/[pcre]/code/branches/pcre16/pcretest.c
ViewVC logotype

Diff of /code/branches/pcre16/pcretest.c

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

revision 79 by nigel, Sat Feb 24 21:40:52 2007 UTC revision 93 by nigel, Sat Feb 24 21:41:42 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 need the internal info for displaying the results of pcre_study() and  /* We include pcre_internal.h because we need the internal info for displaying
73  other internal data; pcretest also uses some of the fixed tables, and generally  the results of pcre_study() and we also need to know about the internal
74  has "inside information" compared to a program that strictly follows the PCRE  macros, structures, and other internal data values; pcretest has "inside
75  API. */  information" compared to a program that strictly follows the PCRE API. */
76    
77  #include "pcre_internal.h"  #include "pcre_internal.h"
78    
79    /* We need access to the data tables that PCRE uses. So as not to have to keep
80    two copies, we include the source file here, changing the names of the external
81    symbols to prevent clashes. */
82    
83    #define _pcre_utf8_table1      utf8_table1
84    #define _pcre_utf8_table1_size utf8_table1_size
85    #define _pcre_utf8_table2      utf8_table2
86    #define _pcre_utf8_table3      utf8_table3
87    #define _pcre_utf8_table4      utf8_table4
88    #define _pcre_utt              utt
89    #define _pcre_utt_size         utt_size
90    #define _pcre_OP_lengths       OP_lengths
91    
92    #include "pcre_tables.c"
93    
94    /* 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
96    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"
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 68  to the DFA matcher (NODFA), and without Line 120  to the DFA matcher (NODFA), and without
120  function (define NOINFOCHECK). */  function (define NOINFOCHECK). */
121    
122    
123    /* Other parameters */
124    
125  #ifndef CLOCKS_PER_SEC  #ifndef CLOCKS_PER_SEC
126  #ifdef CLK_TCK  #ifdef CLK_TCK
127  #define CLOCKS_PER_SEC CLK_TCK  #define CLOCKS_PER_SEC CLK_TCK
# Line 76  function (define NOINFOCHECK). */ Line 130  function (define NOINFOCHECK). */
130  #endif  #endif
131  #endif  #endif
132    
133  #define LOOPREPEAT 500000  /* This is the default loop count for timing. */
134    
135  #define BUFFER_SIZE 30000  #define LOOPREPEAT 500000
 #define PBUFFER_SIZE BUFFER_SIZE  
 #define DBUFFER_SIZE BUFFER_SIZE  
136    
137    /* Static variables */
138    
139  static FILE *outfile;  static FILE *outfile;
140  static int log_store = 0;  static int log_store = 0;
# Line 90  static int callout_extra; Line 143  static int callout_extra;
143  static int callout_fail_count;  static int callout_fail_count;
144  static int callout_fail_id;  static int callout_fail_id;
145  static int first_callout;  static int first_callout;
146    static int locale_set = 0;
147  static int show_malloc;  static int show_malloc;
148  static int use_utf8;  static int use_utf8;
149  static size_t gotten_store;  static size_t gotten_store;
150    
151    /* The buffers grow automatically if very long input lines are encountered. */
152    
153    static int buffer_size = 50000;
154    static uschar *buffer = NULL;
155    static uschar *dbuffer = NULL;
156  static uschar *pbuffer = NULL;  static uschar *pbuffer = NULL;
157    
158    
159    
160  /*************************************************  /*************************************************
161    *        Read or extend an input line            *
162    *************************************************/
163    
164    /* Input lines are read into buffer, but both patterns and data lines can be
165    continued over multiple input lines. In addition, if the buffer fills up, we
166    want to automatically expand it so as to be able to handle extremely large
167    lines that are needed for certain stress tests. When the input buffer is
168    expanded, the other two buffers must also be expanded likewise, and the
169    contents of pbuffer, which are a copy of the input for callouts, must be
170    preserved (for when expansion happens for a data line). This is not the most
171    optimal way of handling this, but hey, this is just a test program!
172    
173    Arguments:
174      f            the file to read
175      start        where in buffer to start (this *must* be within buffer)
176    
177    Returns:       pointer to the start of new data
178                   could be a copy of start, or could be moved
179                   NULL if no data read and EOF reached
180    */
181    
182    static uschar *
183    extend_inputline(FILE *f, uschar *start)
184    {
185    uschar *here = start;
186    
187    for (;;)
188      {
189      int rlen = buffer_size - (here - buffer);
190    
191      if (rlen > 1000)
192        {
193        int dlen;
194        if (fgets((char *)here, rlen,  f) == NULL)
195          return (here == start)? NULL : start;
196        dlen = (int)strlen((char *)here);
197        if (dlen > 0 && here[dlen - 1] == '\n') return start;
198        here += dlen;
199        }
200    
201      else
202        {
203        int new_buffer_size = 2*buffer_size;
204        uschar *new_buffer = (unsigned char *)malloc(new_buffer_size);
205        uschar *new_dbuffer = (unsigned char *)malloc(new_buffer_size);
206        uschar *new_pbuffer = (unsigned char *)malloc(new_buffer_size);
207    
208        if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
209          {
210          fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
211          exit(1);
212          }
213    
214        memcpy(new_buffer, buffer, buffer_size);
215        memcpy(new_pbuffer, pbuffer, buffer_size);
216    
217        buffer_size = new_buffer_size;
218    
219        start = new_buffer + (start - buffer);
220        here = new_buffer + (here - buffer);
221    
222        free(buffer);
223        free(dbuffer);
224        free(pbuffer);
225    
226        buffer = new_buffer;
227        dbuffer = new_dbuffer;
228        pbuffer = new_pbuffer;
229        }
230      }
231    
232    return NULL;  /* Control never gets here */
233    }
234    
235    
236    
237    
238    
239    
240    
241    /*************************************************
242  *          Read number from string               *  *          Read number from string               *
243  *************************************************/  *************************************************/
244    
245  /* 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
246  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
247  for unpicking the -o argument, so just keep it simple.  for unpicking arguments, so just keep it simple.
248    
249  Arguments:  Arguments:
250    str           string to be converted    str           string to be converted
# Line 134  return(result); Line 274  return(result);
274  and returns the value of the character.  and returns the value of the character.
275    
276  Argument:  Argument:
277    buffer   a pointer to the byte vector    utf8bytes   a pointer to the byte vector
278    vptr     a pointer to an int to receive the value    vptr        a pointer to an int to receive the value
279    
280  Returns:   >  0 => the number of bytes consumed  Returns:      >  0 => the number of bytes consumed
281             -6 to 0 => malformed UTF-8 character at offset = (-return)                -6 to 0 => malformed UTF-8 character at offset = (-return)
282  */  */
283    
284  #if !defined NOUTF8  #if !defined NOUTF8
285    
286  static int  static int
287  utf82ord(unsigned char *buffer, int *vptr)  utf82ord(unsigned char *utf8bytes, int *vptr)
288  {  {
289  int c = *buffer++;  int c = *utf8bytes++;
290  int d = c;  int d = c;
291  int i, j, s;  int i, j, s;
292    
# Line 162  if (i == 0 || i == 6) return 0; / Line 302  if (i == 0 || i == 6) return 0; /
302  /* i now has a value in the range 1-5 */  /* i now has a value in the range 1-5 */
303    
304  s = 6*i;  s = 6*i;
305  d = (c & _pcre_utf8_table3[i]) << s;  d = (c & utf8_table3[i]) << s;
306    
307  for (j = 0; j < i; j++)  for (j = 0; j < i; j++)
308    {    {
309    c = *buffer++;    c = *utf8bytes++;
310    if ((c & 0xc0) != 0x80) return -(j+1);    if ((c & 0xc0) != 0x80) return -(j+1);
311    s -= 6;    s -= 6;
312    d |= (c & 0x3f) << s;    d |= (c & 0x3f) << s;
# Line 174  for (j = 0; j < i; j++) Line 314  for (j = 0; j < i; j++)
314    
315  /* Check that encoding was the correct unique one */  /* Check that encoding was the correct unique one */
316    
317  for (j = 0; j < _pcre_utf8_table1_size; j++)  for (j = 0; j < utf8_table1_size; j++)
318    if (d <= _pcre_utf8_table1[j]) break;    if (d <= utf8_table1[j]) break;
319  if (j != i) return -(i+1);  if (j != i) return -(i+1);
320    
321  /* Valid value */  /* Valid value */
# Line 189  return i+1; Line 329  return i+1;
329    
330    
331  /*************************************************  /*************************************************
332    *       Convert character value to UTF-8         *
333    *************************************************/
334    
335    /* This function takes an integer value in the range 0 - 0x7fffffff
336    and encodes it as a UTF-8 character in 0 to 6 bytes.
337    
338    Arguments:
339      cvalue     the character value
340      utf8bytes  pointer to buffer for result - at least 6 bytes long
341    
342    Returns:     number of characters placed in the buffer
343    */
344    
345    #if !defined NOUTF8
346    
347    static int
348    ord2utf8(int cvalue, uschar *utf8bytes)
349    {
350    register int i, j;
351    for (i = 0; i < utf8_table1_size; i++)
352      if (cvalue <= utf8_table1[i]) break;
353    utf8bytes += i;
354    for (j = i; j > 0; j--)
355     {
356     *utf8bytes-- = 0x80 | (cvalue & 0x3f);
357     cvalue >>= 6;
358     }
359    *utf8bytes = utf8_table2[i] | cvalue;
360    return i + 1;
361    }
362    
363    #endif
364    
365    
366    
367    /*************************************************
368  *             Print character string             *  *             Print character string             *
369  *************************************************/  *************************************************/
370    
# Line 198  chars without printing. */ Line 374  chars without printing. */
374    
375  static int pchars(unsigned char *p, int length, FILE *f)  static int pchars(unsigned char *p, int length, FILE *f)
376  {  {
377  int c;  int c = 0;
378  int yield = 0;  int yield = 0;
379    
380  while (length-- > 0)  while (length-- > 0)
# Line 212  while (length-- > 0) Line 388  while (length-- > 0)
388        {        {
389        length -= rc - 1;        length -= rc - 1;
390        p += rc;        p += rc;
391        if (c < 256 && isprint(c))        if (PRINTHEX(c))
392          {          {
393          if (f != NULL) fprintf(f, "%c", c);          if (f != NULL) fprintf(f, "%c", c);
394          yield++;          yield++;
395          }          }
396        else        else
397          {          {
398          int n;          int n = 4;
399          if (f != NULL) fprintf(f, "\\x{%02x}%n", c, &n);          if (f != NULL) fprintf(f, "\\x{%02x}", c);
400          yield += n;          yield += (n <= 0x000000ff)? 2 :
401                     (n <= 0x00000fff)? 3 :
402                     (n <= 0x0000ffff)? 4 :
403                     (n <= 0x000fffff)? 5 : 6;
404          }          }
405        continue;        continue;
406        }        }
# Line 230  while (length-- > 0) Line 409  while (length-- > 0)
409    
410     /* Not UTF-8, or malformed UTF-8  */     /* Not UTF-8, or malformed UTF-8  */
411    
412    if (isprint(c = *(p++)))    c = *p++;
413      if (PRINTHEX(c))
414      {      {
415      if (f != NULL) fprintf(f, "%c", c);      if (f != NULL) fprintf(f, "%c", c);
416      yield++;      yield++;
# Line 404  if ((rc = pcre_fullinfo(re, study, optio Line 584  if ((rc = pcre_fullinfo(re, study, optio
584  *         Byte flipping function                 *  *         Byte flipping function                 *
585  *************************************************/  *************************************************/
586    
587  static long int  static unsigned long int
588  byteflip(long int value, int n)  byteflip(unsigned long int value, int n)
589  {  {
590  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);
591  return ((value & 0x000000ff) << 24) |  return ((value & 0x000000ff) << 24) |
# Line 418  return ((value & 0x000000ff) << 24) | Line 598  return ((value & 0x000000ff) << 24) |
598    
599    
600  /*************************************************  /*************************************************
601    *        Check match or recursion limit          *
602    *************************************************/
603    
604    static int
605    check_match_limit(pcre *re, pcre_extra *extra, uschar *bptr, int len,
606      int start_offset, int options, int *use_offsets, int use_size_offsets,
607      int flag, unsigned long int *limit, int errnumber, const char *msg)
608    {
609    int count;
610    int min = 0;
611    int mid = 64;
612    int max = -1;
613    
614    extra->flags |= flag;
615    
616    for (;;)
617      {
618      *limit = mid;
619    
620      count = pcre_exec(re, extra, (char *)bptr, len, start_offset, options,
621        use_offsets, use_size_offsets);
622    
623      if (count == errnumber)
624        {
625        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
626        min = mid;
627        mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;
628        }
629    
630      else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||
631                             count == PCRE_ERROR_PARTIAL)
632        {
633        if (mid == min + 1)
634          {
635          fprintf(outfile, "Minimum %s limit = %d\n", msg, mid);
636          break;
637          }
638        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
639        max = mid;
640        mid = (min + mid)/2;
641        }
642      else break;    /* Some other error */
643      }
644    
645    extra->flags &= ~flag;
646    return count;
647    }
648    
649    
650    
651    /*************************************************
652    *         Check newline indicator                *
653    *************************************************/
654    
655    /* This is used both at compile and run-time to check for <xxx> escapes, where
656    xxx is LF, CR, CRLF, or ANY. Print a message and return 0 if there is no match.
657    
658    Arguments:
659      p           points after the leading '<'
660      f           file for error message
661    
662    Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
663    */
664    
665    static int
666    check_newline(uschar *p, FILE *f)
667    {
668    if (strncmp((char *)p, "cr>", 3) == 0) return PCRE_NEWLINE_CR;
669    if (strncmp((char *)p, "lf>", 3) == 0) return PCRE_NEWLINE_LF;
670    if (strncmp((char *)p, "crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
671    if (strncmp((char *)p, "any>", 4) == 0) return PCRE_NEWLINE_ANY;
672    fprintf(f, "Unknown newline type at: <%s\n", p);
673    return 0;
674    }
675    
676    
677    
678    /*************************************************
679    *             Usage function                     *
680    *************************************************/
681    
682    static void
683    usage(void)
684    {
685    printf("Usage:     pcretest [options] [<input> [<output>]]\n");
686    printf("  -b       show compiled code (bytecode)\n");
687    printf("  -C       show PCRE compile-time options and exit\n");
688    printf("  -d       debug: show compiled code and information (-b and -i)\n");
689    #if !defined NODFA
690    printf("  -dfa     force DFA matching for all subjects\n");
691    #endif
692    printf("  -help    show usage information\n");
693    printf("  -i       show information about compiled patterns\n"
694           "  -m       output memory used information\n"
695           "  -o <n>   set size of offsets vector to <n>\n");
696    #if !defined NOPOSIX
697    printf("  -p       use POSIX interface\n");
698    #endif
699    printf("  -q       quiet: do not output PCRE version number at start\n");
700    printf("  -S <n>   set stack size to <n> megabytes\n");
701    printf("  -s       output store (memory) used information\n"
702           "  -t       time compilation and execution\n");
703    printf("  -t <n>   time compilation and execution, repeating <n> times\n");
704    printf("  -tm      time execution (matching) only\n");
705    printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
706    }
707    
708    
709    
710    /*************************************************
711  *                Main Program                    *  *                Main Program                    *
712  *************************************************/  *************************************************/
713    
# Line 432  int options = 0; Line 722  int options = 0;
722  int study_options = 0;  int study_options = 0;
723  int op = 1;  int op = 1;
724  int timeit = 0;  int timeit = 0;
725    int timeitm = 0;
726  int showinfo = 0;  int showinfo = 0;
727  int showstore = 0;  int showstore = 0;
728    int quiet = 0;
729  int size_offsets = 45;  int size_offsets = 45;
730  int size_offsets_max;  int size_offsets_max;
731  int *offsets = NULL;  int *offsets = NULL;
# Line 444  int debug = 0; Line 736  int debug = 0;
736  int done = 0;  int done = 0;
737  int all_use_dfa = 0;  int all_use_dfa = 0;
738  int yield = 0;  int yield = 0;
739    int stack_size;
740    
741    /* These vectors store, end-to-end, a list of captured substring names. Assume
742    that 1024 is plenty long enough for the few names we'll be testing. */
743    
744    uschar copynames[1024];
745    uschar getnames[1024];
746    
747  unsigned char *buffer;  uschar *copynamesptr;
748  unsigned char *dbuffer;  uschar *getnamesptr;
749    
750  /* Get buffers from malloc() so that Electric Fence will check their misuse  /* Get buffers from malloc() so that Electric Fence will check their misuse
751  when I am debugging. */  when I am debugging. They grow automatically when very long lines are read. */
752    
753  buffer = (unsigned char *)malloc(BUFFER_SIZE);  buffer = (unsigned char *)malloc(buffer_size);
754  dbuffer = (unsigned char *)malloc(DBUFFER_SIZE);  dbuffer = (unsigned char *)malloc(buffer_size);
755  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.  */  
756    
757  #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) */  
758    
759  outfile = stdout;  outfile = stdout;
760    
761    /* The following  _setmode() stuff is some Windows magic that tells its runtime
762    library to translate CRLF into a single LF character. At least, that's what
763    I've been told: never having used Windows I take this all on trust. Originally
764    it set 0x8000, but then I was advised that _O_BINARY was better. */
765    
766    #if defined(_WIN32) || defined(WIN32)
767    _setmode( _fileno( stdout ), _O_BINARY );
768    #endif
769    
770  /* Scan options */  /* Scan options */
771    
772  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
# Line 473  while (argc > 1 && argv[op][0] == '-') Line 775  while (argc > 1 && argv[op][0] == '-')
775    
776    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
777      showstore = 1;      showstore = 1;
778    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-q") == 0) quiet = 1;
779      else if (strcmp(argv[op], "-b") == 0) debug = 1;
780    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
781    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
782  #if !defined NODFA  #if !defined NODFA
# Line 486  while (argc > 1 && argv[op][0] == '-') Line 789  while (argc > 1 && argv[op][0] == '-')
789      op++;      op++;
790      argc--;      argc--;
791      }      }
792      else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
793        {
794        int both = argv[op][2] == 0;
795        int temp;
796        if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr),
797                         *endptr == 0))
798          {
799          timeitm = temp;
800          op++;
801          argc--;
802          }
803        else timeitm = LOOPREPEAT;
804        if (both) timeit = timeitm;
805        }
806      else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
807          ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
808            *endptr == 0))
809        {
810    #if defined(_WIN32) || defined(WIN32)
811        printf("PCRE: -S not supported on this OS\n");
812        exit(1);
813    #else
814        int rc;
815        struct rlimit rlim;
816        getrlimit(RLIMIT_STACK, &rlim);
817        rlim.rlim_cur = stack_size * 1024 * 1024;
818        rc = setrlimit(RLIMIT_STACK, &rlim);
819        if (rc != 0)
820          {
821        printf("PCRE: setrlimit() failed with error %d\n", rc);
822        exit(1);
823          }
824        op++;
825        argc--;
826    #endif
827        }
828  #if !defined NOPOSIX  #if !defined NOPOSIX
829    else if (strcmp(argv[op], "-p") == 0) posix = 1;    else if (strcmp(argv[op], "-p") == 0) posix = 1;
830  #endif  #endif
# Line 499  while (argc > 1 && argv[op][0] == '-') Line 838  while (argc > 1 && argv[op][0] == '-')
838      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
839      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
840      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
841      printf("  Newline character is %s\n", (rc == '\r')? "CR" : "LF");      printf("  Newline sequence is %s\n", (rc == '\r')? "CR" :
842          (rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" :
843          (rc == -1)? "ANY" : "???");
844      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);
845      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
846      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
847      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
848      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);
849      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %d\n", rc);
850        (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &rc);
851        printf("  Default recursion depth limit = %d\n", rc);
852      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
853      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
854      exit(0);      exit(0);
855      }      }
856      else if (strcmp(argv[op], "-help") == 0 ||
857               strcmp(argv[op], "--help") == 0)
858        {
859        usage();
860        goto EXIT;
861        }
862    else    else
863      {      {
864      printf("** Unknown or malformed option %s\n", argv[op]);      printf("** Unknown or malformed option %s\n", argv[op]);
865      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");  
866      yield = 1;      yield = 1;
867      goto EXIT;      goto EXIT;
868      }      }
# Line 550  if (offsets == NULL) Line 886  if (offsets == NULL)
886    
887  if (argc > 1)  if (argc > 1)
888    {    {
889    infile = fopen(argv[op], "rb");    infile = fopen(argv[op], INPUT_MODE);
890    if (infile == NULL)    if (infile == NULL)
891      {      {
892      printf("** Failed to open %s\n", argv[op]);      printf("** Failed to open %s\n", argv[op]);
# Line 561  if (argc > 1) Line 897  if (argc > 1)
897    
898  if (argc > 2)  if (argc > 2)
899    {    {
900    outfile = fopen(argv[op+1], "wb");    outfile = fopen(argv[op+1], OUTPUT_MODE);
901    if (outfile == NULL)    if (outfile == NULL)
902      {      {
903      printf("** Failed to open %s\n", argv[op+1]);      printf("** Failed to open %s\n", argv[op+1]);
# Line 577  pcre_free = new_free; Line 913  pcre_free = new_free;
913  pcre_stack_malloc = stack_malloc;  pcre_stack_malloc = stack_malloc;
914  pcre_stack_free = stack_free;  pcre_stack_free = stack_free;
915    
916  /* Heading line, then prompt for first regex if stdin */  /* Heading line unless quiet, then prompt for first regex if stdin */
917    
918  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  if (!quiet) fprintf(outfile, "PCRE version %s\n\n", pcre_version());
919    
920  /* Main loop */  /* Main loop */
921    
# Line 606  while (!done) Line 942  while (!done)
942    int do_showinfo = showinfo;    int do_showinfo = showinfo;
943    int do_showrest = 0;    int do_showrest = 0;
944    int do_flip = 0;    int do_flip = 0;
945    int erroroffset, len, delimiter;    int erroroffset, len, delimiter, poffset;
946    
947    use_utf8 = 0;    use_utf8 = 0;
948    
949    if (infile == stdin) printf("  re> ");    if (infile == stdin) printf("  re> ");
950    if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break;    if (extend_inputline(infile, buffer) == NULL) break;
951    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
952    fflush(outfile);    fflush(outfile);
953    
# Line 623  while (!done) Line 959  while (!done)
959    
960    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
961      {      {
962      unsigned long int magic;      unsigned long int magic, get_options;
963      uschar sbuf[8];      uschar sbuf[8];
964      FILE *f;      FILE *f;
965    
# Line 671  while (!done) Line 1007  while (!done)
1007    
1008      /* Need to know if UTF-8 for printing data strings */      /* Need to know if UTF-8 for printing data strings */
1009    
1010      new_info(re, NULL, PCRE_INFO_OPTIONS, &options);      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1011      use_utf8 = (options & PCRE_UTF8) != 0;      use_utf8 = (get_options & PCRE_UTF8) != 0;
1012    
1013      /* Now see if there is any following study data */      /* Now see if there is any following study data */
1014    
# Line 716  while (!done) Line 1052  while (!done)
1052      }      }
1053    
1054    pp = p;    pp = p;
1055      poffset = p - buffer;
1056    
1057    for(;;)    for(;;)
1058      {      {
# Line 726  while (!done) Line 1063  while (!done)
1063        pp++;        pp++;
1064        }        }
1065      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;  
       }  
   
1066      if (infile == stdin) printf("    > ");      if (infile == stdin) printf("    > ");
1067      if (fgets((char *)pp, len, infile) == NULL)      if ((pp = extend_inputline(infile, pp)) == NULL)
1068        {        {
1069        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
1070        done = 1;        done = 1;
# Line 744  while (!done) Line 1073  while (!done)
1073      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
1074      }      }
1075    
1076      /* The buffer may have moved while being extended; reset the start of data
1077      pointer to the correct relative point in the buffer. */
1078    
1079      p = buffer + poffset;
1080    
1081    /* If the first character after the delimiter is backslash, make    /* If the first character after the delimiter is backslash, make
1082    the pattern end with backslash. This is purely to provide a way    the pattern end with backslash. This is purely to provide a way
1083    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 775  while (!done) Line 1109  while (!done)
1109    
1110        case '+': do_showrest = 1; break;        case '+': do_showrest = 1; break;
1111        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
1112          case 'B': do_debug = 1; break;
1113        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
1114        case 'D': do_debug = do_showinfo = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
1115        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
1116        case 'F': do_flip = 1; break;        case 'F': do_flip = 1; break;
1117        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1118        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1119          case 'J': options |= PCRE_DUPNAMES; break;
1120        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1121        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1122    
# Line 796  while (!done) Line 1132  while (!done)
1132    
1133        case 'L':        case 'L':
1134        ppp = pp;        ppp = pp;
1135        /* The '\r' test here is so that it works on Windows */        /* The '\r' test here is so that it works on Windows. */
1136        while (*ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;        /* The '0' test is just in case this is an unterminated line. */
1137          while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
1138        *ppp = 0;        *ppp = 0;
1139        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
1140          {          {
1141          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
1142          goto SKIP_DATA;          goto SKIP_DATA;
1143          }          }
1144          locale_set = 1;
1145        tables = pcre_maketables();        tables = pcre_maketables();
1146        pp = ppp;        pp = ppp;
1147        break;        break;
# Line 815  while (!done) Line 1153  while (!done)
1153        *pp = 0;        *pp = 0;
1154        break;        break;
1155    
1156          case '<':
1157            {
1158            int x = check_newline(pp, outfile);
1159            if (x == 0) goto SKIP_DATA;
1160            options |= x;
1161            while (*pp++ != '>');
1162            }
1163          break;
1164    
1165        case '\r':                      /* So that it works in Windows */        case '\r':                      /* So that it works in Windows */
1166        case '\n':        case '\n':
1167        case ' ':        case ' ':
# Line 839  while (!done) Line 1186  while (!done)
1186      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
1187      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
1188      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1189        if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1190        if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1191    
1192      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1193    
1194      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
# Line 846  while (!done) Line 1196  while (!done)
1196    
1197      if (rc != 0)      if (rc != 0)
1198        {        {
1199        (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);        (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1200        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
1201        goto SKIP_DATA;        goto SKIP_DATA;
1202        }        }
# Line 858  while (!done) Line 1208  while (!done)
1208  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1209    
1210      {      {
1211      if (timeit)      if (timeit > 0)
1212        {        {
1213        register int i;        register int i;
1214        clock_t time_taken;        clock_t time_taken;
1215        clock_t start_time = clock();        clock_t start_time = clock();
1216        for (i = 0; i < LOOPREPEAT; i++)        for (i = 0; i < timeit; i++)
1217          {          {
1218          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1219          if (re != NULL) free(re);          if (re != NULL) free(re);
1220          }          }
1221        time_taken = clock() - start_time;        time_taken = clock() - start_time;
1222        fprintf(outfile, "Compile time %.3f milliseconds\n",        fprintf(outfile, "Compile time %.4f milliseconds\n",
1223          (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /          (((double)time_taken * 1000.0) / (double)timeit) /
1224            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
1225        }        }
1226    
# Line 887  while (!done) Line 1237  while (!done)
1237          {          {
1238          for (;;)          for (;;)
1239            {            {
1240            if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)            if (extend_inputline(infile, buffer) == NULL)
1241              {              {
1242              done = 1;              done = 1;
1243              goto CONTINUE;              goto CONTINUE;
# Line 922  while (!done) Line 1272  while (!done)
1272    
1273      if (do_study)      if (do_study)
1274        {        {
1275        if (timeit)        if (timeit > 0)
1276          {          {
1277          register int i;          register int i;
1278          clock_t time_taken;          clock_t time_taken;
1279          clock_t start_time = clock();          clock_t start_time = clock();
1280          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeit; i++)
1281            extra = pcre_study(re, study_options, &error);            extra = pcre_study(re, study_options, &error);
1282          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1283          if (extra != NULL) free(extra);          if (extra != NULL) free(extra);
1284          fprintf(outfile, "  Study time %.3f milliseconds\n",          fprintf(outfile, "  Study time %.4f milliseconds\n",
1285            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeit) /
1286              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
1287          }          }
1288        extra = pcre_study(re, study_options, &error);        extra = pcre_study(re, study_options, &error);
# Line 975  while (!done) Line 1325  while (!done)
1325    
1326      SHOW_INFO:      SHOW_INFO:
1327    
1328        if (do_debug)
1329          {
1330          fprintf(outfile, "------------------------------------------------------------------\n");
1331          pcre_printint(re, outfile);
1332          }
1333    
1334      if (do_showinfo)      if (do_showinfo)
1335        {        {
1336        unsigned long int get_options, all_options;        unsigned long int get_options, all_options;
# Line 985  while (!done) Line 1341  while (!done)
1341        int nameentrysize, namecount;        int nameentrysize, namecount;
1342        const uschar *nametable;        const uschar *nametable;
1343    
       if (do_debug)  
         {  
         fprintf(outfile, "------------------------------------------------------------------\n");  
         _pcre_printint(re, outfile);  
         }  
   
1344        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1345        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1346        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
# Line 1048  while (!done) Line 1398  while (!done)
1398        if (do_flip)        if (do_flip)
1399          {          {
1400          all_options = byteflip(all_options, sizeof(all_options));          all_options = byteflip(all_options, sizeof(all_options));
1401          }           }
1402    
1403        if ((all_options & PCRE_NOPARTIAL) != 0)        if ((all_options & PCRE_NOPARTIAL) != 0)
1404          fprintf(outfile, "Partial matching not supported\n");          fprintf(outfile, "Partial matching not supported\n");
1405    
1406        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1407          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",
1408            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1409            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1410            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
# Line 1064  while (!done) Line 1414  while (!done)
1414            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
1415            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
1416            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1417              ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1418            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1419            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "");            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1420              ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1421    
1422        if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0)        switch (get_options & PCRE_NEWLINE_BITS)
1423          fprintf(outfile, "Case state changes\n");          {
1424            case PCRE_NEWLINE_CR:
1425            fprintf(outfile, "Forced newline sequence: CR\n");
1426            break;
1427    
1428            case PCRE_NEWLINE_LF:
1429            fprintf(outfile, "Forced newline sequence: LF\n");
1430            break;
1431    
1432            case PCRE_NEWLINE_CRLF:
1433            fprintf(outfile, "Forced newline sequence: CRLF\n");
1434            break;
1435    
1436            case PCRE_NEWLINE_ANY:
1437            fprintf(outfile, "Forced newline sequence: ANY\n");
1438            break;
1439    
1440            default:
1441            break;
1442            }
1443    
1444        if (first_char == -1)        if (first_char == -1)
1445          {          {
1446          fprintf(outfile, "First char at start or follows \\n\n");          fprintf(outfile, "First char at start or follows newline\n");
1447          }          }
1448        else if (first_char < 0)        else if (first_char < 0)
1449          {          {
# Line 1083  while (!done) Line 1454  while (!done)
1454          int ch = first_char & 255;          int ch = first_char & 255;
1455          const char *caseless = ((first_char & REQ_CASELESS) == 0)?          const char *caseless = ((first_char & REQ_CASELESS) == 0)?
1456            "" : " (caseless)";            "" : " (caseless)";
1457          if (isprint(ch))          if (PRINTHEX(ch))
1458            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);
1459          else          else
1460            fprintf(outfile, "First char = %d%s\n", ch, caseless);            fprintf(outfile, "First char = %d%s\n", ch, caseless);
# Line 1098  while (!done) Line 1469  while (!done)
1469          int ch = need_char & 255;          int ch = need_char & 255;
1470          const char *caseless = ((need_char & REQ_CASELESS) == 0)?          const char *caseless = ((need_char & REQ_CASELESS) == 0)?
1471            "" : " (caseless)";            "" : " (caseless)";
1472          if (isprint(ch))          if (PRINTHEX(ch))
1473            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);
1474          else          else
1475            fprintf(outfile, "Need char = %d%s\n", ch, caseless);            fprintf(outfile, "Need char = %d%s\n", ch, caseless);
# Line 1134  while (!done) Line 1505  while (!done)
1505                    fprintf(outfile, "\n  ");                    fprintf(outfile, "\n  ");
1506                    c = 2;                    c = 2;
1507                    }                    }
1508                  if (isprint(i) && i != ' ')                  if (PRINTHEX(i) && i != ' ')
1509                    {                    {
1510                    fprintf(outfile, "%c ", i);                    fprintf(outfile, "%c ", i);
1511                    c += 2;                    c += 2;
# Line 1193  while (!done) Line 1564  while (!done)
1564                  strerror(errno));                  strerror(errno));
1565                }                }
1566              else fprintf(outfile, "Study data written to %s\n", to_file);              else fprintf(outfile, "Study data written to %s\n", to_file);
1567    
1568              }              }
1569            }            }
1570          fclose(f);          fclose(f);
# Line 1209  while (!done) Line 1581  while (!done)
1581    
1582    for (;;)    for (;;)
1583      {      {
1584      unsigned char *q;      uschar *q;
1585      unsigned char *bptr = dbuffer;      uschar *bptr = dbuffer;
1586      int *use_offsets = offsets;      int *use_offsets = offsets;
1587      int use_size_offsets = size_offsets;      int use_size_offsets = size_offsets;
1588      int callout_data = 0;      int callout_data = 0;
# Line 1227  while (!done) Line 1599  while (!done)
1599    
1600      options = 0;      options = 0;
1601    
1602        *copynames = 0;
1603        *getnames = 0;
1604    
1605        copynamesptr = copynames;
1606        getnamesptr = getnames;
1607    
1608      pcre_callout = callout;      pcre_callout = callout;
1609      first_callout = 1;      first_callout = 1;
1610      callout_extra = 0;      callout_extra = 0;
# Line 1235  while (!done) Line 1613  while (!done)
1613      callout_fail_id = -1;      callout_fail_id = -1;
1614      show_malloc = 0;      show_malloc = 0;
1615    
1616      if (infile == stdin) printf("data> ");      if (extra != NULL) extra->flags &=
1617      if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)        ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
1618    
1619        len = 0;
1620        for (;;)
1621        {        {
1622        done = 1;        if (infile == stdin) printf("data> ");
1623        goto CONTINUE;        if (extend_inputline(infile, buffer + len) == NULL)
1624            {
1625            if (len > 0) break;
1626            done = 1;
1627            goto CONTINUE;
1628            }
1629          if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1630          len = (int)strlen((char *)buffer);
1631          if (buffer[len-1] == '\n') break;
1632        }        }
     if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);  
1633    
     len = (int)strlen((char *)buffer);  
1634      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
1635      buffer[len] = 0;      buffer[len] = 0;
1636      if (len == 0) break;      if (len == 0) break;
# Line 1273  while (!done) Line 1660  while (!done)
1660          c -= '0';          c -= '0';
1661          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')
1662            c = c * 8 + *p++ - '0';            c = c * 8 + *p++ - '0';
1663    
1664    #if !defined NOUTF8
1665            if (use_utf8 && c > 255)
1666              {
1667              unsigned char buff8[8];
1668              int ii, utn;
1669              utn = ord2utf8(c, buff8);
1670              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1671              c = buff8[ii];   /* Last byte */
1672              }
1673    #endif
1674          break;          break;
1675    
1676          case 'x':          case 'x':
# Line 1290  while (!done) Line 1688  while (!done)
1688              {              {
1689              unsigned char buff8[8];              unsigned char buff8[8];
1690              int ii, utn;              int ii, utn;
1691              utn = _pcre_ord2utf8(c, buff8);              utn = ord2utf8(c, buff8);
1692              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1693              c = buff8[ii];   /* Last byte */              c = buff8[ii];   /* Last byte */
1694              p = pt + 1;              p = pt + 1;
# Line 1334  while (!done) Line 1732  while (!done)
1732            }            }
1733          else if (isalnum(*p))          else if (isalnum(*p))
1734            {            {
1735            uschar name[256];            uschar *npp = copynamesptr;
           uschar *npp = name;  
1736            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1737              *npp++ = 0;
1738            *npp = 0;            *npp = 0;
1739            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)copynamesptr);
1740            if (n < 0)            if (n < 0)
1741              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", copynamesptr);
1742            else copystrings |= 1 << n;            copynamesptr = npp;
1743            }            }
1744          else if (*p == '+')          else if (*p == '+')
1745            {            {
# Line 1402  while (!done) Line 1800  while (!done)
1800            }            }
1801          else if (isalnum(*p))          else if (isalnum(*p))
1802            {            {
1803            uschar name[256];            uschar *npp = getnamesptr;
           uschar *npp = name;  
1804            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1805              *npp++ = 0;
1806            *npp = 0;            *npp = 0;
1807            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)getnamesptr);
1808            if (n < 0)            if (n < 0)
1809              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", getnamesptr);
1810            else getstrings |= 1 << n;            getnamesptr = npp;
1811            }            }
1812          continue;          continue;
1813    
# Line 1448  while (!done) Line 1846  while (!done)
1846          options |= PCRE_PARTIAL;          options |= PCRE_PARTIAL;
1847          continue;          continue;
1848    
1849            case 'Q':
1850            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1851            if (extra == NULL)
1852              {
1853              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1854              extra->flags = 0;
1855              }
1856            extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
1857            extra->match_limit_recursion = n;
1858            continue;
1859    
1860            case 'q':
1861            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1862            if (extra == NULL)
1863              {
1864              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1865              extra->flags = 0;
1866              }
1867            extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
1868            extra->match_limit = n;
1869            continue;
1870    
1871  #if !defined NODFA  #if !defined NODFA
1872          case 'R':          case 'R':
1873          options |= PCRE_DFA_RESTART;          options |= PCRE_DFA_RESTART;
# Line 1465  while (!done) Line 1885  while (!done)
1885          case '?':          case '?':
1886          options |= PCRE_NO_UTF8_CHECK;          options |= PCRE_NO_UTF8_CHECK;
1887          continue;          continue;
1888    
1889            case '<':
1890              {
1891              int x = check_newline(p, outfile);
1892              if (x == 0) goto NEXT_DATA;
1893              options |= x;
1894              while (*p++ != '>');
1895              }
1896            continue;
1897          }          }
1898        *q++ = c;        *q++ = c;
1899        }        }
# Line 1495  while (!done) Line 1924  while (!done)
1924    
1925        if (rc != 0)        if (rc != 0)
1926          {          {
1927          (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);          (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1928          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
1929          }          }
1930          else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
1931                  != 0)
1932            {
1933            fprintf(outfile, "Matched with REG_NOSUB\n");
1934            }
1935        else        else
1936          {          {
1937          size_t i;          size_t i;
# Line 1529  while (!done) Line 1963  while (!done)
1963    
1964      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
1965        {        {
1966        if (timeit)        if (timeitm > 0)
1967          {          {
1968          register int i;          register int i;
1969          clock_t time_taken;          clock_t time_taken;
# Line 1539  while (!done) Line 1973  while (!done)
1973          if (all_use_dfa || use_dfa)          if (all_use_dfa || use_dfa)
1974            {            {
1975            int workspace[1000];            int workspace[1000];
1976            for (i = 0; i < LOOPREPEAT; i++)            for (i = 0; i < timeitm; i++)
1977              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
1978                options | g_notempty, use_offsets, use_size_offsets, workspace,                options | g_notempty, use_offsets, use_size_offsets, workspace,
1979                sizeof(workspace)/sizeof(int));                sizeof(workspace)/sizeof(int));
# Line 1547  while (!done) Line 1981  while (!done)
1981          else          else
1982  #endif  #endif
1983    
1984          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeitm; i++)
1985            count = pcre_exec(re, extra, (char *)bptr, len,            count = pcre_exec(re, extra, (char *)bptr, len,
1986              start_offset, options | g_notempty, use_offsets, use_size_offsets);              start_offset, options | g_notempty, use_offsets, use_size_offsets);
1987    
1988          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1989          fprintf(outfile, "Execute time %.3f milliseconds\n",          fprintf(outfile, "Execute time %.4f milliseconds\n",
1990            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeitm) /
1991              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
1992          }          }
1993    
1994        /* 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
1995        varying limits in order to find the minimum value. */        varying limits in order to find the minimum value for the match limit and
1996          for the recursion limit. */
1997    
1998        if (find_match_limit)        if (find_match_limit)
1999          {          {
         int min = 0;  
         int mid = 64;  
         int max = -1;  
   
2000          if (extra == NULL)          if (extra == NULL)
2001            {            {
2002            extra = (pcre_extra *)malloc(sizeof(pcre_extra));            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2003            extra->flags = 0;            extra->flags = 0;
2004            }            }
         extra->flags |= PCRE_EXTRA_MATCH_LIMIT;  
2005    
2006          for (;;)          (void)check_match_limit(re, extra, bptr, len, start_offset,
2007            {            options|g_notempty, use_offsets, use_size_offsets,
2008            extra->match_limit = mid;            PCRE_EXTRA_MATCH_LIMIT, &(extra->match_limit),
2009            count = pcre_exec(re, extra, (char *)bptr, len, start_offset,            PCRE_ERROR_MATCHLIMIT, "match()");
2010              options | g_notempty, use_offsets, use_size_offsets);  
2011            if (count == PCRE_ERROR_MATCHLIMIT)          count = check_match_limit(re, extra, bptr, len, start_offset,
2012              {            options|g_notempty, use_offsets, use_size_offsets,
2013              /* fprintf(outfile, "Testing match limit = %d\n", mid); */            PCRE_EXTRA_MATCH_LIMIT_RECURSION, &(extra->match_limit_recursion),
2014              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;  
2015          }          }
2016    
2017        /* If callout_data is set, use the interface with additional data */        /* If callout_data is set, use the interface with additional data */
# Line 1651  while (!done) Line 2063  while (!done)
2063    
2064        if (count >= 0)        if (count >= 0)
2065          {          {
2066          int i;          int i, maxcount;
2067    
2068    #if !defined NODFA
2069            if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else
2070    #endif
2071              maxcount = use_size_offsets/3;
2072    
2073            /* This is a check against a lunatic return value. */
2074    
2075            if (count > maxcount)
2076              {
2077              fprintf(outfile,
2078                "** PCRE error: returned count %d is too big for offset size %d\n",
2079                count, use_size_offsets);
2080              count = use_size_offsets/3;
2081              if (do_g || do_G)
2082                {
2083                fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G');
2084                do_g = do_G = FALSE;        /* Break g/G loop */
2085                }
2086              }
2087    
2088          for (i = 0; i < count * 2; i += 2)          for (i = 0; i < count * 2; i += 2)
2089            {            {
2090            if (use_offsets[i] < 0)            if (use_offsets[i] < 0)
# Line 1679  while (!done) Line 2112  while (!done)
2112            {            {
2113            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
2114              {              {
2115              char copybuffer[16];              char copybuffer[256];
2116              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,
2117                i, copybuffer, sizeof(copybuffer));                i, copybuffer, sizeof(copybuffer));
2118              if (rc < 0)              if (rc < 0)
# Line 1689  while (!done) Line 2122  while (!done)
2122              }              }
2123            }            }
2124    
2125            for (copynamesptr = copynames;
2126                 *copynamesptr != 0;
2127                 copynamesptr += (int)strlen((char*)copynamesptr) + 1)
2128              {
2129              char copybuffer[256];
2130              int rc = pcre_copy_named_substring(re, (char *)bptr, use_offsets,
2131                count, (char *)copynamesptr, copybuffer, sizeof(copybuffer));
2132              if (rc < 0)
2133                fprintf(outfile, "copy substring %s failed %d\n", copynamesptr, rc);
2134              else
2135                fprintf(outfile, "  C %s (%d) %s\n", copybuffer, rc, copynamesptr);
2136              }
2137    
2138          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2139            {            {
2140            if ((getstrings & (1 << i)) != 0)            if ((getstrings & (1 << i)) != 0)
# Line 1701  while (!done) Line 2147  while (!done)
2147              else              else
2148                {                {
2149                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
               /* free((void *)substring); */  
2150                pcre_free_substring(substring);                pcre_free_substring(substring);
2151                }                }
2152              }              }
2153            }            }
2154    
2155            for (getnamesptr = getnames;
2156                 *getnamesptr != 0;
2157                 getnamesptr += (int)strlen((char*)getnamesptr) + 1)
2158              {
2159              const char *substring;
2160              int rc = pcre_get_named_substring(re, (char *)bptr, use_offsets,
2161                count, (char *)getnamesptr, &substring);
2162              if (rc < 0)
2163                fprintf(outfile, "copy substring %s failed %d\n", getnamesptr, rc);
2164              else
2165                {
2166                fprintf(outfile, "  G %s (%d) %s\n", substring, rc, getnamesptr);
2167                pcre_free_substring(substring);
2168                }
2169              }
2170    
2171          if (getlist)          if (getlist)
2172            {            {
2173            const char **stringlist;            const char **stringlist;
# Line 1806  while (!done) Line 2267  while (!done)
2267          len -= use_offsets[1];          len -= use_offsets[1];
2268          }          }
2269        }  /* End of loop for /g and /G */        }  /* End of loop for /g and /G */
2270    
2271        NEXT_DATA: continue;
2272      }    /* End of loop for data lines */      }    /* End of loop for data lines */
2273    
2274    CONTINUE:    CONTINUE:
# Line 1820  while (!done) Line 2283  while (!done)
2283      {      {
2284      new_free((void *)tables);      new_free((void *)tables);
2285      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");
2286        locale_set = 0;
2287      }      }
2288    }    }
2289    

Legend:
Removed from v.79  
changed lines
  Added in v.93

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12