/[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 77 by nigel, Sat Feb 24 21:40:45 2007 UTC revision 391 by ph10, Tue Mar 17 21:16:01 2009 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    #define isatty _isatty         /* This is what Windows calls them, I'm told */
75    #define fileno _fileno
76    
77    #else
78    #include <sys/time.h>          /* These two includes are needed */
79    #include <sys/resource.h>      /* for setrlimit(). */
80    #define INPUT_MODE   "rb"
81    #define OUTPUT_MODE  "wb"
82    #endif
83    
84    
85  /* We need the internal info for displaying the results of pcre_study() and  /* We have to include pcre_internal.h because we need the internal info for
86  other internal data; pcretest also uses some of the fixed tables, and generally  displaying the results of pcre_study() and we also need to know about the
87  has "inside information" compared to a program that strictly follows the PCRE  internal macros, structures, and other internal data values; pcretest has
88  API. */  "inside information" compared to a program that strictly follows the PCRE API.
89    
90    Although pcre_internal.h does itself include pcre.h, we explicitly include it
91    here before pcre_internal.h so that the PCRE_EXP_xxx macros get set
92    appropriately for an application, not for building PCRE. */
93    
94    #include "pcre.h"
95  #include "pcre_internal.h"  #include "pcre_internal.h"
96    
97    /* We need access to some of the data tables that PCRE uses. So as not to have
98    to keep two copies, we include the source file here, changing the names of the
99    external symbols to prevent clashes. */
100    
101    #define _pcre_ucp_gentype      ucp_gentype
102    #define _pcre_utf8_table1      utf8_table1
103    #define _pcre_utf8_table1_size utf8_table1_size
104    #define _pcre_utf8_table2      utf8_table2
105    #define _pcre_utf8_table3      utf8_table3
106    #define _pcre_utf8_table4      utf8_table4
107    #define _pcre_utt              utt
108    #define _pcre_utt_size         utt_size
109    #define _pcre_utt_names        utt_names
110    #define _pcre_OP_lengths       OP_lengths
111    
112    #include "pcre_tables.c"
113    
114    /* We also need the pcre_printint() function for printing out compiled
115    patterns. This function is in a separate file so that it can be included in
116    pcre_compile.c when that module is compiled with debugging enabled.
117    
118    The definition of the macro PRINTABLE, which determines whether to print an
119    output character as-is or as a hex value when showing compiled patterns, is
120    contained in this file. We uses it here also, in cases when the locale has not
121    been explicitly changed, so as to get consistent output from systems that
122    differ in their output from isprint() even in the "C" locale. */
123    
124    #include "pcre_printint.src"
125    
126    #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
127    
128    
129  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
130  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 134  Makefile. */
134  #include "pcreposix.h"  #include "pcreposix.h"
135  #endif  #endif
136    
137    /* It is also possible, for the benefit of the version currently imported into
138    Exim, to build pcretest without support for UTF8 (define NOUTF8), without the
139    interface to the DFA matcher (NODFA), and without the doublecheck of the old
140    "info" function (define NOINFOCHECK). In fact, we automatically cut out the
141    UTF8 support if PCRE is built without it. */
142    
143    #ifndef SUPPORT_UTF8
144    #ifndef NOUTF8
145    #define NOUTF8
146    #endif
147    #endif
148    
149    
150    /* Other parameters */
151    
152  #ifndef CLOCKS_PER_SEC  #ifndef CLOCKS_PER_SEC
153  #ifdef CLK_TCK  #ifdef CLK_TCK
154  #define CLOCKS_PER_SEC CLK_TCK  #define CLOCKS_PER_SEC CLK_TCK
# Line 70  Makefile. */ Line 157  Makefile. */
157  #endif  #endif
158  #endif  #endif
159    
160  #define LOOPREPEAT 500000  /* This is the default loop count for timing. */
161    
162  #define BUFFER_SIZE 30000  #define LOOPREPEAT 500000
 #define PBUFFER_SIZE BUFFER_SIZE  
 #define DBUFFER_SIZE BUFFER_SIZE  
163    
164    /* Static variables */
165    
166  static FILE *outfile;  static FILE *outfile;
167  static int log_store = 0;  static int log_store = 0;
# Line 83  static int callout_count; Line 169  static int callout_count;
169  static int callout_extra;  static int callout_extra;
170  static int callout_fail_count;  static int callout_fail_count;
171  static int callout_fail_id;  static int callout_fail_id;
172    static int debug_lengths;
173  static int first_callout;  static int first_callout;
174    static int locale_set = 0;
175  static int show_malloc;  static int show_malloc;
176  static int use_utf8;  static int use_utf8;
177  static size_t gotten_store;  static size_t gotten_store;
178    
179    /* The buffers grow automatically if very long input lines are encountered. */
180    
181    static int buffer_size = 50000;
182    static uschar *buffer = NULL;
183    static uschar *dbuffer = NULL;
184  static uschar *pbuffer = NULL;  static uschar *pbuffer = NULL;
185    
186    
187    
188  /*************************************************  /*************************************************
189    *        Read or extend an input line            *
190    *************************************************/
191    
192    /* Input lines are read into buffer, but both patterns and data lines can be
193    continued over multiple input lines. In addition, if the buffer fills up, we
194    want to automatically expand it so as to be able to handle extremely large
195    lines that are needed for certain stress tests. When the input buffer is
196    expanded, the other two buffers must also be expanded likewise, and the
197    contents of pbuffer, which are a copy of the input for callouts, must be
198    preserved (for when expansion happens for a data line). This is not the most
199    optimal way of handling this, but hey, this is just a test program!
200    
201    Arguments:
202      f            the file to read
203      start        where in buffer to start (this *must* be within buffer)
204      prompt       for stdin or readline()
205    
206    Returns:       pointer to the start of new data
207                   could be a copy of start, or could be moved
208                   NULL if no data read and EOF reached
209    */
210    
211    static uschar *
212    extend_inputline(FILE *f, uschar *start, const char *prompt)
213    {
214    uschar *here = start;
215    
216    for (;;)
217      {
218      int rlen = buffer_size - (here - buffer);
219    
220      if (rlen > 1000)
221        {
222        int dlen;
223    
224        /* If libreadline support is required, use readline() to read a line if the
225        input is a terminal. Note that readline() removes the trailing newline, so
226        we must put it back again, to be compatible with fgets(). */
227    
228    #ifdef SUPPORT_LIBREADLINE
229        if (isatty(fileno(f)))
230          {
231          size_t len;
232          char *s = readline(prompt);
233          if (s == NULL) return (here == start)? NULL : start;
234          len = strlen(s);
235          if (len > 0) add_history(s);
236          if (len > rlen - 1) len = rlen - 1;
237          memcpy(here, s, len);
238          here[len] = '\n';
239          here[len+1] = 0;
240          free(s);
241          }
242        else
243    #endif
244    
245        /* Read the next line by normal means, prompting if the file is stdin. */
246    
247          {
248          if (f == stdin) printf(prompt);
249          if (fgets((char *)here, rlen,  f) == NULL)
250            return (here == start)? NULL : start;
251          }
252    
253        dlen = (int)strlen((char *)here);
254        if (dlen > 0 && here[dlen - 1] == '\n') return start;
255        here += dlen;
256        }
257    
258      else
259        {
260        int new_buffer_size = 2*buffer_size;
261        uschar *new_buffer = (unsigned char *)malloc(new_buffer_size);
262        uschar *new_dbuffer = (unsigned char *)malloc(new_buffer_size);
263        uschar *new_pbuffer = (unsigned char *)malloc(new_buffer_size);
264    
265        if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
266          {
267          fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
268          exit(1);
269          }
270    
271        memcpy(new_buffer, buffer, buffer_size);
272        memcpy(new_pbuffer, pbuffer, buffer_size);
273    
274        buffer_size = new_buffer_size;
275    
276        start = new_buffer + (start - buffer);
277        here = new_buffer + (here - buffer);
278    
279        free(buffer);
280        free(dbuffer);
281        free(pbuffer);
282    
283        buffer = new_buffer;
284        dbuffer = new_dbuffer;
285        pbuffer = new_pbuffer;
286        }
287      }
288    
289    return NULL;  /* Control never gets here */
290    }
291    
292    
293    
294    
295    
296    
297    
298    /*************************************************
299  *          Read number from string               *  *          Read number from string               *
300  *************************************************/  *************************************************/
301    
302  /* 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
303  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
304  for unpicking the -o argument, so just keep it simple.  for unpicking arguments, so just keep it simple.
305    
306  Arguments:  Arguments:
307    str           string to be converted    str           string to be converted
# Line 128  return(result); Line 331  return(result);
331  and returns the value of the character.  and returns the value of the character.
332    
333  Argument:  Argument:
334    buffer   a pointer to the byte vector    utf8bytes   a pointer to the byte vector
335    vptr     a pointer to an int to receive the value    vptr        a pointer to an int to receive the value
336    
337  Returns:   >  0 => the number of bytes consumed  Returns:      >  0 => the number of bytes consumed
338             -6 to 0 => malformed UTF-8 character at offset = (-return)                -6 to 0 => malformed UTF-8 character at offset = (-return)
339  */  */
340    
341    #if !defined NOUTF8
342    
343  static int  static int
344  utf82ord(unsigned char *buffer, int *vptr)  utf82ord(unsigned char *utf8bytes, int *vptr)
345  {  {
346  int c = *buffer++;  int c = *utf8bytes++;
347  int d = c;  int d = c;
348  int i, j, s;  int i, j, s;
349    
# Line 154  if (i == 0 || i == 6) return 0;        / Line 359  if (i == 0 || i == 6) return 0;        /
359  /* i now has a value in the range 1-5 */  /* i now has a value in the range 1-5 */
360    
361  s = 6*i;  s = 6*i;
362  d = (c & _pcre_utf8_table3[i]) << s;  d = (c & utf8_table3[i]) << s;
363    
364  for (j = 0; j < i; j++)  for (j = 0; j < i; j++)
365    {    {
366    c = *buffer++;    c = *utf8bytes++;
367    if ((c & 0xc0) != 0x80) return -(j+1);    if ((c & 0xc0) != 0x80) return -(j+1);
368    s -= 6;    s -= 6;
369    d |= (c & 0x3f) << s;    d |= (c & 0x3f) << s;
# Line 166  for (j = 0; j < i; j++) Line 371  for (j = 0; j < i; j++)
371    
372  /* Check that encoding was the correct unique one */  /* Check that encoding was the correct unique one */
373    
374  for (j = 0; j < _pcre_utf8_table1_size; j++)  for (j = 0; j < utf8_table1_size; j++)
375    if (d <= _pcre_utf8_table1[j]) break;    if (d <= utf8_table1[j]) break;
376  if (j != i) return -(i+1);  if (j != i) return -(i+1);
377    
378  /* Valid value */  /* Valid value */
# Line 176  if (j != i) return -(i+1); Line 381  if (j != i) return -(i+1);
381  return i+1;  return i+1;
382  }  }
383    
384    #endif
385    
386    
387    
388    /*************************************************
389    *       Convert character value to UTF-8         *
390    *************************************************/
391    
392    /* This function takes an integer value in the range 0 - 0x7fffffff
393    and encodes it as a UTF-8 character in 0 to 6 bytes.
394    
395    Arguments:
396      cvalue     the character value
397      utf8bytes  pointer to buffer for result - at least 6 bytes long
398    
399    Returns:     number of characters placed in the buffer
400    */
401    
402    #if !defined NOUTF8
403    
404    static int
405    ord2utf8(int cvalue, uschar *utf8bytes)
406    {
407    register int i, j;
408    for (i = 0; i < utf8_table1_size; i++)
409      if (cvalue <= utf8_table1[i]) break;
410    utf8bytes += i;
411    for (j = i; j > 0; j--)
412     {
413     *utf8bytes-- = 0x80 | (cvalue & 0x3f);
414     cvalue >>= 6;
415     }
416    *utf8bytes = utf8_table2[i] | cvalue;
417    return i + 1;
418    }
419    
420    #endif
421    
422    
423    
424  /*************************************************  /*************************************************
# Line 188  chars without printing. */ Line 431  chars without printing. */
431    
432  static int pchars(unsigned char *p, int length, FILE *f)  static int pchars(unsigned char *p, int length, FILE *f)
433  {  {
434  int c;  int c = 0;
435  int yield = 0;  int yield = 0;
436    
437  while (length-- > 0)  while (length-- > 0)
438    {    {
439    #if !defined NOUTF8
440    if (use_utf8)    if (use_utf8)
441      {      {
442      int rc = utf82ord(p, &c);      int rc = utf82ord(p, &c);
# Line 201  while (length-- > 0) Line 445  while (length-- > 0)
445        {        {
446        length -= rc - 1;        length -= rc - 1;
447        p += rc;        p += rc;
448        if (c < 256 && isprint(c))        if (PRINTHEX(c))
449          {          {
450          if (f != NULL) fprintf(f, "%c", c);          if (f != NULL) fprintf(f, "%c", c);
451          yield++;          yield++;
452          }          }
453        else        else
454          {          {
455          int n;          int n = 4;
456          if (f != NULL) fprintf(f, "\\x{%02x}%n", c, &n);          if (f != NULL) fprintf(f, "\\x{%02x}", c);
457          yield += n;          yield += (n <= 0x000000ff)? 2 :
458                     (n <= 0x00000fff)? 3 :
459                     (n <= 0x0000ffff)? 4 :
460                     (n <= 0x000fffff)? 5 : 6;
461          }          }
462        continue;        continue;
463        }        }
464      }      }
465    #endif
466    
467     /* Not UTF-8, or malformed UTF-8  */     /* Not UTF-8, or malformed UTF-8  */
468    
469    if (isprint(c = *(p++)))    c = *p++;
470      if (PRINTHEX(c))
471      {      {
472      if (f != NULL) fprintf(f, "%c", c);      if (f != NULL) fprintf(f, "%c", c);
473      yield++;      yield++;
# Line 392  if ((rc = pcre_fullinfo(re, study, optio Line 641  if ((rc = pcre_fullinfo(re, study, optio
641  *         Byte flipping function                 *  *         Byte flipping function                 *
642  *************************************************/  *************************************************/
643    
644  static long int  static unsigned long int
645  byteflip(long int value, int n)  byteflip(unsigned long int value, int n)
646  {  {
647  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);
648  return ((value & 0x000000ff) << 24) |  return ((value & 0x000000ff) << 24) |
# Line 406  return ((value & 0x000000ff) << 24) | Line 655  return ((value & 0x000000ff) << 24) |
655    
656    
657  /*************************************************  /*************************************************
658    *        Check match or recursion limit          *
659    *************************************************/
660    
661    static int
662    check_match_limit(pcre *re, pcre_extra *extra, uschar *bptr, int len,
663      int start_offset, int options, int *use_offsets, int use_size_offsets,
664      int flag, unsigned long int *limit, int errnumber, const char *msg)
665    {
666    int count;
667    int min = 0;
668    int mid = 64;
669    int max = -1;
670    
671    extra->flags |= flag;
672    
673    for (;;)
674      {
675      *limit = mid;
676    
677      count = pcre_exec(re, extra, (char *)bptr, len, start_offset, options,
678        use_offsets, use_size_offsets);
679    
680      if (count == errnumber)
681        {
682        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
683        min = mid;
684        mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;
685        }
686    
687      else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||
688                             count == PCRE_ERROR_PARTIAL)
689        {
690        if (mid == min + 1)
691          {
692          fprintf(outfile, "Minimum %s limit = %d\n", msg, mid);
693          break;
694          }
695        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
696        max = mid;
697        mid = (min + mid)/2;
698        }
699      else break;    /* Some other error */
700      }
701    
702    extra->flags &= ~flag;
703    return count;
704    }
705    
706    
707    
708    /*************************************************
709    *         Case-independent strncmp() function    *
710    *************************************************/
711    
712    /*
713    Arguments:
714      s         first string
715      t         second string
716      n         number of characters to compare
717    
718    Returns:    < 0, = 0, or > 0, according to the comparison
719    */
720    
721    static int
722    strncmpic(uschar *s, uschar *t, int n)
723    {
724    while (n--)
725      {
726      int c = tolower(*s++) - tolower(*t++);
727      if (c) return c;
728      }
729    return 0;
730    }
731    
732    
733    
734    /*************************************************
735    *         Check newline indicator                *
736    *************************************************/
737    
738    /* This is used both at compile and run-time to check for <xxx> escapes, where
739    xxx is LF, CR, CRLF, ANYCRLF, or ANY. Print a message and return 0 if there is
740    no match.
741    
742    Arguments:
743      p           points after the leading '<'
744      f           file for error message
745    
746    Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
747    */
748    
749    static int
750    check_newline(uschar *p, FILE *f)
751    {
752    if (strncmpic(p, (uschar *)"cr>", 3) == 0) return PCRE_NEWLINE_CR;
753    if (strncmpic(p, (uschar *)"lf>", 3) == 0) return PCRE_NEWLINE_LF;
754    if (strncmpic(p, (uschar *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
755    if (strncmpic(p, (uschar *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF;
756    if (strncmpic(p, (uschar *)"any>", 4) == 0) return PCRE_NEWLINE_ANY;
757    if (strncmpic(p, (uschar *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF;
758    if (strncmpic(p, (uschar *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE;
759    fprintf(f, "Unknown newline type at: <%s\n", p);
760    return 0;
761    }
762    
763    
764    
765    /*************************************************
766    *             Usage function                     *
767    *************************************************/
768    
769    static void
770    usage(void)
771    {
772    printf("Usage:     pcretest [options] [<input file> [<output file>]]\n\n");
773    printf("Input and output default to stdin and stdout.\n");
774    #ifdef SUPPORT_LIBREADLINE
775    printf("If input is a terminal, readline() is used to read from it.\n");
776    #else
777    printf("This version of pcretest is not linked with readline().\n");
778    #endif
779    printf("\nOptions:\n");
780    printf("  -b       show compiled code (bytecode)\n");
781    printf("  -C       show PCRE compile-time options and exit\n");
782    printf("  -d       debug: show compiled code and information (-b and -i)\n");
783    #if !defined NODFA
784    printf("  -dfa     force DFA matching for all subjects\n");
785    #endif
786    printf("  -help    show usage information\n");
787    printf("  -i       show information about compiled patterns\n"
788           "  -M       find MATCH_LIMIT minimum for each subject\n"
789           "  -m       output memory used information\n"
790           "  -o <n>   set size of offsets vector to <n>\n");
791    #if !defined NOPOSIX
792    printf("  -p       use POSIX interface\n");
793    #endif
794    printf("  -q       quiet: do not output PCRE version number at start\n");
795    printf("  -S <n>   set stack size to <n> megabytes\n");
796    printf("  -s       output store (memory) used information\n"
797           "  -t       time compilation and execution\n");
798    printf("  -t <n>   time compilation and execution, repeating <n> times\n");
799    printf("  -tm      time execution (matching) only\n");
800    printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
801    }
802    
803    
804    
805    /*************************************************
806  *                Main Program                    *  *                Main Program                    *
807  *************************************************/  *************************************************/
808    
# Line 418  int main(int argc, char **argv) Line 815  int main(int argc, char **argv)
815  FILE *infile = stdin;  FILE *infile = stdin;
816  int options = 0;  int options = 0;
817  int study_options = 0;  int study_options = 0;
818    int default_find_match_limit = FALSE;
819  int op = 1;  int op = 1;
820  int timeit = 0;  int timeit = 0;
821    int timeitm = 0;
822  int showinfo = 0;  int showinfo = 0;
823  int showstore = 0;  int showstore = 0;
824    int quiet = 0;
825  int size_offsets = 45;  int size_offsets = 45;
826  int size_offsets_max;  int size_offsets_max;
827  int *offsets = NULL;  int *offsets = NULL;
# Line 432  int debug = 0; Line 832  int debug = 0;
832  int done = 0;  int done = 0;
833  int all_use_dfa = 0;  int all_use_dfa = 0;
834  int yield = 0;  int yield = 0;
835    int stack_size;
836    
837  unsigned char *buffer;  /* These vectors store, end-to-end, a list of captured substring names. Assume
838  unsigned char *dbuffer;  that 1024 is plenty long enough for the few names we'll be testing. */
839    
840    uschar copynames[1024];
841    uschar getnames[1024];
842    
843    uschar *copynamesptr;
844    uschar *getnamesptr;
845    
846  /* Get buffers from malloc() so that Electric Fence will check their misuse  /* Get buffers from malloc() so that Electric Fence will check their misuse
847  when I am debugging. */  when I am debugging. They grow automatically when very long lines are read. */
848    
849  buffer = (unsigned char *)malloc(BUFFER_SIZE);  buffer = (unsigned char *)malloc(buffer_size);
850  dbuffer = (unsigned char *)malloc(DBUFFER_SIZE);  dbuffer = (unsigned char *)malloc(buffer_size);
851  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.  */  
852    
853  #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) */  
854    
855  outfile = stdout;  outfile = stdout;
856    
857    /* The following  _setmode() stuff is some Windows magic that tells its runtime
858    library to translate CRLF into a single LF character. At least, that's what
859    I've been told: never having used Windows I take this all on trust. Originally
860    it set 0x8000, but then I was advised that _O_BINARY was better. */
861    
862    #if defined(_WIN32) || defined(WIN32)
863    _setmode( _fileno( stdout ), _O_BINARY );
864    #endif
865    
866  /* Scan options */  /* Scan options */
867    
868  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
# Line 461  while (argc > 1 && argv[op][0] == '-') Line 871  while (argc > 1 && argv[op][0] == '-')
871    
872    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
873      showstore = 1;      showstore = 1;
874    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-q") == 0) quiet = 1;
875      else if (strcmp(argv[op], "-b") == 0) debug = 1;
876    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
877    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
878      else if (strcmp(argv[op], "-M") == 0) default_find_match_limit = TRUE;
879    #if !defined NODFA
880    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
881    #endif
882    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&
883        ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),        ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),
884          *endptr == 0))          *endptr == 0))
# Line 472  while (argc > 1 && argv[op][0] == '-') Line 886  while (argc > 1 && argv[op][0] == '-')
886      op++;      op++;
887      argc--;      argc--;
888      }      }
889      else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
890        {
891        int both = argv[op][2] == 0;
892        int temp;
893        if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr),
894                         *endptr == 0))
895          {
896          timeitm = temp;
897          op++;
898          argc--;
899          }
900        else timeitm = LOOPREPEAT;
901        if (both) timeit = timeitm;
902        }
903      else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
904          ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
905            *endptr == 0))
906        {
907    #if defined(_WIN32) || defined(WIN32)
908        printf("PCRE: -S not supported on this OS\n");
909        exit(1);
910    #else
911        int rc;
912        struct rlimit rlim;
913        getrlimit(RLIMIT_STACK, &rlim);
914        rlim.rlim_cur = stack_size * 1024 * 1024;
915        rc = setrlimit(RLIMIT_STACK, &rlim);
916        if (rc != 0)
917          {
918        printf("PCRE: setrlimit() failed with error %d\n", rc);
919        exit(1);
920          }
921        op++;
922        argc--;
923    #endif
924        }
925  #if !defined NOPOSIX  #if !defined NOPOSIX
926    else if (strcmp(argv[op], "-p") == 0) posix = 1;    else if (strcmp(argv[op], "-p") == 0) posix = 1;
927  #endif  #endif
928    else if (strcmp(argv[op], "-C") == 0)    else if (strcmp(argv[op], "-C") == 0)
929      {      {
930      int rc;      int rc;
931        unsigned long int lrc;
932      printf("PCRE version %s\n", pcre_version());      printf("PCRE version %s\n", pcre_version());
933      printf("Compiled with\n");      printf("Compiled with\n");
934      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
# Line 485  while (argc > 1 && argv[op][0] == '-') Line 936  while (argc > 1 && argv[op][0] == '-')
936      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
937      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
938      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
939      printf("  Newline character is %s\n", (rc == '\r')? "CR" : "LF");      /* Note that these values are always the ASCII values, even
940        in EBCDIC environments. CR is 13 and NL is 10. */
941        printf("  Newline sequence is %s\n", (rc == 13)? "CR" :
942          (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
943          (rc == -2)? "ANYCRLF" :
944          (rc == -1)? "ANY" : "???");
945        (void)pcre_config(PCRE_CONFIG_BSR, &rc);
946        printf("  \\R matches %s\n", rc? "CR, LF, or CRLF only" :
947                                         "all Unicode newlines");
948      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);
949      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
950      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
951      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
952      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &lrc);
953      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %ld\n", lrc);
954        (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &lrc);
955        printf("  Default recursion depth limit = %ld\n", lrc);
956      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
957      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
958      exit(0);      goto EXIT;
959        }
960      else if (strcmp(argv[op], "-help") == 0 ||
961               strcmp(argv[op], "--help") == 0)
962        {
963        usage();
964        goto EXIT;
965      }      }
966    else    else
967      {      {
968      printf("** Unknown or malformed option %s\n", argv[op]);      printf("** Unknown or malformed option %s\n", argv[op]);
969      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");  
     printf("  -dfa   force DFA matching for all subjects\n");  
     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");  
970      yield = 1;      yield = 1;
971      goto EXIT;      goto EXIT;
972      }      }
# Line 525  offsets = (int *)malloc(size_offsets_max Line 981  offsets = (int *)malloc(size_offsets_max
981  if (offsets == NULL)  if (offsets == NULL)
982    {    {
983    printf("** Failed to get %d bytes of memory for offsets vector\n",    printf("** Failed to get %d bytes of memory for offsets vector\n",
984      size_offsets_max * sizeof(int));      (int)(size_offsets_max * sizeof(int)));
985    yield = 1;    yield = 1;
986    goto EXIT;    goto EXIT;
987    }    }
# Line 534  if (offsets == NULL) Line 990  if (offsets == NULL)
990    
991  if (argc > 1)  if (argc > 1)
992    {    {
993    infile = fopen(argv[op], "rb");    infile = fopen(argv[op], INPUT_MODE);
994    if (infile == NULL)    if (infile == NULL)
995      {      {
996      printf("** Failed to open %s\n", argv[op]);      printf("** Failed to open %s\n", argv[op]);
# Line 545  if (argc > 1) Line 1001  if (argc > 1)
1001    
1002  if (argc > 2)  if (argc > 2)
1003    {    {
1004    outfile = fopen(argv[op+1], "wb");    outfile = fopen(argv[op+1], OUTPUT_MODE);
1005    if (outfile == NULL)    if (outfile == NULL)
1006      {      {
1007      printf("** Failed to open %s\n", argv[op+1]);      printf("** Failed to open %s\n", argv[op+1]);
# Line 561  pcre_free = new_free; Line 1017  pcre_free = new_free;
1017  pcre_stack_malloc = stack_malloc;  pcre_stack_malloc = stack_malloc;
1018  pcre_stack_free = stack_free;  pcre_stack_free = stack_free;
1019    
1020  /* Heading line, then prompt for first regex if stdin */  /* Heading line unless quiet, then prompt for first regex if stdin */
1021    
1022  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  if (!quiet) fprintf(outfile, "PCRE version %s\n\n", pcre_version());
1023    
1024  /* Main loop */  /* Main loop */
1025    
# Line 590  while (!done) Line 1046  while (!done)
1046    int do_showinfo = showinfo;    int do_showinfo = showinfo;
1047    int do_showrest = 0;    int do_showrest = 0;
1048    int do_flip = 0;    int do_flip = 0;
1049    int erroroffset, len, delimiter;    int erroroffset, len, delimiter, poffset;
1050    
1051    use_utf8 = 0;    use_utf8 = 0;
1052      debug_lengths = 1;
1053    
1054    if (infile == stdin) printf("  re> ");    if (extend_inputline(infile, buffer, "  re> ") == NULL) break;
   if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break;  
1055    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1056    fflush(outfile);    fflush(outfile);
1057    
# Line 607  while (!done) Line 1063  while (!done)
1063    
1064    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
1065      {      {
1066      unsigned long int magic;      unsigned long int magic, get_options;
1067      uschar sbuf[8];      uschar sbuf[8];
1068      FILE *f;      FILE *f;
1069    
# Line 655  while (!done) Line 1111  while (!done)
1111    
1112      /* Need to know if UTF-8 for printing data strings */      /* Need to know if UTF-8 for printing data strings */
1113    
1114      new_info(re, NULL, PCRE_INFO_OPTIONS, &options);      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1115      use_utf8 = (options & PCRE_UTF8) != 0;      use_utf8 = (get_options & PCRE_UTF8) != 0;
1116    
1117      /* Now see if there is any following study data */      /* Now see if there is any following study data */
1118    
# Line 695  while (!done) Line 1151  while (!done)
1151    
1152    if (isalnum(delimiter) || delimiter == '\\')    if (isalnum(delimiter) || delimiter == '\\')
1153      {      {
1154      fprintf(outfile, "** Delimiter must not be alphameric or \\\n");      fprintf(outfile, "** Delimiter must not be alphanumeric or \\\n");
1155      goto SKIP_DATA;      goto SKIP_DATA;
1156      }      }
1157    
1158    pp = p;    pp = p;
1159      poffset = p - buffer;
1160    
1161    for(;;)    for(;;)
1162      {      {
# Line 710  while (!done) Line 1167  while (!done)
1167        pp++;        pp++;
1168        }        }
1169      if (*pp != 0) break;      if (*pp != 0) break;
1170        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)  
1171        {        {
1172        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
1173        done = 1;        done = 1;
# Line 728  while (!done) Line 1176  while (!done)
1176      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
1177      }      }
1178    
1179      /* The buffer may have moved while being extended; reset the start of data
1180      pointer to the correct relative point in the buffer. */
1181    
1182      p = buffer + poffset;
1183    
1184    /* If the first character after the delimiter is backslash, make    /* If the first character after the delimiter is backslash, make
1185    the pattern end with backslash. This is purely to provide a way    the pattern end with backslash. This is purely to provide a way
1186    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 759  while (!done) Line 1212  while (!done)
1212    
1213        case '+': do_showrest = 1; break;        case '+': do_showrest = 1; break;
1214        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
1215          case 'B': do_debug = 1; break;
1216        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
1217        case 'D': do_debug = do_showinfo = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
1218        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
1219        case 'F': do_flip = 1; break;        case 'F': do_flip = 1; break;
1220        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1221        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1222          case 'J': options |= PCRE_DUPNAMES; break;
1223        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1224        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1225    
# Line 775  while (!done) Line 1230  while (!done)
1230        case 'S': do_study = 1; break;        case 'S': do_study = 1; break;
1231        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1232        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
1233          case 'Z': debug_lengths = 0; break;
1234        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1235        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
1236    
1237        case 'L':        case 'L':
1238        ppp = pp;        ppp = pp;
1239        /* The '\r' test here is so that it works on Windows */        /* The '\r' test here is so that it works on Windows. */
1240        while (*ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;        /* The '0' test is just in case this is an unterminated line. */
1241          while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
1242        *ppp = 0;        *ppp = 0;
1243        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
1244          {          {
1245          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
1246          goto SKIP_DATA;          goto SKIP_DATA;
1247          }          }
1248          locale_set = 1;
1249        tables = pcre_maketables();        tables = pcre_maketables();
1250        pp = ppp;        pp = ppp;
1251        break;        break;
# Line 799  while (!done) Line 1257  while (!done)
1257        *pp = 0;        *pp = 0;
1258        break;        break;
1259    
1260          case '<':
1261            {
1262            if (strncmp((char *)pp, "JS>", 3) == 0)
1263              {
1264              options |= PCRE_JAVASCRIPT_COMPAT;
1265              pp += 3;
1266              }
1267            else
1268              {
1269              int x = check_newline(pp, outfile);
1270              if (x == 0) goto SKIP_DATA;
1271              options |= x;
1272              while (*pp++ != '>');
1273              }
1274            }
1275          break;
1276    
1277        case '\r':                      /* So that it works in Windows */        case '\r':                      /* So that it works in Windows */
1278        case '\n':        case '\n':
1279        case ' ':        case ' ':
# Line 823  while (!done) Line 1298  while (!done)
1298      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
1299      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
1300      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1301        if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1302        if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1303    
1304      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1305    
1306      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
# Line 830  while (!done) Line 1308  while (!done)
1308    
1309      if (rc != 0)      if (rc != 0)
1310        {        {
1311        (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);        (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1312        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
1313        goto SKIP_DATA;        goto SKIP_DATA;
1314        }        }
# Line 842  while (!done) Line 1320  while (!done)
1320  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1321    
1322      {      {
1323      if (timeit)      if (timeit > 0)
1324        {        {
1325        register int i;        register int i;
1326        clock_t time_taken;        clock_t time_taken;
1327        clock_t start_time = clock();        clock_t start_time = clock();
1328        for (i = 0; i < LOOPREPEAT; i++)        for (i = 0; i < timeit; i++)
1329          {          {
1330          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1331          if (re != NULL) free(re);          if (re != NULL) free(re);
1332          }          }
1333        time_taken = clock() - start_time;        time_taken = clock() - start_time;
1334        fprintf(outfile, "Compile time %.3f milliseconds\n",        fprintf(outfile, "Compile time %.4f milliseconds\n",
1335          (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /          (((double)time_taken * 1000.0) / (double)timeit) /
1336            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
1337        }        }
1338    
# Line 871  while (!done) Line 1349  while (!done)
1349          {          {
1350          for (;;)          for (;;)
1351            {            {
1352            if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)            if (extend_inputline(infile, buffer, NULL) == NULL)
1353              {              {
1354              done = 1;              done = 1;
1355              goto CONTINUE;              goto CONTINUE;
# Line 906  while (!done) Line 1384  while (!done)
1384    
1385      if (do_study)      if (do_study)
1386        {        {
1387        if (timeit)        if (timeit > 0)
1388          {          {
1389          register int i;          register int i;
1390          clock_t time_taken;          clock_t time_taken;
1391          clock_t start_time = clock();          clock_t start_time = clock();
1392          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeit; i++)
1393            extra = pcre_study(re, study_options, &error);            extra = pcre_study(re, study_options, &error);
1394          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1395          if (extra != NULL) free(extra);          if (extra != NULL) free(extra);
1396          fprintf(outfile, "  Study time %.3f milliseconds\n",          fprintf(outfile, "  Study time %.4f milliseconds\n",
1397            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeit) /
1398              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
1399          }          }
1400        extra = pcre_study(re, study_options, &error);        extra = pcre_study(re, study_options, &error);
# Line 934  while (!done) Line 1412  while (!done)
1412      if (do_flip)      if (do_flip)
1413        {        {
1414        real_pcre *rre = (real_pcre *)re;        real_pcre *rre = (real_pcre *)re;
1415        rre->magic_number = byteflip(rre->magic_number, sizeof(rre->magic_number));        rre->magic_number =
1416            byteflip(rre->magic_number, sizeof(rre->magic_number));
1417        rre->size = byteflip(rre->size, sizeof(rre->size));        rre->size = byteflip(rre->size, sizeof(rre->size));
1418        rre->options = byteflip(rre->options, sizeof(rre->options));        rre->options = byteflip(rre->options, sizeof(rre->options));
1419        rre->top_bracket = byteflip(rre->top_bracket, sizeof(rre->top_bracket));        rre->flags = (pcre_uint16)byteflip(rre->flags, sizeof(rre->flags));
1420        rre->top_backref = byteflip(rre->top_backref, sizeof(rre->top_backref));        rre->top_bracket =
1421        rre->first_byte = byteflip(rre->first_byte, sizeof(rre->first_byte));          (pcre_uint16)byteflip(rre->top_bracket, sizeof(rre->top_bracket));
1422        rre->req_byte = byteflip(rre->req_byte, sizeof(rre->req_byte));        rre->top_backref =
1423        rre->name_table_offset = byteflip(rre->name_table_offset,          (pcre_uint16)byteflip(rre->top_backref, sizeof(rre->top_backref));
1424          rre->first_byte =
1425            (pcre_uint16)byteflip(rre->first_byte, sizeof(rre->first_byte));
1426          rre->req_byte =
1427            (pcre_uint16)byteflip(rre->req_byte, sizeof(rre->req_byte));
1428          rre->name_table_offset = (pcre_uint16)byteflip(rre->name_table_offset,
1429          sizeof(rre->name_table_offset));          sizeof(rre->name_table_offset));
1430        rre->name_entry_size = byteflip(rre->name_entry_size,        rre->name_entry_size = (pcre_uint16)byteflip(rre->name_entry_size,
1431          sizeof(rre->name_entry_size));          sizeof(rre->name_entry_size));
1432        rre->name_count = byteflip(rre->name_count, sizeof(rre->name_count));        rre->name_count = (pcre_uint16)byteflip(rre->name_count,
1433            sizeof(rre->name_count));
1434    
1435        if (extra != NULL)        if (extra != NULL)
1436          {          {
# Line 959  while (!done) Line 1444  while (!done)
1444    
1445      SHOW_INFO:      SHOW_INFO:
1446    
1447        if (do_debug)
1448          {
1449          fprintf(outfile, "------------------------------------------------------------------\n");
1450          pcre_printint(re, outfile, debug_lengths);
1451          }
1452    
1453      if (do_showinfo)      if (do_showinfo)
1454        {        {
1455        unsigned long int get_options, all_options;        unsigned long int get_options, all_options;
1456    #if !defined NOINFOCHECK
1457        int old_first_char, old_options, old_count;        int old_first_char, old_options, old_count;
1458        int count, backrefmax, first_char, need_char;  #endif
1459          int count, backrefmax, first_char, need_char, okpartial, jchanged,
1460            hascrorlf;
1461        int nameentrysize, namecount;        int nameentrysize, namecount;
1462        const uschar *nametable;        const uschar *nametable;
1463    
       if (do_debug)  
         {  
         fprintf(outfile, "------------------------------------------------------------------\n");  
         _pcre_printint(re, outfile);  
         }  
   
1464        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1465        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1466        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
# Line 982  while (!done) Line 1470  while (!done)
1470        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);
1471        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);
1472        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);
1473          new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial);
1474          new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged);
1475          new_info(re, NULL, PCRE_INFO_HASCRORLF, &hascrorlf);
1476    
1477    #if !defined NOINFOCHECK
1478        old_count = pcre_info(re, &old_options, &old_first_char);        old_count = pcre_info(re, &old_options, &old_first_char);
1479        if (count < 0) fprintf(outfile,        if (count < 0) fprintf(outfile,
1480          "Error %d from pcre_info()\n", count);          "Error %d from pcre_info()\n", count);
# Line 1000  while (!done) Line 1492  while (!done)
1492            "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",            "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",
1493              get_options, old_options);              get_options, old_options);
1494          }          }
1495    #endif
1496    
1497        if (size != regex_gotten_store) fprintf(outfile,        if (size != regex_gotten_store) fprintf(outfile,
1498          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",
# Line 1021  while (!done) Line 1514  while (!done)
1514            }            }
1515          }          }
1516    
1517        /* The NOPARTIAL bit is a private bit in the options, so we have        if (!okpartial) fprintf(outfile, "Partial matching not supported\n");
1518        to fish it out via out back door */        if (hascrorlf) fprintf(outfile, "Contains explicit CR or LF match\n");
1519    
1520        all_options = ((real_pcre *)re)->options;        all_options = ((real_pcre *)re)->options;
1521        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");  
1522    
1523        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1524          else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s\n",          else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s\n",
1525            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1526            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1527            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
1528            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",
1529            ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",            ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",
1530            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",
1531              ((get_options & PCRE_BSR_ANYCRLF) != 0)? " bsr_anycrlf" : "",
1532              ((get_options & PCRE_BSR_UNICODE) != 0)? " bsr_unicode" : "",
1533            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
1534            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
1535            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1536              ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1537            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1538            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "");            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1539              ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1540    
1541          if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
1542    
1543          switch (get_options & PCRE_NEWLINE_BITS)
1544            {
1545            case PCRE_NEWLINE_CR:
1546            fprintf(outfile, "Forced newline sequence: CR\n");
1547            break;
1548    
1549            case PCRE_NEWLINE_LF:
1550            fprintf(outfile, "Forced newline sequence: LF\n");
1551            break;
1552    
1553            case PCRE_NEWLINE_CRLF:
1554            fprintf(outfile, "Forced newline sequence: CRLF\n");
1555            break;
1556    
1557        if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0)          case PCRE_NEWLINE_ANYCRLF:
1558          fprintf(outfile, "Case state changes\n");          fprintf(outfile, "Forced newline sequence: ANYCRLF\n");
1559            break;
1560    
1561            case PCRE_NEWLINE_ANY:
1562            fprintf(outfile, "Forced newline sequence: ANY\n");
1563            break;
1564    
1565            default:
1566            break;
1567            }
1568    
1569        if (first_char == -1)        if (first_char == -1)
1570          {          {
1571          fprintf(outfile, "First char at start or follows \\n\n");          fprintf(outfile, "First char at start or follows newline\n");
1572          }          }
1573        else if (first_char < 0)        else if (first_char < 0)
1574          {          {
# Line 1063  while (!done) Line 1579  while (!done)
1579          int ch = first_char & 255;          int ch = first_char & 255;
1580          const char *caseless = ((first_char & REQ_CASELESS) == 0)?          const char *caseless = ((first_char & REQ_CASELESS) == 0)?
1581            "" : " (caseless)";            "" : " (caseless)";
1582          if (isprint(ch))          if (PRINTHEX(ch))
1583            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);
1584          else          else
1585            fprintf(outfile, "First char = %d%s\n", ch, caseless);            fprintf(outfile, "First char = %d%s\n", ch, caseless);
# Line 1078  while (!done) Line 1594  while (!done)
1594          int ch = need_char & 255;          int ch = need_char & 255;
1595          const char *caseless = ((need_char & REQ_CASELESS) == 0)?          const char *caseless = ((need_char & REQ_CASELESS) == 0)?
1596            "" : " (caseless)";            "" : " (caseless)";
1597          if (isprint(ch))          if (PRINTHEX(ch))
1598            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);
1599          else          else
1600            fprintf(outfile, "Need char = %d%s\n", ch, caseless);            fprintf(outfile, "Need char = %d%s\n", ch, caseless);
# Line 1114  while (!done) Line 1630  while (!done)
1630                    fprintf(outfile, "\n  ");                    fprintf(outfile, "\n  ");
1631                    c = 2;                    c = 2;
1632                    }                    }
1633                  if (isprint(i) && i != ' ')                  if (PRINTHEX(i) && i != ' ')
1634                    {                    {
1635                    fprintf(outfile, "%c ", i);                    fprintf(outfile, "%c ", i);
1636                    c += 2;                    c += 2;
# Line 1146  while (!done) Line 1662  while (!done)
1662        else        else
1663          {          {
1664          uschar sbuf[8];          uschar sbuf[8];
1665          sbuf[0] = (true_size >> 24)  & 255;          sbuf[0] = (uschar)((true_size >> 24) & 255);
1666          sbuf[1] = (true_size >> 16)  & 255;          sbuf[1] = (uschar)((true_size >> 16) & 255);
1667          sbuf[2] = (true_size >>  8)  & 255;          sbuf[2] = (uschar)((true_size >>  8) & 255);
1668          sbuf[3] = (true_size)  & 255;          sbuf[3] = (uschar)((true_size) & 255);
1669    
1670          sbuf[4] = (true_study_size >> 24)  & 255;          sbuf[4] = (uschar)((true_study_size >> 24) & 255);
1671          sbuf[5] = (true_study_size >> 16)  & 255;          sbuf[5] = (uschar)((true_study_size >> 16) & 255);
1672          sbuf[6] = (true_study_size >>  8)  & 255;          sbuf[6] = (uschar)((true_study_size >>  8) & 255);
1673          sbuf[7] = (true_study_size)  & 255;          sbuf[7] = (uschar)((true_study_size) & 255);
1674    
1675          if (fwrite(sbuf, 1, 8, f) < 8 ||          if (fwrite(sbuf, 1, 8, f) < 8 ||
1676              fwrite(re, 1, true_size, f) < true_size)              fwrite(re, 1, true_size, f) < true_size)
# Line 1173  while (!done) Line 1689  while (!done)
1689                  strerror(errno));                  strerror(errno));
1690                }                }
1691              else fprintf(outfile, "Study data written to %s\n", to_file);              else fprintf(outfile, "Study data written to %s\n", to_file);
1692    
1693              }              }
1694            }            }
1695          fclose(f);          fclose(f);
# Line 1189  while (!done) Line 1706  while (!done)
1706    
1707    for (;;)    for (;;)
1708      {      {
1709      unsigned char *q;      uschar *q;
1710      unsigned char *bptr = dbuffer;      uschar *bptr;
1711      int *use_offsets = offsets;      int *use_offsets = offsets;
1712      int use_size_offsets = size_offsets;      int use_size_offsets = size_offsets;
1713      int callout_data = 0;      int callout_data = 0;
1714      int callout_data_set = 0;      int callout_data_set = 0;
1715      int count, c;      int count, c;
1716      int copystrings = 0;      int copystrings = 0;
1717      int find_match_limit = 0;      int find_match_limit = default_find_match_limit;
1718      int getstrings = 0;      int getstrings = 0;
1719      int getlist = 0;      int getlist = 0;
1720      int gmatched = 0;      int gmatched = 0;
# Line 1207  while (!done) Line 1724  while (!done)
1724    
1725      options = 0;      options = 0;
1726    
1727        *copynames = 0;
1728        *getnames = 0;
1729    
1730        copynamesptr = copynames;
1731        getnamesptr = getnames;
1732    
1733      pcre_callout = callout;      pcre_callout = callout;
1734      first_callout = 1;      first_callout = 1;
1735      callout_extra = 0;      callout_extra = 0;
# Line 1215  while (!done) Line 1738  while (!done)
1738      callout_fail_id = -1;      callout_fail_id = -1;
1739      show_malloc = 0;      show_malloc = 0;
1740    
1741      if (infile == stdin) printf("data> ");      if (extra != NULL) extra->flags &=
1742      if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)        ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
1743    
1744        len = 0;
1745        for (;;)
1746        {        {
1747        done = 1;        if (extend_inputline(infile, buffer + len, "data> ") == NULL)
1748        goto CONTINUE;          {
1749            if (len > 0) break;
1750            done = 1;
1751            goto CONTINUE;
1752            }
1753          if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1754          len = (int)strlen((char *)buffer);
1755          if (buffer[len-1] == '\n') break;
1756        }        }
     if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);  
1757    
     len = (int)strlen((char *)buffer);  
1758      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
1759      buffer[len] = 0;      buffer[len] = 0;
1760      if (len == 0) break;      if (len == 0) break;
# Line 1231  while (!done) Line 1762  while (!done)
1762      p = buffer;      p = buffer;
1763      while (isspace(*p)) p++;      while (isspace(*p)) p++;
1764    
1765      q = dbuffer;      bptr = q = dbuffer;
1766      while ((c = *p++) != 0)      while ((c = *p++) != 0)
1767        {        {
1768        int i = 0;        int i = 0;
# Line 1253  while (!done) Line 1784  while (!done)
1784          c -= '0';          c -= '0';
1785          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')
1786            c = c * 8 + *p++ - '0';            c = c * 8 + *p++ - '0';
1787    
1788    #if !defined NOUTF8
1789            if (use_utf8 && c > 255)
1790              {
1791              unsigned char buff8[8];
1792              int ii, utn;
1793              utn = ord2utf8(c, buff8);
1794              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1795              c = buff8[ii];   /* Last byte */
1796              }
1797    #endif
1798          break;          break;
1799    
1800          case 'x':          case 'x':
1801    
1802          /* Handle \x{..} specially - new Perl thing for utf8 */          /* Handle \x{..} specially - new Perl thing for utf8 */
1803    
1804    #if !defined NOUTF8
1805          if (*p == '{')          if (*p == '{')
1806            {            {
1807            unsigned char *pt = p;            unsigned char *pt = p;
# Line 1269  while (!done) Line 1812  while (!done)
1812              {              {
1813              unsigned char buff8[8];              unsigned char buff8[8];
1814              int ii, utn;              int ii, utn;
1815              utn = _pcre_ord2utf8(c, buff8);              if (use_utf8)
1816              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];                {
1817              c = buff8[ii];   /* Last byte */                utn = ord2utf8(c, buff8);
1818                  for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1819                  c = buff8[ii];   /* Last byte */
1820                  }
1821                else
1822                 {
1823                 if (c > 255)
1824                   fprintf(outfile, "** Character \\x{%x} is greater than 255 and "
1825                     "UTF-8 mode is not enabled.\n"
1826                     "** Truncation will probably give the wrong result.\n", c);
1827                 }
1828              p = pt + 1;              p = pt + 1;
1829              break;              break;
1830              }              }
1831            /* Not correct form; fall through */            /* Not correct form; fall through */
1832            }            }
1833    #endif
1834    
1835          /* Ordinary \x */          /* Ordinary \x */
1836    
# Line 1312  while (!done) Line 1866  while (!done)
1866            }            }
1867          else if (isalnum(*p))          else if (isalnum(*p))
1868            {            {
1869            uschar name[256];            uschar *npp = copynamesptr;
           uschar *npp = name;  
1870            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1871              *npp++ = 0;
1872            *npp = 0;            *npp = 0;
1873            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)copynamesptr);
1874            if (n < 0)            if (n < 0)
1875              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", copynamesptr);
1876            else copystrings |= 1 << n;            copynamesptr = npp;
1877            }            }
1878          else if (*p == '+')          else if (*p == '+')
1879            {            {
# Line 1357  while (!done) Line 1911  while (!done)
1911            }            }
1912          continue;          continue;
1913    
1914    #if !defined NODFA
1915          case 'D':          case 'D':
1916    #if !defined NOPOSIX
1917          if (posix || do_posix)          if (posix || do_posix)
1918            printf("** Can't use dfa matching in POSIX mode: \\D ignored\n");            printf("** Can't use dfa matching in POSIX mode: \\D ignored\n");
1919          else          else
1920    #endif
1921            use_dfa = 1;            use_dfa = 1;
1922          continue;          continue;
1923    
1924          case 'F':          case 'F':
1925          options |= PCRE_DFA_SHORTEST;          options |= PCRE_DFA_SHORTEST;
1926          continue;          continue;
1927    #endif
1928    
1929          case 'G':          case 'G':
1930          if (isdigit(*p))          if (isdigit(*p))
# Line 1376  while (!done) Line 1934  while (!done)
1934            }            }
1935          else if (isalnum(*p))          else if (isalnum(*p))
1936            {            {
1937            uschar name[256];            uschar *npp = getnamesptr;
           uschar *npp = name;  
1938            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1939              *npp++ = 0;
1940            *npp = 0;            *npp = 0;
1941            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)getnamesptr);
1942            if (n < 0)            if (n < 0)
1943              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", getnamesptr);
1944            else getstrings |= 1 << n;            getnamesptr = npp;
1945            }            }
1946          continue;          continue;
1947    
# Line 1409  while (!done) Line 1967  while (!done)
1967            if (offsets == NULL)            if (offsets == NULL)
1968              {              {
1969              printf("** Failed to get %d bytes of memory for offsets vector\n",              printf("** Failed to get %d bytes of memory for offsets vector\n",
1970                size_offsets_max * sizeof(int));                (int)(size_offsets_max * sizeof(int)));
1971              yield = 1;              yield = 1;
1972              goto EXIT;              goto EXIT;
1973              }              }
# Line 1422  while (!done) Line 1980  while (!done)
1980          options |= PCRE_PARTIAL;          options |= PCRE_PARTIAL;
1981          continue;          continue;
1982    
1983            case 'Q':
1984            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1985            if (extra == NULL)
1986              {
1987              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1988              extra->flags = 0;
1989              }
1990            extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
1991            extra->match_limit_recursion = n;
1992            continue;
1993    
1994            case 'q':
1995            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1996            if (extra == NULL)
1997              {
1998              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1999              extra->flags = 0;
2000              }
2001            extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
2002            extra->match_limit = n;
2003            continue;
2004    
2005    #if !defined NODFA
2006          case 'R':          case 'R':
2007          options |= PCRE_DFA_RESTART;          options |= PCRE_DFA_RESTART;
2008          continue;          continue;
2009    #endif
2010    
2011          case 'S':          case 'S':
2012          show_malloc = 1;          show_malloc = 1;
2013          continue;          continue;
2014    
2015            case 'Y':
2016            options |= PCRE_NO_START_OPTIMIZE;
2017            continue;
2018    
2019          case 'Z':          case 'Z':
2020          options |= PCRE_NOTEOL;          options |= PCRE_NOTEOL;
# Line 1437  while (!done) Line 2023  while (!done)
2023          case '?':          case '?':
2024          options |= PCRE_NO_UTF8_CHECK;          options |= PCRE_NO_UTF8_CHECK;
2025          continue;          continue;
2026    
2027            case '<':
2028              {
2029              int x = check_newline(p, outfile);
2030              if (x == 0) goto NEXT_DATA;
2031              options |= x;
2032              while (*p++ != '>');
2033              }
2034            continue;
2035          }          }
2036        *q++ = c;        *q++ = c;
2037        }        }
2038      *q = 0;      *q = 0;
2039      len = q - dbuffer;      len = q - dbuffer;
2040    
2041        /* Move the data to the end of the buffer so that a read over the end of
2042        the buffer will be seen by valgrind, even if it doesn't cause a crash. If
2043        we are using the POSIX interface, we must include the terminating zero. */
2044    
2045    #if !defined NOPOSIX
2046        if (posix || do_posix)
2047          {
2048          memmove(bptr + buffer_size - len - 1, bptr, len + 1);
2049          bptr += buffer_size - len - 1;
2050          }
2051        else
2052    #endif
2053          {
2054          memmove(bptr + buffer_size - len, bptr, len);
2055          bptr += buffer_size - len;
2056          }
2057    
2058      if ((all_use_dfa || use_dfa) && find_match_limit)      if ((all_use_dfa || use_dfa) && find_match_limit)
2059        {        {
2060        printf("**Match limit not relevant for DFA matching: ignored\n");        printf("**Match limit not relevant for DFA matching: ignored\n");
# Line 1462  while (!done) Line 2074  while (!done)
2074          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);
2075        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;
2076        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;
2077          if ((options & PCRE_NOTEMPTY) != 0) eflags |= REG_NOTEMPTY;
2078    
2079        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);
2080    
2081        if (rc != 0)        if (rc != 0)
2082          {          {
2083          (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);          (void)regerror(rc, &preg, (char *)buffer, buffer_size);
2084          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
2085          }          }
2086          else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
2087                  != 0)
2088            {
2089            fprintf(outfile, "Matched with REG_NOSUB\n");
2090            }
2091        else        else
2092          {          {
2093          size_t i;          size_t i;
# Line 1501  while (!done) Line 2119  while (!done)
2119    
2120      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
2121        {        {
2122        if (timeit)        if (timeitm > 0)
2123          {          {
2124          register int i;          register int i;
2125          clock_t time_taken;          clock_t time_taken;
2126          clock_t start_time = clock();          clock_t start_time = clock();
2127    
2128    #if !defined NODFA
2129          if (all_use_dfa || use_dfa)          if (all_use_dfa || use_dfa)
2130            {            {
2131            int workspace[1000];            int workspace[1000];
2132            for (i = 0; i < LOOPREPEAT; i++)            for (i = 0; i < timeitm; i++)
2133              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
2134                options | g_notempty, use_offsets, use_size_offsets, workspace,                options | g_notempty, use_offsets, use_size_offsets, workspace,
2135                sizeof(workspace)/sizeof(int));                sizeof(workspace)/sizeof(int));
2136            }            }
2137          else          else
2138    #endif
2139    
2140          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeitm; i++)
2141            count = pcre_exec(re, extra, (char *)bptr, len,            count = pcre_exec(re, extra, (char *)bptr, len,
2142              start_offset, options | g_notempty, use_offsets, use_size_offsets);              start_offset, options | g_notempty, use_offsets, use_size_offsets);
2143    
2144          time_taken = clock() - start_time;          time_taken = clock() - start_time;
2145          fprintf(outfile, "Execute time %.3f milliseconds\n",          fprintf(outfile, "Execute time %.4f milliseconds\n",
2146            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeitm) /
2147              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
2148          }          }
2149    
2150        /* 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
2151        varying limits in order to find the minimum value. */        varying limits in order to find the minimum value for the match limit and
2152          for the recursion limit. */
2153    
2154        if (find_match_limit)        if (find_match_limit)
2155          {          {
         int min = 0;  
         int mid = 64;  
         int max = -1;  
   
2156          if (extra == NULL)          if (extra == NULL)
2157            {            {
2158            extra = (pcre_extra *)malloc(sizeof(pcre_extra));            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2159            extra->flags = 0;            extra->flags = 0;
2160            }            }
         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 */  
           }  
2161    
2162          extra->flags &= ~PCRE_EXTRA_MATCH_LIMIT;          (void)check_match_limit(re, extra, bptr, len, start_offset,
2163              options|g_notempty, use_offsets, use_size_offsets,
2164              PCRE_EXTRA_MATCH_LIMIT, &(extra->match_limit),
2165              PCRE_ERROR_MATCHLIMIT, "match()");
2166    
2167            count = check_match_limit(re, extra, bptr, len, start_offset,
2168              options|g_notempty, use_offsets, use_size_offsets,
2169              PCRE_EXTRA_MATCH_LIMIT_RECURSION, &(extra->match_limit_recursion),
2170              PCRE_ERROR_RECURSIONLIMIT, "match() recursion");
2171          }          }
2172    
2173        /* If callout_data is set, use the interface with additional data */        /* If callout_data is set, use the interface with additional data */
# Line 1591  while (!done) Line 2189  while (!done)
2189        /* The normal case is just to do the match once, with the default        /* The normal case is just to do the match once, with the default
2190        value of match_limit. */        value of match_limit. */
2191    
2192    #if !defined NODFA
2193        else if (all_use_dfa || use_dfa)        else if (all_use_dfa || use_dfa)
2194          {          {
2195          int workspace[1000];          int workspace[1000];
# Line 1603  while (!done) Line 2202  while (!done)
2202            count = use_size_offsets/2;            count = use_size_offsets/2;
2203            }            }
2204          }          }
2205    #endif
2206    
2207        else        else
2208          {          {
# Line 1619  while (!done) Line 2219  while (!done)
2219    
2220        if (count >= 0)        if (count >= 0)
2221          {          {
2222          int i;          int i, maxcount;
2223    
2224    #if !defined NODFA
2225            if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else
2226    #endif
2227              maxcount = use_size_offsets/3;
2228    
2229            /* This is a check against a lunatic return value. */
2230    
2231            if (count > maxcount)
2232              {
2233              fprintf(outfile,
2234                "** PCRE error: returned count %d is too big for offset size %d\n",
2235                count, use_size_offsets);
2236              count = use_size_offsets/3;
2237              if (do_g || do_G)
2238                {
2239                fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G');
2240                do_g = do_G = FALSE;        /* Break g/G loop */
2241                }
2242              }
2243    
2244          for (i = 0; i < count * 2; i += 2)          for (i = 0; i < count * 2; i += 2)
2245            {            {
2246            if (use_offsets[i] < 0)            if (use_offsets[i] < 0)
# Line 1647  while (!done) Line 2268  while (!done)
2268            {            {
2269            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
2270              {              {
2271              char copybuffer[16];              char copybuffer[256];
2272              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,
2273                i, copybuffer, sizeof(copybuffer));                i, copybuffer, sizeof(copybuffer));
2274              if (rc < 0)              if (rc < 0)
# Line 1657  while (!done) Line 2278  while (!done)
2278              }              }
2279            }            }
2280    
2281            for (copynamesptr = copynames;
2282                 *copynamesptr != 0;
2283                 copynamesptr += (int)strlen((char*)copynamesptr) + 1)
2284              {
2285              char copybuffer[256];
2286              int rc = pcre_copy_named_substring(re, (char *)bptr, use_offsets,
2287                count, (char *)copynamesptr, copybuffer, sizeof(copybuffer));
2288              if (rc < 0)
2289                fprintf(outfile, "copy substring %s failed %d\n", copynamesptr, rc);
2290              else
2291                fprintf(outfile, "  C %s (%d) %s\n", copybuffer, rc, copynamesptr);
2292              }
2293    
2294          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2295            {            {
2296            if ((getstrings & (1 << i)) != 0)            if ((getstrings & (1 << i)) != 0)
# Line 1669  while (!done) Line 2303  while (!done)
2303              else              else
2304                {                {
2305                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
               /* free((void *)substring); */  
2306                pcre_free_substring(substring);                pcre_free_substring(substring);
2307                }                }
2308              }              }
2309            }            }
2310    
2311            for (getnamesptr = getnames;
2312                 *getnamesptr != 0;
2313                 getnamesptr += (int)strlen((char*)getnamesptr) + 1)
2314              {
2315              const char *substring;
2316              int rc = pcre_get_named_substring(re, (char *)bptr, use_offsets,
2317                count, (char *)getnamesptr, &substring);
2318              if (rc < 0)
2319                fprintf(outfile, "copy substring %s failed %d\n", getnamesptr, rc);
2320              else
2321                {
2322                fprintf(outfile, "  G %s (%d) %s\n", substring, rc, getnamesptr);
2323                pcre_free_substring(substring);
2324                }
2325              }
2326    
2327          if (getlist)          if (getlist)
2328            {            {
2329            const char **stringlist;            const char **stringlist;
# Line 1699  while (!done) Line 2348  while (!done)
2348        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2349          {          {
2350          fprintf(outfile, "Partial match");          fprintf(outfile, "Partial match");
2351    #if !defined NODFA
2352          if ((all_use_dfa || use_dfa) && use_size_offsets > 2)          if ((all_use_dfa || use_dfa) && use_size_offsets > 2)
2353            fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],            fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],
2354              bptr + use_offsets[0]);              bptr + use_offsets[0]);
2355    #endif
2356          fprintf(outfile, "\n");          fprintf(outfile, "\n");
2357          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2358          }          }
2359    
2360        /* 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
2361        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
2362        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
2363        matching, the advance must be one character, not one byte. Fudge the        string - that was checked before setting g_notempty.
2364        offset values to achieve this. We won't be at the end of the string -  
2365        that was checked before setting g_notempty. */        Complication arises in the case when the newline option is "any" or
2366          "anycrlf". If the previous match was at the end of a line terminated by
2367          CRLF, an advance of one character just passes the \r, whereas we should
2368          prefer the longer newline sequence, as does the code in pcre_exec().
2369          Fudge the offset value to achieve this.
2370    
2371          Otherwise, in the case of UTF-8 matching, the advance must be one
2372          character, not one byte. */
2373    
2374        else        else
2375          {          {
2376          if (g_notempty != 0)          if (g_notempty != 0)
2377            {            {
2378            int onechar = 1;            int onechar = 1;
2379              unsigned int obits = ((real_pcre *)re)->options;
2380            use_offsets[0] = start_offset;            use_offsets[0] = start_offset;
2381            if (use_utf8)            if ((obits & PCRE_NEWLINE_BITS) == 0)
2382                {
2383                int d;
2384                (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);
2385                /* Note that these values are always the ASCII ones, even in
2386                EBCDIC environments. CR = 13, NL = 10. */
2387                obits = (d == 13)? PCRE_NEWLINE_CR :
2388                        (d == 10)? PCRE_NEWLINE_LF :
2389                        (d == (13<<8 | 10))? PCRE_NEWLINE_CRLF :
2390                        (d == -2)? PCRE_NEWLINE_ANYCRLF :
2391                        (d == -1)? PCRE_NEWLINE_ANY : 0;
2392                }
2393              if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2394                   (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2395                  &&
2396                  start_offset < len - 1 &&
2397                  bptr[start_offset] == '\r' &&
2398                  bptr[start_offset+1] == '\n')
2399                onechar++;
2400              else if (use_utf8)
2401              {              {
2402              while (start_offset + onechar < len)              while (start_offset + onechar < len)
2403                {                {
# Line 1754  while (!done) Line 2432  while (!done)
2432        character. */        character. */
2433    
2434        g_notempty = 0;        g_notempty = 0;
2435    
2436        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2437          {          {
2438          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
# Line 1772  while (!done) Line 2451  while (!done)
2451          len -= use_offsets[1];          len -= use_offsets[1];
2452          }          }
2453        }  /* End of loop for /g and /G */        }  /* End of loop for /g and /G */
2454    
2455        NEXT_DATA: continue;
2456      }    /* End of loop for data lines */      }    /* End of loop for data lines */
2457    
2458    CONTINUE:    CONTINUE:
# Line 1786  while (!done) Line 2467  while (!done)
2467      {      {
2468      new_free((void *)tables);      new_free((void *)tables);
2469      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");
2470        locale_set = 0;
2471      }      }
2472    }    }
2473    

Legend:
Removed from v.77  
changed lines
  Added in v.391

  ViewVC Help
Powered by ViewVC 1.1.5