/[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 537 by ph10, Tue Jun 8 15:28:26 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 = (int)(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. Print
745    a message and return 0 if there is no match.
746    
747    Arguments:
748      p           points after the leading '<'
749      f           file for error message
750    
751    Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
752    */
753    
754    static int
755    check_newline(uschar *p, FILE *f)
756    {
757    if (strncmpic(p, (uschar *)"cr>", 3) == 0) return PCRE_NEWLINE_CR;
758    if (strncmpic(p, (uschar *)"lf>", 3) == 0) return PCRE_NEWLINE_LF;
759    if (strncmpic(p, (uschar *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
760    if (strncmpic(p, (uschar *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF;
761    if (strncmpic(p, (uschar *)"any>", 4) == 0) return PCRE_NEWLINE_ANY;
762    if (strncmpic(p, (uschar *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF;
763    if (strncmpic(p, (uschar *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE;
764    fprintf(f, "Unknown newline type at: <%s\n", p);
765    return 0;
766    }
767    
768    
769    
770    /*************************************************
771    *             Usage function                     *
772    *************************************************/
773    
774    static void
775    usage(void)
776    {
777    printf("Usage:     pcretest [options] [<input file> [<output file>]]\n\n");
778    printf("Input and output default to stdin and stdout.\n");
779    #ifdef SUPPORT_LIBREADLINE
780    printf("If input is a terminal, readline() is used to read from it.\n");
781    #else
782    printf("This version of pcretest is not linked with readline().\n");
783    #endif
784    printf("\nOptions:\n");
785    printf("  -b       show compiled code (bytecode)\n");
786    printf("  -C       show PCRE compile-time options and exit\n");
787    printf("  -d       debug: show compiled code and information (-b and -i)\n");
788    #if !defined NODFA
789    printf("  -dfa     force DFA matching for all subjects\n");
790    #endif
791    printf("  -help    show usage information\n");
792    printf("  -i       show information about compiled patterns\n"
793           "  -M       find MATCH_LIMIT minimum for each subject\n"
794           "  -m       output memory used information\n"
795           "  -o <n>   set size of offsets vector to <n>\n");
796    #if !defined NOPOSIX
797    printf("  -p       use POSIX interface\n");
798    #endif
799    printf("  -q       quiet: do not output PCRE version number at start\n");
800    printf("  -S <n>   set stack size to <n> megabytes\n");
801    printf("  -s       output store (memory) used information\n"
802           "  -t       time compilation and execution\n");
803    printf("  -t <n>   time compilation and execution, repeating <n> times\n");
804    printf("  -tm      time execution (matching) only\n");
805    printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
806    }
807    
808    
809    
810    /*************************************************
811  *                Main Program                    *  *                Main Program                    *
812  *************************************************/  *************************************************/
813    
# Line 430  int main(int argc, char **argv) Line 820  int main(int argc, char **argv)
820  FILE *infile = stdin;  FILE *infile = stdin;
821  int options = 0;  int options = 0;
822  int study_options = 0;  int study_options = 0;
823    int default_find_match_limit = FALSE;
824  int op = 1;  int op = 1;
825  int timeit = 0;  int timeit = 0;
826    int timeitm = 0;
827  int showinfo = 0;  int showinfo = 0;
828  int showstore = 0;  int showstore = 0;
829    int quiet = 0;
830  int size_offsets = 45;  int size_offsets = 45;
831  int size_offsets_max;  int size_offsets_max;
832  int *offsets = NULL;  int *offsets = NULL;
# Line 444  int debug = 0; Line 837  int debug = 0;
837  int done = 0;  int done = 0;
838  int all_use_dfa = 0;  int all_use_dfa = 0;
839  int yield = 0;  int yield = 0;
840    int stack_size;
841    
842    /* These vectors store, end-to-end, a list of captured substring names. Assume
843    that 1024 is plenty long enough for the few names we'll be testing. */
844    
845    uschar copynames[1024];
846    uschar getnames[1024];
847    
848  unsigned char *buffer;  uschar *copynamesptr;
849  unsigned char *dbuffer;  uschar *getnamesptr;
850    
851  /* Get buffers from malloc() so that Electric Fence will check their misuse  /* Get buffers from malloc() so that Electric Fence will check their misuse
852  when I am debugging. */  when I am debugging. They grow automatically when very long lines are read. */
853    
854  buffer = (unsigned char *)malloc(BUFFER_SIZE);  buffer = (unsigned char *)malloc(buffer_size);
855  dbuffer = (unsigned char *)malloc(DBUFFER_SIZE);  dbuffer = (unsigned char *)malloc(buffer_size);
856  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.  */  
857    
858  #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) */  
859    
860  outfile = stdout;  outfile = stdout;
861    
862    /* The following  _setmode() stuff is some Windows magic that tells its runtime
863    library to translate CRLF into a single LF character. At least, that's what
864    I've been told: never having used Windows I take this all on trust. Originally
865    it set 0x8000, but then I was advised that _O_BINARY was better. */
866    
867    #if defined(_WIN32) || defined(WIN32)
868    _setmode( _fileno( stdout ), _O_BINARY );
869    #endif
870    
871  /* Scan options */  /* Scan options */
872    
873  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
# Line 473  while (argc > 1 && argv[op][0] == '-') Line 876  while (argc > 1 && argv[op][0] == '-')
876    
877    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
878      showstore = 1;      showstore = 1;
879    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-q") == 0) quiet = 1;
880      else if (strcmp(argv[op], "-b") == 0) debug = 1;
881    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
882    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
883      else if (strcmp(argv[op], "-M") == 0) default_find_match_limit = TRUE;
884  #if !defined NODFA  #if !defined NODFA
885    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
886  #endif  #endif
# Line 486  while (argc > 1 && argv[op][0] == '-') Line 891  while (argc > 1 && argv[op][0] == '-')
891      op++;      op++;
892      argc--;      argc--;
893      }      }
894      else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
895        {
896        int both = argv[op][2] == 0;
897        int temp;
898        if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr),
899                         *endptr == 0))
900          {
901          timeitm = temp;
902          op++;
903          argc--;
904          }
905        else timeitm = LOOPREPEAT;
906        if (both) timeit = timeitm;
907        }
908      else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
909          ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
910            *endptr == 0))
911        {
912    #if defined(_WIN32) || defined(WIN32)
913        printf("PCRE: -S not supported on this OS\n");
914        exit(1);
915    #else
916        int rc;
917        struct rlimit rlim;
918        getrlimit(RLIMIT_STACK, &rlim);
919        rlim.rlim_cur = stack_size * 1024 * 1024;
920        rc = setrlimit(RLIMIT_STACK, &rlim);
921        if (rc != 0)
922          {
923        printf("PCRE: setrlimit() failed with error %d\n", rc);
924        exit(1);
925          }
926        op++;
927        argc--;
928    #endif
929        }
930  #if !defined NOPOSIX  #if !defined NOPOSIX
931    else if (strcmp(argv[op], "-p") == 0) posix = 1;    else if (strcmp(argv[op], "-p") == 0) posix = 1;
932  #endif  #endif
933    else if (strcmp(argv[op], "-C") == 0)    else if (strcmp(argv[op], "-C") == 0)
934      {      {
935      int rc;      int rc;
936        unsigned long int lrc;
937      printf("PCRE version %s\n", pcre_version());      printf("PCRE version %s\n", pcre_version());
938      printf("Compiled with\n");      printf("Compiled with\n");
939      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
# Line 499  while (argc > 1 && argv[op][0] == '-') Line 941  while (argc > 1 && argv[op][0] == '-')
941      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
942      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
943      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
944      printf("  Newline character is %s\n", (rc == '\r')? "CR" : "LF");      /* Note that these values are always the ASCII values, even
945        in EBCDIC environments. CR is 13 and NL is 10. */
946        printf("  Newline sequence is %s\n", (rc == 13)? "CR" :
947          (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
948          (rc == -2)? "ANYCRLF" :
949          (rc == -1)? "ANY" : "???");
950        (void)pcre_config(PCRE_CONFIG_BSR, &rc);
951        printf("  \\R matches %s\n", rc? "CR, LF, or CRLF only" :
952                                         "all Unicode newlines");
953      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);
954      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
955      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
956      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
957      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &lrc);
958      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %ld\n", lrc);
959        (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &lrc);
960        printf("  Default recursion depth limit = %ld\n", lrc);
961      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
962      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
963      exit(0);      goto EXIT;
964        }
965      else if (strcmp(argv[op], "-help") == 0 ||
966               strcmp(argv[op], "--help") == 0)
967        {
968        usage();
969        goto EXIT;
970      }      }
971    else    else
972      {      {
973      printf("** Unknown or malformed option %s\n", argv[op]);      printf("** Unknown or malformed option %s\n", argv[op]);
974      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");  
975      yield = 1;      yield = 1;
976      goto EXIT;      goto EXIT;
977      }      }
# Line 541  offsets = (int *)malloc(size_offsets_max Line 986  offsets = (int *)malloc(size_offsets_max
986  if (offsets == NULL)  if (offsets == NULL)
987    {    {
988    printf("** Failed to get %d bytes of memory for offsets vector\n",    printf("** Failed to get %d bytes of memory for offsets vector\n",
989      size_offsets_max * sizeof(int));      (int)(size_offsets_max * sizeof(int)));
990    yield = 1;    yield = 1;
991    goto EXIT;    goto EXIT;
992    }    }
# Line 550  if (offsets == NULL) Line 995  if (offsets == NULL)
995    
996  if (argc > 1)  if (argc > 1)
997    {    {
998    infile = fopen(argv[op], "rb");    infile = fopen(argv[op], INPUT_MODE);
999    if (infile == NULL)    if (infile == NULL)
1000      {      {
1001      printf("** Failed to open %s\n", argv[op]);      printf("** Failed to open %s\n", argv[op]);
# Line 561  if (argc > 1) Line 1006  if (argc > 1)
1006    
1007  if (argc > 2)  if (argc > 2)
1008    {    {
1009    outfile = fopen(argv[op+1], "wb");    outfile = fopen(argv[op+1], OUTPUT_MODE);
1010    if (outfile == NULL)    if (outfile == NULL)
1011      {      {
1012      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 1022  pcre_free = new_free;
1022  pcre_stack_malloc = stack_malloc;  pcre_stack_malloc = stack_malloc;
1023  pcre_stack_free = stack_free;  pcre_stack_free = stack_free;
1024    
1025  /* Heading line, then prompt for first regex if stdin */  /* Heading line unless quiet, then prompt for first regex if stdin */
1026    
1027  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  if (!quiet) fprintf(outfile, "PCRE version %s\n\n", pcre_version());
1028    
1029  /* Main loop */  /* Main loop */
1030    
# Line 594  while (!done) Line 1039  while (!done)
1039  #endif  #endif
1040    
1041    const char *error;    const char *error;
1042      unsigned char *markptr;
1043    unsigned char *p, *pp, *ppp;    unsigned char *p, *pp, *ppp;
1044    unsigned char *to_file = NULL;    unsigned char *to_file = NULL;
1045    const unsigned char *tables = NULL;    const unsigned char *tables = NULL;
1046    unsigned long int true_size, true_study_size = 0;    unsigned long int true_size, true_study_size = 0;
1047    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
1048      int do_mark = 0;
1049    int do_study = 0;    int do_study = 0;
1050    int do_debug = debug;    int do_debug = debug;
1051    int do_G = 0;    int do_G = 0;
# Line 606  while (!done) Line 1053  while (!done)
1053    int do_showinfo = showinfo;    int do_showinfo = showinfo;
1054    int do_showrest = 0;    int do_showrest = 0;
1055    int do_flip = 0;    int do_flip = 0;
1056    int erroroffset, len, delimiter;    int erroroffset, len, delimiter, poffset;
1057    
1058    use_utf8 = 0;    use_utf8 = 0;
1059      debug_lengths = 1;
1060    
1061    if (infile == stdin) printf("  re> ");    if (extend_inputline(infile, buffer, "  re> ") == NULL) break;
   if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break;  
1062    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1063    fflush(outfile);    fflush(outfile);
1064    
# Line 623  while (!done) Line 1070  while (!done)
1070    
1071    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
1072      {      {
1073      unsigned long int magic;      unsigned long int magic, get_options;
1074      uschar sbuf[8];      uschar sbuf[8];
1075      FILE *f;      FILE *f;
1076    
# Line 671  while (!done) Line 1118  while (!done)
1118    
1119      /* Need to know if UTF-8 for printing data strings */      /* Need to know if UTF-8 for printing data strings */
1120    
1121      new_info(re, NULL, PCRE_INFO_OPTIONS, &options);      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1122      use_utf8 = (options & PCRE_UTF8) != 0;      use_utf8 = (get_options & PCRE_UTF8) != 0;
1123    
1124      /* Now see if there is any following study data */      /* Now see if there is any following study data */
1125    
# Line 711  while (!done) Line 1158  while (!done)
1158    
1159    if (isalnum(delimiter) || delimiter == '\\')    if (isalnum(delimiter) || delimiter == '\\')
1160      {      {
1161      fprintf(outfile, "** Delimiter must not be alphameric or \\\n");      fprintf(outfile, "** Delimiter must not be alphanumeric or \\\n");
1162      goto SKIP_DATA;      goto SKIP_DATA;
1163      }      }
1164    
1165    pp = p;    pp = p;
1166      poffset = (int)(p - buffer);
1167    
1168    for(;;)    for(;;)
1169      {      {
# Line 726  while (!done) Line 1174  while (!done)
1174        pp++;        pp++;
1175        }        }
1176      if (*pp != 0) break;      if (*pp != 0) break;
1177        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)  
1178        {        {
1179        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
1180        done = 1;        done = 1;
# Line 744  while (!done) Line 1183  while (!done)
1183      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
1184      }      }
1185    
1186      /* The buffer may have moved while being extended; reset the start of data
1187      pointer to the correct relative point in the buffer. */
1188    
1189      p = buffer + poffset;
1190    
1191    /* If the first character after the delimiter is backslash, make    /* If the first character after the delimiter is backslash, make
1192    the pattern end with backslash. This is purely to provide a way    the pattern end with backslash. This is purely to provide a way
1193    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 1219  while (!done)
1219    
1220        case '+': do_showrest = 1; break;        case '+': do_showrest = 1; break;
1221        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
1222          case 'B': do_debug = 1; break;
1223        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
1224        case 'D': do_debug = do_showinfo = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
1225        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
1226        case 'F': do_flip = 1; break;        case 'F': do_flip = 1; break;
1227        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1228        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1229          case 'J': options |= PCRE_DUPNAMES; break;
1230          case 'K': do_mark = 1; break;
1231        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1232        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1233    
# Line 790  while (!done) Line 1237  while (!done)
1237    
1238        case 'S': do_study = 1; break;        case 'S': do_study = 1; break;
1239        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1240          case 'W': options |= PCRE_UCP; 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 (strncmpic(pp, (uschar *)"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_UCP) != 0) cflags |= REG_UCP;
1313        if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY;
1314    
1315      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1316    
1317      /* 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 1319  while (!done)
1319    
1320      if (rc != 0)      if (rc != 0)
1321        {        {
1322        (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);        (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1323        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
1324        goto SKIP_DATA;        goto SKIP_DATA;
1325        }        }
# Line 858  while (!done) Line 1331  while (!done)
1331  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1332    
1333      {      {
1334      if (timeit)      unsigned long int get_options;
1335    
1336        if (timeit > 0)
1337        {        {
1338        register int i;        register int i;
1339        clock_t time_taken;        clock_t time_taken;
1340        clock_t start_time = clock();        clock_t start_time = clock();
1341        for (i = 0; i < LOOPREPEAT; i++)        for (i = 0; i < timeit; i++)
1342          {          {
1343          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1344          if (re != NULL) free(re);          if (re != NULL) free(re);
1345          }          }
1346        time_taken = clock() - start_time;        time_taken = clock() - start_time;
1347        fprintf(outfile, "Compile time %.3f milliseconds\n",        fprintf(outfile, "Compile time %.4f milliseconds\n",
1348          (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /          (((double)time_taken * 1000.0) / (double)timeit) /
1349            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
1350        }        }
1351    
# Line 887  while (!done) Line 1362  while (!done)
1362          {          {
1363          for (;;)          for (;;)
1364            {            {
1365            if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)            if (extend_inputline(infile, buffer, NULL) == NULL)
1366              {              {
1367              done = 1;              done = 1;
1368              goto CONTINUE;              goto CONTINUE;
# Line 901  while (!done) Line 1376  while (!done)
1376        goto CONTINUE;        goto CONTINUE;
1377        }        }
1378    
1379      /* Compilation succeeded; print data if required. There are now two      /* Compilation succeeded. It is now possible to set the UTF-8 option from
1380      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
1381      returns only limited data. Check that it agrees with the newer one. */      lines. */
1382    
1383        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1384        if ((get_options & PCRE_UTF8) != 0) use_utf8 = 1;
1385    
1386        /* Print information if required. There are now two info-returning
1387        functions. The old one has a limited interface and returns only limited
1388        data. Check that it agrees with the newer one. */
1389    
1390      if (log_store)      if (log_store)
1391        fprintf(outfile, "Memory allocation (code space): %d\n",        fprintf(outfile, "Memory allocation (code space): %d\n",
# Line 922  while (!done) Line 1404  while (!done)
1404    
1405      if (do_study)      if (do_study)
1406        {        {
1407        if (timeit)        if (timeit > 0)
1408          {          {
1409          register int i;          register int i;
1410          clock_t time_taken;          clock_t time_taken;
1411          clock_t start_time = clock();          clock_t start_time = clock();
1412          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeit; i++)
1413            extra = pcre_study(re, study_options, &error);            extra = pcre_study(re, study_options, &error);
1414          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1415          if (extra != NULL) free(extra);          if (extra != NULL) free(extra);
1416          fprintf(outfile, "  Study time %.3f milliseconds\n",          fprintf(outfile, "  Study time %.4f milliseconds\n",
1417            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeit) /
1418              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
1419          }          }
1420        extra = pcre_study(re, study_options, &error);        extra = pcre_study(re, study_options, &error);
# Line 942  while (!done) Line 1424  while (!done)
1424          true_study_size = ((pcre_study_data *)(extra->study_data))->size;          true_study_size = ((pcre_study_data *)(extra->study_data))->size;
1425        }        }
1426    
1427        /* If /K was present, we set up for handling MARK data. */
1428    
1429        if (do_mark)
1430          {
1431          if (extra == NULL)
1432            {
1433            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1434            extra->flags = 0;
1435            }
1436          extra->mark = &markptr;
1437          extra->flags |= PCRE_EXTRA_MARK;
1438          }
1439    
1440      /* 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
1441      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
1442      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 1445  while (!done)
1445      if (do_flip)      if (do_flip)
1446        {        {
1447        real_pcre *rre = (real_pcre *)re;        real_pcre *rre = (real_pcre *)re;
1448        rre->magic_number = byteflip(rre->magic_number, sizeof(rre->magic_number));        rre->magic_number =
1449            byteflip(rre->magic_number, sizeof(rre->magic_number));
1450        rre->size = byteflip(rre->size, sizeof(rre->size));        rre->size = byteflip(rre->size, sizeof(rre->size));
1451        rre->options = byteflip(rre->options, sizeof(rre->options));        rre->options = byteflip(rre->options, sizeof(rre->options));
1452        rre->top_bracket = byteflip(rre->top_bracket, sizeof(rre->top_bracket));        rre->flags = (pcre_uint16)byteflip(rre->flags, sizeof(rre->flags));
1453        rre->top_backref = byteflip(rre->top_backref, sizeof(rre->top_backref));        rre->top_bracket =
1454        rre->first_byte = byteflip(rre->first_byte, sizeof(rre->first_byte));          (pcre_uint16)byteflip(rre->top_bracket, sizeof(rre->top_bracket));
1455        rre->req_byte = byteflip(rre->req_byte, sizeof(rre->req_byte));        rre->top_backref =
1456        rre->name_table_offset = byteflip(rre->name_table_offset,          (pcre_uint16)byteflip(rre->top_backref, sizeof(rre->top_backref));
1457          rre->first_byte =
1458            (pcre_uint16)byteflip(rre->first_byte, sizeof(rre->first_byte));
1459          rre->req_byte =
1460            (pcre_uint16)byteflip(rre->req_byte, sizeof(rre->req_byte));
1461          rre->name_table_offset = (pcre_uint16)byteflip(rre->name_table_offset,
1462          sizeof(rre->name_table_offset));          sizeof(rre->name_table_offset));
1463        rre->name_entry_size = byteflip(rre->name_entry_size,        rre->name_entry_size = (pcre_uint16)byteflip(rre->name_entry_size,
1464          sizeof(rre->name_entry_size));          sizeof(rre->name_entry_size));
1465        rre->name_count = byteflip(rre->name_count, sizeof(rre->name_count));        rre->name_count = (pcre_uint16)byteflip(rre->name_count,
1466            sizeof(rre->name_count));
1467    
1468        if (extra != NULL)        if (extra != NULL)
1469          {          {
1470          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
1471          rsd->size = byteflip(rsd->size, sizeof(rsd->size));          rsd->size = byteflip(rsd->size, sizeof(rsd->size));
1472          rsd->options = byteflip(rsd->options, sizeof(rsd->options));          rsd->flags = byteflip(rsd->flags, sizeof(rsd->flags));
1473            rsd->minlength = byteflip(rsd->minlength, sizeof(rsd->minlength));
1474          }          }
1475        }        }
1476    
# Line 975  while (!done) Line 1478  while (!done)
1478    
1479      SHOW_INFO:      SHOW_INFO:
1480    
1481        if (do_debug)
1482          {
1483          fprintf(outfile, "------------------------------------------------------------------\n");
1484          pcre_printint(re, outfile, debug_lengths);
1485          }
1486    
1487        /* We already have the options in get_options (see above) */
1488    
1489      if (do_showinfo)      if (do_showinfo)
1490        {        {
1491        unsigned long int get_options, all_options;        unsigned long int all_options;
1492  #if !defined NOINFOCHECK  #if !defined NOINFOCHECK
1493        int old_first_char, old_options, old_count;        int old_first_char, old_options, old_count;
1494  #endif  #endif
1495        int count, backrefmax, first_char, need_char;        int count, backrefmax, first_char, need_char, okpartial, jchanged,
1496            hascrorlf;
1497        int nameentrysize, namecount;        int nameentrysize, namecount;
1498        const uschar *nametable;        const uschar *nametable;
1499    
       if (do_debug)  
         {  
         fprintf(outfile, "------------------------------------------------------------------\n");  
         _pcre_printint(re, outfile);  
         }  
   
       new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);  
1500        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1501        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
1502        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);
# Line 1000  while (!done) Line 1505  while (!done)
1505        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);
1506        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);
1507        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);
1508          new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial);
1509          new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged);
1510          new_info(re, NULL, PCRE_INFO_HASCRORLF, &hascrorlf);
1511    
1512  #if !defined NOINFOCHECK  #if !defined NOINFOCHECK
1513        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 1549  while (!done)
1549            }            }
1550          }          }
1551    
1552        /* The NOPARTIAL bit is a private bit in the options, so we have        if (!okpartial) fprintf(outfile, "Partial matching not supported\n");
1553        to fish it out via out back door */        if (hascrorlf) fprintf(outfile, "Contains explicit CR or LF match\n");
1554    
1555        all_options = ((real_pcre *)re)->options;        all_options = ((real_pcre *)re)->options;
1556        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");  
1557    
1558        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1559          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%s\n",
1560            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1561            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1562            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
1563            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",
1564            ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",            ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",
1565            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",
1566              ((get_options & PCRE_BSR_ANYCRLF) != 0)? " bsr_anycrlf" : "",
1567              ((get_options & PCRE_BSR_UNICODE) != 0)? " bsr_unicode" : "",
1568            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
1569            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
1570            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1571              ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1572            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1573            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "");            ((get_options & PCRE_UCP) != 0)? " ucp" : "",
1574              ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1575              ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1576    
1577        if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0)        if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
1578          fprintf(outfile, "Case state changes\n");  
1579          switch (get_options & PCRE_NEWLINE_BITS)
1580            {
1581            case PCRE_NEWLINE_CR:
1582            fprintf(outfile, "Forced newline sequence: CR\n");
1583            break;
1584    
1585            case PCRE_NEWLINE_LF:
1586            fprintf(outfile, "Forced newline sequence: LF\n");
1587            break;
1588    
1589            case PCRE_NEWLINE_CRLF:
1590            fprintf(outfile, "Forced newline sequence: CRLF\n");
1591            break;
1592    
1593            case PCRE_NEWLINE_ANYCRLF:
1594            fprintf(outfile, "Forced newline sequence: ANYCRLF\n");
1595            break;
1596    
1597            case PCRE_NEWLINE_ANY:
1598            fprintf(outfile, "Forced newline sequence: ANY\n");
1599            break;
1600    
1601            default:
1602            break;
1603            }
1604    
1605        if (first_char == -1)        if (first_char == -1)
1606          {          {
1607          fprintf(outfile, "First char at start or follows \\n\n");          fprintf(outfile, "First char at start or follows newline\n");
1608          }          }
1609        else if (first_char < 0)        else if (first_char < 0)
1610          {          {
# Line 1083  while (!done) Line 1615  while (!done)
1615          int ch = first_char & 255;          int ch = first_char & 255;
1616          const char *caseless = ((first_char & REQ_CASELESS) == 0)?          const char *caseless = ((first_char & REQ_CASELESS) == 0)?
1617            "" : " (caseless)";            "" : " (caseless)";
1618          if (isprint(ch))          if (PRINTHEX(ch))
1619            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);
1620          else          else
1621            fprintf(outfile, "First char = %d%s\n", ch, caseless);            fprintf(outfile, "First char = %d%s\n", ch, caseless);
# Line 1098  while (!done) Line 1630  while (!done)
1630          int ch = need_char & 255;          int ch = need_char & 255;
1631          const char *caseless = ((need_char & REQ_CASELESS) == 0)?          const char *caseless = ((need_char & REQ_CASELESS) == 0)?
1632            "" : " (caseless)";            "" : " (caseless)";
1633          if (isprint(ch))          if (PRINTHEX(ch))
1634            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);
1635          else          else
1636            fprintf(outfile, "Need char = %d%s\n", ch, caseless);            fprintf(outfile, "Need char = %d%s\n", ch, caseless);
# Line 1116  while (!done) Line 1648  while (!done)
1648          else          else
1649            {            {
1650            uschar *start_bits = NULL;            uschar *start_bits = NULL;
1651            new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);            int minlength;
1652    
1653              new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength);
1654              fprintf(outfile, "Subject length lower bound = %d\n", minlength);
1655    
1656              new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);
1657            if (start_bits == NULL)            if (start_bits == NULL)
1658              fprintf(outfile, "No starting byte set\n");              fprintf(outfile, "No set of starting bytes\n");
1659            else            else
1660              {              {
1661              int i;              int i;
# Line 1134  while (!done) Line 1670  while (!done)
1670                    fprintf(outfile, "\n  ");                    fprintf(outfile, "\n  ");
1671                    c = 2;                    c = 2;
1672                    }                    }
1673                  if (isprint(i) && i != ' ')                  if (PRINTHEX(i) && i != ' ')
1674                    {                    {
1675                    fprintf(outfile, "%c ", i);                    fprintf(outfile, "%c ", i);
1676                    c += 2;                    c += 2;
# Line 1166  while (!done) Line 1702  while (!done)
1702        else        else
1703          {          {
1704          uschar sbuf[8];          uschar sbuf[8];
1705          sbuf[0] = (true_size >> 24)  & 255;          sbuf[0] = (uschar)((true_size >> 24) & 255);
1706          sbuf[1] = (true_size >> 16)  & 255;          sbuf[1] = (uschar)((true_size >> 16) & 255);
1707          sbuf[2] = (true_size >>  8)  & 255;          sbuf[2] = (uschar)((true_size >>  8) & 255);
1708          sbuf[3] = (true_size)  & 255;          sbuf[3] = (uschar)((true_size) & 255);
1709    
1710          sbuf[4] = (true_study_size >> 24)  & 255;          sbuf[4] = (uschar)((true_study_size >> 24) & 255);
1711          sbuf[5] = (true_study_size >> 16)  & 255;          sbuf[5] = (uschar)((true_study_size >> 16) & 255);
1712          sbuf[6] = (true_study_size >>  8)  & 255;          sbuf[6] = (uschar)((true_study_size >>  8) & 255);
1713          sbuf[7] = (true_study_size)  & 255;          sbuf[7] = (uschar)((true_study_size) & 255);
1714    
1715          if (fwrite(sbuf, 1, 8, f) < 8 ||          if (fwrite(sbuf, 1, 8, f) < 8 ||
1716              fwrite(re, 1, true_size, f) < true_size)              fwrite(re, 1, true_size, f) < true_size)
# Line 1193  while (!done) Line 1729  while (!done)
1729                  strerror(errno));                  strerror(errno));
1730                }                }
1731              else fprintf(outfile, "Study data written to %s\n", to_file);              else fprintf(outfile, "Study data written to %s\n", to_file);
1732    
1733              }              }
1734            }            }
1735          fclose(f);          fclose(f);
# Line 1209  while (!done) Line 1746  while (!done)
1746    
1747    for (;;)    for (;;)
1748      {      {
1749      unsigned char *q;      uschar *q;
1750      unsigned char *bptr = dbuffer;      uschar *bptr;
1751      int *use_offsets = offsets;      int *use_offsets = offsets;
1752      int use_size_offsets = size_offsets;      int use_size_offsets = size_offsets;
1753      int callout_data = 0;      int callout_data = 0;
1754      int callout_data_set = 0;      int callout_data_set = 0;
1755      int count, c;      int count, c;
1756      int copystrings = 0;      int copystrings = 0;
1757      int find_match_limit = 0;      int find_match_limit = default_find_match_limit;
1758      int getstrings = 0;      int getstrings = 0;
1759      int getlist = 0;      int getlist = 0;
1760      int gmatched = 0;      int gmatched = 0;
# Line 1227  while (!done) Line 1764  while (!done)
1764    
1765      options = 0;      options = 0;
1766    
1767        *copynames = 0;
1768        *getnames = 0;
1769    
1770        copynamesptr = copynames;
1771        getnamesptr = getnames;
1772    
1773      pcre_callout = callout;      pcre_callout = callout;
1774      first_callout = 1;      first_callout = 1;
1775      callout_extra = 0;      callout_extra = 0;
# Line 1235  while (!done) Line 1778  while (!done)
1778      callout_fail_id = -1;      callout_fail_id = -1;
1779      show_malloc = 0;      show_malloc = 0;
1780    
1781      if (infile == stdin) printf("data> ");      if (extra != NULL) extra->flags &=
1782      if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)        ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
1783    
1784        len = 0;
1785        for (;;)
1786        {        {
1787        done = 1;        if (extend_inputline(infile, buffer + len, "data> ") == NULL)
1788        goto CONTINUE;          {
1789            if (len > 0)    /* Reached EOF without hitting a newline */
1790              {
1791              fprintf(outfile, "\n");
1792              break;
1793              }
1794            done = 1;
1795            goto CONTINUE;
1796            }
1797          if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1798          len = (int)strlen((char *)buffer);
1799          if (buffer[len-1] == '\n') break;
1800        }        }
     if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);  
1801    
     len = (int)strlen((char *)buffer);  
1802      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
1803      buffer[len] = 0;      buffer[len] = 0;
1804      if (len == 0) break;      if (len == 0) break;
# Line 1251  while (!done) Line 1806  while (!done)
1806      p = buffer;      p = buffer;
1807      while (isspace(*p)) p++;      while (isspace(*p)) p++;
1808    
1809      q = dbuffer;      bptr = q = dbuffer;
1810      while ((c = *p++) != 0)      while ((c = *p++) != 0)
1811        {        {
1812        int i = 0;        int i = 0;
# Line 1273  while (!done) Line 1828  while (!done)
1828          c -= '0';          c -= '0';
1829          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')
1830            c = c * 8 + *p++ - '0';            c = c * 8 + *p++ - '0';
1831    
1832    #if !defined NOUTF8
1833            if (use_utf8 && c > 255)
1834              {
1835              unsigned char buff8[8];
1836              int ii, utn;
1837              utn = ord2utf8(c, buff8);
1838              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1839              c = buff8[ii];   /* Last byte */
1840              }
1841    #endif
1842          break;          break;
1843    
1844          case 'x':          case 'x':
# Line 1290  while (!done) Line 1856  while (!done)
1856              {              {
1857              unsigned char buff8[8];              unsigned char buff8[8];
1858              int ii, utn;              int ii, utn;
1859              utn = _pcre_ord2utf8(c, buff8);              if (use_utf8)
1860              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];                {
1861              c = buff8[ii];   /* Last byte */                utn = ord2utf8(c, buff8);
1862                  for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1863                  c = buff8[ii];   /* Last byte */
1864                  }
1865                else
1866                 {
1867                 if (c > 255)
1868                   fprintf(outfile, "** Character \\x{%x} is greater than 255 and "
1869                     "UTF-8 mode is not enabled.\n"
1870                     "** Truncation will probably give the wrong result.\n", c);
1871                 }
1872              p = pt + 1;              p = pt + 1;
1873              break;              break;
1874              }              }
# Line 1334  while (!done) Line 1910  while (!done)
1910            }            }
1911          else if (isalnum(*p))          else if (isalnum(*p))
1912            {            {
1913            uschar name[256];            uschar *npp = copynamesptr;
           uschar *npp = name;  
1914            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1915              *npp++ = 0;
1916            *npp = 0;            *npp = 0;
1917            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)copynamesptr);
1918            if (n < 0)            if (n < 0)
1919              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", copynamesptr);
1920            else copystrings |= 1 << n;            copynamesptr = npp;
1921            }            }
1922          else if (*p == '+')          else if (*p == '+')
1923            {            {
# Line 1402  while (!done) Line 1978  while (!done)
1978            }            }
1979          else if (isalnum(*p))          else if (isalnum(*p))
1980            {            {
1981            uschar name[256];            uschar *npp = getnamesptr;
           uschar *npp = name;  
1982            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1983              *npp++ = 0;
1984            *npp = 0;            *npp = 0;
1985            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)getnamesptr);
1986            if (n < 0)            if (n < 0)
1987              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", getnamesptr);
1988            else getstrings |= 1 << n;            getnamesptr = npp;
1989            }            }
1990          continue;          continue;
1991    
# Line 1422  while (!done) Line 1998  while (!done)
1998          continue;          continue;
1999    
2000          case 'N':          case 'N':
2001          options |= PCRE_NOTEMPTY;          if ((options & PCRE_NOTEMPTY) != 0)
2002              options = (options & ~PCRE_NOTEMPTY) | PCRE_NOTEMPTY_ATSTART;
2003            else
2004              options |= PCRE_NOTEMPTY;
2005          continue;          continue;
2006    
2007          case 'O':          case 'O':
# Line 1435  while (!done) Line 2014  while (!done)
2014            if (offsets == NULL)            if (offsets == NULL)
2015              {              {
2016              printf("** Failed to get %d bytes of memory for offsets vector\n",              printf("** Failed to get %d bytes of memory for offsets vector\n",
2017                size_offsets_max * sizeof(int));                (int)(size_offsets_max * sizeof(int)));
2018              yield = 1;              yield = 1;
2019              goto EXIT;              goto EXIT;
2020              }              }
# Line 1445  while (!done) Line 2024  while (!done)
2024          continue;          continue;
2025    
2026          case 'P':          case 'P':
2027          options |= PCRE_PARTIAL;          options |= ((options & PCRE_PARTIAL_SOFT) == 0)?
2028              PCRE_PARTIAL_SOFT : PCRE_PARTIAL_HARD;
2029            continue;
2030    
2031            case 'Q':
2032            while(isdigit(*p)) n = n * 10 + *p++ - '0';
2033            if (extra == NULL)
2034              {
2035              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2036              extra->flags = 0;
2037              }
2038            extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
2039            extra->match_limit_recursion = n;
2040            continue;
2041    
2042            case 'q':
2043            while(isdigit(*p)) n = n * 10 + *p++ - '0';
2044            if (extra == NULL)
2045              {
2046              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2047              extra->flags = 0;
2048              }
2049            extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
2050            extra->match_limit = n;
2051          continue;          continue;
2052    
2053  #if !defined NODFA  #if !defined NODFA
# Line 1458  while (!done) Line 2060  while (!done)
2060          show_malloc = 1;          show_malloc = 1;
2061          continue;          continue;
2062    
2063            case 'Y':
2064            options |= PCRE_NO_START_OPTIMIZE;
2065            continue;
2066    
2067          case 'Z':          case 'Z':
2068          options |= PCRE_NOTEOL;          options |= PCRE_NOTEOL;
2069          continue;          continue;
# Line 1465  while (!done) Line 2071  while (!done)
2071          case '?':          case '?':
2072          options |= PCRE_NO_UTF8_CHECK;          options |= PCRE_NO_UTF8_CHECK;
2073          continue;          continue;
2074    
2075            case '<':
2076              {
2077              int x = check_newline(p, outfile);
2078              if (x == 0) goto NEXT_DATA;
2079              options |= x;
2080              while (*p++ != '>');
2081              }
2082            continue;
2083          }          }
2084        *q++ = c;        *q++ = c;
2085        }        }
2086      *q = 0;      *q = 0;
2087      len = q - dbuffer;      len = (int)(q - dbuffer);
2088    
2089        /* Move the data to the end of the buffer so that a read over the end of
2090        the buffer will be seen by valgrind, even if it doesn't cause a crash. If
2091        we are using the POSIX interface, we must include the terminating zero. */
2092    
2093    #if !defined NOPOSIX
2094        if (posix || do_posix)
2095          {
2096          memmove(bptr + buffer_size - len - 1, bptr, len + 1);
2097          bptr += buffer_size - len - 1;
2098          }
2099        else
2100    #endif
2101          {
2102          memmove(bptr + buffer_size - len, bptr, len);
2103          bptr += buffer_size - len;
2104          }
2105    
2106      if ((all_use_dfa || use_dfa) && find_match_limit)      if ((all_use_dfa || use_dfa) && find_match_limit)
2107        {        {
# Line 1490  while (!done) Line 2122  while (!done)
2122          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);
2123        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;
2124        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;
2125          if ((options & PCRE_NOTEMPTY) != 0) eflags |= REG_NOTEMPTY;
2126    
2127        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);
2128    
2129        if (rc != 0)        if (rc != 0)
2130          {          {
2131          (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);          (void)regerror(rc, &preg, (char *)buffer, buffer_size);
2132          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
2133          }          }
2134          else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
2135                  != 0)
2136            {
2137            fprintf(outfile, "Matched with REG_NOSUB\n");
2138            }
2139        else        else
2140          {          {
2141          size_t i;          size_t i;
# Line 1529  while (!done) Line 2167  while (!done)
2167    
2168      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
2169        {        {
2170        if (timeit)        markptr = NULL;
2171    
2172          if (timeitm > 0)
2173          {          {
2174          register int i;          register int i;
2175          clock_t time_taken;          clock_t time_taken;
# Line 1539  while (!done) Line 2179  while (!done)
2179          if (all_use_dfa || use_dfa)          if (all_use_dfa || use_dfa)
2180            {            {
2181            int workspace[1000];            int workspace[1000];
2182            for (i = 0; i < LOOPREPEAT; i++)            for (i = 0; i < timeitm; i++)
2183              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,              count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2184                options | g_notempty, use_offsets, use_size_offsets, workspace,                options | g_notempty, use_offsets, use_size_offsets, workspace,
2185                sizeof(workspace)/sizeof(int));                sizeof(workspace)/sizeof(int));
2186            }            }
2187          else          else
2188  #endif  #endif
2189    
2190          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeitm; i++)
2191            count = pcre_exec(re, extra, (char *)bptr, len,            count = pcre_exec(re, extra, (char *)bptr, len,
2192              start_offset, options | g_notempty, use_offsets, use_size_offsets);              start_offset, options | g_notempty, use_offsets, use_size_offsets);
2193    
2194          time_taken = clock() - start_time;          time_taken = clock() - start_time;
2195          fprintf(outfile, "Execute time %.3f milliseconds\n",          fprintf(outfile, "Execute time %.4f milliseconds\n",
2196            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeitm) /
2197              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
2198          }          }
2199    
2200        /* 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
2201        varying limits in order to find the minimum value. */        varying limits in order to find the minimum value for the match limit and
2202          for the recursion limit. */
2203    
2204        if (find_match_limit)        if (find_match_limit)
2205          {          {
         int min = 0;  
         int mid = 64;  
         int max = -1;  
   
2206          if (extra == NULL)          if (extra == NULL)
2207            {            {
2208            extra = (pcre_extra *)malloc(sizeof(pcre_extra));            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2209            extra->flags = 0;            extra->flags = 0;
2210            }            }
         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 */  
           }  
2211    
2212          extra->flags &= ~PCRE_EXTRA_MATCH_LIMIT;          (void)check_match_limit(re, extra, bptr, len, start_offset,
2213              options|g_notempty, use_offsets, use_size_offsets,
2214              PCRE_EXTRA_MATCH_LIMIT, &(extra->match_limit),
2215              PCRE_ERROR_MATCHLIMIT, "match()");
2216    
2217            count = check_match_limit(re, extra, bptr, len, start_offset,
2218              options|g_notempty, use_offsets, use_size_offsets,
2219              PCRE_EXTRA_MATCH_LIMIT_RECURSION, &(extra->match_limit_recursion),
2220              PCRE_ERROR_RECURSIONLIMIT, "match() recursion");
2221          }          }
2222    
2223        /* 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 2243  while (!done)
2243        else if (all_use_dfa || use_dfa)        else if (all_use_dfa || use_dfa)
2244          {          {
2245          int workspace[1000];          int workspace[1000];
2246          count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,          count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2247            options | g_notempty, use_offsets, use_size_offsets, workspace,            options | g_notempty, use_offsets, use_size_offsets, workspace,
2248            sizeof(workspace)/sizeof(int));            sizeof(workspace)/sizeof(int));
2249          if (count == 0)          if (count == 0)
# Line 1651  while (!done) Line 2269  while (!done)
2269    
2270        if (count >= 0)        if (count >= 0)
2271          {          {
2272          int i;          int i, maxcount;
2273    
2274    #if !defined NODFA
2275            if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else
2276    #endif
2277              maxcount = use_size_offsets/3;
2278    
2279            /* This is a check against a lunatic return value. */
2280    
2281            if (count > maxcount)
2282              {
2283              fprintf(outfile,
2284                "** PCRE error: returned count %d is too big for offset size %d\n",
2285                count, use_size_offsets);
2286              count = use_size_offsets/3;
2287              if (do_g || do_G)
2288                {
2289                fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G');
2290                do_g = do_G = FALSE;        /* Break g/G loop */
2291                }
2292              }
2293    
2294          for (i = 0; i < count * 2; i += 2)          for (i = 0; i < count * 2; i += 2)
2295            {            {
2296            if (use_offsets[i] < 0)            if (use_offsets[i] < 0)
# Line 1675  while (!done) Line 2314  while (!done)
2314              }              }
2315            }            }
2316    
2317            if (markptr != NULL) fprintf(outfile, "MK: %s\n", markptr);
2318    
2319          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2320            {            {
2321            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
2322              {              {
2323              char copybuffer[16];              char copybuffer[256];
2324              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,
2325                i, copybuffer, sizeof(copybuffer));                i, copybuffer, sizeof(copybuffer));
2326              if (rc < 0)              if (rc < 0)
# Line 1689  while (!done) Line 2330  while (!done)
2330              }              }
2331            }            }
2332    
2333            for (copynamesptr = copynames;
2334                 *copynamesptr != 0;
2335                 copynamesptr += (int)strlen((char*)copynamesptr) + 1)
2336              {
2337              char copybuffer[256];
2338              int rc = pcre_copy_named_substring(re, (char *)bptr, use_offsets,
2339                count, (char *)copynamesptr, copybuffer, sizeof(copybuffer));
2340              if (rc < 0)
2341                fprintf(outfile, "copy substring %s failed %d\n", copynamesptr, rc);
2342              else
2343                fprintf(outfile, "  C %s (%d) %s\n", copybuffer, rc, copynamesptr);
2344              }
2345    
2346          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2347            {            {
2348            if ((getstrings & (1 << i)) != 0)            if ((getstrings & (1 << i)) != 0)
# Line 1701  while (!done) Line 2355  while (!done)
2355              else              else
2356                {                {
2357                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
               /* free((void *)substring); */  
2358                pcre_free_substring(substring);                pcre_free_substring(substring);
2359                }                }
2360              }              }
2361            }            }
2362    
2363            for (getnamesptr = getnames;
2364                 *getnamesptr != 0;
2365                 getnamesptr += (int)strlen((char*)getnamesptr) + 1)
2366              {
2367              const char *substring;
2368              int rc = pcre_get_named_substring(re, (char *)bptr, use_offsets,
2369                count, (char *)getnamesptr, &substring);
2370              if (rc < 0)
2371                fprintf(outfile, "copy substring %s failed %d\n", getnamesptr, rc);
2372              else
2373                {
2374                fprintf(outfile, "  G %s (%d) %s\n", substring, rc, getnamesptr);
2375                pcre_free_substring(substring);
2376                }
2377              }
2378    
2379          if (getlist)          if (getlist)
2380            {            {
2381            const char **stringlist;            const char **stringlist;
# Line 1730  while (!done) Line 2399  while (!done)
2399    
2400        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2401          {          {
2402          fprintf(outfile, "Partial match");          if (markptr == NULL) fprintf(outfile, "Partial match");
2403  #if !defined NODFA            else fprintf(outfile, "Partial match, mark=%s", markptr);
2404          if ((all_use_dfa || use_dfa) && use_size_offsets > 2)          if (use_size_offsets > 1)
2405            fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],            {
2406              bptr + use_offsets[0]);            fprintf(outfile, ": ");
2407  #endif            pchars(bptr + use_offsets[0], use_offsets[1] - use_offsets[0],
2408                outfile);
2409              }
2410          fprintf(outfile, "\n");          fprintf(outfile, "\n");
2411          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2412          }          }
2413    
2414        /* 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
2415        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
2416        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
2417        matching, the advance must be one character, not one byte. Fudge the        string - that was checked before setting g_notempty.
2418        offset values to achieve this. We won't be at the end of the string -  
2419        that was checked before setting g_notempty. */        Complication arises in the case when the newline option is "any" or
2420          "anycrlf". If the previous match was at the end of a line terminated by
2421          CRLF, an advance of one character just passes the \r, whereas we should
2422          prefer the longer newline sequence, as does the code in pcre_exec().
2423          Fudge the offset value to achieve this.
2424    
2425          Otherwise, in the case of UTF-8 matching, the advance must be one
2426          character, not one byte. */
2427    
2428        else        else
2429          {          {
2430          if (g_notempty != 0)          if (g_notempty != 0)
2431            {            {
2432            int onechar = 1;            int onechar = 1;
2433              unsigned int obits = ((real_pcre *)re)->options;
2434            use_offsets[0] = start_offset;            use_offsets[0] = start_offset;
2435            if (use_utf8)            if ((obits & PCRE_NEWLINE_BITS) == 0)
2436                {
2437                int d;
2438                (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);
2439                /* Note that these values are always the ASCII ones, even in
2440                EBCDIC environments. CR = 13, NL = 10. */
2441                obits = (d == 13)? PCRE_NEWLINE_CR :
2442                        (d == 10)? PCRE_NEWLINE_LF :
2443                        (d == (13<<8 | 10))? PCRE_NEWLINE_CRLF :
2444                        (d == -2)? PCRE_NEWLINE_ANYCRLF :
2445                        (d == -1)? PCRE_NEWLINE_ANY : 0;
2446                }
2447              if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2448                   (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2449                  &&
2450                  start_offset < len - 1 &&
2451                  bptr[start_offset] == '\r' &&
2452                  bptr[start_offset+1] == '\n')
2453                onechar++;
2454              else if (use_utf8)
2455              {              {
2456              while (start_offset + onechar < len)              while (start_offset + onechar < len)
2457                {                {
# Line 1769  while (!done) Line 2467  while (!done)
2467            {            {
2468            if (count == PCRE_ERROR_NOMATCH)            if (count == PCRE_ERROR_NOMATCH)
2469              {              {
2470              if (gmatched == 0) fprintf(outfile, "No match\n");              if (gmatched == 0)
2471                  {
2472                  if (markptr == NULL) fprintf(outfile, "No match\n");
2473                    else fprintf(outfile, "No match, mark = %s\n", markptr);
2474                  }
2475              }              }
2476            else fprintf(outfile, "Error %d\n", count);            else fprintf(outfile, "Error %d\n", count);
2477            break;  /* Out of the /g loop */            break;  /* Out of the /g loop */
# Line 1781  while (!done) Line 2483  while (!done)
2483        if (!do_g && !do_G) break;        if (!do_g && !do_G) break;
2484    
2485        /* 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
2486        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
2487        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
2488        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
2489        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
2490        character. */        character. */
2491    
2492        g_notempty = 0;        g_notempty = 0;
2493    
2494        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2495          {          {
2496          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
2497          g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED;          g_notempty = PCRE_NOTEMPTY_ATSTART | PCRE_ANCHORED;
2498          }          }
2499    
2500        /* 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 2509  while (!done)
2509          len -= use_offsets[1];          len -= use_offsets[1];
2510          }          }
2511        }  /* End of loop for /g and /G */        }  /* End of loop for /g and /G */
2512    
2513        NEXT_DATA: continue;
2514      }    /* End of loop for data lines */      }    /* End of loop for data lines */
2515    
2516    CONTINUE:    CONTINUE:
# Line 1820  while (!done) Line 2525  while (!done)
2525      {      {
2526      new_free((void *)tables);      new_free((void *)tables);
2527      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");
2528        locale_set = 0;
2529      }      }
2530    }    }
2531    

Legend:
Removed from v.79  
changed lines
  Added in v.537

  ViewVC Help
Powered by ViewVC 1.1.5