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

Diff of /code/trunk/pcretest.c

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

revision 85 by nigel, Sat Feb 24 21:41:13 2007 UTC revision 144 by ph10, Mon Apr 2 13:32:07 2007 UTC
# Line 44  POSSIBILITY OF SUCH DAMAGE. Line 44  POSSIBILITY OF SUCH DAMAGE.
44  #include <locale.h>  #include <locale.h>
45  #include <errno.h>  #include <errno.h>
46    
47    
48    /* A number of things vary for Windows builds. Originally, pcretest opened its
49    input and output without "b"; then I was told that "b" was needed in some
50    environments, so it was added for release 5.0 to both the input and output. (It
51    makes no difference on Unix-like systems.) Later I was told that it is wrong
52    for the input on Windows. I've now abstracted the modes into two macros that
53    are set here, to make it easier to fiddle with them, and removed "b" from the
54    input mode under Windows. */
55    
56    #if defined(_WIN32) || defined(WIN32)
57    #include <io.h>                /* For _setmode() */
58    #include <fcntl.h>             /* For _O_BINARY */
59    #define INPUT_MODE   "r"
60    #define OUTPUT_MODE  "wb"
61    
62    #else
63    #include <sys/time.h>          /* These two includes are needed */
64    #include <sys/resource.h>      /* for setrlimit(). */
65    #define INPUT_MODE   "rb"
66    #define OUTPUT_MODE  "wb"
67    #endif
68    
69    
70  #define PCRE_SPY        /* For Win32 build, import data, not export */  #define PCRE_SPY        /* For Win32 build, import data, not export */
71    
72  /* We include pcre_internal.h because we need the internal info for displaying  /* We include pcre_internal.h because we need the internal info for displaying
# Line 70  symbols to prevent clashes. */ Line 93  symbols to prevent clashes. */
93    
94  /* We also need the pcre_printint() function for printing out compiled  /* We also need the pcre_printint() function for printing out compiled
95  patterns. This function is in a separate file so that it can be included in  patterns. This function is in a separate file so that it can be included in
96  pcre_compile.c when that module is compiled with debugging enabled. */  pcre_compile.c when that module is compiled with debugging enabled.
97    
98    The definition of the macro PRINTABLE, which determines whether to print an
99    output character as-is or as a hex value when showing compiled patterns, is
100    contained in this file. We uses it here also, in cases when the locale has not
101    been explicitly changed, so as to get consistent output from systems that
102    differ in their output from isprint() even in the "C" locale. */
103    
104  #include "pcre_printint.src"  #include "pcre_printint.src"
105    
106    #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
107    
108    
109  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
110  testing the POSIX interface, though this is not available via the standard  testing the POSIX interface, though this is not available via the standard
# Line 83  Makefile. */ Line 114  Makefile. */
114  #include "pcreposix.h"  #include "pcreposix.h"
115  #endif  #endif
116    
117  /* It is also possible, for the benefit of the version imported into Exim, to  /* It is also possible, for the benefit of the version currently imported into
118  build pcretest without support for UTF8 (define NOUTF8), without the interface  Exim, to build pcretest without support for UTF8 (define NOUTF8), without the
119  to the DFA matcher (NODFA), and without the doublecheck of the old "info"  interface to the DFA matcher (NODFA), and without the doublecheck of the old
120  function (define NOINFOCHECK). */  "info" function (define NOINFOCHECK). In fact, we automatically cut out the
121    UTF8 support if PCRE is built without it. */
122    
123    #ifndef SUPPORT_UTF8
124    #ifndef NOUTF8
125    #define NOUTF8
126    #endif
127    #endif
128    
129    
130  /* Other parameters */  /* Other parameters */
# Line 99  function (define NOINFOCHECK). */ Line 137  function (define NOINFOCHECK). */
137  #endif  #endif
138  #endif  #endif
139    
140  #define LOOPREPEAT 500000  /* This is the default loop count for timing. */
   
 #define BUFFER_SIZE 30000  
 #define PBUFFER_SIZE BUFFER_SIZE  
 #define DBUFFER_SIZE BUFFER_SIZE  
141    
142    #define LOOPREPEAT 500000
143    
144  /* Static variables */  /* Static variables */
145    
# Line 115  static int callout_extra; Line 150  static int callout_extra;
150  static int callout_fail_count;  static int callout_fail_count;
151  static int callout_fail_id;  static int callout_fail_id;
152  static int first_callout;  static int first_callout;
153    static int locale_set = 0;
154  static int show_malloc;  static int show_malloc;
155  static int use_utf8;  static int use_utf8;
156  static size_t gotten_store;  static size_t gotten_store;
157    
158    /* The buffers grow automatically if very long input lines are encountered. */
159    
160    static int buffer_size = 50000;
161    static uschar *buffer = NULL;
162    static uschar *dbuffer = NULL;
163  static uschar *pbuffer = NULL;  static uschar *pbuffer = NULL;
164    
165    
166    
167  /*************************************************  /*************************************************
168    *        Read or extend an input line            *
169    *************************************************/
170    
171    /* Input lines are read into buffer, but both patterns and data lines can be
172    continued over multiple input lines. In addition, if the buffer fills up, we
173    want to automatically expand it so as to be able to handle extremely large
174    lines that are needed for certain stress tests. When the input buffer is
175    expanded, the other two buffers must also be expanded likewise, and the
176    contents of pbuffer, which are a copy of the input for callouts, must be
177    preserved (for when expansion happens for a data line). This is not the most
178    optimal way of handling this, but hey, this is just a test program!
179    
180    Arguments:
181      f            the file to read
182      start        where in buffer to start (this *must* be within buffer)
183    
184    Returns:       pointer to the start of new data
185                   could be a copy of start, or could be moved
186                   NULL if no data read and EOF reached
187    */
188    
189    static uschar *
190    extend_inputline(FILE *f, uschar *start)
191    {
192    uschar *here = start;
193    
194    for (;;)
195      {
196      int rlen = buffer_size - (here - buffer);
197    
198      if (rlen > 1000)
199        {
200        int dlen;
201        if (fgets((char *)here, rlen,  f) == NULL)
202          return (here == start)? NULL : start;
203        dlen = (int)strlen((char *)here);
204        if (dlen > 0 && here[dlen - 1] == '\n') return start;
205        here += dlen;
206        }
207    
208      else
209        {
210        int new_buffer_size = 2*buffer_size;
211        uschar *new_buffer = (unsigned char *)malloc(new_buffer_size);
212        uschar *new_dbuffer = (unsigned char *)malloc(new_buffer_size);
213        uschar *new_pbuffer = (unsigned char *)malloc(new_buffer_size);
214    
215        if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
216          {
217          fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
218          exit(1);
219          }
220    
221        memcpy(new_buffer, buffer, buffer_size);
222        memcpy(new_pbuffer, pbuffer, buffer_size);
223    
224        buffer_size = new_buffer_size;
225    
226        start = new_buffer + (start - buffer);
227        here = new_buffer + (here - buffer);
228    
229        free(buffer);
230        free(dbuffer);
231        free(pbuffer);
232    
233        buffer = new_buffer;
234        dbuffer = new_dbuffer;
235        pbuffer = new_pbuffer;
236        }
237      }
238    
239    return NULL;  /* Control never gets here */
240    }
241    
242    
243    
244    
245    
246    
247    
248    /*************************************************
249  *          Read number from string               *  *          Read number from string               *
250  *************************************************/  *************************************************/
251    
252  /* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess  /* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess
253  around with conditional compilation, just do the job by hand. It is only used  around with conditional compilation, just do the job by hand. It is only used
254  for unpicking the -o argument, so just keep it simple.  for unpicking arguments, so just keep it simple.
255    
256  Arguments:  Arguments:
257    str           string to be converted    str           string to be converted
# Line 159  return(result); Line 281  return(result);
281  and returns the value of the character.  and returns the value of the character.
282    
283  Argument:  Argument:
284    buffer   a pointer to the byte vector    utf8bytes   a pointer to the byte vector
285    vptr     a pointer to an int to receive the value    vptr        a pointer to an int to receive the value
286    
287  Returns:   >  0 => the number of bytes consumed  Returns:      >  0 => the number of bytes consumed
288             -6 to 0 => malformed UTF-8 character at offset = (-return)                -6 to 0 => malformed UTF-8 character at offset = (-return)
289  */  */
290    
291  #if !defined NOUTF8  #if !defined NOUTF8
292    
293  static int  static int
294  utf82ord(unsigned char *buffer, int *vptr)  utf82ord(unsigned char *utf8bytes, int *vptr)
295  {  {
296  int c = *buffer++;  int c = *utf8bytes++;
297  int d = c;  int d = c;
298  int i, j, s;  int i, j, s;
299    
# Line 191  d = (c & utf8_table3[i]) << s; Line 313  d = (c & utf8_table3[i]) << s;
313    
314  for (j = 0; j < i; j++)  for (j = 0; j < i; j++)
315    {    {
316    c = *buffer++;    c = *utf8bytes++;
317    if ((c & 0xc0) != 0x80) return -(j+1);    if ((c & 0xc0) != 0x80) return -(j+1);
318    s -= 6;    s -= 6;
319    d |= (c & 0x3f) << s;    d |= (c & 0x3f) << s;
# Line 222  and encodes it as a UTF-8 character in 0 Line 344  and encodes it as a UTF-8 character in 0
344    
345  Arguments:  Arguments:
346    cvalue     the character value    cvalue     the character value
347    buffer     pointer to buffer for result - at least 6 bytes long    utf8bytes  pointer to buffer for result - at least 6 bytes long
348    
349  Returns:     number of characters placed in the buffer  Returns:     number of characters placed in the buffer
350  */  */
351    
352    #if !defined NOUTF8
353    
354  static int  static int
355  ord2utf8(int cvalue, uschar *buffer)  ord2utf8(int cvalue, uschar *utf8bytes)
356  {  {
357  register int i, j;  register int i, j;
358  for (i = 0; i < utf8_table1_size; i++)  for (i = 0; i < utf8_table1_size; i++)
359    if (cvalue <= utf8_table1[i]) break;    if (cvalue <= utf8_table1[i]) break;
360  buffer += i;  utf8bytes += i;
361  for (j = i; j > 0; j--)  for (j = i; j > 0; j--)
362   {   {
363   *buffer-- = 0x80 | (cvalue & 0x3f);   *utf8bytes-- = 0x80 | (cvalue & 0x3f);
364   cvalue >>= 6;   cvalue >>= 6;
365   }   }
366  *buffer = utf8_table2[i] | cvalue;  *utf8bytes = utf8_table2[i] | cvalue;
367  return i + 1;  return i + 1;
368  }  }
369    
370    #endif
371    
372    
373    
374  /*************************************************  /*************************************************
# Line 269  while (length-- > 0) Line 395  while (length-- > 0)
395        {        {
396        length -= rc - 1;        length -= rc - 1;
397        p += rc;        p += rc;
398        if (c < 256 && isprint(c))        if (PRINTHEX(c))
399          {          {
400          if (f != NULL) fprintf(f, "%c", c);          if (f != NULL) fprintf(f, "%c", c);
401          yield++;          yield++;
402          }          }
403        else        else
404          {          {
405          int n;          int n = 4;
406          if (f != NULL) fprintf(f, "\\x{%02x}%n", c, &n);          if (f != NULL) fprintf(f, "\\x{%02x}", c);
407          yield += n;          yield += (n <= 0x000000ff)? 2 :
408                     (n <= 0x00000fff)? 3 :
409                     (n <= 0x0000ffff)? 4 :
410                     (n <= 0x000fffff)? 5 : 6;
411          }          }
412        continue;        continue;
413        }        }
# Line 287  while (length-- > 0) Line 416  while (length-- > 0)
416    
417     /* Not UTF-8, or malformed UTF-8  */     /* Not UTF-8, or malformed UTF-8  */
418    
419    if (isprint(c = *(p++)))    c = *p++;
420      if (PRINTHEX(c))
421      {      {
422      if (f != NULL) fprintf(f, "%c", c);      if (f != NULL) fprintf(f, "%c", c);
423      yield++;      yield++;
# Line 461  if ((rc = pcre_fullinfo(re, study, optio Line 591  if ((rc = pcre_fullinfo(re, study, optio
591  *         Byte flipping function                 *  *         Byte flipping function                 *
592  *************************************************/  *************************************************/
593    
594  static long int  static unsigned long int
595  byteflip(long int value, int n)  byteflip(unsigned long int value, int n)
596  {  {
597  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);
598  return ((value & 0x000000ff) << 24) |  return ((value & 0x000000ff) << 24) |
# Line 475  return ((value & 0x000000ff) << 24) | Line 605  return ((value & 0x000000ff) << 24) |
605    
606    
607  /*************************************************  /*************************************************
608    *        Check match or recursion limit          *
609    *************************************************/
610    
611    static int
612    check_match_limit(pcre *re, pcre_extra *extra, uschar *bptr, int len,
613      int start_offset, int options, int *use_offsets, int use_size_offsets,
614      int flag, unsigned long int *limit, int errnumber, const char *msg)
615    {
616    int count;
617    int min = 0;
618    int mid = 64;
619    int max = -1;
620    
621    extra->flags |= flag;
622    
623    for (;;)
624      {
625      *limit = mid;
626    
627      count = pcre_exec(re, extra, (char *)bptr, len, start_offset, options,
628        use_offsets, use_size_offsets);
629    
630      if (count == errnumber)
631        {
632        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
633        min = mid;
634        mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;
635        }
636    
637      else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||
638                             count == PCRE_ERROR_PARTIAL)
639        {
640        if (mid == min + 1)
641          {
642          fprintf(outfile, "Minimum %s limit = %d\n", msg, mid);
643          break;
644          }
645        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
646        max = mid;
647        mid = (min + mid)/2;
648        }
649      else break;    /* Some other error */
650      }
651    
652    extra->flags &= ~flag;
653    return count;
654    }
655    
656    
657    
658    /*************************************************
659    *         Check newline indicator                *
660    *************************************************/
661    
662    /* This is used both at compile and run-time to check for <xxx> escapes, where
663    xxx is LF, CR, CRLF, or ANY. Print a message and return 0 if there is no match.
664    
665    Arguments:
666      p           points after the leading '<'
667      f           file for error message
668    
669    Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
670    */
671    
672    static int
673    check_newline(uschar *p, FILE *f)
674    {
675    if (strncmp((char *)p, "cr>", 3) == 0) return PCRE_NEWLINE_CR;
676    if (strncmp((char *)p, "lf>", 3) == 0) return PCRE_NEWLINE_LF;
677    if (strncmp((char *)p, "crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
678    if (strncmp((char *)p, "any>", 4) == 0) return PCRE_NEWLINE_ANY;
679    fprintf(f, "Unknown newline type at: <%s\n", p);
680    return 0;
681    }
682    
683    
684    
685    /*************************************************
686    *             Usage function                     *
687    *************************************************/
688    
689    static void
690    usage(void)
691    {
692    printf("Usage:     pcretest [options] [<input> [<output>]]\n");
693    printf("  -b       show compiled code (bytecode)\n");
694    printf("  -C       show PCRE compile-time options and exit\n");
695    printf("  -d       debug: show compiled code and information (-b and -i)\n");
696    #if !defined NODFA
697    printf("  -dfa     force DFA matching for all subjects\n");
698    #endif
699    printf("  -help    show usage information\n");
700    printf("  -i       show information about compiled patterns\n"
701           "  -m       output memory used information\n"
702           "  -o <n>   set size of offsets vector to <n>\n");
703    #if !defined NOPOSIX
704    printf("  -p       use POSIX interface\n");
705    #endif
706    printf("  -q       quiet: do not output PCRE version number at start\n");
707    printf("  -S <n>   set stack size to <n> megabytes\n");
708    printf("  -s       output store (memory) used information\n"
709           "  -t       time compilation and execution\n");
710    printf("  -t <n>   time compilation and execution, repeating <n> times\n");
711    printf("  -tm      time execution (matching) only\n");
712    printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
713    }
714    
715    
716    
717    /*************************************************
718  *                Main Program                    *  *                Main Program                    *
719  *************************************************/  *************************************************/
720    
# Line 489  int options = 0; Line 729  int options = 0;
729  int study_options = 0;  int study_options = 0;
730  int op = 1;  int op = 1;
731  int timeit = 0;  int timeit = 0;
732    int timeitm = 0;
733  int showinfo = 0;  int showinfo = 0;
734  int showstore = 0;  int showstore = 0;
735    int quiet = 0;
736  int size_offsets = 45;  int size_offsets = 45;
737  int size_offsets_max;  int size_offsets_max;
738  int *offsets = NULL;  int *offsets = NULL;
# Line 501  int debug = 0; Line 743  int debug = 0;
743  int done = 0;  int done = 0;
744  int all_use_dfa = 0;  int all_use_dfa = 0;
745  int yield = 0;  int yield = 0;
746    int stack_size;
747    
748    /* These vectors store, end-to-end, a list of captured substring names. Assume
749    that 1024 is plenty long enough for the few names we'll be testing. */
750    
751  unsigned char *buffer;  uschar copynames[1024];
752  unsigned char *dbuffer;  uschar getnames[1024];
753    
754    uschar *copynamesptr;
755    uschar *getnamesptr;
756    
757  /* Get buffers from malloc() so that Electric Fence will check their misuse  /* Get buffers from malloc() so that Electric Fence will check their misuse
758  when I am debugging. */  when I am debugging. They grow automatically when very long lines are read. */
759    
760  buffer = (unsigned char *)malloc(BUFFER_SIZE);  buffer = (unsigned char *)malloc(buffer_size);
761  dbuffer = (unsigned char *)malloc(DBUFFER_SIZE);  dbuffer = (unsigned char *)malloc(buffer_size);
762  pbuffer = (unsigned char *)malloc(PBUFFER_SIZE);  pbuffer = (unsigned char *)malloc(buffer_size);
   
 /* The outfile variable is static so that new_malloc can use it. The _setmode()  
 stuff is some magic that I don't understand, but which apparently does good  
 things in Windows. It's related to line terminations.  */  
763    
764  #if defined(_WIN32) || defined(WIN32)  /* The outfile variable is static so that new_malloc can use it. */
 _setmode( _fileno( stdout ), 0x8000 );  
 #endif  /* defined(_WIN32) || defined(WIN32) */  
765    
766  outfile = stdout;  outfile = stdout;
767    
768    /* The following  _setmode() stuff is some Windows magic that tells its runtime
769    library to translate CRLF into a single LF character. At least, that's what
770    I've been told: never having used Windows I take this all on trust. Originally
771    it set 0x8000, but then I was advised that _O_BINARY was better. */
772    
773    #if defined(_WIN32) || defined(WIN32)
774    _setmode( _fileno( stdout ), _O_BINARY );
775    #endif
776    
777  /* Scan options */  /* Scan options */
778    
779  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
# Line 530  while (argc > 1 && argv[op][0] == '-') Line 782  while (argc > 1 && argv[op][0] == '-')
782    
783    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
784      showstore = 1;      showstore = 1;
785    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-q") == 0) quiet = 1;
786      else if (strcmp(argv[op], "-b") == 0) debug = 1;
787    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
788    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
789  #if !defined NODFA  #if !defined NODFA
# Line 543  while (argc > 1 && argv[op][0] == '-') Line 796  while (argc > 1 && argv[op][0] == '-')
796      op++;      op++;
797      argc--;      argc--;
798      }      }
799      else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
800        {
801        int both = argv[op][2] == 0;
802        int temp;
803        if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr),
804                         *endptr == 0))
805          {
806          timeitm = temp;
807          op++;
808          argc--;
809          }
810        else timeitm = LOOPREPEAT;
811        if (both) timeit = timeitm;
812        }
813      else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
814          ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
815            *endptr == 0))
816        {
817    #if defined(_WIN32) || defined(WIN32)
818        printf("PCRE: -S not supported on this OS\n");
819        exit(1);
820    #else
821        int rc;
822        struct rlimit rlim;
823        getrlimit(RLIMIT_STACK, &rlim);
824        rlim.rlim_cur = stack_size * 1024 * 1024;
825        rc = setrlimit(RLIMIT_STACK, &rlim);
826        if (rc != 0)
827          {
828        printf("PCRE: setrlimit() failed with error %d\n", rc);
829        exit(1);
830          }
831        op++;
832        argc--;
833    #endif
834        }
835  #if !defined NOPOSIX  #if !defined NOPOSIX
836    else if (strcmp(argv[op], "-p") == 0) posix = 1;    else if (strcmp(argv[op], "-p") == 0) posix = 1;
837  #endif  #endif
# Line 556  while (argc > 1 && argv[op][0] == '-') Line 845  while (argc > 1 && argv[op][0] == '-')
845      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
846      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
847      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
848      printf("  Newline character is %s\n", (rc == '\r')? "CR" : "LF");      printf("  Newline sequence is %s\n", (rc == '\r')? "CR" :
849          (rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" :
850          (rc == -1)? "ANY" : "???");
851      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);
852      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
853      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
854      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
855      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);
856      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %d\n", rc);
857        (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &rc);
858        printf("  Default recursion depth limit = %d\n", rc);
859      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
860      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
861      exit(0);      goto EXIT;
862        }
863      else if (strcmp(argv[op], "-help") == 0 ||
864               strcmp(argv[op], "--help") == 0)
865        {
866        usage();
867        goto EXIT;
868      }      }
869    else    else
870      {      {
871      printf("** Unknown or malformed option %s\n", argv[op]);      printf("** Unknown or malformed option %s\n", argv[op]);
872      printf("Usage:   pcretest [-d] [-i] [-o <n>] [-p] [-s] [-t] [<input> [<output>]]\n");      usage();
     printf("  -C     show PCRE compile-time options and exit\n");  
     printf("  -d     debug: show compiled code; implies -i\n");  
 #if !defined NODFA  
     printf("  -dfa   force DFA matching for all subjects\n");  
 #endif  
     printf("  -i     show information about compiled pattern\n"  
            "  -m     output memory used information\n"  
            "  -o <n> set size of offsets vector to <n>\n");  
 #if !defined NOPOSIX  
     printf("  -p     use POSIX interface\n");  
 #endif  
     printf("  -s     output store (memory) used information\n"  
            "  -t     time compilation and execution\n");  
873      yield = 1;      yield = 1;
874      goto EXIT;      goto EXIT;
875      }      }
# Line 607  if (offsets == NULL) Line 893  if (offsets == NULL)
893    
894  if (argc > 1)  if (argc > 1)
895    {    {
896    infile = fopen(argv[op], "rb");    infile = fopen(argv[op], INPUT_MODE);
897    if (infile == NULL)    if (infile == NULL)
898      {      {
899      printf("** Failed to open %s\n", argv[op]);      printf("** Failed to open %s\n", argv[op]);
# Line 618  if (argc > 1) Line 904  if (argc > 1)
904    
905  if (argc > 2)  if (argc > 2)
906    {    {
907    outfile = fopen(argv[op+1], "wb");    outfile = fopen(argv[op+1], OUTPUT_MODE);
908    if (outfile == NULL)    if (outfile == NULL)
909      {      {
910      printf("** Failed to open %s\n", argv[op+1]);      printf("** Failed to open %s\n", argv[op+1]);
# Line 634  pcre_free = new_free; Line 920  pcre_free = new_free;
920  pcre_stack_malloc = stack_malloc;  pcre_stack_malloc = stack_malloc;
921  pcre_stack_free = stack_free;  pcre_stack_free = stack_free;
922    
923  /* Heading line, then prompt for first regex if stdin */  /* Heading line unless quiet, then prompt for first regex if stdin */
924    
925  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  if (!quiet) fprintf(outfile, "PCRE version %s\n\n", pcre_version());
926    
927  /* Main loop */  /* Main loop */
928    
# Line 658  while (!done) Line 944  while (!done)
944    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
945    int do_study = 0;    int do_study = 0;
946    int do_debug = debug;    int do_debug = debug;
947      int debug_lengths = 1;
948    int do_G = 0;    int do_G = 0;
949    int do_g = 0;    int do_g = 0;
950    int do_showinfo = showinfo;    int do_showinfo = showinfo;
951    int do_showrest = 0;    int do_showrest = 0;
952    int do_flip = 0;    int do_flip = 0;
953    int erroroffset, len, delimiter;    int erroroffset, len, delimiter, poffset;
954    
955    use_utf8 = 0;    use_utf8 = 0;
956    
957    if (infile == stdin) printf("  re> ");    if (infile == stdin) printf("  re> ");
958    if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break;    if (extend_inputline(infile, buffer) == NULL) break;
959    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
960    fflush(outfile);    fflush(outfile);
961    
# Line 680  while (!done) Line 967  while (!done)
967    
968    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
969      {      {
970      unsigned long int magic;      unsigned long int magic, get_options;
971      uschar sbuf[8];      uschar sbuf[8];
972      FILE *f;      FILE *f;
973    
# Line 728  while (!done) Line 1015  while (!done)
1015    
1016      /* Need to know if UTF-8 for printing data strings */      /* Need to know if UTF-8 for printing data strings */
1017    
1018      new_info(re, NULL, PCRE_INFO_OPTIONS, &options);      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1019      use_utf8 = (options & PCRE_UTF8) != 0;      use_utf8 = (get_options & PCRE_UTF8) != 0;
1020    
1021      /* Now see if there is any following study data */      /* Now see if there is any following study data */
1022    
# Line 773  while (!done) Line 1060  while (!done)
1060      }      }
1061    
1062    pp = p;    pp = p;
1063      poffset = p - buffer;
1064    
1065    for(;;)    for(;;)
1066      {      {
# Line 783  while (!done) Line 1071  while (!done)
1071        pp++;        pp++;
1072        }        }
1073      if (*pp != 0) break;      if (*pp != 0) break;
   
     len = BUFFER_SIZE - (pp - buffer);  
     if (len < 256)  
       {  
       fprintf(outfile, "** Expression too long - missing delimiter?\n");  
       goto SKIP_DATA;  
       }  
   
1074      if (infile == stdin) printf("    > ");      if (infile == stdin) printf("    > ");
1075      if (fgets((char *)pp, len, infile) == NULL)      if ((pp = extend_inputline(infile, pp)) == NULL)
1076        {        {
1077        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
1078        done = 1;        done = 1;
# Line 801  while (!done) Line 1081  while (!done)
1081      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
1082      }      }
1083    
1084      /* The buffer may have moved while being extended; reset the start of data
1085      pointer to the correct relative point in the buffer. */
1086    
1087      p = buffer + poffset;
1088    
1089    /* If the first character after the delimiter is backslash, make    /* If the first character after the delimiter is backslash, make
1090    the pattern end with backslash. This is purely to provide a way    the pattern end with backslash. This is purely to provide a way
1091    of testing for the error message when a pattern ends with backslash. */    of testing for the error message when a pattern ends with backslash. */
# Line 832  while (!done) Line 1117  while (!done)
1117    
1118        case '+': do_showrest = 1; break;        case '+': do_showrest = 1; break;
1119        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
1120          case 'B': do_debug = 1; break;
1121        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
1122        case 'D': do_debug = do_showinfo = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
1123        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
1124        case 'F': do_flip = 1; break;        case 'F': do_flip = 1; break;
1125        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1126        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1127          case 'J': options |= PCRE_DUPNAMES; break;
1128        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1129        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1130    
# Line 848  while (!done) Line 1135  while (!done)
1135        case 'S': do_study = 1; break;        case 'S': do_study = 1; break;
1136        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1137        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
1138          case 'Z': debug_lengths = 0; break;
1139        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1140        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
1141    
1142        case 'L':        case 'L':
1143        ppp = pp;        ppp = pp;
1144        /* The '\r' test here is so that it works on Windows */        /* The '\r' test here is so that it works on Windows. */
1145        while (*ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;        /* The '0' test is just in case this is an unterminated line. */
1146          while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
1147        *ppp = 0;        *ppp = 0;
1148        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
1149          {          {
1150          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
1151          goto SKIP_DATA;          goto SKIP_DATA;
1152          }          }
1153          locale_set = 1;
1154        tables = pcre_maketables();        tables = pcre_maketables();
1155        pp = ppp;        pp = ppp;
1156        break;        break;
# Line 872  while (!done) Line 1162  while (!done)
1162        *pp = 0;        *pp = 0;
1163        break;        break;
1164    
1165          case '<':
1166            {
1167            int x = check_newline(pp, outfile);
1168            if (x == 0) goto SKIP_DATA;
1169            options |= x;
1170            while (*pp++ != '>');
1171            }
1172          break;
1173    
1174        case '\r':                      /* So that it works in Windows */        case '\r':                      /* So that it works in Windows */
1175        case '\n':        case '\n':
1176        case ' ':        case ' ':
# Line 896  while (!done) Line 1195  while (!done)
1195      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
1196      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
1197      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1198        if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1199        if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1200    
1201      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1202    
1203      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
# Line 903  while (!done) Line 1205  while (!done)
1205    
1206      if (rc != 0)      if (rc != 0)
1207        {        {
1208        (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);        (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1209        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
1210        goto SKIP_DATA;        goto SKIP_DATA;
1211        }        }
# Line 915  while (!done) Line 1217  while (!done)
1217  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1218    
1219      {      {
1220      if (timeit)      if (timeit > 0)
1221        {        {
1222        register int i;        register int i;
1223        clock_t time_taken;        clock_t time_taken;
1224        clock_t start_time = clock();        clock_t start_time = clock();
1225        for (i = 0; i < LOOPREPEAT; i++)        for (i = 0; i < timeit; i++)
1226          {          {
1227          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1228          if (re != NULL) free(re);          if (re != NULL) free(re);
1229          }          }
1230        time_taken = clock() - start_time;        time_taken = clock() - start_time;
1231        fprintf(outfile, "Compile time %.3f milliseconds\n",        fprintf(outfile, "Compile time %.4f milliseconds\n",
1232          (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /          (((double)time_taken * 1000.0) / (double)timeit) /
1233            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
1234        }        }
1235    
# Line 944  while (!done) Line 1246  while (!done)
1246          {          {
1247          for (;;)          for (;;)
1248            {            {
1249            if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)            if (extend_inputline(infile, buffer) == NULL)
1250              {              {
1251              done = 1;              done = 1;
1252              goto CONTINUE;              goto CONTINUE;
# Line 979  while (!done) Line 1281  while (!done)
1281    
1282      if (do_study)      if (do_study)
1283        {        {
1284        if (timeit)        if (timeit > 0)
1285          {          {
1286          register int i;          register int i;
1287          clock_t time_taken;          clock_t time_taken;
1288          clock_t start_time = clock();          clock_t start_time = clock();
1289          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeit; i++)
1290            extra = pcre_study(re, study_options, &error);            extra = pcre_study(re, study_options, &error);
1291          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1292          if (extra != NULL) free(extra);          if (extra != NULL) free(extra);
1293          fprintf(outfile, "  Study time %.3f milliseconds\n",          fprintf(outfile, "  Study time %.4f milliseconds\n",
1294            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeit) /
1295              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
1296          }          }
1297        extra = pcre_study(re, study_options, &error);        extra = pcre_study(re, study_options, &error);
# Line 1032  while (!done) Line 1334  while (!done)
1334    
1335      SHOW_INFO:      SHOW_INFO:
1336    
1337        if (do_debug)
1338          {
1339          fprintf(outfile, "------------------------------------------------------------------\n");
1340          pcre_printint(re, outfile, debug_lengths);
1341          }
1342    
1343      if (do_showinfo)      if (do_showinfo)
1344        {        {
1345        unsigned long int get_options, all_options;        unsigned long int get_options, all_options;
# Line 1042  while (!done) Line 1350  while (!done)
1350        int nameentrysize, namecount;        int nameentrysize, namecount;
1351        const uschar *nametable;        const uschar *nametable;
1352    
       if (do_debug)  
         {  
         fprintf(outfile, "------------------------------------------------------------------\n");  
         pcre_printint(re, outfile);  
         }  
   
1353        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1354        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1355        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
# Line 1105  while (!done) Line 1407  while (!done)
1407        if (do_flip)        if (do_flip)
1408          {          {
1409          all_options = byteflip(all_options, sizeof(all_options));          all_options = byteflip(all_options, sizeof(all_options));
1410          }           }
1411    
1412        if ((all_options & PCRE_NOPARTIAL) != 0)        if ((all_options & PCRE_NOPARTIAL) != 0)
1413          fprintf(outfile, "Partial matching not supported\n");          fprintf(outfile, "Partial matching not supported\n");
1414    
1415        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1416          else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s\n",          else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s%s%s\n",
1417            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1418            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1419            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
# Line 1121  while (!done) Line 1423  while (!done)
1423            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
1424            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
1425            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1426              ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1427            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1428            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "");            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1429              ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1430    
1431        if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0)        switch (get_options & PCRE_NEWLINE_BITS)
1432          fprintf(outfile, "Case state changes\n");          {
1433            case PCRE_NEWLINE_CR:
1434            fprintf(outfile, "Forced newline sequence: CR\n");
1435            break;
1436    
1437            case PCRE_NEWLINE_LF:
1438            fprintf(outfile, "Forced newline sequence: LF\n");
1439            break;
1440    
1441            case PCRE_NEWLINE_CRLF:
1442            fprintf(outfile, "Forced newline sequence: CRLF\n");
1443            break;
1444    
1445            case PCRE_NEWLINE_ANY:
1446            fprintf(outfile, "Forced newline sequence: ANY\n");
1447            break;
1448    
1449            default:
1450            break;
1451            }
1452    
1453        if (first_char == -1)        if (first_char == -1)
1454          {          {
1455          fprintf(outfile, "First char at start or follows \\n\n");          fprintf(outfile, "First char at start or follows newline\n");
1456          }          }
1457        else if (first_char < 0)        else if (first_char < 0)
1458          {          {
# Line 1140  while (!done) Line 1463  while (!done)
1463          int ch = first_char & 255;          int ch = first_char & 255;
1464          const char *caseless = ((first_char & REQ_CASELESS) == 0)?          const char *caseless = ((first_char & REQ_CASELESS) == 0)?
1465            "" : " (caseless)";            "" : " (caseless)";
1466          if (isprint(ch))          if (PRINTHEX(ch))
1467            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);
1468          else          else
1469            fprintf(outfile, "First char = %d%s\n", ch, caseless);            fprintf(outfile, "First char = %d%s\n", ch, caseless);
# Line 1155  while (!done) Line 1478  while (!done)
1478          int ch = need_char & 255;          int ch = need_char & 255;
1479          const char *caseless = ((need_char & REQ_CASELESS) == 0)?          const char *caseless = ((need_char & REQ_CASELESS) == 0)?
1480            "" : " (caseless)";            "" : " (caseless)";
1481          if (isprint(ch))          if (PRINTHEX(ch))
1482            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);
1483          else          else
1484            fprintf(outfile, "Need char = %d%s\n", ch, caseless);            fprintf(outfile, "Need char = %d%s\n", ch, caseless);
# Line 1191  while (!done) Line 1514  while (!done)
1514                    fprintf(outfile, "\n  ");                    fprintf(outfile, "\n  ");
1515                    c = 2;                    c = 2;
1516                    }                    }
1517                  if (isprint(i) && i != ' ')                  if (PRINTHEX(i) && i != ' ')
1518                    {                    {
1519                    fprintf(outfile, "%c ", i);                    fprintf(outfile, "%c ", i);
1520                    c += 2;                    c += 2;
# Line 1250  while (!done) Line 1573  while (!done)
1573                  strerror(errno));                  strerror(errno));
1574                }                }
1575              else fprintf(outfile, "Study data written to %s\n", to_file);              else fprintf(outfile, "Study data written to %s\n", to_file);
1576    
1577              }              }
1578            }            }
1579          fclose(f);          fclose(f);
# Line 1266  while (!done) Line 1590  while (!done)
1590    
1591    for (;;)    for (;;)
1592      {      {
1593      unsigned char *q;      uschar *q;
1594      unsigned char *bptr = dbuffer;      uschar *bptr = dbuffer;
1595      int *use_offsets = offsets;      int *use_offsets = offsets;
1596      int use_size_offsets = size_offsets;      int use_size_offsets = size_offsets;
1597      int callout_data = 0;      int callout_data = 0;
# Line 1284  while (!done) Line 1608  while (!done)
1608    
1609      options = 0;      options = 0;
1610    
1611        *copynames = 0;
1612        *getnames = 0;
1613    
1614        copynamesptr = copynames;
1615        getnamesptr = getnames;
1616    
1617      pcre_callout = callout;      pcre_callout = callout;
1618      first_callout = 1;      first_callout = 1;
1619      callout_extra = 0;      callout_extra = 0;
# Line 1292  while (!done) Line 1622  while (!done)
1622      callout_fail_id = -1;      callout_fail_id = -1;
1623      show_malloc = 0;      show_malloc = 0;
1624    
1625      if (infile == stdin) printf("data> ");      if (extra != NULL) extra->flags &=
1626      if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)        ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
1627    
1628        len = 0;
1629        for (;;)
1630        {        {
1631        done = 1;        if (infile == stdin) printf("data> ");
1632        goto CONTINUE;        if (extend_inputline(infile, buffer + len) == NULL)
1633            {
1634            if (len > 0) break;
1635            done = 1;
1636            goto CONTINUE;
1637            }
1638          if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1639          len = (int)strlen((char *)buffer);
1640          if (buffer[len-1] == '\n') break;
1641        }        }
     if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);  
1642    
     len = (int)strlen((char *)buffer);  
1643      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
1644      buffer[len] = 0;      buffer[len] = 0;
1645      if (len == 0) break;      if (len == 0) break;
# Line 1330  while (!done) Line 1669  while (!done)
1669          c -= '0';          c -= '0';
1670          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')
1671            c = c * 8 + *p++ - '0';            c = c * 8 + *p++ - '0';
1672    
1673    #if !defined NOUTF8
1674            if (use_utf8 && c > 255)
1675              {
1676              unsigned char buff8[8];
1677              int ii, utn;
1678              utn = ord2utf8(c, buff8);
1679              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1680              c = buff8[ii];   /* Last byte */
1681              }
1682    #endif
1683          break;          break;
1684    
1685          case 'x':          case 'x':
# Line 1391  while (!done) Line 1741  while (!done)
1741            }            }
1742          else if (isalnum(*p))          else if (isalnum(*p))
1743            {            {
1744            uschar name[256];            uschar *npp = copynamesptr;
           uschar *npp = name;  
1745            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1746              *npp++ = 0;
1747            *npp = 0;            *npp = 0;
1748            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)copynamesptr);
1749            if (n < 0)            if (n < 0)
1750              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", copynamesptr);
1751            else copystrings |= 1 << n;            copynamesptr = npp;
1752            }            }
1753          else if (*p == '+')          else if (*p == '+')
1754            {            {
# Line 1459  while (!done) Line 1809  while (!done)
1809            }            }
1810          else if (isalnum(*p))          else if (isalnum(*p))
1811            {            {
1812            uschar name[256];            uschar *npp = getnamesptr;
           uschar *npp = name;  
1813            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1814              *npp++ = 0;
1815            *npp = 0;            *npp = 0;
1816            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)getnamesptr);
1817            if (n < 0)            if (n < 0)
1818              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", getnamesptr);
1819            else getstrings |= 1 << n;            getnamesptr = npp;
1820            }            }
1821          continue;          continue;
1822    
# Line 1505  while (!done) Line 1855  while (!done)
1855          options |= PCRE_PARTIAL;          options |= PCRE_PARTIAL;
1856          continue;          continue;
1857    
1858            case 'Q':
1859            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1860            if (extra == NULL)
1861              {
1862              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1863              extra->flags = 0;
1864              }
1865            extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
1866            extra->match_limit_recursion = n;
1867            continue;
1868    
1869            case 'q':
1870            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1871            if (extra == NULL)
1872              {
1873              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1874              extra->flags = 0;
1875              }
1876            extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
1877            extra->match_limit = n;
1878            continue;
1879    
1880  #if !defined NODFA  #if !defined NODFA
1881          case 'R':          case 'R':
1882          options |= PCRE_DFA_RESTART;          options |= PCRE_DFA_RESTART;
# Line 1522  while (!done) Line 1894  while (!done)
1894          case '?':          case '?':
1895          options |= PCRE_NO_UTF8_CHECK;          options |= PCRE_NO_UTF8_CHECK;
1896          continue;          continue;
1897    
1898            case '<':
1899              {
1900              int x = check_newline(p, outfile);
1901              if (x == 0) goto NEXT_DATA;
1902              options |= x;
1903              while (*p++ != '>');
1904              }
1905            continue;
1906          }          }
1907        *q++ = c;        *q++ = c;
1908        }        }
# Line 1552  while (!done) Line 1933  while (!done)
1933    
1934        if (rc != 0)        if (rc != 0)
1935          {          {
1936          (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);          (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1937          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
1938          }          }
1939          else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
1940                  != 0)
1941            {
1942            fprintf(outfile, "Matched with REG_NOSUB\n");
1943            }
1944        else        else
1945          {          {
1946          size_t i;          size_t i;
# Line 1586  while (!done) Line 1972  while (!done)
1972    
1973      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
1974        {        {
1975        if (timeit)        if (timeitm > 0)
1976          {          {
1977          register int i;          register int i;
1978          clock_t time_taken;          clock_t time_taken;
# Line 1596  while (!done) Line 1982  while (!done)
1982          if (all_use_dfa || use_dfa)          if (all_use_dfa || use_dfa)
1983            {            {
1984            int workspace[1000];            int workspace[1000];
1985            for (i = 0; i < LOOPREPEAT; i++)            for (i = 0; i < timeitm; i++)
1986              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
1987                options | g_notempty, use_offsets, use_size_offsets, workspace,                options | g_notempty, use_offsets, use_size_offsets, workspace,
1988                sizeof(workspace)/sizeof(int));                sizeof(workspace)/sizeof(int));
# Line 1604  while (!done) Line 1990  while (!done)
1990          else          else
1991  #endif  #endif
1992    
1993          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeitm; i++)
1994            count = pcre_exec(re, extra, (char *)bptr, len,            count = pcre_exec(re, extra, (char *)bptr, len,
1995              start_offset, options | g_notempty, use_offsets, use_size_offsets);              start_offset, options | g_notempty, use_offsets, use_size_offsets);
1996    
1997          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1998          fprintf(outfile, "Execute time %.3f milliseconds\n",          fprintf(outfile, "Execute time %.4f milliseconds\n",
1999            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeitm) /
2000              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
2001          }          }
2002    
2003        /* If find_match_limit is set, we want to do repeated matches with        /* If find_match_limit is set, we want to do repeated matches with
2004        varying limits in order to find the minimum value. */        varying limits in order to find the minimum value for the match limit and
2005          for the recursion limit. */
2006    
2007        if (find_match_limit)        if (find_match_limit)
2008          {          {
         int min = 0;  
         int mid = 64;  
         int max = -1;  
   
2009          if (extra == NULL)          if (extra == NULL)
2010            {            {
2011            extra = (pcre_extra *)malloc(sizeof(pcre_extra));            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2012            extra->flags = 0;            extra->flags = 0;
2013            }            }
         extra->flags |= PCRE_EXTRA_MATCH_LIMIT;  
2014    
2015          for (;;)          (void)check_match_limit(re, extra, bptr, len, start_offset,
2016            {            options|g_notempty, use_offsets, use_size_offsets,
2017            extra->match_limit = mid;            PCRE_EXTRA_MATCH_LIMIT, &(extra->match_limit),
2018            count = pcre_exec(re, extra, (char *)bptr, len, start_offset,            PCRE_ERROR_MATCHLIMIT, "match()");
2019              options | g_notempty, use_offsets, use_size_offsets);  
2020            if (count == PCRE_ERROR_MATCHLIMIT)          count = check_match_limit(re, extra, bptr, len, start_offset,
2021              {            options|g_notempty, use_offsets, use_size_offsets,
2022              /* fprintf(outfile, "Testing match limit = %d\n", mid); */            PCRE_EXTRA_MATCH_LIMIT_RECURSION, &(extra->match_limit_recursion),
2023              min = mid;            PCRE_ERROR_RECURSIONLIMIT, "match() recursion");
             mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;  
             }  
           else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||  
                                  count == PCRE_ERROR_PARTIAL)  
             {  
             if (mid == min + 1)  
               {  
               fprintf(outfile, "Minimum match limit = %d\n", mid);  
               break;  
               }  
             /* fprintf(outfile, "Testing match limit = %d\n", mid); */  
             max = mid;  
             mid = (min + mid)/2;  
             }  
           else break;    /* Some other error */  
           }  
   
         extra->flags &= ~PCRE_EXTRA_MATCH_LIMIT;  
2024          }          }
2025    
2026        /* If callout_data is set, use the interface with additional data */        /* If callout_data is set, use the interface with additional data */
# Line 1708  while (!done) Line 2072  while (!done)
2072    
2073        if (count >= 0)        if (count >= 0)
2074          {          {
2075          int i;          int i, maxcount;
2076    
2077    #if !defined NODFA
2078            if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else
2079    #endif
2080              maxcount = use_size_offsets/3;
2081    
2082            /* This is a check against a lunatic return value. */
2083    
2084            if (count > maxcount)
2085              {
2086              fprintf(outfile,
2087                "** PCRE error: returned count %d is too big for offset size %d\n",
2088                count, use_size_offsets);
2089              count = use_size_offsets/3;
2090              if (do_g || do_G)
2091                {
2092                fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G');
2093                do_g = do_G = FALSE;        /* Break g/G loop */
2094                }
2095              }
2096    
2097          for (i = 0; i < count * 2; i += 2)          for (i = 0; i < count * 2; i += 2)
2098            {            {
2099            if (use_offsets[i] < 0)            if (use_offsets[i] < 0)
# Line 1736  while (!done) Line 2121  while (!done)
2121            {            {
2122            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
2123              {              {
2124              char copybuffer[16];              char copybuffer[256];
2125              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,
2126                i, copybuffer, sizeof(copybuffer));                i, copybuffer, sizeof(copybuffer));
2127              if (rc < 0)              if (rc < 0)
# Line 1746  while (!done) Line 2131  while (!done)
2131              }              }
2132            }            }
2133    
2134            for (copynamesptr = copynames;
2135                 *copynamesptr != 0;
2136                 copynamesptr += (int)strlen((char*)copynamesptr) + 1)
2137              {
2138              char copybuffer[256];
2139              int rc = pcre_copy_named_substring(re, (char *)bptr, use_offsets,
2140                count, (char *)copynamesptr, copybuffer, sizeof(copybuffer));
2141              if (rc < 0)
2142                fprintf(outfile, "copy substring %s failed %d\n", copynamesptr, rc);
2143              else
2144                fprintf(outfile, "  C %s (%d) %s\n", copybuffer, rc, copynamesptr);
2145              }
2146    
2147          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2148            {            {
2149            if ((getstrings & (1 << i)) != 0)            if ((getstrings & (1 << i)) != 0)
# Line 1758  while (!done) Line 2156  while (!done)
2156              else              else
2157                {                {
2158                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
               /* free((void *)substring); */  
2159                pcre_free_substring(substring);                pcre_free_substring(substring);
2160                }                }
2161              }              }
2162            }            }
2163    
2164            for (getnamesptr = getnames;
2165                 *getnamesptr != 0;
2166                 getnamesptr += (int)strlen((char*)getnamesptr) + 1)
2167              {
2168              const char *substring;
2169              int rc = pcre_get_named_substring(re, (char *)bptr, use_offsets,
2170                count, (char *)getnamesptr, &substring);
2171              if (rc < 0)
2172                fprintf(outfile, "copy substring %s failed %d\n", getnamesptr, rc);
2173              else
2174                {
2175                fprintf(outfile, "  G %s (%d) %s\n", substring, rc, getnamesptr);
2176                pcre_free_substring(substring);
2177                }
2178              }
2179    
2180          if (getlist)          if (getlist)
2181            {            {
2182            const char **stringlist;            const char **stringlist;
# Line 1798  while (!done) Line 2211  while (!done)
2211          }          }
2212    
2213        /* Failed to match. If this is a /g or /G loop and we previously set        /* Failed to match. If this is a /g or /G loop and we previously set
2214        g_notempty after a null match, this is not necessarily the end.        g_notempty after a null match, this is not necessarily the end. We want
2215        We want to advance the start offset, and continue. In the case of UTF-8        to advance the start offset, and continue. We won't be at the end of the
2216        matching, the advance must be one character, not one byte. Fudge the        string - that was checked before setting g_notempty.
2217        offset values to achieve this. We won't be at the end of the string -  
2218        that was checked before setting g_notempty. */        Complication arises in the case when the newline option is "any".
2219          If the previous match was at the end of a line terminated by CRLF, an
2220          advance of one character just passes the \r, whereas we should prefer the
2221          longer newline sequence, as does the code in pcre_exec(). Fudge the
2222          offset value to achieve this.
2223    
2224          Otherwise, in the case of UTF-8 matching, the advance must be one
2225          character, not one byte. */
2226    
2227        else        else
2228          {          {
# Line 1810  while (!done) Line 2230  while (!done)
2230            {            {
2231            int onechar = 1;            int onechar = 1;
2232            use_offsets[0] = start_offset;            use_offsets[0] = start_offset;
2233            if (use_utf8)            if ((((real_pcre *)re)->options & PCRE_NEWLINE_BITS) ==
2234                      PCRE_NEWLINE_ANY &&
2235                  start_offset < len - 1 &&
2236                  bptr[start_offset] == '\r' &&
2237                  bptr[start_offset+1] == '\n')
2238                onechar++;
2239              else if (use_utf8)
2240              {              {
2241              while (start_offset + onechar < len)              while (start_offset + onechar < len)
2242                {                {
# Line 1845  while (!done) Line 2271  while (!done)
2271        character. */        character. */
2272    
2273        g_notempty = 0;        g_notempty = 0;
2274    
2275        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2276          {          {
2277          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
# Line 1863  while (!done) Line 2290  while (!done)
2290          len -= use_offsets[1];          len -= use_offsets[1];
2291          }          }
2292        }  /* End of loop for /g and /G */        }  /* End of loop for /g and /G */
2293    
2294        NEXT_DATA: continue;
2295      }    /* End of loop for data lines */      }    /* End of loop for data lines */
2296    
2297    CONTINUE:    CONTINUE:
# Line 1877  while (!done) Line 2306  while (!done)
2306      {      {
2307      new_free((void *)tables);      new_free((void *)tables);
2308      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");
2309        locale_set = 0;
2310      }      }
2311    }    }
2312    

Legend:
Removed from v.85  
changed lines
  Added in v.144

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12