set isinclass from it. */          can set isinclass from it. */
2534    
2535          if (codevalue != OP_XCLASS)          if (codevalue != OP_XCLASS)
2536            {            {
2537            ecode = code + 33;            ecode = code + 1 + (32 / sizeof(pcre_uchar));
2538            if (clen > 0)            if (clen > 0)
2539              {              {
2540              isinclass = (c > 255)? (codevalue == OP_NCLASS) :              isinclass = (c > 255)? (codevalue == OP_NCLASS) :
2541                ((code[1 + c/8] & (1 << (c&7))) != 0);                ((((pcre_uint8 *)(code + 1))[c/8] & (1 << (c&7))) != 0);
2542              }              }
2543            }            }
2544    
# Line 2226  for (;;) Line 2549  for (;;)
2549          else          else
2550           {           {
2551           ecode = code + GET(code, 1);           ecode = code + GET(code, 1);
2552           if (clen > 0) isinclass = _pcre_xclass(c, code + 1 + LINK_SIZE);           if (clen > 0) isinclass = PRIV(xclass)(c, code + 1 + LINK_SIZE, utf);
2553           }           }
2554    
2555          /* 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
2556          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
2557          quantifier, this is where it will be. */          quantifier, this is where it will be. */
2558    
2559          next_state_offset = ecode - start_code;          next_state_offset = (int)(ecode - start_code);
2560    
2561          switch (*ecode)          switch (*ecode)
2562            {            {
# Line 2260  for (;;) Line 2583  for (;;)
2583            case OP_CRMINRANGE:            case OP_CRMINRANGE:
2584            count = current_state->count;  /* Already matched */            count = current_state->count;  /* Already matched */
2585            if (count >= GET2(ecode, 1))            if (count >= GET2(ecode, 1))
2586              { ADD_ACTIVE(next_state_offset + 5, 0); }              { ADD_ACTIVE(next_state_offset + 1 + 2 * IMM2_SIZE, 0); }
2587            if (isinclass)            if (isinclass)
2588              {              {
2589              int max = GET2(ecode, 3);              int max = GET2(ecode, 1 + IMM2_SIZE);
2590              if (++count >= max && max != 0)   /* Max 0 => no limit */              if (++count >= max && max != 0)   /* Max 0 => no limit */
2591                { ADD_NEW(next_state_offset + 5, 0); }                { ADD_NEW(next_state_offset + 1 + 2 * IMM2_SIZE, 0); }
2592              else              else
2593                { ADD_NEW(state_offset, count); }                { ADD_NEW(state_offset, count); }
2594              }              }
# Line 2296  for (;;) Line 2619  for (;;)
2619          int rc;          int rc;
2620          int local_offsets[2];          int local_offsets[2];
2621          int local_workspace[1000];          int local_workspace[1000];
2622          const uschar *endasscode = code + GET(code, 1);          const pcre_uchar *endasscode = code + GET(code, 1);
2623    
2624          while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);          while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);
2625    
# Line 2304  for (;;) Line 2627  for (;;)
2627            md,                                   /* static match data */            md,                                   /* static match data */
2628            code,                                 /* this subexpression's code */            code,                                 /* this subexpression's code */
2629            ptr,                                  /* where we currently are */            ptr,                                  /* where we currently are */
2630            ptr - start_subject,                  /* start offset */            (int)(ptr - start_subject),           /* start offset */
2631            local_offsets,                        /* offset vector */            local_offsets,                        /* offset vector */
2632            sizeof(local_offsets)/sizeof(int),    /* size of same */            sizeof(local_offsets)/sizeof(int),    /* size of same */
2633            local_workspace,                      /* workspace vector */            local_workspace,                      /* workspace vector */
2634            sizeof(local_workspace)/sizeof(int),  /* size of same */            sizeof(local_workspace)/sizeof(int),  /* size of same */
2635            ims,                                  /* the current ims flags */            rlevel);                              /* function recursion level */
           rlevel,                               /* function recursion level */  
           recursing);                           /* pass on regex recursion */  
2636    
2637          if (rc == PCRE_ERROR_DFA_UITEM) return rc;          if (rc == PCRE_ERROR_DFA_UITEM) return rc;
2638          if ((rc >= 0) == (codevalue == OP_ASSERT || codevalue == OP_ASSERTBACK))          if ((rc >= 0) == (codevalue == OP_ASSERT || codevalue == OP_ASSERTBACK))
2639              { ADD_ACTIVE(endasscode + LINK_SIZE + 1 - start_code, 0); }              { ADD_ACTIVE((int)(endasscode + LINK_SIZE + 1 - start_code), 0); }
2640          }          }
2641        break;        break;
2642    
# Line 2335  for (;;) Line 2656  for (;;)
2656          if (code[LINK_SIZE+1] == OP_CALLOUT)          if (code[LINK_SIZE+1] == OP_CALLOUT)
2657            {            {
2658            rrc = 0;            rrc = 0;
2659            if (pcre_callout != NULL)            if (PUBL(callout) != NULL)
2660              {              {
2661              pcre_callout_block cb;              PUBL(callout_block) cb;
2662              cb.version          = 1;   /* Version 1 of the callout block */              cb.version          = 1;   /* Version 1 of the callout block */
2663              cb.callout_number   = code[LINK_SIZE+2];              cb.callout_number   = code[LINK_SIZE+2];
2664              cb.offset_vector    = offsets;              cb.offset_vector    = offsets;
2665    #ifdef COMPILE_PCRE8
2666              cb.subject          = (PCRE_SPTR)start_subject;              cb.subject          = (PCRE_SPTR)start_subject;
2667              cb.subject_length   = end_subject - start_subject;  #else
2668              cb.start_match      = current_subject - start_subject;              cb.subject          = (PCRE_SPTR16)start_subject;
2669              cb.current_position = ptr - start_subject;  #endif
2670                cb.subject_length   = (int)(end_subject - start_subject);
2671                cb.start_match      = (int)(current_subject - start_subject);
2672                cb.current_position = (int)(ptr - start_subject);
2673              cb.pattern_position = GET(code, LINK_SIZE + 3);              cb.pattern_position = GET(code, LINK_SIZE + 3);
2674              cb.next_item_length = GET(code, 3 + 2*LINK_SIZE);              cb.next_item_length = GET(code, 3 + 2*LINK_SIZE);
2675              cb.capture_top      = 1;              cb.capture_top      = 1;
2676              cb.capture_last     = -1;              cb.capture_last     = -1;
2677              cb.callout_data     = md->callout_data;              cb.callout_data     = md->callout_data;
2678              if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc;   /* Abandon */              cb.mark             = NULL;   /* No (*MARK) support */
2679                if ((rrc = (*PUBL(callout))(&cb)) < 0) return rrc;   /* Abandon */
2680              }              }
2681            if (rrc > 0) break;                      /* Fail this thread */            if (rrc > 0) break;                      /* Fail this thread */
2682            code += _pcre_OP_lengths[OP_CALLOUT];    /* Skip callout data */            code += PRIV(OP_lengths)[OP_CALLOUT];    /* Skip callout data */
2683            }            }
2684    
2685          condcode = code[LINK_SIZE+1];          condcode = code[LINK_SIZE+1];
# Line 2374  for (;;) Line 2700  for (;;)
2700    
2701          else if (condcode == OP_RREF || condcode == OP_NRREF)          else if (condcode == OP_RREF || condcode == OP_NRREF)
2702            {            {
2703            int value = GET2(code, LINK_SIZE+2);            int value = GET2(code, LINK_SIZE + 2);
2704            if (value != RREF_ANY) return PCRE_ERROR_DFA_UCOND;            if (value != RREF_ANY) return PCRE_ERROR_DFA_UCOND;
2705            if (recursing > 0)            if (md->recursive != NULL)
2706              { ADD_ACTIVE(state_offset + LINK_SIZE + 4, 0); }              { ADD_ACTIVE(state_offset + LINK_SIZE + 2 + IMM2_SIZE, 0); }
2707            else { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }            else { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
2708            }            }
2709    
# Line 2386  for (;;) Line 2712  for (;;)
2712          else          else
2713            {            {
2714            int rc;            int rc;
2715            const uschar *asscode = code + LINK_SIZE + 1;            const pcre_uchar *asscode = code + LINK_SIZE + 1;
2716            const uschar *endasscode = asscode + GET(asscode, 1);            const pcre_uchar *endasscode = asscode + GET(asscode, 1);
2717    
2718            while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);            while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);
2719    
# Line 2395  for (;;) Line 2721  for (;;)
2721              md,                                   /* fixed match data */              md,                                   /* fixed match data */
2722              asscode,                              /* this subexpression's code */              asscode,                              /* this subexpression's code */
2723              ptr,                                  /* where we currently are */              ptr,                                  /* where we currently are */
2724              ptr - start_subject,                  /* start offset */              (int)(ptr - start_subject),           /* start offset */
2725              local_offsets,                        /* offset vector */              local_offsets,                        /* offset vector */
2726              sizeof(local_offsets)/sizeof(int),    /* size of same */              sizeof(local_offsets)/sizeof(int),    /* size of same */
2727              local_workspace,                      /* workspace vector */              local_workspace,                      /* workspace vector */
2728              sizeof(local_workspace)/sizeof(int),  /* size of same */              sizeof(local_workspace)/sizeof(int),  /* size of same */
2729              ims,                                  /* the current ims flags */              rlevel);                              /* function recursion level */
             rlevel,                               /* function recursion level */  
             recursing);                           /* pass on regex recursion */  
2730    
2731            if (rc == PCRE_ERROR_DFA_UITEM) return rc;            if (rc == PCRE_ERROR_DFA_UITEM) return rc;
2732            if ((rc >= 0) ==            if ((rc >= 0) ==
2733                  (condcode == OP_ASSERT || condcode == OP_ASSERTBACK))                  (condcode == OP_ASSERT || condcode == OP_ASSERTBACK))
2734              { ADD_ACTIVE(endasscode + LINK_SIZE + 1 - start_code, 0); }              { ADD_ACTIVE((int)(endasscode + LINK_SIZE + 1 - start_code), 0); }
2735            else            else
2736              { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }              { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
2737            }            }
# Line 2417  for (;;) Line 2741  for (;;)
2741        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2742        case OP_RECURSE:        case OP_RECURSE:
2743          {          {
2744            dfa_recursion_info *ri;
2745          int local_offsets[1000];          int local_offsets[1000];
2746          int local_workspace[1000];          int local_workspace[1000];
2747            const pcre_uchar *callpat = start_code + GET(code, 1);
2748            int recno = (callpat == md->start_code)? 0 :
2749              GET2(callpat, 1 + LINK_SIZE);
2750          int rc;          int rc;
2751    
2752          DPRINTF(("%.*sStarting regex recursion %d\n", rlevel*2-2, SP,          DPRINTF(("%.*sStarting regex recursion\n", rlevel*2-2, SP));
2753            recursing + 1));  
2754            /* Check for repeating a recursion without advancing the subject
2755            pointer. This should catch convoluted mutual recursions. (Some simple
2756            cases are caught at compile time.) */
2757    
2758            for (ri = md->recursive; ri != NULL; ri = ri->prevrec)
2759              if (recno == ri->group_num && ptr == ri->subject_position)
2760                return PCRE_ERROR_RECURSELOOP;
2761    
2762            /* Remember this recursion and where we started it so as to
2763            catch infinite loops. */
2764    
2765            new_recursive.group_num = recno;
2766            new_recursive.subject_position = ptr;
2767            new_recursive.prevrec = md->recursive;
2768            md->recursive = &new_recursive;
2769    
2770          rc = internal_dfa_exec(          rc = internal_dfa_exec(
2771            md,                                   /* fixed match data */            md,                                   /* fixed match data */
2772            start_code + GET(code, 1),            /* this subexpression's code */            callpat,                              /* this subexpression's code */
2773            ptr,                                  /* where we currently are */            ptr,                                  /* where we currently are */
2774            ptr - start_subject,                  /* start offset */            (int)(ptr - start_subject),           /* start offset */
2775            local_offsets,                        /* offset vector */            local_offsets,                        /* offset vector */
2776            sizeof(local_offsets)/sizeof(int),    /* size of same */            sizeof(local_offsets)/sizeof(int),    /* size of same */
2777            local_workspace,                      /* workspace vector */            local_workspace,                      /* workspace vector */
2778            sizeof(local_workspace)/sizeof(int),  /* size of same */            sizeof(local_workspace)/sizeof(int),  /* size of same */
2779            ims,                                  /* the current ims flags */            rlevel);                              /* function recursion level */
2780            rlevel,                               /* function recursion level */  
2781            recursing + 1);                       /* regex recurse level */          md->recursive = new_recursive.prevrec;  /* Done this recursion */
2782    
2783          DPRINTF(("%.*sReturn from regex recursion %d: rc=%d\n", rlevel*2-2, SP,          DPRINTF(("%.*sReturn from regex recursion: rc=%d\n", rlevel*2-2, SP,
2784            recursing + 1, rc));            rc));
2785    
2786          /* Ran out of internal offsets */          /* Ran out of internal offsets */
2787    
# Line 2452  for (;;) Line 2795  for (;;)
2795            {            {
2796            for (rc = rc*2 - 2; rc >= 0; rc -= 2)            for (rc = rc*2 - 2; rc >= 0; rc -= 2)
2797              {              {
             const uschar *p = start_subject + local_offsets[rc];  
             const uschar *pp = start_subject + local_offsets[rc+1];  
2798              int charcount = local_offsets[rc+1] - local_offsets[rc];              int charcount = local_offsets[rc+1] - local_offsets[rc];
2799              while (p < pp) if ((*p++ & 0xc0) == 0x80) charcount--;  #ifdef SUPPORT_UTF
2800                if (utf)
2801                  {
2802                  const pcre_uchar *p = start_subject + local_offsets[rc];
2803                  const pcre_uchar *pp = start_subject + local_offsets[rc+1];
2804                  while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
2805                  }
2806    #endif
2807              if (charcount > 0)              if (charcount > 0)
2808                {                {
2809                ADD_NEW_DATA(-(state_offset + LINK_SIZE + 1), 0, (charcount - 1));                ADD_NEW_DATA(-(state_offset + LINK_SIZE + 1), 0, (charcount - 1));
# Line 2471  for (;;) Line 2819  for (;;)
2819        break;        break;
2820    
2821        /*-----------------------------------------------------------------*/        /*-----------------------------------------------------------------*/
2822          case OP_BRAPOS:
2823          case OP_SBRAPOS:
2824          case OP_CBRAPOS:
2825          case OP_SCBRAPOS:
2826          case OP_BRAPOSZERO:
2827            {
2828            int charcount, matched_count;
2829            const pcre_uchar *local_ptr = ptr;
2830            BOOL allow_zero;
2831    
2832            if (codevalue == OP_BRAPOSZERO)
2833              {
2834              allow_zero = TRUE;
2835              codevalue = *(++code);  /* Codevalue will be one of above BRAs */
2836              }
2837            else allow_zero = FALSE;
2838    
2839            /* Loop to match the subpattern as many times as possible as if it were
2840            a complete pattern. */
2841    
2842            for (matched_count = 0;; matched_count++)
2843              {
2844              int local_offsets[2];
2845              int local_workspace[1000];
2846    
2847              int rc = internal_dfa_exec(
2848                md,                                   /* fixed match data */
2849                code,                                 /* this subexpression's code */
2850                local_ptr,                            /* where we currently are */
2851                (int)(ptr - start_subject),           /* start offset */
2852                local_offsets,                        /* offset vector */
2853                sizeof(local_offsets)/sizeof(int),    /* size of same */
2854                local_workspace,                      /* workspace vector */
2855                sizeof(local_workspace)/sizeof(int),  /* size of same */
2856                rlevel);                              /* function recursion level */
2857    
2858              /* Failed to match */
2859    
2860              if (rc < 0)
2861                {
2862                if (rc != PCRE_ERROR_NOMATCH) return rc;
2863                break;
2864                }
2865    
2866              /* Matched: break the loop if zero characters matched. */
2867    
2868              charcount = local_offsets[1] - local_offsets[0];
2869              if (charcount == 0) break;
2870              local_ptr += charcount;    /* Advance temporary position ptr */
2871              }
2872    
2873            /* At this point we have matched the subpattern matched_count
2874            times, and local_ptr is pointing to the character after the end of the
2875            last match. */
2876    
2877            if (matched_count > 0 || allow_zero)
2878              {
2879              const pcre_uchar *end_subpattern = code;
2880              int next_state_offset;
2881    
2882              do { end_subpattern += GET(end_subpattern, 1); }
2883                while (*end_subpattern == OP_ALT);
2884              next_state_offset =
2885                (int)(end_subpattern - start_code + LINK_SIZE + 1);
2886    
2887              /* Optimization: if there are no more active states, and there
2888              are no new states yet set up, then skip over the subject string
2889              right here, to save looping. Otherwise, set up the new state to swing
2890              into action when the end of the matched substring is reached. */
2891    
2892              if (i + 1 >= active_count && new_count == 0)
2893                {
2894                ptr = local_ptr;
2895                clen = 0;
2896                ADD_NEW(next_state_offset, 0);
2897                }
2898              else
2899                {
2900                const pcre_uchar *p = ptr;
2901                const pcre_uchar *pp = local_ptr;
2902                charcount = (int)(pp - p);
2903    #ifdef SUPPORT_UTF
2904                if (utf) while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
2905    #endif
2906                ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));
2907                }
2908              }
2909            }
2910          break;
2911    
2912          /*-----------------------------------------------------------------*/
2913        case OP_ONCE:        case OP_ONCE:
2914          case OP_ONCE_NC:
2915          {          {
2916          int local_offsets[2];          int local_offsets[2];
2917          int local_workspace[1000];          int local_workspace[1000];
# Line 2480  for (;;) Line 2920  for (;;)
2920            md,                                   /* fixed match data */            md,                                   /* fixed match data */
2921            code,                                 /* this subexpression's code */            code,                                 /* this subexpression's code */
2922            ptr,                                  /* where we currently are */            ptr,                                  /* where we currently are */
2923            ptr - start_subject,                  /* start offset */            (int)(ptr - start_subject),           /* start offset */
2924            local_offsets,                        /* offset vector */            local_offsets,                        /* offset vector */
2925            sizeof(local_offsets)/sizeof(int),    /* size of same */            sizeof(local_offsets)/sizeof(int),    /* size of same */
2926            local_workspace,                      /* workspace vector */            local_workspace,                      /* workspace vector */
2927            sizeof(local_workspace)/sizeof(int),  /* size of same */            sizeof(local_workspace)/sizeof(int),  /* size of same */
2928            ims,                                  /* the current ims flags */            rlevel);                              /* function recursion level */
           rlevel,                               /* function recursion level */  
           recursing);                           /* pass on regex recursion */  
2929    
2930          if (rc >= 0)          if (rc >= 0)
2931            {            {
2932            const uschar *end_subpattern = code;            const pcre_uchar *end_subpattern = code;
2933            int charcount = local_offsets[1] - local_offsets[0];            int charcount = local_offsets[1] - local_offsets[0];
2934            int next_state_offset, repeat_state_offset;            int next_state_offset, repeat_state_offset;
2935    
2936            do { end_subpattern += GET(end_subpattern, 1); }            do { end_subpattern += GET(end_subpattern, 1); }
2937              while (*end_subpattern == OP_ALT);              while (*end_subpattern == OP_ALT);
2938            next_state_offset = end_subpattern - start_code + LINK_SIZE + 1;            next_state_offset =
2939                (int)(end_subpattern - start_code + LINK_SIZE + 1);
2940    
2941            /* If the end of this subpattern is KETRMAX or KETRMIN, we must            /* If the end of this subpattern is KETRMAX or KETRMIN, we must
2942            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 2505  for (;;) Line 2944  for (;;)
2944    
2945            repeat_state_offset = (*end_subpattern == OP_KETRMAX ||            repeat_state_offset = (*end_subpattern == OP_KETRMAX ||
2946                                   *end_subpattern == OP_KETRMIN)?                                   *end_subpattern == OP_KETRMIN)?
2947              end_subpattern - start_code - GET(end_subpattern, 1) : -1;              (int)(end_subpattern - start_code - GET(end_subpattern, 1)) : -1;
2948    
2949            /* 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
2950            current character pointer. This is important so that the duplicate            current character pointer. This is important so that the duplicate
# Line 2520  for (;;) Line 2959  for (;;)
2959            /* Optimization: if there are no more active states, and there            /* Optimization: if there are no more active states, and there
2960            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
2961            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
2962            into action when the end of the substring is reached. */            into action when the end of the matched substring is reached. */
2963    
2964            else if (i + 1 >= active_count && new_count == 0)            else if (i + 1 >= active_count && new_count == 0)
2965              {              {
# Line 2543  for (;;) Line 2982  for (;;)
2982              }              }
2983            else            else
2984              {              {
2985              const uschar *p = start_subject + local_offsets[0];  #ifdef SUPPORT_UTF
2986              const uschar *pp = start_subject + local_offsets[1];              if (utf)
2987              while (p < pp) if ((*p++ & 0xc0) == 0x80) charcount--;                {
2988                  const pcre_uchar *p = start_subject + local_offsets[0];
2989                  const pcre_uchar *pp = start_subject + local_offsets[1];
2990                  while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
2991                  }
2992    #endif
2993              ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));              ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));
2994              if (repeat_state_offset >= 0)              if (repeat_state_offset >= 0)
2995                { ADD_NEW_DATA(-repeat_state_offset, 0, (charcount - 1)); }                { ADD_NEW_DATA(-repeat_state_offset, 0, (charcount - 1)); }
2996              }              }
   
2997            }            }
2998          else if (rc != PCRE_ERROR_NOMATCH) return rc;          else if (rc != PCRE_ERROR_NOMATCH) return rc;
2999          }          }
# Line 2562  for (;;) Line 3005  for (;;)
3005    
3006        case OP_CALLOUT:        case OP_CALLOUT:
3007        rrc = 0;        rrc = 0;
3008        if (pcre_callout != NULL)        if (PUBL(callout) != NULL)
3009          {          {
3010          pcre_callout_block cb;          PUBL(callout_block) cb;
3011          cb.version          = 1;   /* Version 1 of the callout block */          cb.version          = 1;   /* Version 1 of the callout block */
3012          cb.callout_number   = code[1];          cb.callout_number   = code[1];
3013          cb.offset_vector    = offsets;          cb.offset_vector    = offsets;
3014    #ifdef COMPILE_PCRE8
3015          cb.subject          = (PCRE_SPTR)start_subject;          cb.subject          = (PCRE_SPTR)start_subject;
3016          cb.subject_length   = end_subject - start_subject;  #else
3017          cb.start_match      = current_subject - start_subject;          cb.subject          = (PCRE_SPTR16)start_subject;
3018          cb.current_position = ptr - start_subject;  #endif
3019            cb.subject_length   = (int)(end_subject - start_subject);
3020            cb.start_match      = (int)(current_subject - start_subject);
3021            cb.current_position = (int)(ptr - start_subject);
3022          cb.pattern_position = GET(code, 2);          cb.pattern_position = GET(code, 2);
3023          cb.next_item_length = GET(code, 2 + LINK_SIZE);          cb.next_item_length = GET(code, 2 + LINK_SIZE);
3024          cb.capture_top      = 1;          cb.capture_top      = 1;
3025          cb.capture_last     = -1;          cb.capture_last     = -1;
3026          cb.callout_data     = md->callout_data;          cb.callout_data     = md->callout_data;
3027          if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc;   /* Abandon */          cb.mark             = NULL;   /* No (*MARK) support */
3028            if ((rrc = (*PUBL(callout))(&cb)) < 0) return rrc;   /* Abandon */
3029          }          }
3030        if (rrc == 0)        if (rrc == 0)
3031          { ADD_ACTIVE(state_offset + _pcre_OP_lengths[OP_CALLOUT], 0); }          { ADD_ACTIVE(state_offset + PRIV(OP_lengths)[OP_CALLOUT], 0); }
3032        break;        break;
3033    
3034    
# Line 2609  for (;;) Line 3057  for (;;)
3057    if (new_count <= 0)    if (new_count <= 0)
3058      {      {
3059      if (rlevel == 1 &&                               /* Top level, and */      if (rlevel == 1 &&                               /* Top level, and */
3060          could_continue &&                            /* Some could go on */          could_continue &&                            /* Some could go on, and */
3061          forced_fail != workspace[1] &&               /* Not all forced fail & */          forced_fail != workspace[1] &&               /* Not all forced fail & */
3062          (                                            /* either... */          (                                            /* either... */
3063          (md->moptions & PCRE_PARTIAL_HARD) != 0      /* Hard partial */          (md->moptions & PCRE_PARTIAL_HARD) != 0      /* Hard partial */
# Line 2617  for (;;) Line 3065  for (;;)
3065          ((md->moptions & PCRE_PARTIAL_SOFT) != 0 &&  /* Soft partial and */          ((md->moptions & PCRE_PARTIAL_SOFT) != 0 &&  /* Soft partial and */
3066           match_count < 0)                            /* no matches */           match_count < 0)                            /* no matches */
3067          ) &&                                         /* And... */          ) &&                                         /* And... */
3068          ptr >= end_subject &&                     /* Reached end of subject */          (
3069          ptr > current_subject)                    /* Matched non-empty string */          partial_newline ||                           /* Either partial NL */
3070              (                                          /* or ... */
3071              ptr >= end_subject &&                /* End of subject and */
3072              ptr > md->start_used_ptr)            /* Inspected non-empty string */
3073              )
3074            )
3075        {        {
3076        if (offsetcount >= 2)        if (offsetcount >= 2)
3077          {          {
3078          offsets[0] = md->start_used_ptr - start_subject;          offsets[0] = (int)(md->start_used_ptr - start_subject);
3079          offsets[1] = end_subject - start_subject;          offsets[1] = (int)(end_subject - start_subject);
3080          }          }
3081        match_count = PCRE_ERROR_PARTIAL;        match_count = PCRE_ERROR_PARTIAL;
3082        }        }
# Line 2677  Returns:          > 0 => number of match Line 3130  Returns:          > 0 => number of match
3130                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
3131  */  */
3132    
3133    #ifdef COMPILE_PCRE8
3134  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
3135  pcre_dfa_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_dfa_exec(const pcre *argument_re, const pcre_extra *extra_data,
3136    const char *subject, int length, int start_offset, int options, int *offsets,    const char *subject, int length, int start_offset, int options, int *offsets,
3137    int offsetcount, int *workspace, int wscount)    int offsetcount, int *workspace, int wscount)
3138    #else
3139    PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
3140    pcre16_dfa_exec(const pcre16 *argument_re, const pcre16_extra *extra_data,
3141      PCRE_SPTR16 subject, int length, int start_offset, int options, int *offsets,
3142      int offsetcount, int *workspace, int wscount)
3143    #endif
3144  {  {
3145  real_pcre *re = (real_pcre *)argument_re;  REAL_PCRE *re = (REAL_PCRE *)argument_re;
3146  dfa_match_data match_block;  dfa_match_data match_block;
3147  dfa_match_data *md = &match_block;  dfa_match_data *md = &match_block;
3148  BOOL utf8, anchored, startline, firstline;  BOOL utf, anchored, startline, firstline;
3149  const uschar *current_subject, *end_subject, *lcc;  const pcre_uchar *current_subject, *end_subject;
   
 pcre_study_data internal_study;  
3150  const pcre_study_data *study = NULL;  const pcre_study_data *study = NULL;
 real_pcre internal_re;  
3151    
3152  const uschar *req_byte_ptr;  const pcre_uchar *req_char_ptr;
3153  const uschar *start_bits = NULL;  const pcre_uint8 *start_bits = NULL;
3154  BOOL first_byte_caseless = FALSE;  BOOL has_first_char = FALSE;
3155  BOOL req_byte_caseless = FALSE;  BOOL has_req_char = FALSE;
3156  int first_byte = -1;  pcre_uchar first_char = 0;
3157  int req_byte = -1;  pcre_uchar first_char2 = 0;
3158  int req_byte2 = -1;  pcre_uchar req_char = 0;
3159    pcre_uchar req_char2 = 0;
3160  int newline;  int newline;
3161    
3162  /* Plausibility checks */  /* Plausibility checks */
# Line 2708  if (re == NULL || subject == NULL || wor Line 3166  if (re == NULL || subject == NULL || wor
3166     (offsets == NULL && offsetcount > 0)) return PCRE_ERROR_NULL;     (offsets == NULL && offsetcount > 0)) return PCRE_ERROR_NULL;
3167  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;
3168  if (wscount < 20) return PCRE_ERROR_DFA_WSSIZE;  if (wscount < 20) return PCRE_ERROR_DFA_WSSIZE;
3169    if (start_offset < 0 || start_offset > length) return PCRE_ERROR_BADOFFSET;
3170    
3171  /* We need to find the pointer to any study data before we test for byte  /* Check that the first field in the block is the magic number. If it is not,
3172  flipping, so we scan the extra_data block first. This may set two fields in the  return with PCRE_ERROR_BADMAGIC. However, if the magic number is equal to
3173  match block, so we must initialize them beforehand. However, the other fields  REVERSED_MAGIC_NUMBER we return with PCRE_ERROR_BADENDIANNESS, which
3174  in the match block must not be set until after the byte flipping. */  means that the pattern is likely compiled with different endianness. */
3175    
3176    if (re->magic_number != MAGIC_NUMBER)
3177      return re->magic_number == REVERSED_MAGIC_NUMBER?
3178        PCRE_ERROR_BADENDIANNESS:PCRE_ERROR_BADMAGIC;
3179    if ((re->flags & PCRE_MODE) == 0) return PCRE_ERROR_BADMODE;
3180    
3181    /* If restarting after a partial match, do some sanity checks on the contents
3182    of the workspace. */
3183    
3184    if ((options & PCRE_DFA_RESTART) != 0)
3185      {
3186      if ((workspace[0] & (-2)) != 0 || workspace[1] < 1 ||
3187        workspace[1] > (wscount - 2)/INTS_PER_STATEBLOCK)
3188          return PCRE_ERROR_DFA_BADRESTART;
3189      }
3190    
3191    /* Set up study, callout, and table data */
3192    
3193  md->tables = re->tables;  md->tables = re->tables;
3194  md->callout_data = NULL;  md->callout_data = NULL;
# Line 2731  if (extra_data != NULL) Line 3207  if (extra_data != NULL)
3207      md->tables = extra_data->tables;      md->tables = extra_data->tables;
3208    }    }
3209    
 /* 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;  
   }  
   
3210  /* Set some local values */  /* Set some local values */
3211    
3212  current_subject = (const unsigned char *)subject + start_offset;  current_subject = (const pcre_uchar *)subject + start_offset;
3213  end_subject = (const unsigned char *)subject + length;  end_subject = (const pcre_uchar *)subject + length;
3214  req_byte_ptr = current_subject - 1;  req_char_ptr = current_subject - 1;
3215    
3216  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3217  utf8 = (re->options & PCRE_UTF8) != 0;  /* PCRE_UTF16 has the same value as PCRE_UTF8. */
3218    utf = (re->options & PCRE_UTF8) != 0;
3219  #else  #else
3220  utf8 = FALSE;  utf = FALSE;
3221  #endif  #endif
3222    
3223  anchored = (options & (PCRE_ANCHORED|PCRE_DFA_RESTART)) != 0 ||  anchored = (options & (PCRE_ANCHORED|PCRE_DFA_RESTART)) != 0 ||
# Line 2760  anchored = (options & (PCRE_ANCHORED|PCR Line 3225  anchored = (options & (PCRE_ANCHORED|PCR
3225    
3226  /* The remaining fixed data for passing around. */  /* The remaining fixed data for passing around. */
3227    
3228  md->start_code = (const uschar *)argument_re +  md->start_code = (const pcre_uchar *)argument_re +
3229      re->name_table_offset + re->name_count * re->name_entry_size;      re->name_table_offset + re->name_count * re->name_entry_size;
3230  md->start_subject = (const unsigned char *)subject;  md->start_subject = (const pcre_uchar *)subject;
3231  md->end_subject = end_subject;  md->end_subject = end_subject;
3232  md->start_offset = start_offset;  md->start_offset = start_offset;
3233  md->moptions = options;  md->moptions = options;
# Line 2823  else Line 3288  else
3288  /* 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
3289  back the character offset. */  back the character offset. */
3290    
3291  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3292  if (utf8 && (options & PCRE_NO_UTF8_CHECK) == 0)  if (utf && (options & PCRE_NO_UTF8_CHECK) == 0)
3293    {    {
3294    if (_pcre_valid_utf8((uschar *)subject, length) >= 0)    int erroroffset;
3295      return PCRE_ERROR_BADUTF8;    int errorcode = PRIV(valid_utf)((pcre_uchar *)subject, length, &erroroffset);
3296    if (start_offset > 0 && start_offset < length)    if (errorcode != 0)
3297      {      {
3298      int tb = ((uschar *)subject)[start_offset];      if (offsetcount >= 2)
     if (tb > 127)  
3299        {        {
3300        tb &= 0xc0;        offsets[0] = erroroffset;
3301        if (tb != 0 && tb != 0xc0) return PCRE_ERROR_BADUTF8_OFFSET;        offsets[1] = errorcode;
3302        }        }
3303        return (errorcode <= PCRE_UTF8_ERR5 && (options & PCRE_PARTIAL_HARD) != 0)?
3304          PCRE_ERROR_SHORTUTF8 : PCRE_ERROR_BADUTF8;
3305      }      }
3306      if (start_offset > 0 && start_offset < length &&
3307            NOT_FIRSTCHAR(((PCRE_PUCHAR)subject)[start_offset]))
3308        return PCRE_ERROR_BADUTF8_OFFSET;
3309    }    }
3310  #endif  #endif
3311    
# Line 2844  if (utf8 && (options & PCRE_NO_UTF8_CHEC Line 3313  if (utf8 && (options & PCRE_NO_UTF8_CHEC
3313  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
3314  in other programs later. */  in other programs later. */
3315    
3316  if (md->tables == NULL) md->tables = _pcre_default_tables;  if (md->tables == NULL) md->tables = PRIV(default_tables);
3317    
3318  /* 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
3319  used in a loop when finding where to start. */  where to start. */
3320    
 lcc = md->tables + lcc_offset;  
3321  startline = (re->flags & PCRE_STARTLINE) != 0;  startline = (re->flags & PCRE_STARTLINE) != 0;
3322  firstline = (re->options & PCRE_FIRSTLINE) != 0;  firstline = (re->options & PCRE_FIRSTLINE) != 0;
3323    
# Line 2863  if (!anchored) Line 3331  if (!anchored)
3331    {    {
3332    if ((re->flags & PCRE_FIRSTSET) != 0)    if ((re->flags & PCRE_FIRSTSET) != 0)
3333      {      {
3334      first_byte = re->first_byte & 255;      has_first_char = TRUE;
3335      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)      first_char = first_char2 = (pcre_uchar)(re->first_char);
3336        first_byte = lcc[first_byte];      if ((re->flags & PCRE_FCH_CASELESS) != 0)
3337          {
3338          first_char2 = TABLE_GET(first_char, md->tables + fcc_offset, first_char);
3339    #if defined SUPPORT_UCP && !(defined COMPILE_PCRE8)
3340          if (utf && first_char > 127)
3341            first_char2 = UCD_OTHERCASE(first_char);
3342    #endif
3343          }
3344      }      }
3345    else    else
3346      {      {
# Line 2880  character" set. */ Line 3355  character" set. */
3355    
3356  if ((re->flags & PCRE_REQCHSET) != 0)  if ((re->flags & PCRE_REQCHSET) != 0)
3357    {    {
3358    req_byte = re->req_byte & 255;    has_req_char = TRUE;
3359    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;    req_char = req_char2 = (pcre_uchar)(re->req_char);
3360    req_byte2 = (md->tables + fcc_offset)[req_byte];  /* case flipped */    if ((re->flags & PCRE_RCH_CASELESS) != 0)
3361        {
3362        req_char2 = TABLE_GET(req_char, md->tables + fcc_offset, req_char);
3363    #if defined SUPPORT_UCP && !(defined COMPILE_PCRE8)
3364        if (utf && req_char > 127)
3365          req_char2 = UCD_OTHERCASE(req_char);
3366    #endif
3367        }
3368    }    }
3369    
3370  /* 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 2895  for (;;) Line 3377  for (;;)
3377    
3378    if ((options & PCRE_DFA_RESTART) == 0)    if ((options & PCRE_DFA_RESTART) == 0)
3379      {      {
3380      const uschar *save_end_subject = end_subject;      const pcre_uchar *save_end_subject = end_subject;
3381    
3382      /* 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
3383      line of a multiline string. Implement this by temporarily adjusting      line of a multiline string. Implement this by temporarily adjusting
# Line 2904  for (;;) Line 3386  for (;;)
3386    
3387      if (firstline)      if (firstline)
3388        {        {
3389        USPTR t = current_subject;        PCRE_PUCHAR t = current_subject;
3390  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3391        if (utf8)        if (utf)
3392          {          {
3393          while (t < md->end_subject && !IS_NEWLINE(t))          while (t < md->end_subject && !IS_NEWLINE(t))
3394            {            {
3395            t++;            t++;
3396            while (t < end_subject && (*t & 0xc0) == 0x80) t++;            ACROSSCHAR(t < end_subject, *t, t++);
3397            }            }
3398          }          }
3399        else        else
# Line 2922  for (;;) Line 3404  for (;;)
3404    
3405      /* There are some optimizations that avoid running the match if a known      /* There are some optimizations that avoid running the match if a known
3406      starting point is not found. However, there is an option that disables      starting point is not found. However, there is an option that disables
3407      these, for testing and for ensuring that all callouts do actually occur. */      these, for testing and for ensuring that all callouts do actually occur.
3408        The option can be set in the regex by (*NO_START_OPT) or passed in
3409        match-time options. */
3410    
3411      if ((options & PCRE_NO_START_OPTIMIZE) == 0)      if (((options | re->options) & PCRE_NO_START_OPTIMIZE) == 0)
3412        {        {
3413        /* Advance to a known first byte. */        /* Advance to a known first char. */
3414    
3415        if (first_byte >= 0)        if (has_first_char)
3416          {          {
3417          if (first_byte_caseless)          if (first_char != first_char2)
3418            while (current_subject < end_subject &&            while (current_subject < end_subject &&
3419                   lcc[*current_subject] != first_byte)                *current_subject != first_char && *current_subject != first_char2)
3420              current_subject++;              current_subject++;
3421          else          else
3422            while (current_subject < end_subject &&            while (current_subject < end_subject &&
3423                   *current_subject != first_byte)                   *current_subject != first_char)
3424              current_subject++;              current_subject++;
3425          }          }
3426    
# Line 2946  for (;;) Line 3430  for (;;)
3430          {          {
3431          if (current_subject > md->start_subject + start_offset)          if (current_subject > md->start_subject + start_offset)
3432            {            {
3433  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3434            if (utf8)            if (utf)
3435              {              {
3436              while (current_subject < end_subject &&              while (current_subject < end_subject &&
3437                     !WAS_NEWLINE(current_subject))                     !WAS_NEWLINE(current_subject))
3438                {                {
3439                current_subject++;                current_subject++;
3440                while(current_subject < end_subject &&                ACROSSCHAR(current_subject < end_subject, *current_subject,
3441                      (*current_subject & 0xc0) == 0x80)                  current_subject++);
                 current_subject++;  
3442                }                }
3443              }              }
3444            else            else
# Line 2982  for (;;) Line 3465  for (;;)
3465          while (current_subject < end_subject)          while (current_subject < end_subject)
3466            {            {
3467            register unsigned int c = *current_subject;            register unsigned int c = *current_subject;
3468            if ((start_bits[c/8] & (1 << (c&7))) == 0) current_subject++;  #ifndef COMPILE_PCRE8
3469              else break;            if (c > 255) c = 255;
3470    #endif
3471              if ((start_bits[c/8] & (1 << (c&7))) == 0)
3472                {
3473                current_subject++;
3474    #if defined SUPPORT_UTF && defined COMPILE_PCRE8
3475                /* In non 8-bit mode, the iteration will stop for
3476                characters > 255 at the beginning or not stop at all. */
3477                if (utf)
3478                  ACROSSCHAR(current_subject < end_subject, *current_subject,
3479                    current_subject++);
3480    #endif
3481                }
3482              else break;
3483            }            }
3484          }          }
3485        }        }
# Line 2996  for (;;) Line 3492  for (;;)
3492      disabling is explicitly requested (and of course, by the test above, this      disabling is explicitly requested (and of course, by the test above, this
3493      code is not obeyed when restarting after a partial match). */      code is not obeyed when restarting after a partial match). */
3494    
3495      if ((options & PCRE_NO_START_OPTIMIZE) == 0 &&      if (((options | re->options) & PCRE_NO_START_OPTIMIZE) == 0 &&
3496          (options & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) == 0)          (options & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) == 0)
3497        {        {
3498        /* If the pattern was studied, a minimum subject length may be set. This        /* If the pattern was studied, a minimum subject length may be set. This
# Line 3008  for (;;) Line 3504  for (;;)
3504            (pcre_uint32)(end_subject - current_subject) < study->minlength)            (pcre_uint32)(end_subject - current_subject) < study->minlength)
3505          return PCRE_ERROR_NOMATCH;          return PCRE_ERROR_NOMATCH;
3506    
3507        /* If req_byte is set, we know that that character must appear in the        /* If req_char is set, we know that that character must appear in the
3508        subject for the match to succeed. If the first character is set, req_byte        subject for the match to succeed. If the first character is set, req_char
3509        must be later in the subject; otherwise the test starts at the match        must be later in the subject; otherwise the test starts at the match
3510        point. This optimization can save a huge amount of work in patterns with        point. This optimization can save a huge amount of work in patterns with
3511        nested unlimited repeats that aren't going to match. Writing separate        nested unlimited repeats that aren't going to match. Writing separate
# Line 3021  for (;;) Line 3517  for (;;)
3517        patterns. This showed up when somebody was matching /^C/ on a 32-megabyte        patterns. This showed up when somebody was matching /^C/ on a 32-megabyte
3518        string... so we don't do this when the string is sufficiently long. */        string... so we don't do this when the string is sufficiently long. */
3519    
3520        if (req_byte >= 0 && end_subject - current_subject < REQ_BYTE_MAX)        if (has_req_char && end_subject - current_subject < REQ_BYTE_MAX)
3521          {          {
3522          register const uschar *p = current_subject + ((first_byte >= 0)? 1 : 0);          register PCRE_PUCHAR p = current_subject + (has_first_char? 1:0);
3523    
3524          /* We don't need to repeat the search if we haven't yet reached the          /* We don't need to repeat the search if we haven't yet reached the
3525          place we found it at last time. */          place we found it at last time. */
3526    
3527          if (p > req_byte_ptr)          if (p > req_char_ptr)
3528            {            {
3529            if (req_byte_caseless)            if (req_char != req_char2)
3530              {              {
3531              while (p < end_subject)              while (p < end_subject)
3532                {                {
3533                register int pp = *p++;                register int pp = *p++;
3534                if (pp == req_byte || pp == req_byte2) { p--; break; }                if (pp == req_char || pp == req_char2) { p--; break; }
3535                }                }
3536              }              }
3537            else            else
3538              {              {
3539              while (p < end_subject)              while (p < end_subject)
3540                {                {
3541                if (*p++ == req_byte) { p--; break; }                if (*p++ == req_char) { p--; break; }
3542                }                }
3543              }              }
3544    
# Line 3055  for (;;) Line 3551  for (;;)
3551            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
3552            the start hasn't passed this character yet. */            the start hasn't passed this character yet. */
3553    
3554            req_byte_ptr = p;            req_char_ptr = p;
3555            }            }
3556          }          }
3557        }        }
# Line 3064  for (;;) Line 3560  for (;;)
3560    /* OK, now we can do the business */    /* OK, now we can do the business */
3561    
3562    md->start_used_ptr = current_subject;    md->start_used_ptr = current_subject;
3563      md->recursive = NULL;
3564    
3565    rc = internal_dfa_exec(    rc = internal_dfa_exec(
3566      md,                                /* fixed match data */      md,                                /* fixed match data */
# Line 3074  for (;;) Line 3571  for (;;)
3571      offsetcount,                       /* size of same */      offsetcount,                       /* size of same */
3572      workspace,                         /* workspace vector */      workspace,                         /* workspace vector */
3573      wscount,                           /* size of same */      wscount,                           /* size of same */
3574      re->options & (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL), /* ims flags */      0);                                /* function recurse level */
     0,                                 /* function recurse level */  
     0);                                /* regex recurse level */  
3575    
3576    /* Anything other than "no match" means we are done, always; otherwise, carry    /* Anything other than "no match" means we are done, always; otherwise, carry
3577    on only if not anchored. */    on only if not anchored. */
# Line 3088  for (;;) Line 3583  for (;;)
3583    
3584    if (firstline && IS_NEWLINE(current_subject)) break;    if (firstline && IS_NEWLINE(current_subject)) break;
3585    current_subject++;    current_subject++;
3586    if (utf8)  #ifdef SUPPORT_UTF
3587      if (utf)
3588      {      {
3589      while (current_subject < end_subject && (*current_subject & 0xc0) == 0x80)      ACROSSCHAR(current_subject < end_subject, *current_subject,
3590        current_subject++;        current_subject++);
3591      }      }
3592    #endif
3593    if (current_subject > end_subject) break;    if (current_subject > end_subject) break;
3594    
3595    /* If we have just passed a CR and we are now at a LF, and the pattern does    /* If we have just passed a CR and we are now at a LF, and the pattern does

Legend:
Removed from v.498  
changed lines
  Added in v.982

  ViewVC Help
Powered by ViewVC 1.1.5