/[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 79 by nigel, Sat Feb 24 21:40:52 2007 UTC revision 516 by ph10, Tue May 4 15:51:35 2010 UTC
# Line 36  POSSIBILITY OF SUCH DAMAGE. Line 36  POSSIBILITY OF SUCH DAMAGE.
36  */  */
37    
38    
39    #ifdef HAVE_CONFIG_H
40    #include "config.h"
41    #endif
42    
43  #include <ctype.h>  #include <ctype.h>
44  #include <stdio.h>  #include <stdio.h>
45  #include <string.h>  #include <string.h>
# Line 44  POSSIBILITY OF SUCH DAMAGE. Line 48  POSSIBILITY OF SUCH DAMAGE.
48  #include <locale.h>  #include <locale.h>
49  #include <errno.h>  #include <errno.h>
50    
51  #define PCRE_SPY        /* For Win32 build, import data, not export */  #ifdef SUPPORT_LIBREADLINE
52    #ifdef HAVE_UNISTD_H
53    #include <unistd.h>
54    #endif
55    #include <readline/readline.h>
56    #include <readline/history.h>
57    #endif
58    
59    
60    /* A number of things vary for Windows builds. Originally, pcretest opened its
61    input and output without "b"; then I was told that "b" was needed in some
62    environments, so it was added for release 5.0 to both the input and output. (It
63    makes no difference on Unix-like systems.) Later I was told that it is wrong
64    for the input on Windows. I've now abstracted the modes into two macros that
65    are set here, to make it easier to fiddle with them, and removed "b" from the
66    input mode under Windows. */
67    
68    #if defined(_WIN32) || defined(WIN32)
69    #include <io.h>                /* For _setmode() */
70    #include <fcntl.h>             /* For _O_BINARY */
71    #define INPUT_MODE   "r"
72    #define OUTPUT_MODE  "wb"
73    
74    #ifndef isatty
75    #define isatty _isatty         /* This is what Windows calls them, I'm told, */
76    #endif                         /* though in some environments they seem to   */
77                                   /* be already defined, hence the #ifndefs.    */
78    #ifndef fileno
79    #define fileno _fileno
80    #endif
81    
82    #else
83    #include <sys/time.h>          /* These two includes are needed */
84    #include <sys/resource.h>      /* for setrlimit(). */
85    #define INPUT_MODE   "rb"
86    #define OUTPUT_MODE  "wb"
87    #endif
88    
 /* We need the internal info for displaying the results of pcre_study() and  
 other internal data; pcretest also uses some of the fixed tables, and generally  
 has "inside information" compared to a program that strictly follows the PCRE  
 API. */  
89    
90    /* We have to include pcre_internal.h because we need the internal info for
91    displaying the results of pcre_study() and we also need to know about the
92    internal macros, structures, and other internal data values; pcretest has
93    "inside information" compared to a program that strictly follows the PCRE API.
94    
95    Although pcre_internal.h does itself include pcre.h, we explicitly include it
96    here before pcre_internal.h so that the PCRE_EXP_xxx macros get set
97    appropriately for an application, not for building PCRE. */
98    
99    #include "pcre.h"
100  #include "pcre_internal.h"  #include "pcre_internal.h"
101    
102    /* We need access to some of the data tables that PCRE uses. So as not to have
103    to keep two copies, we include the source file here, changing the names of the
104    external symbols to prevent clashes. */
105    
106    #define _pcre_ucp_gentype      ucp_gentype
107    #define _pcre_utf8_table1      utf8_table1
108    #define _pcre_utf8_table1_size utf8_table1_size
109    #define _pcre_utf8_table2      utf8_table2
110    #define _pcre_utf8_table3      utf8_table3
111    #define _pcre_utf8_table4      utf8_table4
112    #define _pcre_utt              utt
113    #define _pcre_utt_size         utt_size
114    #define _pcre_utt_names        utt_names
115    #define _pcre_OP_lengths       OP_lengths
116    
117    #include "pcre_tables.c"
118    
119    /* We also need the pcre_printint() function for printing out compiled
120    patterns. This function is in a separate file so that it can be included in
121    pcre_compile.c when that module is compiled with debugging enabled. It needs to
122    know which case is being compiled. */
123    
124    #define COMPILING_PCRETEST
125    #include "pcre_printint.src"
126    
127    /* The definition of the macro PRINTABLE, which determines whether to print an
128    output character as-is or as a hex value when showing compiled patterns, is
129    contained in the printint.src file. We uses it here also, in cases when the
130    locale has not been explicitly changed, so as to get consistent output from
131    systems that differ in their output from isprint() even in the "C" locale. */
132    
133    #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
134    
135  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
136  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 62  Makefile. */ Line 140  Makefile. */
140  #include "pcreposix.h"  #include "pcreposix.h"
141  #endif  #endif
142    
143  /* 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
144  build pcretest without support for UTF8 (define NOUTF8), without the interface  Exim, to build pcretest without support for UTF8 (define NOUTF8), without the
145  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
146  function (define NOINFOCHECK). */  "info" function (define NOINFOCHECK). In fact, we automatically cut out the
147    UTF8 support if PCRE is built without it. */
148    
149    #ifndef SUPPORT_UTF8
150    #ifndef NOUTF8
151    #define NOUTF8
152    #endif
153    #endif
154    
155    
156    /* Other parameters */
157    
158  #ifndef CLOCKS_PER_SEC  #ifndef CLOCKS_PER_SEC
159  #ifdef CLK_TCK  #ifdef CLK_TCK
# Line 76  function (define NOINFOCHECK). */ Line 163  function (define NOINFOCHECK). */
163  #endif  #endif
164  #endif  #endif
165    
166  #define LOOPREPEAT 500000  /* This is the default loop count for timing. */
167    
168  #define BUFFER_SIZE 30000  #define LOOPREPEAT 500000
 #define PBUFFER_SIZE BUFFER_SIZE  
 #define DBUFFER_SIZE BUFFER_SIZE  
169    
170    /* Static variables */
171    
172  static FILE *outfile;  static FILE *outfile;
173  static int log_store = 0;  static int log_store = 0;
# Line 89  static int callout_count; Line 175  static int callout_count;
175  static int callout_extra;  static int callout_extra;
176  static int callout_fail_count;  static int callout_fail_count;
177  static int callout_fail_id;  static int callout_fail_id;
178    static int debug_lengths;
179  static int first_callout;  static int first_callout;
180    static int locale_set = 0;
181  static int show_malloc;  static int show_malloc;
182  static int use_utf8;  static int use_utf8;
183  static size_t gotten_store;  static size_t gotten_store;
184    
185    /* The buffers grow automatically if very long input lines are encountered. */
186    
187    static int buffer_size = 50000;
188    static uschar *buffer = NULL;
189    static uschar *dbuffer = NULL;
190  static uschar *pbuffer = NULL;  static uschar *pbuffer = NULL;
191    
192    
193    
194  /*************************************************  /*************************************************
195    *        Read or extend an input line            *
196    *************************************************/
197    
198    /* Input lines are read into buffer, but both patterns and data lines can be
199    continued over multiple input lines. In addition, if the buffer fills up, we
200    want to automatically expand it so as to be able to handle extremely large
201    lines that are needed for certain stress tests. When the input buffer is
202    expanded, the other two buffers must also be expanded likewise, and the
203    contents of pbuffer, which are a copy of the input for callouts, must be
204    preserved (for when expansion happens for a data line). This is not the most
205    optimal way of handling this, but hey, this is just a test program!
206    
207    Arguments:
208      f            the file to read
209      start        where in buffer to start (this *must* be within buffer)
210      prompt       for stdin or readline()
211    
212    Returns:       pointer to the start of new data
213                   could be a copy of start, or could be moved
214                   NULL if no data read and EOF reached
215    */
216    
217    static uschar *
218    extend_inputline(FILE *f, uschar *start, const char *prompt)
219    {
220    uschar *here = start;
221    
222    for (;;)
223      {
224      int rlen = buffer_size - (here - buffer);
225    
226      if (rlen > 1000)
227        {
228        int dlen;
229    
230        /* If libreadline support is required, use readline() to read a line if the
231        input is a terminal. Note that readline() removes the trailing newline, so
232        we must put it back again, to be compatible with fgets(). */
233    
234    #ifdef SUPPORT_LIBREADLINE
235        if (isatty(fileno(f)))
236          {
237          size_t len;
238          char *s = readline(prompt);
239          if (s == NULL) return (here == start)? NULL : start;
240          len = strlen(s);
241          if (len > 0) add_history(s);
242          if (len > rlen - 1) len = rlen - 1;
243          memcpy(here, s, len);
244          here[len] = '\n';
245          here[len+1] = 0;
246          free(s);
247          }
248        else
249    #endif
250    
251        /* Read the next line by normal means, prompting if the file is stdin. */
252    
253          {
254          if (f == stdin) printf("%s", prompt);
255          if (fgets((char *)here, rlen,  f) == NULL)
256            return (here == start)? NULL : start;
257          }
258    
259        dlen = (int)strlen((char *)here);
260        if (dlen > 0 && here[dlen - 1] == '\n') return start;
261        here += dlen;
262        }
263    
264      else
265        {
266        int new_buffer_size = 2*buffer_size;
267        uschar *new_buffer = (unsigned char *)malloc(new_buffer_size);
268        uschar *new_dbuffer = (unsigned char *)malloc(new_buffer_size);
269        uschar *new_pbuffer = (unsigned char *)malloc(new_buffer_size);
270    
271        if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
272          {
273          fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
274          exit(1);
275          }
276    
277        memcpy(new_buffer, buffer, buffer_size);
278        memcpy(new_pbuffer, pbuffer, buffer_size);
279    
280        buffer_size = new_buffer_size;
281    
282        start = new_buffer + (start - buffer);
283        here = new_buffer + (here - buffer);
284    
285        free(buffer);
286        free(dbuffer);
287        free(pbuffer);
288    
289        buffer = new_buffer;
290        dbuffer = new_dbuffer;
291        pbuffer = new_pbuffer;
292        }
293      }
294    
295    return NULL;  /* Control never gets here */
296    }
297    
298    
299    
300    
301    
302    
303    
304    /*************************************************
305  *          Read number from string               *  *          Read number from string               *
306  *************************************************/  *************************************************/
307    
308  /* 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
309  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
310  for unpicking the -o argument, so just keep it simple.  for unpicking arguments, so just keep it simple.
311    
312  Arguments:  Arguments:
313    str           string to be converted    str           string to be converted
# Line 134  return(result); Line 337  return(result);
337  and returns the value of the character.  and returns the value of the character.
338    
339  Argument:  Argument:
340    buffer   a pointer to the byte vector    utf8bytes   a pointer to the byte vector
341    vptr     a pointer to an int to receive the value    vptr        a pointer to an int to receive the value
342    
343  Returns:   >  0 => the number of bytes consumed  Returns:      >  0 => the number of bytes consumed
344             -6 to 0 => malformed UTF-8 character at offset = (-return)                -6 to 0 => malformed UTF-8 character at offset = (-return)
345  */  */
346    
347  #if !defined NOUTF8  #if !defined NOUTF8
348    
349  static int  static int
350  utf82ord(unsigned char *buffer, int *vptr)  utf82ord(unsigned char *utf8bytes, int *vptr)
351  {  {
352  int c = *buffer++;  int c = *utf8bytes++;
353  int d = c;  int d = c;
354  int i, j, s;  int i, j, s;
355    
# Line 162  if (i == 0 || i == 6) return 0;        / Line 365  if (i == 0 || i == 6) return 0;        /
365  /* i now has a value in the range 1-5 */  /* i now has a value in the range 1-5 */
366    
367  s = 6*i;  s = 6*i;
368  d = (c & _pcre_utf8_table3[i]) << s;  d = (c & utf8_table3[i]) << s;
369    
370  for (j = 0; j < i; j++)  for (j = 0; j < i; j++)
371    {    {
372    c = *buffer++;    c = *utf8bytes++;
373    if ((c & 0xc0) != 0x80) return -(j+1);    if ((c & 0xc0) != 0x80) return -(j+1);
374    s -= 6;    s -= 6;
375    d |= (c & 0x3f) << s;    d |= (c & 0x3f) << s;
# Line 174  for (j = 0; j < i; j++) Line 377  for (j = 0; j < i; j++)
377    
378  /* Check that encoding was the correct unique one */  /* Check that encoding was the correct unique one */
379    
380  for (j = 0; j < _pcre_utf8_table1_size; j++)  for (j = 0; j < utf8_table1_size; j++)
381    if (d <= _pcre_utf8_table1[j]) break;    if (d <= utf8_table1[j]) break;
382  if (j != i) return -(i+1);  if (j != i) return -(i+1);
383    
384  /* Valid value */  /* Valid value */
# Line 189  return i+1; Line 392  return i+1;
392    
393    
394  /*************************************************  /*************************************************
395    *       Convert character value to UTF-8         *
396    *************************************************/
397    
398    /* This function takes an integer value in the range 0 - 0x7fffffff
399    and encodes it as a UTF-8 character in 0 to 6 bytes.
400    
401    Arguments:
402      cvalue     the character value
403      utf8bytes  pointer to buffer for result - at least 6 bytes long
404    
405    Returns:     number of characters placed in the buffer
406    */
407    
408    #if !defined NOUTF8
409    
410    static int
411    ord2utf8(int cvalue, uschar *utf8bytes)
412    {
413    register int i, j;
414    for (i = 0; i < utf8_table1_size; i++)
415      if (cvalue <= utf8_table1[i]) break;
416    utf8bytes += i;
417    for (j = i; j > 0; j--)
418     {
419     *utf8bytes-- = 0x80 | (cvalue & 0x3f);
420     cvalue >>= 6;
421     }
422    *utf8bytes = utf8_table2[i] | cvalue;
423    return i + 1;
424    }
425    
426    #endif
427    
428    
429    
430    /*************************************************
431  *             Print character string             *  *             Print character string             *
432  *************************************************/  *************************************************/
433    
# Line 198  chars without printing. */ Line 437  chars without printing. */
437    
438  static int pchars(unsigned char *p, int length, FILE *f)  static int pchars(unsigned char *p, int length, FILE *f)
439  {  {
440  int c;  int c = 0;
441  int yield = 0;  int yield = 0;
442    
443  while (length-- > 0)  while (length-- > 0)
# Line 212  while (length-- > 0) Line 451  while (length-- > 0)
451        {        {
452        length -= rc - 1;        length -= rc - 1;
453        p += rc;        p += rc;
454        if (c < 256 && isprint(c))        if (PRINTHEX(c))
455          {          {
456          if (f != NULL) fprintf(f, "%c", c);          if (f != NULL) fprintf(f, "%c", c);
457          yield++;          yield++;
458          }          }
459        else        else
460          {          {
461          int n;          int n = 4;
462          if (f != NULL) fprintf(f, "\\x{%02x}%n", c, &n);          if (f != NULL) fprintf(f, "\\x{%02x}", c);
463          yield += n;          yield += (n <= 0x000000ff)? 2 :
464                     (n <= 0x00000fff)? 3 :
465                     (n <= 0x0000ffff)? 4 :
466                     (n <= 0x000fffff)? 5 : 6;
467          }          }
468        continue;        continue;
469        }        }
# Line 230  while (length-- > 0) Line 472  while (length-- > 0)
472    
473     /* Not UTF-8, or malformed UTF-8  */     /* Not UTF-8, or malformed UTF-8  */
474    
475    if (isprint(c = *(p++)))    c = *p++;
476      if (PRINTHEX(c))
477      {      {
478      if (f != NULL) fprintf(f, "%c", c);      if (f != NULL) fprintf(f, "%c", c);
479      yield++;      yield++;
# Line 404  if ((rc = pcre_fullinfo(re, study, optio Line 647  if ((rc = pcre_fullinfo(re, study, optio
647  *         Byte flipping function                 *  *         Byte flipping function                 *
648  *************************************************/  *************************************************/
649    
650  static long int  static unsigned long int
651  byteflip(long int value, int n)  byteflip(unsigned long int value, int n)
652  {  {
653  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);
654  return ((value & 0x000000ff) << 24) |  return ((value & 0x000000ff) << 24) |
# Line 418  return ((value & 0x000000ff) << 24) | Line 661  return ((value & 0x000000ff) << 24) |
661    
662    
663  /*************************************************  /*************************************************
664    *        Check match or recursion limit          *
665    *************************************************/
666    
667    static int
668    check_match_limit(pcre *re, pcre_extra *extra, uschar *bptr, int len,
669      int start_offset, int options, int *use_offsets, int use_size_offsets,
670      int flag, unsigned long int *limit, int errnumber, const char *msg)
671    {
672    int count;
673    int min = 0;
674    int mid = 64;
675    int max = -1;
676    
677    extra->flags |= flag;
678    
679    for (;;)
680      {
681      *limit = mid;
682    
683      count = pcre_exec(re, extra, (char *)bptr, len, start_offset, options,
684        use_offsets, use_size_offsets);
685    
686      if (count == errnumber)
687        {
688        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
689        min = mid;
690        mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;
691        }
692    
693      else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||
694                             count == PCRE_ERROR_PARTIAL)
695        {
696        if (mid == min + 1)
697          {
698          fprintf(outfile, "Minimum %s limit = %d\n", msg, mid);
699          break;
700          }
701        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
702        max = mid;
703        mid = (min + mid)/2;
704        }
705      else break;    /* Some other error */
706      }
707    
708    extra->flags &= ~flag;
709    return count;
710    }
711    
712    
713    
714    /*************************************************
715    *         Case-independent strncmp() function    *
716    *************************************************/
717    
718    /*
719    Arguments:
720      s         first string
721      t         second string
722      n         number of characters to compare
723    
724    Returns:    < 0, = 0, or > 0, according to the comparison
725    */
726    
727    static int
728    strncmpic(uschar *s, uschar *t, int n)
729    {
730    while (n--)
731      {
732      int c = tolower(*s++) - tolower(*t++);
733      if (c) return c;
734      }
735    return 0;
736    }
737    
738    
739    
740    /*************************************************
741    *         Check newline indicator                *
742    *************************************************/
743    
744    /* This is used both at compile and run-time to check for <xxx> escapes, where
745    xxx is LF, CR, CRLF, ANYCRLF, or ANY. Print a message and return 0 if there is
746    no match.
747    
748    Arguments:
749      p           points after the leading '<'
750      f           file for error message
751    
752    Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
753    */
754    
755    static int
756    check_newline(uschar *p, FILE *f)
757    {
758    if (strncmpic(p, (uschar *)"cr>", 3) == 0) return PCRE_NEWLINE_CR;
759    if (strncmpic(p, (uschar *)"lf>", 3) == 0) return PCRE_NEWLINE_LF;
760    if (strncmpic(p, (uschar *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
761    if (strncmpic(p, (uschar *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF;
762    if (strncmpic(p, (uschar *)"any>", 4) == 0) return PCRE_NEWLINE_ANY;
763    if (strncmpic(p, (uschar *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF;
764    if (strncmpic(p, (uschar *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE;
765    fprintf(f, "Unknown newline type at: <%s\n", p);
766    return 0;
767    }
768    
769    
770    
771    /*************************************************
772    *             Usage function                     *
773    *************************************************/
774    
775    static void
776    usage(void)
777    {
778    printf("Usage:     pcretest [options] [<input file> [<output file>]]\n\n");
779    printf("Input and output default to stdin and stdout.\n");
780    #ifdef SUPPORT_LIBREADLINE
781    printf("If input is a terminal, readline() is used to read from it.\n");
782    #else
783    printf("This version of pcretest is not linked with readline().\n");
784    #endif
785    printf("\nOptions:\n");
786    printf("  -b       show compiled code (bytecode)\n");
787    printf("  -C       show PCRE compile-time options and exit\n");
788    printf("  -d       debug: show compiled code and information (-b and -i)\n");
789    #if !defined NODFA
790    printf("  -dfa     force DFA matching for all subjects\n");
791    #endif
792    printf("  -help    show usage information\n");
793    printf("  -i       show information about compiled patterns\n"
794           "  -M       find MATCH_LIMIT minimum for each subject\n"
795           "  -m       output memory used information\n"
796           "  -o <n>   set size of offsets vector to <n>\n");
797    #if !defined NOPOSIX
798    printf("  -p       use POSIX interface\n");
799    #endif
800    printf("  -q       quiet: do not output PCRE version number at start\n");
801    printf("  -S <n>   set stack size to <n> megabytes\n");
802    printf("  -s       output store (memory) used information\n"
803           "  -t       time compilation and execution\n");
804    printf("  -t <n>   time compilation and execution, repeating <n> times\n");
805    printf("  -tm      time execution (matching) only\n");
806    printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
807    }
808    
809    
810    
811    /*************************************************
812  *                Main Program                    *  *                Main Program                    *
813  *************************************************/  *************************************************/
814    
# Line 430  int main(int argc, char **argv) Line 821  int main(int argc, char **argv)
821  FILE *infile = stdin;  FILE *infile = stdin;
822  int options = 0;  int options = 0;
823  int study_options = 0;  int study_options = 0;
824    int default_find_match_limit = FALSE;
825  int op = 1;  int op = 1;
826  int timeit = 0;  int timeit = 0;
827    int timeitm = 0;
828  int showinfo = 0;  int showinfo = 0;
829  int showstore = 0;  int showstore = 0;
830    int quiet = 0;
831  int size_offsets = 45;  int size_offsets = 45;
832  int size_offsets_max;  int size_offsets_max;
833  int *offsets = NULL;  int *offsets = NULL;
# Line 444  int debug = 0; Line 838  int debug = 0;
838  int done = 0;  int done = 0;
839  int all_use_dfa = 0;  int all_use_dfa = 0;
840  int yield = 0;  int yield = 0;
841    int stack_size;
842    
843    /* These vectors store, end-to-end, a list of captured substring names. Assume
844    that 1024 is plenty long enough for the few names we'll be testing. */
845    
846    uschar copynames[1024];
847    uschar getnames[1024];
848    
849  unsigned char *buffer;  uschar *copynamesptr;
850  unsigned char *dbuffer;  uschar *getnamesptr;
851    
852  /* Get buffers from malloc() so that Electric Fence will check their misuse  /* Get buffers from malloc() so that Electric Fence will check their misuse
853  when I am debugging. */  when I am debugging. They grow automatically when very long lines are read. */
854    
855  buffer = (unsigned char *)malloc(BUFFER_SIZE);  buffer = (unsigned char *)malloc(buffer_size);
856  dbuffer = (unsigned char *)malloc(DBUFFER_SIZE);  dbuffer = (unsigned char *)malloc(buffer_size);
857  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.  */  
858    
859  #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) */  
860    
861  outfile = stdout;  outfile = stdout;
862    
863    /* The following  _setmode() stuff is some Windows magic that tells its runtime
864    library to translate CRLF into a single LF character. At least, that's what
865    I've been told: never having used Windows I take this all on trust. Originally
866    it set 0x8000, but then I was advised that _O_BINARY was better. */
867    
868    #if defined(_WIN32) || defined(WIN32)
869    _setmode( _fileno( stdout ), _O_BINARY );
870    #endif
871    
872  /* Scan options */  /* Scan options */
873    
874  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
# Line 473  while (argc > 1 && argv[op][0] == '-') Line 877  while (argc > 1 && argv[op][0] == '-')
877    
878    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
879      showstore = 1;      showstore = 1;
880    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-q") == 0) quiet = 1;
881      else if (strcmp(argv[op], "-b") == 0) debug = 1;
882    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
883    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
884      else if (strcmp(argv[op], "-M") == 0) default_find_match_limit = TRUE;
885  #if !defined NODFA  #if !defined NODFA
886    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
887  #endif  #endif
# Line 486  while (argc > 1 && argv[op][0] == '-') Line 892  while (argc > 1 && argv[op][0] == '-')
892      op++;      op++;
893      argc--;      argc--;
894      }      }
895      else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
896        {
897        int both = argv[op][2] == 0;
898        int temp;
899        if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr),
900                         *endptr == 0))
901          {
902          timeitm = temp;
903          op++;
904          argc--;
905          }
906        else timeitm = LOOPREPEAT;
907        if (both) timeit = timeitm;
908        }
909      else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
910          ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
911            *endptr == 0))
912        {
913    #if defined(_WIN32) || defined(WIN32)
914        printf("PCRE: -S not supported on this OS\n");
915        exit(1);
916    #else
917        int rc;
918        struct rlimit rlim;
919        getrlimit(RLIMIT_STACK, &rlim);
920        rlim.rlim_cur = stack_size * 1024 * 1024;
921        rc = setrlimit(RLIMIT_STACK, &rlim);
922        if (rc != 0)
923          {
924        printf("PCRE: setrlimit() failed with error %d\n", rc);
925        exit(1);
926          }
927        op++;
928        argc--;
929    #endif
930        }
931  #if !defined NOPOSIX  #if !defined NOPOSIX
932    else if (strcmp(argv[op], "-p") == 0) posix = 1;    else if (strcmp(argv[op], "-p") == 0) posix = 1;
933  #endif  #endif
934    else if (strcmp(argv[op], "-C") == 0)    else if (strcmp(argv[op], "-C") == 0)
935      {      {
936      int rc;      int rc;
937        unsigned long int lrc;
938      printf("PCRE version %s\n", pcre_version());      printf("PCRE version %s\n", pcre_version());
939      printf("Compiled with\n");      printf("Compiled with\n");
940      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
# Line 499  while (argc > 1 && argv[op][0] == '-') Line 942  while (argc > 1 && argv[op][0] == '-')
942      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
943      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
944      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
945      printf("  Newline character is %s\n", (rc == '\r')? "CR" : "LF");      /* Note that these values are always the ASCII values, even
946        in EBCDIC environments. CR is 13 and NL is 10. */
947        printf("  Newline sequence is %s\n", (rc == 13)? "CR" :
948          (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
949          (rc == -2)? "ANYCRLF" :
950          (rc == -1)? "ANY" : "???");
951        (void)pcre_config(PCRE_CONFIG_BSR, &rc);
952        printf("  \\R matches %s\n", rc? "CR, LF, or CRLF only" :
953                                         "all Unicode newlines");
954      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);
955      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
956      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
957      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
958      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &lrc);
959      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %ld\n", lrc);
960        (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &lrc);
961        printf("  Default recursion depth limit = %ld\n", lrc);
962      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
963      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
964      exit(0);      goto EXIT;
965        }
966      else if (strcmp(argv[op], "-help") == 0 ||
967               strcmp(argv[op], "--help") == 0)
968        {
969        usage();
970        goto EXIT;
971      }      }
972    else    else
973      {      {
974      printf("** Unknown or malformed option %s\n", argv[op]);      printf("** Unknown or malformed option %s\n", argv[op]);
975      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");  
976      yield = 1;      yield = 1;
977      goto EXIT;      goto EXIT;
978      }      }
# Line 541  offsets = (int *)malloc(size_offsets_max Line 987  offsets = (int *)malloc(size_offsets_max
987  if (offsets == NULL)  if (offsets == NULL)
988    {    {
989    printf("** Failed to get %d bytes of memory for offsets vector\n",    printf("** Failed to get %d bytes of memory for offsets vector\n",
990      size_offsets_max * sizeof(int));      (int)(size_offsets_max * sizeof(int)));
991    yield = 1;    yield = 1;
992    goto EXIT;    goto EXIT;
993    }    }
# Line 550  if (offsets == NULL) Line 996  if (offsets == NULL)
996    
997  if (argc > 1)  if (argc > 1)
998    {    {
999    infile = fopen(argv[op], "rb");    infile = fopen(argv[op], INPUT_MODE);
1000    if (infile == NULL)    if (infile == NULL)
1001      {      {
1002      printf("** Failed to open %s\n", argv[op]);      printf("** Failed to open %s\n", argv[op]);
# Line 561  if (argc > 1) Line 1007  if (argc > 1)
1007    
1008  if (argc > 2)  if (argc > 2)
1009    {    {
1010    outfile = fopen(argv[op+1], "wb");    outfile = fopen(argv[op+1], OUTPUT_MODE);
1011    if (outfile == NULL)    if (outfile == NULL)
1012      {      {
1013      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 1023  pcre_free = new_free;
1023  pcre_stack_malloc = stack_malloc;  pcre_stack_malloc = stack_malloc;
1024  pcre_stack_free = stack_free;  pcre_stack_free = stack_free;
1025    
1026  /* Heading line, then prompt for first regex if stdin */  /* Heading line unless quiet, then prompt for first regex if stdin */
1027    
1028  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  if (!quiet) fprintf(outfile, "PCRE version %s\n\n", pcre_version());
1029    
1030  /* Main loop */  /* Main loop */
1031    
# Line 594  while (!done) Line 1040  while (!done)
1040  #endif  #endif
1041    
1042    const char *error;    const char *error;
1043      unsigned char *markptr;
1044    unsigned char *p, *pp, *ppp;    unsigned char *p, *pp, *ppp;
1045    unsigned char *to_file = NULL;    unsigned char *to_file = NULL;
1046    const unsigned char *tables = NULL;    const unsigned char *tables = NULL;
1047    unsigned long int true_size, true_study_size = 0;    unsigned long int true_size, true_study_size = 0;
1048    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
1049      int do_mark = 0;
1050    int do_study = 0;    int do_study = 0;
1051    int do_debug = debug;    int do_debug = debug;
1052    int do_G = 0;    int do_G = 0;
# Line 606  while (!done) Line 1054  while (!done)
1054    int do_showinfo = showinfo;    int do_showinfo = showinfo;
1055    int do_showrest = 0;    int do_showrest = 0;
1056    int do_flip = 0;    int do_flip = 0;
1057    int erroroffset, len, delimiter;    int erroroffset, len, delimiter, poffset;
1058    
1059    use_utf8 = 0;    use_utf8 = 0;
1060      debug_lengths = 1;
1061    
1062    if (infile == stdin) printf("  re> ");    if (extend_inputline(infile, buffer, "  re> ") == NULL) break;
   if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break;  
1063    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1064    fflush(outfile);    fflush(outfile);
1065    
# Line 623  while (!done) Line 1071  while (!done)
1071    
1072    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
1073      {      {
1074      unsigned long int magic;      unsigned long int magic, get_options;
1075      uschar sbuf[8];      uschar sbuf[8];
1076      FILE *f;      FILE *f;
1077    
# Line 671  while (!done) Line 1119  while (!done)
1119    
1120      /* Need to know if UTF-8 for printing data strings */      /* Need to know if UTF-8 for printing data strings */
1121    
1122      new_info(re, NULL, PCRE_INFO_OPTIONS, &options);      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1123      use_utf8 = (options & PCRE_UTF8) != 0;      use_utf8 = (get_options & PCRE_UTF8) != 0;
1124    
1125      /* Now see if there is any following study data */      /* Now see if there is any following study data */
1126    
# Line 711  while (!done) Line 1159  while (!done)
1159    
1160    if (isalnum(delimiter) || delimiter == '\\')    if (isalnum(delimiter) || delimiter == '\\')
1161      {      {
1162      fprintf(outfile, "** Delimiter must not be alphameric or \\\n");      fprintf(outfile, "** Delimiter must not be alphanumeric or \\\n");
1163      goto SKIP_DATA;      goto SKIP_DATA;
1164      }      }
1165    
1166    pp = p;    pp = p;
1167      poffset = p - buffer;
1168    
1169    for(;;)    for(;;)
1170      {      {
# Line 726  while (!done) Line 1175  while (!done)
1175        pp++;        pp++;
1176        }        }
1177      if (*pp != 0) break;      if (*pp != 0) break;
1178        if ((pp = extend_inputline(infile, pp, "    > ")) == NULL)
     len = BUFFER_SIZE - (pp - buffer);  
     if (len < 256)  
       {  
       fprintf(outfile, "** Expression too long - missing delimiter?\n");  
       goto SKIP_DATA;  
       }  
   
     if (infile == stdin) printf("    > ");  
     if (fgets((char *)pp, len, infile) == NULL)  
1179        {        {
1180        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
1181        done = 1;        done = 1;
# Line 744  while (!done) Line 1184  while (!done)
1184      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
1185      }      }
1186    
1187      /* The buffer may have moved while being extended; reset the start of data
1188      pointer to the correct relative point in the buffer. */
1189    
1190      p = buffer + poffset;
1191    
1192    /* If the first character after the delimiter is backslash, make    /* If the first character after the delimiter is backslash, make
1193    the pattern end with backslash. This is purely to provide a way    the pattern end with backslash. This is purely to provide a way
1194    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 1220  while (!done)
1220    
1221        case '+': do_showrest = 1; break;        case '+': do_showrest = 1; break;
1222        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
1223          case 'B': do_debug = 1; break;
1224        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
1225        case 'D': do_debug = do_showinfo = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
1226        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
1227        case 'F': do_flip = 1; break;        case 'F': do_flip = 1; break;
1228        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1229        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1230          case 'J': options |= PCRE_DUPNAMES; break;
1231          case 'K': do_mark = 1; break;
1232        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1233        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1234    
# Line 791  while (!done) Line 1239  while (!done)
1239        case 'S': do_study = 1; break;        case 'S': do_study = 1; break;
1240        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1241        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
1242          case 'Z': debug_lengths = 0; break;
1243        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1244        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
1245    
1246        case 'L':        case 'L':
1247        ppp = pp;        ppp = pp;
1248        /* The '\r' test here is so that it works on Windows */        /* The '\r' test here is so that it works on Windows. */
1249        while (*ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;        /* The '0' test is just in case this is an unterminated line. */
1250          while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
1251        *ppp = 0;        *ppp = 0;
1252        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
1253          {          {
1254          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
1255          goto SKIP_DATA;          goto SKIP_DATA;
1256          }          }
1257          locale_set = 1;
1258        tables = pcre_maketables();        tables = pcre_maketables();
1259        pp = ppp;        pp = ppp;
1260        break;        break;
# Line 815  while (!done) Line 1266  while (!done)
1266        *pp = 0;        *pp = 0;
1267        break;        break;
1268    
1269          case '<':
1270            {
1271            if (strncmp((char *)pp, "JS>", 3) == 0)
1272              {
1273              options |= PCRE_JAVASCRIPT_COMPAT;
1274              pp += 3;
1275              }
1276            else
1277              {
1278              int x = check_newline(pp, outfile);
1279              if (x == 0) goto SKIP_DATA;
1280              options |= x;
1281              while (*pp++ != '>');
1282              }
1283            }
1284          break;
1285    
1286        case '\r':                      /* So that it works in Windows */        case '\r':                      /* So that it works in Windows */
1287        case '\n':        case '\n':
1288        case ' ':        case ' ':
# Line 839  while (!done) Line 1307  while (!done)
1307      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
1308      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
1309      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1310        if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1311        if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1312        if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY;
1313    
1314      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1315    
1316      /* 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 1318  while (!done)
1318    
1319      if (rc != 0)      if (rc != 0)
1320        {        {
1321        (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);        (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1322        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
1323        goto SKIP_DATA;        goto SKIP_DATA;
1324        }        }
# Line 858  while (!done) Line 1330  while (!done)
1330  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1331    
1332      {      {
1333      if (timeit)      unsigned long int get_options;
1334    
1335        if (timeit > 0)
1336        {        {
1337        register int i;        register int i;
1338        clock_t time_taken;        clock_t time_taken;
1339        clock_t start_time = clock();        clock_t start_time = clock();
1340        for (i = 0; i < LOOPREPEAT; i++)        for (i = 0; i < timeit; i++)
1341          {          {
1342          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1343          if (re != NULL) free(re);          if (re != NULL) free(re);
1344          }          }
1345        time_taken = clock() - start_time;        time_taken = clock() - start_time;
1346        fprintf(outfile, "Compile time %.3f milliseconds\n",        fprintf(outfile, "Compile time %.4f milliseconds\n",
1347          (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /          (((double)time_taken * 1000.0) / (double)timeit) /
1348            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
1349        }        }
1350    
# Line 887  while (!done) Line 1361  while (!done)
1361          {          {
1362          for (;;)          for (;;)
1363            {            {
1364            if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)            if (extend_inputline(infile, buffer, NULL) == NULL)
1365              {              {
1366              done = 1;              done = 1;
1367              goto CONTINUE;              goto CONTINUE;
# Line 901  while (!done) Line 1375  while (!done)
1375        goto CONTINUE;        goto CONTINUE;
1376        }        }
1377    
1378      /* Compilation succeeded; print data if required. There are now two      /* Compilation succeeded. It is now possible to set the UTF-8 option from
1379      info-returning functions. The old one has a limited interface and      within the regex; check for this so that we know how to process the data
1380      returns only limited data. Check that it agrees with the newer one. */      lines. */
1381    
1382        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1383        if ((get_options & PCRE_UTF8) != 0) use_utf8 = 1;
1384    
1385        /* Print information if required. There are now two info-returning
1386        functions. The old one has a limited interface and returns only limited
1387        data. Check that it agrees with the newer one. */
1388    
1389      if (log_store)      if (log_store)
1390        fprintf(outfile, "Memory allocation (code space): %d\n",        fprintf(outfile, "Memory allocation (code space): %d\n",
# Line 922  while (!done) Line 1403  while (!done)
1403    
1404      if (do_study)      if (do_study)
1405        {        {
1406        if (timeit)        if (timeit > 0)
1407          {          {
1408          register int i;          register int i;
1409          clock_t time_taken;          clock_t time_taken;
1410          clock_t start_time = clock();          clock_t start_time = clock();
1411          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeit; i++)
1412            extra = pcre_study(re, study_options, &error);            extra = pcre_study(re, study_options, &error);
1413          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1414          if (extra != NULL) free(extra);          if (extra != NULL) free(extra);
1415          fprintf(outfile, "  Study time %.3f milliseconds\n",          fprintf(outfile, "  Study time %.4f milliseconds\n",
1416            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeit) /
1417              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
1418          }          }
1419        extra = pcre_study(re, study_options, &error);        extra = pcre_study(re, study_options, &error);
# Line 942  while (!done) Line 1423  while (!done)
1423          true_study_size = ((pcre_study_data *)(extra->study_data))->size;          true_study_size = ((pcre_study_data *)(extra->study_data))->size;
1424        }        }
1425    
1426        /* If /K was present, we set up for handling MARK data. */
1427    
1428        if (do_mark)
1429          {
1430          if (extra == NULL)
1431            {
1432            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1433            extra->flags = 0;
1434            }
1435          extra->mark = &markptr;
1436          extra->flags |= PCRE_EXTRA_MARK;
1437          }
1438    
1439      /* If the 'F' option was present, we flip the bytes of all the integer      /* If the 'F' option was present, we flip the bytes of all the integer
1440      fields in the regex data block and the study block. This is to make it      fields in the regex data block and the study block. This is to make it
1441      possible to test PCRE's handling of byte-flipped patterns, e.g. those      possible to test PCRE's handling of byte-flipped patterns, e.g. those
# Line 950  while (!done) Line 1444  while (!done)
1444      if (do_flip)      if (do_flip)
1445        {        {
1446        real_pcre *rre = (real_pcre *)re;        real_pcre *rre = (real_pcre *)re;
1447        rre->magic_number = byteflip(rre->magic_number, sizeof(rre->magic_number));        rre->magic_number =
1448            byteflip(rre->magic_number, sizeof(rre->magic_number));
1449        rre->size = byteflip(rre->size, sizeof(rre->size));        rre->size = byteflip(rre->size, sizeof(rre->size));
1450        rre->options = byteflip(rre->options, sizeof(rre->options));        rre->options = byteflip(rre->options, sizeof(rre->options));
1451        rre->top_bracket = byteflip(rre->top_bracket, sizeof(rre->top_bracket));        rre->flags = (pcre_uint16)byteflip(rre->flags, sizeof(rre->flags));
1452        rre->top_backref = byteflip(rre->top_backref, sizeof(rre->top_backref));        rre->top_bracket =
1453        rre->first_byte = byteflip(rre->first_byte, sizeof(rre->first_byte));          (pcre_uint16)byteflip(rre->top_bracket, sizeof(rre->top_bracket));
1454        rre->req_byte = byteflip(rre->req_byte, sizeof(rre->req_byte));        rre->top_backref =
1455        rre->name_table_offset = byteflip(rre->name_table_offset,          (pcre_uint16)byteflip(rre->top_backref, sizeof(rre->top_backref));
1456          rre->first_byte =
1457            (pcre_uint16)byteflip(rre->first_byte, sizeof(rre->first_byte));
1458          rre->req_byte =
1459            (pcre_uint16)byteflip(rre->req_byte, sizeof(rre->req_byte));
1460          rre->name_table_offset = (pcre_uint16)byteflip(rre->name_table_offset,
1461          sizeof(rre->name_table_offset));          sizeof(rre->name_table_offset));
1462        rre->name_entry_size = byteflip(rre->name_entry_size,        rre->name_entry_size = (pcre_uint16)byteflip(rre->name_entry_size,
1463          sizeof(rre->name_entry_size));          sizeof(rre->name_entry_size));
1464        rre->name_count = byteflip(rre->name_count, sizeof(rre->name_count));        rre->name_count = (pcre_uint16)byteflip(rre->name_count,
1465            sizeof(rre->name_count));
1466    
1467        if (extra != NULL)        if (extra != NULL)
1468          {          {
1469          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
1470          rsd->size = byteflip(rsd->size, sizeof(rsd->size));          rsd->size = byteflip(rsd->size, sizeof(rsd->size));
1471          rsd->options = byteflip(rsd->options, sizeof(rsd->options));          rsd->flags = byteflip(rsd->flags, sizeof(rsd->flags));
1472            rsd->minlength = byteflip(rsd->minlength, sizeof(rsd->minlength));
1473          }          }
1474        }        }
1475    
# Line 975  while (!done) Line 1477  while (!done)
1477    
1478      SHOW_INFO:      SHOW_INFO:
1479    
1480        if (do_debug)
1481          {
1482          fprintf(outfile, "------------------------------------------------------------------\n");
1483          pcre_printint(re, outfile, debug_lengths);
1484          }
1485    
1486        /* We already have the options in get_options (see above) */
1487    
1488      if (do_showinfo)      if (do_showinfo)
1489        {        {
1490        unsigned long int get_options, all_options;        unsigned long int all_options;
1491  #if !defined NOINFOCHECK  #if !defined NOINFOCHECK
1492        int old_first_char, old_options, old_count;        int old_first_char, old_options, old_count;
1493  #endif  #endif
1494        int count, backrefmax, first_char, need_char;        int count, backrefmax, first_char, need_char, okpartial, jchanged,
1495            hascrorlf;
1496        int nameentrysize, namecount;        int nameentrysize, namecount;
1497        const uschar *nametable;        const uschar *nametable;
1498    
       if (do_debug)  
         {  
         fprintf(outfile, "------------------------------------------------------------------\n");  
         _pcre_printint(re, outfile);  
         }  
   
       new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);  
1499        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1500        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
1501        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);
# Line 1000  while (!done) Line 1504  while (!done)
1504        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);
1505        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);
1506        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);
1507          new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial);
1508          new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged);
1509          new_info(re, NULL, PCRE_INFO_HASCRORLF, &hascrorlf);
1510    
1511  #if !defined NOINFOCHECK  #if !defined NOINFOCHECK
1512        old_count = pcre_info(re, &old_options, &old_first_char);        old_count = pcre_info(re, &old_options, &old_first_char);
# Line 1041  while (!done) Line 1548  while (!done)
1548            }            }
1549          }          }
1550    
1551        /* The NOPARTIAL bit is a private bit in the options, so we have        if (!okpartial) fprintf(outfile, "Partial matching not supported\n");
1552        to fish it out via out back door */        if (hascrorlf) fprintf(outfile, "Contains explicit CR or LF match\n");
1553    
1554        all_options = ((real_pcre *)re)->options;        all_options = ((real_pcre *)re)->options;
1555        if (do_flip)        if (do_flip) all_options = byteflip(all_options, sizeof(all_options));
         {  
         all_options = byteflip(all_options, sizeof(all_options));  
         }  
   
       if ((all_options & PCRE_NOPARTIAL) != 0)  
         fprintf(outfile, "Partial matching not supported\n");  
1556    
1557        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1558          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%s%s\n",
1559            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1560            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1561            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
1562            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",
1563            ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",            ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",
1564            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",
1565              ((get_options & PCRE_BSR_ANYCRLF) != 0)? " bsr_anycrlf" : "",
1566              ((get_options & PCRE_BSR_UNICODE) != 0)? " bsr_unicode" : "",
1567            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
1568            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
1569            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1570              ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1571            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1572            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "");            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1573              ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1574    
1575        if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0)        if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
1576          fprintf(outfile, "Case state changes\n");  
1577          switch (get_options & PCRE_NEWLINE_BITS)
1578            {
1579            case PCRE_NEWLINE_CR:
1580            fprintf(outfile, "Forced newline sequence: CR\n");
1581            break;
1582    
1583            case PCRE_NEWLINE_LF:
1584            fprintf(outfile, "Forced newline sequence: LF\n");
1585            break;
1586    
1587            case PCRE_NEWLINE_CRLF:
1588            fprintf(outfile, "Forced newline sequence: CRLF\n");
1589            break;
1590    
1591            case PCRE_NEWLINE_ANYCRLF:
1592            fprintf(outfile, "Forced newline sequence: ANYCRLF\n");
1593            break;
1594    
1595            case PCRE_NEWLINE_ANY:
1596            fprintf(outfile, "Forced newline sequence: ANY\n");
1597            break;
1598    
1599            default:
1600            break;
1601            }
1602    
1603        if (first_char == -1)        if (first_char == -1)
1604          {          {
1605          fprintf(outfile, "First char at start or follows \\n\n");          fprintf(outfile, "First char at start or follows newline\n");
1606          }          }
1607        else if (first_char < 0)        else if (first_char < 0)
1608          {          {
# Line 1083  while (!done) Line 1613  while (!done)
1613          int ch = first_char & 255;          int ch = first_char & 255;
1614          const char *caseless = ((first_char & REQ_CASELESS) == 0)?          const char *caseless = ((first_char & REQ_CASELESS) == 0)?
1615            "" : " (caseless)";            "" : " (caseless)";
1616          if (isprint(ch))          if (PRINTHEX(ch))
1617            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);
1618          else          else
1619            fprintf(outfile, "First char = %d%s\n", ch, caseless);            fprintf(outfile, "First char = %d%s\n", ch, caseless);
# Line 1098  while (!done) Line 1628  while (!done)
1628          int ch = need_char & 255;          int ch = need_char & 255;
1629          const char *caseless = ((need_char & REQ_CASELESS) == 0)?          const char *caseless = ((need_char & REQ_CASELESS) == 0)?
1630            "" : " (caseless)";            "" : " (caseless)";
1631          if (isprint(ch))          if (PRINTHEX(ch))
1632            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);
1633          else          else
1634            fprintf(outfile, "Need char = %d%s\n", ch, caseless);            fprintf(outfile, "Need char = %d%s\n", ch, caseless);
# Line 1116  while (!done) Line 1646  while (!done)
1646          else          else
1647            {            {
1648            uschar *start_bits = NULL;            uschar *start_bits = NULL;
1649            new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);            int minlength;
1650    
1651              new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength);
1652              fprintf(outfile, "Subject length lower bound = %d\n", minlength);
1653    
1654              new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);
1655            if (start_bits == NULL)            if (start_bits == NULL)
1656              fprintf(outfile, "No starting byte set\n");              fprintf(outfile, "No set of starting bytes\n");
1657            else            else
1658              {              {
1659              int i;              int i;
# Line 1134  while (!done) Line 1668  while (!done)
1668                    fprintf(outfile, "\n  ");                    fprintf(outfile, "\n  ");
1669                    c = 2;                    c = 2;
1670                    }                    }
1671                  if (isprint(i) && i != ' ')                  if (PRINTHEX(i) && i != ' ')
1672                    {                    {
1673                    fprintf(outfile, "%c ", i);                    fprintf(outfile, "%c ", i);
1674                    c += 2;                    c += 2;
# Line 1166  while (!done) Line 1700  while (!done)
1700        else        else
1701          {          {
1702          uschar sbuf[8];          uschar sbuf[8];
1703          sbuf[0] = (true_size >> 24)  & 255;          sbuf[0] = (uschar)((true_size >> 24) & 255);
1704          sbuf[1] = (true_size >> 16)  & 255;          sbuf[1] = (uschar)((true_size >> 16) & 255);
1705          sbuf[2] = (true_size >>  8)  & 255;          sbuf[2] = (uschar)((true_size >>  8) & 255);
1706          sbuf[3] = (true_size)  & 255;          sbuf[3] = (uschar)((true_size) & 255);
1707    
1708          sbuf[4] = (true_study_size >> 24)  & 255;          sbuf[4] = (uschar)((true_study_size >> 24) & 255);
1709          sbuf[5] = (true_study_size >> 16)  & 255;          sbuf[5] = (uschar)((true_study_size >> 16) & 255);
1710          sbuf[6] = (true_study_size >>  8)  & 255;          sbuf[6] = (uschar)((true_study_size >>  8) & 255);
1711          sbuf[7] = (true_study_size)  & 255;          sbuf[7] = (uschar)((true_study_size) & 255);
1712    
1713          if (fwrite(sbuf, 1, 8, f) < 8 ||          if (fwrite(sbuf, 1, 8, f) < 8 ||
1714              fwrite(re, 1, true_size, f) < true_size)              fwrite(re, 1, true_size, f) < true_size)
# Line 1193  while (!done) Line 1727  while (!done)
1727                  strerror(errno));                  strerror(errno));
1728                }                }
1729              else fprintf(outfile, "Study data written to %s\n", to_file);              else fprintf(outfile, "Study data written to %s\n", to_file);
1730    
1731              }              }
1732            }            }
1733          fclose(f);          fclose(f);
# Line 1209  while (!done) Line 1744  while (!done)
1744    
1745    for (;;)    for (;;)
1746      {      {
1747      unsigned char *q;      uschar *q;
1748      unsigned char *bptr = dbuffer;      uschar *bptr;
1749      int *use_offsets = offsets;      int *use_offsets = offsets;
1750      int use_size_offsets = size_offsets;      int use_size_offsets = size_offsets;
1751      int callout_data = 0;      int callout_data = 0;
1752      int callout_data_set = 0;      int callout_data_set = 0;
1753      int count, c;      int count, c;
1754      int copystrings = 0;      int copystrings = 0;
1755      int find_match_limit = 0;      int find_match_limit = default_find_match_limit;
1756      int getstrings = 0;      int getstrings = 0;
1757      int getlist = 0;      int getlist = 0;
1758      int gmatched = 0;      int gmatched = 0;
# Line 1227  while (!done) Line 1762  while (!done)
1762    
1763      options = 0;      options = 0;
1764    
1765        *copynames = 0;
1766        *getnames = 0;
1767    
1768        copynamesptr = copynames;
1769        getnamesptr = getnames;
1770    
1771      pcre_callout = callout;      pcre_callout = callout;
1772      first_callout = 1;      first_callout = 1;
1773      callout_extra = 0;      callout_extra = 0;
# Line 1235  while (!done) Line 1776  while (!done)
1776      callout_fail_id = -1;      callout_fail_id = -1;
1777      show_malloc = 0;      show_malloc = 0;
1778    
1779      if (infile == stdin) printf("data> ");      if (extra != NULL) extra->flags &=
1780      if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)        ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
1781    
1782        len = 0;
1783        for (;;)
1784        {        {
1785        done = 1;        if (extend_inputline(infile, buffer + len, "data> ") == NULL)
1786        goto CONTINUE;          {
1787            if (len > 0) break;
1788            done = 1;
1789            goto CONTINUE;
1790            }
1791          if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1792          len = (int)strlen((char *)buffer);
1793          if (buffer[len-1] == '\n') break;
1794        }        }
     if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);  
1795    
     len = (int)strlen((char *)buffer);  
1796      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
1797      buffer[len] = 0;      buffer[len] = 0;
1798      if (len == 0) break;      if (len == 0) break;
# Line 1251  while (!done) Line 1800  while (!done)
1800      p = buffer;      p = buffer;
1801      while (isspace(*p)) p++;      while (isspace(*p)) p++;
1802    
1803      q = dbuffer;      bptr = q = dbuffer;
1804      while ((c = *p++) != 0)      while ((c = *p++) != 0)
1805        {        {
1806        int i = 0;        int i = 0;
# Line 1273  while (!done) Line 1822  while (!done)
1822          c -= '0';          c -= '0';
1823          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')
1824            c = c * 8 + *p++ - '0';            c = c * 8 + *p++ - '0';
1825    
1826    #if !defined NOUTF8
1827            if (use_utf8 && c > 255)
1828              {
1829              unsigned char buff8[8];
1830              int ii, utn;
1831              utn = ord2utf8(c, buff8);
1832              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1833              c = buff8[ii];   /* Last byte */
1834              }
1835    #endif
1836          break;          break;
1837    
1838          case 'x':          case 'x':
# Line 1290  while (!done) Line 1850  while (!done)
1850              {              {
1851              unsigned char buff8[8];              unsigned char buff8[8];
1852              int ii, utn;              int ii, utn;
1853              utn = _pcre_ord2utf8(c, buff8);              if (use_utf8)
1854              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];                {
1855              c = buff8[ii];   /* Last byte */                utn = ord2utf8(c, buff8);
1856                  for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1857                  c = buff8[ii];   /* Last byte */
1858                  }
1859                else
1860                 {
1861                 if (c > 255)
1862                   fprintf(outfile, "** Character \\x{%x} is greater than 255 and "
1863                     "UTF-8 mode is not enabled.\n"
1864                     "** Truncation will probably give the wrong result.\n", c);
1865                 }
1866              p = pt + 1;              p = pt + 1;
1867              break;              break;
1868              }              }
# Line 1334  while (!done) Line 1904  while (!done)
1904            }            }
1905          else if (isalnum(*p))          else if (isalnum(*p))
1906            {            {
1907            uschar name[256];            uschar *npp = copynamesptr;
           uschar *npp = name;  
1908            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1909              *npp++ = 0;
1910            *npp = 0;            *npp = 0;
1911            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)copynamesptr);
1912            if (n < 0)            if (n < 0)
1913              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", copynamesptr);
1914            else copystrings |= 1 << n;            copynamesptr = npp;
1915            }            }
1916          else if (*p == '+')          else if (*p == '+')
1917            {            {
# Line 1402  while (!done) Line 1972  while (!done)
1972            }            }
1973          else if (isalnum(*p))          else if (isalnum(*p))
1974            {            {
1975            uschar name[256];            uschar *npp = getnamesptr;
           uschar *npp = name;  
1976            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1977              *npp++ = 0;
1978            *npp = 0;            *npp = 0;
1979            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)getnamesptr);
1980            if (n < 0)            if (n < 0)
1981              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", getnamesptr);
1982            else getstrings |= 1 << n;            getnamesptr = npp;
1983            }            }
1984          continue;          continue;
1985    
# Line 1422  while (!done) Line 1992  while (!done)
1992          continue;          continue;
1993    
1994          case 'N':          case 'N':
1995          options |= PCRE_NOTEMPTY;          if ((options & PCRE_NOTEMPTY) != 0)
1996              options = (options & ~PCRE_NOTEMPTY) | PCRE_NOTEMPTY_ATSTART;
1997            else
1998              options |= PCRE_NOTEMPTY;
1999          continue;          continue;
2000    
2001          case 'O':          case 'O':
# Line 1435  while (!done) Line 2008  while (!done)
2008            if (offsets == NULL)            if (offsets == NULL)
2009              {              {
2010              printf("** Failed to get %d bytes of memory for offsets vector\n",              printf("** Failed to get %d bytes of memory for offsets vector\n",
2011                size_offsets_max * sizeof(int));                (int)(size_offsets_max * sizeof(int)));
2012              yield = 1;              yield = 1;
2013              goto EXIT;              goto EXIT;
2014              }              }
# Line 1445  while (!done) Line 2018  while (!done)
2018          continue;          continue;
2019    
2020          case 'P':          case 'P':
2021          options |= PCRE_PARTIAL;          options |= ((options & PCRE_PARTIAL_SOFT) == 0)?
2022              PCRE_PARTIAL_SOFT : PCRE_PARTIAL_HARD;
2023            continue;
2024    
2025            case 'Q':
2026            while(isdigit(*p)) n = n * 10 + *p++ - '0';
2027            if (extra == NULL)
2028              {
2029              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2030              extra->flags = 0;
2031              }
2032            extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
2033            extra->match_limit_recursion = n;
2034            continue;
2035    
2036            case 'q':
2037            while(isdigit(*p)) n = n * 10 + *p++ - '0';
2038            if (extra == NULL)
2039              {
2040              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2041              extra->flags = 0;
2042              }
2043            extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
2044            extra->match_limit = n;
2045          continue;          continue;
2046    
2047  #if !defined NODFA  #if !defined NODFA
# Line 1458  while (!done) Line 2054  while (!done)
2054          show_malloc = 1;          show_malloc = 1;
2055          continue;          continue;
2056    
2057            case 'Y':
2058            options |= PCRE_NO_START_OPTIMIZE;
2059            continue;
2060    
2061          case 'Z':          case 'Z':
2062          options |= PCRE_NOTEOL;          options |= PCRE_NOTEOL;
2063          continue;          continue;
# Line 1465  while (!done) Line 2065  while (!done)
2065          case '?':          case '?':
2066          options |= PCRE_NO_UTF8_CHECK;          options |= PCRE_NO_UTF8_CHECK;
2067          continue;          continue;
2068    
2069            case '<':
2070              {
2071              int x = check_newline(p, outfile);
2072              if (x == 0) goto NEXT_DATA;
2073              options |= x;
2074              while (*p++ != '>');
2075              }
2076            continue;
2077          }          }
2078        *q++ = c;        *q++ = c;
2079        }        }
2080      *q = 0;      *q = 0;
2081      len = q - dbuffer;      len = q - dbuffer;
2082    
2083        /* Move the data to the end of the buffer so that a read over the end of
2084        the buffer will be seen by valgrind, even if it doesn't cause a crash. If
2085        we are using the POSIX interface, we must include the terminating zero. */
2086    
2087    #if !defined NOPOSIX
2088        if (posix || do_posix)
2089          {
2090          memmove(bptr + buffer_size - len - 1, bptr, len + 1);
2091          bptr += buffer_size - len - 1;
2092          }
2093        else
2094    #endif
2095          {
2096          memmove(bptr + buffer_size - len, bptr, len);
2097          bptr += buffer_size - len;
2098          }
2099    
2100      if ((all_use_dfa || use_dfa) && find_match_limit)      if ((all_use_dfa || use_dfa) && find_match_limit)
2101        {        {
2102        printf("**Match limit not relevant for DFA matching: ignored\n");        printf("**Match limit not relevant for DFA matching: ignored\n");
# Line 1490  while (!done) Line 2116  while (!done)
2116          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);
2117        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;
2118        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;
2119          if ((options & PCRE_NOTEMPTY) != 0) eflags |= REG_NOTEMPTY;
2120    
2121        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);
2122    
2123        if (rc != 0)        if (rc != 0)
2124          {          {
2125          (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);          (void)regerror(rc, &preg, (char *)buffer, buffer_size);
2126          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
2127          }          }
2128          else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
2129                  != 0)
2130            {
2131            fprintf(outfile, "Matched with REG_NOSUB\n");
2132            }
2133        else        else
2134          {          {
2135          size_t i;          size_t i;
# Line 1529  while (!done) Line 2161  while (!done)
2161    
2162      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
2163        {        {
2164        if (timeit)        markptr = NULL;
2165    
2166          if (timeitm > 0)
2167          {          {
2168          register int i;          register int i;
2169          clock_t time_taken;          clock_t time_taken;
# Line 1539  while (!done) Line 2173  while (!done)
2173          if (all_use_dfa || use_dfa)          if (all_use_dfa || use_dfa)
2174            {            {
2175            int workspace[1000];            int workspace[1000];
2176            for (i = 0; i < LOOPREPEAT; i++)            for (i = 0; i < timeitm; i++)
2177              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,              count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2178                options | g_notempty, use_offsets, use_size_offsets, workspace,                options | g_notempty, use_offsets, use_size_offsets, workspace,
2179                sizeof(workspace)/sizeof(int));                sizeof(workspace)/sizeof(int));
2180            }            }
2181          else          else
2182  #endif  #endif
2183    
2184          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeitm; i++)
2185            count = pcre_exec(re, extra, (char *)bptr, len,            count = pcre_exec(re, extra, (char *)bptr, len,
2186              start_offset, options | g_notempty, use_offsets, use_size_offsets);              start_offset, options | g_notempty, use_offsets, use_size_offsets);
2187    
2188          time_taken = clock() - start_time;          time_taken = clock() - start_time;
2189          fprintf(outfile, "Execute time %.3f milliseconds\n",          fprintf(outfile, "Execute time %.4f milliseconds\n",
2190            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeitm) /
2191              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
2192          }          }
2193    
2194        /* 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
2195        varying limits in order to find the minimum value. */        varying limits in order to find the minimum value for the match limit and
2196          for the recursion limit. */
2197    
2198        if (find_match_limit)        if (find_match_limit)
2199          {          {
         int min = 0;  
         int mid = 64;  
         int max = -1;  
   
2200          if (extra == NULL)          if (extra == NULL)
2201            {            {
2202            extra = (pcre_extra *)malloc(sizeof(pcre_extra));            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2203            extra->flags = 0;            extra->flags = 0;
2204            }            }
         extra->flags |= PCRE_EXTRA_MATCH_LIMIT;  
   
         for (;;)  
           {  
           extra->match_limit = mid;  
           count = pcre_exec(re, extra, (char *)bptr, len, start_offset,  
             options | g_notempty, use_offsets, use_size_offsets);  
           if (count == PCRE_ERROR_MATCHLIMIT)  
             {  
             /* fprintf(outfile, "Testing match limit = %d\n", mid); */  
             min = mid;  
             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 */  
           }  
2205    
2206          extra->flags &= ~PCRE_EXTRA_MATCH_LIMIT;          (void)check_match_limit(re, extra, bptr, len, start_offset,
2207              options|g_notempty, use_offsets, use_size_offsets,
2208              PCRE_EXTRA_MATCH_LIMIT, &(extra->match_limit),
2209              PCRE_ERROR_MATCHLIMIT, "match()");
2210    
2211            count = check_match_limit(re, extra, bptr, len, start_offset,
2212              options|g_notempty, use_offsets, use_size_offsets,
2213              PCRE_EXTRA_MATCH_LIMIT_RECURSION, &(extra->match_limit_recursion),
2214              PCRE_ERROR_RECURSIONLIMIT, "match() recursion");
2215          }          }
2216    
2217        /* If callout_data is set, use the interface with additional data */        /* If callout_data is set, use the interface with additional data */
# Line 1625  while (!done) Line 2237  while (!done)
2237        else if (all_use_dfa || use_dfa)        else if (all_use_dfa || use_dfa)
2238          {          {
2239          int workspace[1000];          int workspace[1000];
2240          count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,          count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2241            options | g_notempty, use_offsets, use_size_offsets, workspace,            options | g_notempty, use_offsets, use_size_offsets, workspace,
2242            sizeof(workspace)/sizeof(int));            sizeof(workspace)/sizeof(int));
2243          if (count == 0)          if (count == 0)
# Line 1651  while (!done) Line 2263  while (!done)
2263    
2264        if (count >= 0)        if (count >= 0)
2265          {          {
2266          int i;          int i, maxcount;
2267    
2268    #if !defined NODFA
2269            if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else
2270    #endif
2271              maxcount = use_size_offsets/3;
2272    
2273            /* This is a check against a lunatic return value. */
2274    
2275            if (count > maxcount)
2276              {
2277              fprintf(outfile,
2278                "** PCRE error: returned count %d is too big for offset size %d\n",
2279                count, use_size_offsets);
2280              count = use_size_offsets/3;
2281              if (do_g || do_G)
2282                {
2283                fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G');
2284                do_g = do_G = FALSE;        /* Break g/G loop */
2285                }
2286              }
2287    
2288          for (i = 0; i < count * 2; i += 2)          for (i = 0; i < count * 2; i += 2)
2289            {            {
2290            if (use_offsets[i] < 0)            if (use_offsets[i] < 0)
# Line 1675  while (!done) Line 2308  while (!done)
2308              }              }
2309            }            }
2310    
2311            if (markptr != NULL) fprintf(outfile, "MK: %s\n", markptr);
2312    
2313          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2314            {            {
2315            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
2316              {              {
2317              char copybuffer[16];              char copybuffer[256];
2318              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,
2319                i, copybuffer, sizeof(copybuffer));                i, copybuffer, sizeof(copybuffer));
2320              if (rc < 0)              if (rc < 0)
# Line 1689  while (!done) Line 2324  while (!done)
2324              }              }
2325            }            }
2326    
2327            for (copynamesptr = copynames;
2328                 *copynamesptr != 0;
2329                 copynamesptr += (int)strlen((char*)copynamesptr) + 1)
2330              {
2331              char copybuffer[256];
2332              int rc = pcre_copy_named_substring(re, (char *)bptr, use_offsets,
2333                count, (char *)copynamesptr, copybuffer, sizeof(copybuffer));
2334              if (rc < 0)
2335                fprintf(outfile, "copy substring %s failed %d\n", copynamesptr, rc);
2336              else
2337                fprintf(outfile, "  C %s (%d) %s\n", copybuffer, rc, copynamesptr);
2338              }
2339    
2340          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2341            {            {
2342            if ((getstrings & (1 << i)) != 0)            if ((getstrings & (1 << i)) != 0)
# Line 1701  while (!done) Line 2349  while (!done)
2349              else              else
2350                {                {
2351                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
               /* free((void *)substring); */  
2352                pcre_free_substring(substring);                pcre_free_substring(substring);
2353                }                }
2354              }              }
2355            }            }
2356    
2357            for (getnamesptr = getnames;
2358                 *getnamesptr != 0;
2359                 getnamesptr += (int)strlen((char*)getnamesptr) + 1)
2360              {
2361              const char *substring;
2362              int rc = pcre_get_named_substring(re, (char *)bptr, use_offsets,
2363                count, (char *)getnamesptr, &substring);
2364              if (rc < 0)
2365                fprintf(outfile, "copy substring %s failed %d\n", getnamesptr, rc);
2366              else
2367                {
2368                fprintf(outfile, "  G %s (%d) %s\n", substring, rc, getnamesptr);
2369                pcre_free_substring(substring);
2370                }
2371              }
2372    
2373          if (getlist)          if (getlist)
2374            {            {
2375            const char **stringlist;            const char **stringlist;
# Line 1730  while (!done) Line 2393  while (!done)
2393    
2394        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2395          {          {
2396          fprintf(outfile, "Partial match");          if (markptr == NULL) fprintf(outfile, "Partial match");
2397  #if !defined NODFA            else fprintf(outfile, "Partial match, mark=%s", markptr);
2398          if ((all_use_dfa || use_dfa) && use_size_offsets > 2)          if (use_size_offsets > 1)
2399            fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],            {
2400              bptr + use_offsets[0]);            fprintf(outfile, ": ");
2401  #endif            pchars(bptr + use_offsets[0], use_offsets[1] - use_offsets[0],
2402                outfile);
2403              }
2404          fprintf(outfile, "\n");          fprintf(outfile, "\n");
2405          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2406          }          }
2407    
2408        /* 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
2409        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
2410        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
2411        matching, the advance must be one character, not one byte. Fudge the        string - that was checked before setting g_notempty.
2412        offset values to achieve this. We won't be at the end of the string -  
2413        that was checked before setting g_notempty. */        Complication arises in the case when the newline option is "any" or
2414          "anycrlf". If the previous match was at the end of a line terminated by
2415          CRLF, an advance of one character just passes the \r, whereas we should
2416          prefer the longer newline sequence, as does the code in pcre_exec().
2417          Fudge the offset value to achieve this.
2418    
2419          Otherwise, in the case of UTF-8 matching, the advance must be one
2420          character, not one byte. */
2421    
2422        else        else
2423          {          {
2424          if (g_notempty != 0)          if (g_notempty != 0)
2425            {            {
2426            int onechar = 1;            int onechar = 1;
2427              unsigned int obits = ((real_pcre *)re)->options;
2428            use_offsets[0] = start_offset;            use_offsets[0] = start_offset;
2429            if (use_utf8)            if ((obits & PCRE_NEWLINE_BITS) == 0)
2430                {
2431                int d;
2432                (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);
2433                /* Note that these values are always the ASCII ones, even in
2434                EBCDIC environments. CR = 13, NL = 10. */
2435                obits = (d == 13)? PCRE_NEWLINE_CR :
2436                        (d == 10)? PCRE_NEWLINE_LF :
2437                        (d == (13<<8 | 10))? PCRE_NEWLINE_CRLF :
2438                        (d == -2)? PCRE_NEWLINE_ANYCRLF :
2439                        (d == -1)? PCRE_NEWLINE_ANY : 0;
2440                }
2441              if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2442                   (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2443                  &&
2444                  start_offset < len - 1 &&
2445                  bptr[start_offset] == '\r' &&
2446                  bptr[start_offset+1] == '\n')
2447                onechar++;
2448              else if (use_utf8)
2449              {              {
2450              while (start_offset + onechar < len)              while (start_offset + onechar < len)
2451                {                {
# Line 1769  while (!done) Line 2461  while (!done)
2461            {            {
2462            if (count == PCRE_ERROR_NOMATCH)            if (count == PCRE_ERROR_NOMATCH)
2463              {              {
2464              if (gmatched == 0) fprintf(outfile, "No match\n");              if (gmatched == 0)
2465                  {
2466                  if (markptr == NULL) fprintf(outfile, "No match\n");
2467                    else fprintf(outfile, "No match, mark = %s\n", markptr);
2468                  }
2469              }              }
2470            else fprintf(outfile, "Error %d\n", count);            else fprintf(outfile, "Error %d\n", count);
2471            break;  /* Out of the /g loop */            break;  /* Out of the /g loop */
# Line 1781  while (!done) Line 2477  while (!done)
2477        if (!do_g && !do_G) break;        if (!do_g && !do_G) break;
2478    
2479        /* If we have matched an empty string, first check to see if we are at        /* If we have matched an empty string, first check to see if we are at
2480        the end of the subject. If so, the /g loop is over. Otherwise, mimic        the end of the subject. If so, the /g loop is over. Otherwise, mimic what
2481        what Perl's /g options does. This turns out to be rather cunning. First        Perl's /g options does. This turns out to be rather cunning. First we set
2482        we set PCRE_NOTEMPTY and PCRE_ANCHORED and try the match again at the        PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED and try the match again at the
2483        same point. If this fails (picked up above) we advance to the next        same point. If this fails (picked up above) we advance to the next
2484        character. */        character. */
2485    
2486        g_notempty = 0;        g_notempty = 0;
2487    
2488        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2489          {          {
2490          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
2491          g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED;          g_notempty = PCRE_NOTEMPTY_ATSTART | PCRE_ANCHORED;
2492          }          }
2493    
2494        /* For /g, update the start offset, leaving the rest alone */        /* For /g, update the start offset, leaving the rest alone */
# Line 1806  while (!done) Line 2503  while (!done)
2503          len -= use_offsets[1];          len -= use_offsets[1];
2504          }          }
2505        }  /* End of loop for /g and /G */        }  /* End of loop for /g and /G */
2506    
2507        NEXT_DATA: continue;
2508      }    /* End of loop for data lines */      }    /* End of loop for data lines */
2509    
2510    CONTINUE:    CONTINUE:
# Line 1820  while (!done) Line 2519  while (!done)
2519      {      {
2520      new_free((void *)tables);      new_free((void *)tables);
2521      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");
2522        locale_set = 0;
2523      }      }
2524    }    }
2525    

Legend:
Removed from v.79  
changed lines
  Added in v.516

  ViewVC Help
Powered by ViewVC 1.1.5