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

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

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

revision 75 by nigel, Sat Feb 24 21:40:37 2007 UTC revision 343 by ph10, Mon Apr 21 16:30:37 2008 UTC
# Line 4  Line 4 
4    
5  /* This program was hacked up as a tester for PCRE. I really should have  /* This program was hacked up as a tester for PCRE. I really should have
6  written it more tidily in the first place. Will I ever learn? It has grown and  written it more tidily in the first place. Will I ever learn? It has grown and
7  been extended and consequently is now rather untidy in places.  been extended and consequently is now rather, er, *very* untidy in places.
8    
9  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
10  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# 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  /* We need the internal info for displaying the results of pcre_study(). Also  #ifdef SUPPORT_LIBREADLINE
52  for getting the opcodes for showing compiled code. */  #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 have to include pcre_internal.h because we need the internal info for
86    displaying the results of pcre_study() and we also need to know about the
87    internal macros, structures, and other internal data values; pcretest has
88    "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"
96    
97    /* We need access to the data tables that PCRE uses. So as not to have to keep
98    two copies, we include the source file here, changing the names of the external
99    symbols to prevent clashes. */
100    
101    #define _pcre_utf8_table1      utf8_table1
102    #define _pcre_utf8_table1_size utf8_table1_size
103    #define _pcre_utf8_table2      utf8_table2
104    #define _pcre_utf8_table3      utf8_table3
105    #define _pcre_utf8_table4      utf8_table4
106    #define _pcre_utt              utt
107    #define _pcre_utt_size         utt_size
108    #define _pcre_utt_names        utt_names
109    #define _pcre_OP_lengths       OP_lengths
110    
111    #include "pcre_tables.c"
112    
113    /* We also need the pcre_printint() function for printing out compiled
114    patterns. This function is in a separate file so that it can be included in
115    pcre_compile.c when that module is compiled with debugging enabled.
116    
117    The definition of the macro PRINTABLE, which determines whether to print an
118    output character as-is or as a hex value when showing compiled patterns, is
119    contained in this file. We uses it here also, in cases when the locale has not
120    been explicitly changed, so as to get consistent output from systems that
121    differ in their output from isprint() even in the "C" locale. */
122    
123    #include "pcre_printint.src"
124    
125    #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
126    
 #define PCRE_SPY        /* For Win32 build, import data, not export */  
 #include "internal.h"  
127    
128  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
129  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 58  Makefile. */ Line 133  Makefile. */
133  #include "pcreposix.h"  #include "pcreposix.h"
134  #endif  #endif
135    
136    /* It is also possible, for the benefit of the version currently imported into
137    Exim, to build pcretest without support for UTF8 (define NOUTF8), without the
138    interface to the DFA matcher (NODFA), and without the doublecheck of the old
139    "info" function (define NOINFOCHECK). In fact, we automatically cut out the
140    UTF8 support if PCRE is built without it. */
141    
142    #ifndef SUPPORT_UTF8
143    #ifndef NOUTF8
144    #define NOUTF8
145    #endif
146    #endif
147    
148    
149    /* Other parameters */
150    
151  #ifndef CLOCKS_PER_SEC  #ifndef CLOCKS_PER_SEC
152  #ifdef CLK_TCK  #ifdef CLK_TCK
153  #define CLOCKS_PER_SEC CLK_TCK  #define CLOCKS_PER_SEC CLK_TCK
# Line 66  Makefile. */ Line 156  Makefile. */
156  #endif  #endif
157  #endif  #endif
158    
159  #define LOOPREPEAT 500000  /* This is the default loop count for timing. */
160    
161  #define BUFFER_SIZE 30000  #define LOOPREPEAT 500000
 #define PBUFFER_SIZE BUFFER_SIZE  
 #define DBUFFER_SIZE BUFFER_SIZE  
162    
163    /* Static variables */
164    
165  static FILE *outfile;  static FILE *outfile;
166  static int log_store = 0;  static int log_store = 0;
# Line 79  static int callout_count; Line 168  static int callout_count;
168  static int callout_extra;  static int callout_extra;
169  static int callout_fail_count;  static int callout_fail_count;
170  static int callout_fail_id;  static int callout_fail_id;
171    static int debug_lengths;
172  static int first_callout;  static int first_callout;
173    static int locale_set = 0;
174  static int show_malloc;  static int show_malloc;
175  static int use_utf8;  static int use_utf8;
176  static size_t gotten_store;  static size_t gotten_store;
177    
178    /* The buffers grow automatically if very long input lines are encountered. */
179    
180    static int buffer_size = 50000;
181    static uschar *buffer = NULL;
182    static uschar *dbuffer = NULL;
183  static uschar *pbuffer = NULL;  static uschar *pbuffer = NULL;
184    
185    
 static const int utf8_table1[] = {  
   0x0000007f, 0x000007ff, 0x0000ffff, 0x001fffff, 0x03ffffff, 0x7fffffff};  
186    
187  static const int utf8_table2[] = {  /*************************************************
188    0, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc};  *        Read or extend an input line            *
189    *************************************************/
190    
191  static const int utf8_table3[] = {  /* Input lines are read into buffer, but both patterns and data lines can be
192    0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01};  continued over multiple input lines. In addition, if the buffer fills up, we
193    want to automatically expand it so as to be able to handle extremely large
194    lines that are needed for certain stress tests. When the input buffer is
195    expanded, the other two buffers must also be expanded likewise, and the
196    contents of pbuffer, which are a copy of the input for callouts, must be
197    preserved (for when expansion happens for a data line). This is not the most
198    optimal way of handling this, but hey, this is just a test program!
199    
200    Arguments:
201      f            the file to read
202      start        where in buffer to start (this *must* be within buffer)
203      prompt       for stdin or readline()
204    
205    Returns:       pointer to the start of new data
206                   could be a copy of start, or could be moved
207                   NULL if no data read and EOF reached
208    */
209    
210    static uschar *
211    extend_inputline(FILE *f, uschar *start, const char *prompt)
212    {
213    uschar *here = start;
214    
215    for (;;)
216      {
217      int rlen = buffer_size - (here - buffer);
218    
219      if (rlen > 1000)
220        {
221        int dlen;
222    
223        /* If libreadline support is required, use readline() to read a line if the
224        input is a terminal. Note that readline() removes the trailing newline, so
225        we must put it back again, to be compatible with fgets(). */
226    
227    #ifdef SUPPORT_LIBREADLINE
228        if (isatty(fileno(f)))
229          {
230          size_t len;
231          char *s = readline(prompt);
232          if (s == NULL) return (here == start)? NULL : start;
233          len = strlen(s);
234          if (len > 0) add_history(s);
235          if (len > rlen - 1) len = rlen - 1;
236          memcpy(here, s, len);
237          here[len] = '\n';
238          here[len+1] = 0;
239          free(s);
240          }
241        else
242    #endif
243    
244        /* Read the next line by normal means, prompting if the file is stdin. */
245    
246          {
247          if (f == stdin) printf(prompt);
248          if (fgets((char *)here, rlen,  f) == NULL)
249            return (here == start)? NULL : start;
250          }
251    
252        dlen = (int)strlen((char *)here);
253        if (dlen > 0 && here[dlen - 1] == '\n') return start;
254        here += dlen;
255        }
256    
257      else
258        {
259        int new_buffer_size = 2*buffer_size;
260        uschar *new_buffer = (unsigned char *)malloc(new_buffer_size);
261        uschar *new_dbuffer = (unsigned char *)malloc(new_buffer_size);
262        uschar *new_pbuffer = (unsigned char *)malloc(new_buffer_size);
263    
264        if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
265          {
266          fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
267          exit(1);
268          }
269    
270        memcpy(new_buffer, buffer, buffer_size);
271        memcpy(new_pbuffer, pbuffer, buffer_size);
272    
273        buffer_size = new_buffer_size;
274    
275        start = new_buffer + (start - buffer);
276        here = new_buffer + (here - buffer);
277    
278        free(buffer);
279        free(dbuffer);
280        free(pbuffer);
281    
282        buffer = new_buffer;
283        dbuffer = new_dbuffer;
284        pbuffer = new_pbuffer;
285        }
286      }
287    
288    return NULL;  /* Control never gets here */
289    }
290    
291    
292    
 /*************************************************  
 *         Print compiled regex                   *  
 *************************************************/  
293    
 /* The code for doing this is held in a separate file that is also included in  
 pcre.c when it is compiled with the debug switch. It defines a function called  
 print_internals(), which uses a table of opcode lengths defined by the macro  
 OP_LENGTHS, whose name must be OP_lengths. It also uses a table that translates  
 Unicode property names to numbers; this is kept in a separate file. */  
   
 static uschar OP_lengths[] = { OP_LENGTHS };  
   
 #include "ucp.h"  
 #include "ucptypetable.c"  
 #include "printint.c"  
294    
295    
296    
# Line 122  static uschar OP_lengths[] = { OP_LENGTH Line 300  static uschar OP_lengths[] = { OP_LENGTH
300    
301  /* 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
302  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
303  for unpicking the -o argument, so just keep it simple.  for unpicking arguments, so just keep it simple.
304    
305  Arguments:  Arguments:
306    str           string to be converted    str           string to be converted
# Line 143  return(result); Line 321  return(result);
321    
322    
323    
 /*************************************************  
 *       Convert character value to UTF-8         *  
 *************************************************/  
   
 /* This function takes an integer value in the range 0 - 0x7fffffff  
 and encodes it as a UTF-8 character in 0 to 6 bytes.  
   
 Arguments:  
   cvalue     the character value  
   buffer     pointer to buffer for result - at least 6 bytes long  
   
 Returns:     number of characters placed in the buffer  
              -1 if input character is negative  
              0 if input character is positive but too big (only when  
              int is longer than 32 bits)  
 */  
   
 static int  
 ord2utf8(int cvalue, unsigned char *buffer)  
 {  
 register int i, j;  
 for (i = 0; i < sizeof(utf8_table1)/sizeof(int); i++)  
   if (cvalue <= utf8_table1[i]) break;  
 if (i >= sizeof(utf8_table1)/sizeof(int)) return 0;  
 if (cvalue < 0) return -1;  
   
 buffer += i;  
 for (j = i; j > 0; j--)  
  {  
  *buffer-- = 0x80 | (cvalue & 0x3f);  
  cvalue >>= 6;  
  }  
 *buffer = utf8_table2[i] | cvalue;  
 return i + 1;  
 }  
   
324    
325  /*************************************************  /*************************************************
326  *            Convert UTF-8 string to value       *  *            Convert UTF-8 string to value       *
# Line 188  return i + 1; Line 330  return i + 1;
330  and returns the value of the character.  and returns the value of the character.
331    
332  Argument:  Argument:
333    buffer   a pointer to the byte vector    utf8bytes   a pointer to the byte vector
334    vptr     a pointer to an int to receive the value    vptr        a pointer to an int to receive the value
335    
336  Returns:   >  0 => the number of bytes consumed  Returns:      >  0 => the number of bytes consumed
337             -6 to 0 => malformed UTF-8 character at offset = (-return)                -6 to 0 => malformed UTF-8 character at offset = (-return)
338  */  */
339    
340    #if !defined NOUTF8
341    
342  static int  static int
343  utf82ord(unsigned char *buffer, int *vptr)  utf82ord(unsigned char *utf8bytes, int *vptr)
344  {  {
345  int c = *buffer++;  int c = *utf8bytes++;
346  int d = c;  int d = c;
347  int i, j, s;  int i, j, s;
348    
# Line 218  d = (c & utf8_table3[i]) << s; Line 362  d = (c & utf8_table3[i]) << s;
362    
363  for (j = 0; j < i; j++)  for (j = 0; j < i; j++)
364    {    {
365    c = *buffer++;    c = *utf8bytes++;
366    if ((c & 0xc0) != 0x80) return -(j+1);    if ((c & 0xc0) != 0x80) return -(j+1);
367    s -= 6;    s -= 6;
368    d |= (c & 0x3f) << s;    d |= (c & 0x3f) << s;
# Line 226  for (j = 0; j < i; j++) Line 370  for (j = 0; j < i; j++)
370    
371  /* Check that encoding was the correct unique one */  /* Check that encoding was the correct unique one */
372    
373  for (j = 0; j < sizeof(utf8_table1)/sizeof(int); j++)  for (j = 0; j < utf8_table1_size; j++)
374    if (d <= utf8_table1[j]) break;    if (d <= utf8_table1[j]) break;
375  if (j != i) return -(i+1);  if (j != i) return -(i+1);
376    
# Line 236  if (j != i) return -(i+1); Line 380  if (j != i) return -(i+1);
380  return i+1;  return i+1;
381  }  }
382    
383    #endif
384    
385    
386    
387    /*************************************************
388    *       Convert character value to UTF-8         *
389    *************************************************/
390    
391    /* This function takes an integer value in the range 0 - 0x7fffffff
392    and encodes it as a UTF-8 character in 0 to 6 bytes.
393    
394    Arguments:
395      cvalue     the character value
396      utf8bytes  pointer to buffer for result - at least 6 bytes long
397    
398    Returns:     number of characters placed in the buffer
399    */
400    
401    #if !defined NOUTF8
402    
403    static int
404    ord2utf8(int cvalue, uschar *utf8bytes)
405    {
406    register int i, j;
407    for (i = 0; i < utf8_table1_size; i++)
408      if (cvalue <= utf8_table1[i]) break;
409    utf8bytes += i;
410    for (j = i; j > 0; j--)
411     {
412     *utf8bytes-- = 0x80 | (cvalue & 0x3f);
413     cvalue >>= 6;
414     }
415    *utf8bytes = utf8_table2[i] | cvalue;
416    return i + 1;
417    }
418    
419    #endif
420    
421    
422    
423  /*************************************************  /*************************************************
# Line 248  chars without printing. */ Line 430  chars without printing. */
430    
431  static int pchars(unsigned char *p, int length, FILE *f)  static int pchars(unsigned char *p, int length, FILE *f)
432  {  {
433  int c;  int c = 0;
434  int yield = 0;  int yield = 0;
435    
436  while (length-- > 0)  while (length-- > 0)
437    {    {
438    #if !defined NOUTF8
439    if (use_utf8)    if (use_utf8)
440      {      {
441      int rc = utf82ord(p, &c);      int rc = utf82ord(p, &c);
# Line 261  while (length-- > 0) Line 444  while (length-- > 0)
444        {        {
445        length -= rc - 1;        length -= rc - 1;
446        p += rc;        p += rc;
447        if (c < 256 && isprint(c))        if (PRINTHEX(c))
448          {          {
449          if (f != NULL) fprintf(f, "%c", c);          if (f != NULL) fprintf(f, "%c", c);
450          yield++;          yield++;
451          }          }
452        else        else
453          {          {
454          int n;          int n = 4;
455          if (f != NULL) fprintf(f, "\\x{%02x}%n", c, &n);          if (f != NULL) fprintf(f, "\\x{%02x}", c);
456          yield += n;          yield += (n <= 0x000000ff)? 2 :
457                     (n <= 0x00000fff)? 3 :
458                     (n <= 0x0000ffff)? 4 :
459                     (n <= 0x000fffff)? 5 : 6;
460          }          }
461        continue;        continue;
462        }        }
463      }      }
464    #endif
465    
466     /* Not UTF-8, or malformed UTF-8  */     /* Not UTF-8, or malformed UTF-8  */
467    
468    if (isprint(c = *(p++)))    c = *p++;
469      if (PRINTHEX(c))
470      {      {
471      if (f != NULL) fprintf(f, "%c", c);      if (f != NULL) fprintf(f, "%c", c);
472      yield++;      yield++;
# Line 403  static void *new_malloc(size_t size) Line 591  static void *new_malloc(size_t size)
591  void *block = malloc(size);  void *block = malloc(size);
592  gotten_store = size;  gotten_store = size;
593  if (show_malloc)  if (show_malloc)
594    fprintf(outfile, "malloc       %3d %p\n", size, block);    fprintf(outfile, "malloc       %3d %p\n", (int)size, block);
595  return block;  return block;
596  }  }
597    
# Line 421  static void *stack_malloc(size_t size) Line 609  static void *stack_malloc(size_t size)
609  {  {
610  void *block = malloc(size);  void *block = malloc(size);
611  if (show_malloc)  if (show_malloc)
612    fprintf(outfile, "stack_malloc %3d %p\n", size, block);    fprintf(outfile, "stack_malloc %3d %p\n", (int)size, block);
613  return block;  return block;
614  }  }
615    
# Line 452  if ((rc = pcre_fullinfo(re, study, optio Line 640  if ((rc = pcre_fullinfo(re, study, optio
640  *         Byte flipping function                 *  *         Byte flipping function                 *
641  *************************************************/  *************************************************/
642    
643  static long int  static unsigned long int
644  byteflip(long int value, int n)  byteflip(unsigned long int value, int n)
645  {  {
646  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);
647  return ((value & 0x000000ff) << 24) |  return ((value & 0x000000ff) << 24) |
# Line 466  return ((value & 0x000000ff) << 24) | Line 654  return ((value & 0x000000ff) << 24) |
654    
655    
656  /*************************************************  /*************************************************
657    *        Check match or recursion limit          *
658    *************************************************/
659    
660    static int
661    check_match_limit(pcre *re, pcre_extra *extra, uschar *bptr, int len,
662      int start_offset, int options, int *use_offsets, int use_size_offsets,
663      int flag, unsigned long int *limit, int errnumber, const char *msg)
664    {
665    int count;
666    int min = 0;
667    int mid = 64;
668    int max = -1;
669    
670    extra->flags |= flag;
671    
672    for (;;)
673      {
674      *limit = mid;
675    
676      count = pcre_exec(re, extra, (char *)bptr, len, start_offset, options,
677        use_offsets, use_size_offsets);
678    
679      if (count == errnumber)
680        {
681        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
682        min = mid;
683        mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;
684        }
685    
686      else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||
687                             count == PCRE_ERROR_PARTIAL)
688        {
689        if (mid == min + 1)
690          {
691          fprintf(outfile, "Minimum %s limit = %d\n", msg, mid);
692          break;
693          }
694        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
695        max = mid;
696        mid = (min + mid)/2;
697        }
698      else break;    /* Some other error */
699      }
700    
701    extra->flags &= ~flag;
702    return count;
703    }
704    
705    
706    
707    /*************************************************
708    *         Case-independent strncmp() function    *
709    *************************************************/
710    
711    /*
712    Arguments:
713      s         first string
714      t         second string
715      n         number of characters to compare
716    
717    Returns:    < 0, = 0, or > 0, according to the comparison
718    */
719    
720    static int
721    strncmpic(uschar *s, uschar *t, int n)
722    {
723    while (n--)
724      {
725      int c = tolower(*s++) - tolower(*t++);
726      if (c) return c;
727      }
728    return 0;
729    }
730    
731    
732    
733    /*************************************************
734    *         Check newline indicator                *
735    *************************************************/
736    
737    /* This is used both at compile and run-time to check for <xxx> escapes, where
738    xxx is LF, CR, CRLF, ANYCRLF, or ANY. Print a message and return 0 if there is
739    no match.
740    
741    Arguments:
742      p           points after the leading '<'
743      f           file for error message
744    
745    Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
746    */
747    
748    static int
749    check_newline(uschar *p, FILE *f)
750    {
751    if (strncmpic(p, (uschar *)"cr>", 3) == 0) return PCRE_NEWLINE_CR;
752    if (strncmpic(p, (uschar *)"lf>", 3) == 0) return PCRE_NEWLINE_LF;
753    if (strncmpic(p, (uschar *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
754    if (strncmpic(p, (uschar *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF;
755    if (strncmpic(p, (uschar *)"any>", 4) == 0) return PCRE_NEWLINE_ANY;
756    if (strncmpic(p, (uschar *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF;
757    if (strncmpic(p, (uschar *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE;
758    fprintf(f, "Unknown newline type at: <%s\n", p);
759    return 0;
760    }
761    
762    
763    
764    /*************************************************
765    *             Usage function                     *
766    *************************************************/
767    
768    static void
769    usage(void)
770    {
771    printf("Usage:     pcretest [options] [<input file> [<output file>]]\n\n");
772    printf("Input and output default to stdin and stdout.\n");
773    #ifdef SUPPORT_LIBREADLINE
774    printf("If input is a terminal, readline() is used to read from it.\n");
775    #else
776    printf("This version of pcretest is not linked with readline().\n");
777    #endif
778    printf("\nOptions:\n");
779    printf("  -b       show compiled code (bytecode)\n");
780    printf("  -C       show PCRE compile-time options and exit\n");
781    printf("  -d       debug: show compiled code and information (-b and -i)\n");
782    #if !defined NODFA
783    printf("  -dfa     force DFA matching for all subjects\n");
784    #endif
785    printf("  -help    show usage information\n");
786    printf("  -i       show information about compiled patterns\n"
787           "  -m       output memory used information\n"
788           "  -o <n>   set size of offsets vector to <n>\n");
789    #if !defined NOPOSIX
790    printf("  -p       use POSIX interface\n");
791    #endif
792    printf("  -q       quiet: do not output PCRE version number at start\n");
793    printf("  -S <n>   set stack size to <n> megabytes\n");
794    printf("  -s       output store (memory) used information\n"
795           "  -t       time compilation and execution\n");
796    printf("  -t <n>   time compilation and execution, repeating <n> times\n");
797    printf("  -tm      time execution (matching) only\n");
798    printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
799    }
800    
801    
802    
803    /*************************************************
804  *                Main Program                    *  *                Main Program                    *
805  *************************************************/  *************************************************/
806    
# Line 480  int options = 0; Line 815  int options = 0;
815  int study_options = 0;  int study_options = 0;
816  int op = 1;  int op = 1;
817  int timeit = 0;  int timeit = 0;
818    int timeitm = 0;
819  int showinfo = 0;  int showinfo = 0;
820  int showstore = 0;  int showstore = 0;
821    int quiet = 0;
822  int size_offsets = 45;  int size_offsets = 45;
823  int size_offsets_max;  int size_offsets_max;
824  int *offsets;  int *offsets = NULL;
825  #if !defined NOPOSIX  #if !defined NOPOSIX
826  int posix = 0;  int posix = 0;
827  #endif  #endif
828  int debug = 0;  int debug = 0;
829  int done = 0;  int done = 0;
830    int all_use_dfa = 0;
831    int yield = 0;
832    int stack_size;
833    
834    /* These vectors store, end-to-end, a list of captured substring names. Assume
835    that 1024 is plenty long enough for the few names we'll be testing. */
836    
837  unsigned char *buffer;  uschar copynames[1024];
838  unsigned char *dbuffer;  uschar getnames[1024];
839    
840    uschar *copynamesptr;
841    uschar *getnamesptr;
842    
843  /* Get buffers from malloc() so that Electric Fence will check their misuse  /* Get buffers from malloc() so that Electric Fence will check their misuse
844  when I am debugging. */  when I am debugging. They grow automatically when very long lines are read. */
845    
846  buffer = (unsigned char *)malloc(BUFFER_SIZE);  buffer = (unsigned char *)malloc(buffer_size);
847  dbuffer = (unsigned char *)malloc(DBUFFER_SIZE);  dbuffer = (unsigned char *)malloc(buffer_size);
848  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.  */  
849    
850  #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) */  
851    
852  outfile = stdout;  outfile = stdout;
853    
854    /* The following  _setmode() stuff is some Windows magic that tells its runtime
855    library to translate CRLF into a single LF character. At least, that's what
856    I've been told: never having used Windows I take this all on trust. Originally
857    it set 0x8000, but then I was advised that _O_BINARY was better. */
858    
859    #if defined(_WIN32) || defined(WIN32)
860    _setmode( _fileno( stdout ), _O_BINARY );
861    #endif
862    
863  /* Scan options */  /* Scan options */
864    
865  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
# Line 519  while (argc > 1 && argv[op][0] == '-') Line 868  while (argc > 1 && argv[op][0] == '-')
868    
869    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
870      showstore = 1;      showstore = 1;
871    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-q") == 0) quiet = 1;
872      else if (strcmp(argv[op], "-b") == 0) debug = 1;
873    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
874    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
875    #if !defined NODFA
876      else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
877    #endif
878    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&
879        ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),        ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),
880          *endptr == 0))          *endptr == 0))
# Line 529  while (argc > 1 && argv[op][0] == '-') Line 882  while (argc > 1 && argv[op][0] == '-')
882      op++;      op++;
883      argc--;      argc--;
884      }      }
885      else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
886        {
887        int both = argv[op][2] == 0;
888        int temp;
889        if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr),
890                         *endptr == 0))
891          {
892          timeitm = temp;
893          op++;
894          argc--;
895          }
896        else timeitm = LOOPREPEAT;
897        if (both) timeit = timeitm;
898        }
899      else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
900          ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
901            *endptr == 0))
902        {
903    #if defined(_WIN32) || defined(WIN32)
904        printf("PCRE: -S not supported on this OS\n");
905        exit(1);
906    #else
907        int rc;
908        struct rlimit rlim;
909        getrlimit(RLIMIT_STACK, &rlim);
910        rlim.rlim_cur = stack_size * 1024 * 1024;
911        rc = setrlimit(RLIMIT_STACK, &rlim);
912        if (rc != 0)
913          {
914        printf("PCRE: setrlimit() failed with error %d\n", rc);
915        exit(1);
916          }
917        op++;
918        argc--;
919    #endif
920        }
921  #if !defined NOPOSIX  #if !defined NOPOSIX
922    else if (strcmp(argv[op], "-p") == 0) posix = 1;    else if (strcmp(argv[op], "-p") == 0) posix = 1;
923  #endif  #endif
# Line 542  while (argc > 1 && argv[op][0] == '-') Line 931  while (argc > 1 && argv[op][0] == '-')
931      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
932      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
933      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
934      printf("  Newline character is %s\n", (rc == '\r')? "CR" : "LF");      printf("  Newline sequence is %s\n", (rc == '\r')? "CR" :
935          (rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" :
936          (rc == -2)? "ANYCRLF" :
937          (rc == -1)? "ANY" : "???");
938        (void)pcre_config(PCRE_CONFIG_BSR, &rc);
939        printf("  \\R matches %s\n", rc? "CR, LF, or CRLF only" :
940                                         "all Unicode newlines");
941      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);
942      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
943      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
944      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
945      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);
946      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %d\n", rc);
947        (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &rc);
948        printf("  Default recursion depth limit = %d\n", rc);
949      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
950      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
951      exit(0);      goto EXIT;
952        }
953      else if (strcmp(argv[op], "-help") == 0 ||
954               strcmp(argv[op], "--help") == 0)
955        {
956        usage();
957        goto EXIT;
958      }      }
959    else    else
960      {      {
961      printf("** Unknown or malformed option %s\n", argv[op]);      printf("** Unknown or malformed option %s\n", argv[op]);
962      printf("Usage:   pcretest [-d] [-i] [-o <n>] [-p] [-s] [-t] [<input> [<output>]]\n");      usage();
963      printf("  -C     show PCRE compile-time options and exit\n");      yield = 1;
964      printf("  -d     debug: show compiled code; implies -i\n"      goto EXIT;
            "  -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");  
     return 1;  
965      }      }
966    op++;    op++;
967    argc--;    argc--;
# Line 580  offsets = (int *)malloc(size_offsets_max Line 974  offsets = (int *)malloc(size_offsets_max
974  if (offsets == NULL)  if (offsets == NULL)
975    {    {
976    printf("** Failed to get %d bytes of memory for offsets vector\n",    printf("** Failed to get %d bytes of memory for offsets vector\n",
977      size_offsets_max * sizeof(int));      (int)(size_offsets_max * sizeof(int)));
978    return 1;    yield = 1;
979      goto EXIT;
980    }    }
981    
982  /* Sort out the input and output files */  /* Sort out the input and output files */
983    
984  if (argc > 1)  if (argc > 1)
985    {    {
986    infile = fopen(argv[op], "rb");    infile = fopen(argv[op], INPUT_MODE);
987    if (infile == NULL)    if (infile == NULL)
988      {      {
989      printf("** Failed to open %s\n", argv[op]);      printf("** Failed to open %s\n", argv[op]);
990      return 1;      yield = 1;
991        goto EXIT;
992      }      }
993    }    }
994    
995  if (argc > 2)  if (argc > 2)
996    {    {
997    outfile = fopen(argv[op+1], "wb");    outfile = fopen(argv[op+1], OUTPUT_MODE);
998    if (outfile == NULL)    if (outfile == NULL)
999      {      {
1000      printf("** Failed to open %s\n", argv[op+1]);      printf("** Failed to open %s\n", argv[op+1]);
1001      return 1;      yield = 1;
1002        goto EXIT;
1003      }      }
1004    }    }
1005    
# Line 613  pcre_free = new_free; Line 1010  pcre_free = new_free;
1010  pcre_stack_malloc = stack_malloc;  pcre_stack_malloc = stack_malloc;
1011  pcre_stack_free = stack_free;  pcre_stack_free = stack_free;
1012    
1013  /* Heading line, then prompt for first regex if stdin */  /* Heading line unless quiet, then prompt for first regex if stdin */
1014    
1015  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  if (!quiet) fprintf(outfile, "PCRE version %s\n\n", pcre_version());
1016    
1017  /* Main loop */  /* Main loop */
1018    
# Line 642  while (!done) Line 1039  while (!done)
1039    int do_showinfo = showinfo;    int do_showinfo = showinfo;
1040    int do_showrest = 0;    int do_showrest = 0;
1041    int do_flip = 0;    int do_flip = 0;
1042    int erroroffset, len, delimiter;    int erroroffset, len, delimiter, poffset;
1043    
1044    use_utf8 = 0;    use_utf8 = 0;
1045      debug_lengths = 1;
1046    
1047    if (infile == stdin) printf("  re> ");    if (extend_inputline(infile, buffer, "  re> ") == NULL) break;
   if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break;  
1048    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1049    fflush(outfile);    fflush(outfile);
1050    
# Line 659  while (!done) Line 1056  while (!done)
1056    
1057    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
1058      {      {
1059      unsigned long int magic;      unsigned long int magic, get_options;
1060      uschar sbuf[8];      uschar sbuf[8];
1061      FILE *f;      FILE *f;
1062    
# Line 707  while (!done) Line 1104  while (!done)
1104    
1105      /* Need to know if UTF-8 for printing data strings */      /* Need to know if UTF-8 for printing data strings */
1106    
1107      new_info(re, NULL, PCRE_INFO_OPTIONS, &options);      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1108      use_utf8 = (options & PCRE_UTF8) != 0;      use_utf8 = (get_options & PCRE_UTF8) != 0;
1109    
1110      /* Now see if there is any following study data */      /* Now see if there is any following study data */
1111    
# Line 747  while (!done) Line 1144  while (!done)
1144    
1145    if (isalnum(delimiter) || delimiter == '\\')    if (isalnum(delimiter) || delimiter == '\\')
1146      {      {
1147      fprintf(outfile, "** Delimiter must not be alphameric or \\\n");      fprintf(outfile, "** Delimiter must not be alphanumeric or \\\n");
1148      goto SKIP_DATA;      goto SKIP_DATA;
1149      }      }
1150    
1151    pp = p;    pp = p;
1152      poffset = p - buffer;
1153    
1154    for(;;)    for(;;)
1155      {      {
# Line 762  while (!done) Line 1160  while (!done)
1160        pp++;        pp++;
1161        }        }
1162      if (*pp != 0) break;      if (*pp != 0) break;
1163        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)  
1164        {        {
1165        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
1166        done = 1;        done = 1;
# Line 780  while (!done) Line 1169  while (!done)
1169      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
1170      }      }
1171    
1172      /* The buffer may have moved while being extended; reset the start of data
1173      pointer to the correct relative point in the buffer. */
1174    
1175      p = buffer + poffset;
1176    
1177    /* If the first character after the delimiter is backslash, make    /* If the first character after the delimiter is backslash, make
1178    the pattern end with backslash. This is purely to provide a way    the pattern end with backslash. This is purely to provide a way
1179    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 802  while (!done) Line 1196  while (!done)
1196      {      {
1197      switch (*pp++)      switch (*pp++)
1198        {        {
1199          case 'f': options |= PCRE_FIRSTLINE; break;
1200        case 'g': do_g = 1; break;        case 'g': do_g = 1; break;
1201        case 'i': options |= PCRE_CASELESS; break;        case 'i': options |= PCRE_CASELESS; break;
1202        case 'm': options |= PCRE_MULTILINE; break;        case 'm': options |= PCRE_MULTILINE; break;
# Line 810  while (!done) Line 1205  while (!done)
1205    
1206        case '+': do_showrest = 1; break;        case '+': do_showrest = 1; break;
1207        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
1208          case 'B': do_debug = 1; break;
1209        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
1210        case 'D': do_debug = do_showinfo = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
1211        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
1212        case 'F': do_flip = 1; break;        case 'F': do_flip = 1; break;
1213        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1214        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1215          case 'J': options |= PCRE_DUPNAMES; break;
1216        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1217        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1218    
# Line 826  while (!done) Line 1223  while (!done)
1223        case 'S': do_study = 1; break;        case 'S': do_study = 1; break;
1224        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1225        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
1226          case 'Z': debug_lengths = 0; break;
1227        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1228        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
1229    
1230        case 'L':        case 'L':
1231        ppp = pp;        ppp = pp;
1232        while (*ppp != '\n' && *ppp != ' ') ppp++;        /* The '\r' test here is so that it works on Windows. */
1233          /* The '0' test is just in case this is an unterminated line. */
1234          while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
1235        *ppp = 0;        *ppp = 0;
1236        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
1237          {          {
1238          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
1239          goto SKIP_DATA;          goto SKIP_DATA;
1240          }          }
1241          locale_set = 1;
1242        tables = pcre_maketables();        tables = pcre_maketables();
1243        pp = ppp;        pp = ppp;
1244        break;        break;
# Line 849  while (!done) Line 1250  while (!done)
1250        *pp = 0;        *pp = 0;
1251        break;        break;
1252    
1253        case '\n': case ' ': break;        case '<':
1254            {
1255            if (strncmp((char *)pp, "JS>", 3) == 0)
1256              {
1257              options |= PCRE_JAVASCRIPT_COMPAT;
1258              pp += 3;
1259              }
1260            else
1261              {
1262              int x = check_newline(pp, outfile);
1263              if (x == 0) goto SKIP_DATA;
1264              options |= x;
1265              while (*pp++ != '>');
1266              }
1267            }
1268          break;
1269    
1270          case '\r':                      /* So that it works in Windows */
1271          case '\n':
1272          case ' ':
1273          break;
1274    
1275        default:        default:
1276        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);
# Line 869  while (!done) Line 1290  while (!done)
1290    
1291      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
1292      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
1293        if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1294        if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1295        if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1296    
1297      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1298    
1299      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
# Line 876  while (!done) Line 1301  while (!done)
1301    
1302      if (rc != 0)      if (rc != 0)
1303        {        {
1304        (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);        (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1305        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
1306        goto SKIP_DATA;        goto SKIP_DATA;
1307        }        }
# Line 888  while (!done) Line 1313  while (!done)
1313  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1314    
1315      {      {
1316      if (timeit)      if (timeit > 0)
1317        {        {
1318        register int i;        register int i;
1319        clock_t time_taken;        clock_t time_taken;
1320        clock_t start_time = clock();        clock_t start_time = clock();
1321        for (i = 0; i < LOOPREPEAT; i++)        for (i = 0; i < timeit; i++)
1322          {          {
1323          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1324          if (re != NULL) free(re);          if (re != NULL) free(re);
1325          }          }
1326        time_taken = clock() - start_time;        time_taken = clock() - start_time;
1327        fprintf(outfile, "Compile time %.3f milliseconds\n",        fprintf(outfile, "Compile time %.4f milliseconds\n",
1328          (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /          (((double)time_taken * 1000.0) / (double)timeit) /
1329            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
1330        }        }
1331    
# Line 917  while (!done) Line 1342  while (!done)
1342          {          {
1343          for (;;)          for (;;)
1344            {            {
1345            if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)            if (extend_inputline(infile, buffer, NULL) == NULL)
1346              {              {
1347              done = 1;              done = 1;
1348              goto CONTINUE;              goto CONTINUE;
# Line 952  while (!done) Line 1377  while (!done)
1377    
1378      if (do_study)      if (do_study)
1379        {        {
1380        if (timeit)        if (timeit > 0)
1381          {          {
1382          register int i;          register int i;
1383          clock_t time_taken;          clock_t time_taken;
1384          clock_t start_time = clock();          clock_t start_time = clock();
1385          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeit; i++)
1386            extra = pcre_study(re, study_options, &error);            extra = pcre_study(re, study_options, &error);
1387          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1388          if (extra != NULL) free(extra);          if (extra != NULL) free(extra);
1389          fprintf(outfile, "  Study time %.3f milliseconds\n",          fprintf(outfile, "  Study time %.4f milliseconds\n",
1390            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeit) /
1391              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
1392          }          }
1393        extra = pcre_study(re, study_options, &error);        extra = pcre_study(re, study_options, &error);
# Line 980  while (!done) Line 1405  while (!done)
1405      if (do_flip)      if (do_flip)
1406        {        {
1407        real_pcre *rre = (real_pcre *)re;        real_pcre *rre = (real_pcre *)re;
1408        rre->magic_number = byteflip(rre->magic_number, sizeof(rre->magic_number));        rre->magic_number =
1409            byteflip(rre->magic_number, sizeof(rre->magic_number));
1410        rre->size = byteflip(rre->size, sizeof(rre->size));        rre->size = byteflip(rre->size, sizeof(rre->size));
1411        rre->options = byteflip(rre->options, sizeof(rre->options));        rre->options = byteflip(rre->options, sizeof(rre->options));
1412        rre->top_bracket = byteflip(rre->top_bracket, sizeof(rre->top_bracket));        rre->flags = (pcre_uint16)byteflip(rre->flags, sizeof(rre->flags));
1413        rre->top_backref = byteflip(rre->top_backref, sizeof(rre->top_backref));        rre->top_bracket =
1414        rre->first_byte = byteflip(rre->first_byte, sizeof(rre->first_byte));          (pcre_uint16)byteflip(rre->top_bracket, sizeof(rre->top_bracket));
1415        rre->req_byte = byteflip(rre->req_byte, sizeof(rre->req_byte));        rre->top_backref =
1416        rre->name_table_offset = byteflip(rre->name_table_offset,          (pcre_uint16)byteflip(rre->top_backref, sizeof(rre->top_backref));
1417          rre->first_byte =
1418            (pcre_uint16)byteflip(rre->first_byte, sizeof(rre->first_byte));
1419          rre->req_byte =
1420            (pcre_uint16)byteflip(rre->req_byte, sizeof(rre->req_byte));
1421          rre->name_table_offset = (pcre_uint16)byteflip(rre->name_table_offset,
1422          sizeof(rre->name_table_offset));          sizeof(rre->name_table_offset));
1423        rre->name_entry_size = byteflip(rre->name_entry_size,        rre->name_entry_size = (pcre_uint16)byteflip(rre->name_entry_size,
1424          sizeof(rre->name_entry_size));          sizeof(rre->name_entry_size));
1425        rre->name_count = byteflip(rre->name_count, sizeof(rre->name_count));        rre->name_count = (pcre_uint16)byteflip(rre->name_count,
1426            sizeof(rre->name_count));
1427    
1428        if (extra != NULL)        if (extra != NULL)
1429          {          {
# Line 1005  while (!done) Line 1437  while (!done)
1437    
1438      SHOW_INFO:      SHOW_INFO:
1439    
1440        if (do_debug)
1441          {
1442          fprintf(outfile, "------------------------------------------------------------------\n");
1443          pcre_printint(re, outfile, debug_lengths);
1444          }
1445    
1446      if (do_showinfo)      if (do_showinfo)
1447        {        {
1448        unsigned long int get_options, all_options;        unsigned long int get_options, all_options;
1449    #if !defined NOINFOCHECK
1450        int old_first_char, old_options, old_count;        int old_first_char, old_options, old_count;
1451        int count, backrefmax, first_char, need_char;  #endif
1452          int count, backrefmax, first_char, need_char, okpartial, jchanged,
1453            hascrorlf;
1454        int nameentrysize, namecount;        int nameentrysize, namecount;
1455        const uschar *nametable;        const uschar *nametable;
1456    
       if (do_debug)  
         {  
         fprintf(outfile, "------------------------------------------------------------------\n");  
         print_internals(re, outfile);  
         }  
   
1457        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1458        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1459        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
# Line 1028  while (!done) Line 1463  while (!done)
1463        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);
1464        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);
1465        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);
1466          new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial);
1467          new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged);
1468          new_info(re, NULL, PCRE_INFO_HASCRORLF, &hascrorlf);
1469    
1470    #if !defined NOINFOCHECK
1471        old_count = pcre_info(re, &old_options, &old_first_char);        old_count = pcre_info(re, &old_options, &old_first_char);
1472        if (count < 0) fprintf(outfile,        if (count < 0) fprintf(outfile,
1473          "Error %d from pcre_info()\n", count);          "Error %d from pcre_info()\n", count);
# Line 1046  while (!done) Line 1485  while (!done)
1485            "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",            "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",
1486              get_options, old_options);              get_options, old_options);
1487          }          }
1488    #endif
1489    
1490        if (size != regex_gotten_store) fprintf(outfile,        if (size != regex_gotten_store) fprintf(outfile,
1491          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",
1492          size, regex_gotten_store);          (int)size, (int)regex_gotten_store);
1493    
1494        fprintf(outfile, "Capturing subpattern count = %d\n", count);        fprintf(outfile, "Capturing subpattern count = %d\n", count);
1495        if (backrefmax > 0)        if (backrefmax > 0)
# Line 1067  while (!done) Line 1507  while (!done)
1507            }            }
1508          }          }
1509    
1510        /* The NOPARTIAL bit is a private bit in the options, so we have        if (!okpartial) fprintf(outfile, "Partial matching not supported\n");
1511        to fish it out via out back door */        if (hascrorlf) fprintf(outfile, "Contains explicit CR or LF match\n");
1512    
1513        all_options = ((real_pcre *)re)->options;        all_options = ((real_pcre *)re)->options;
1514        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");  
1515    
1516        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1517          else fprintf(outfile, "Options:%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",
1518            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1519            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1520            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
1521            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",
1522              ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",
1523            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",
1524              ((get_options & PCRE_BSR_ANYCRLF) != 0)? " bsr_anycrlf" : "",
1525              ((get_options & PCRE_BSR_UNICODE) != 0)? " bsr_unicode" : "",
1526            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
1527            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
1528            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1529              ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1530            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1531            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "");            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1532              ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1533    
1534          if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
1535    
1536          switch (get_options & PCRE_NEWLINE_BITS)
1537            {
1538            case PCRE_NEWLINE_CR:
1539            fprintf(outfile, "Forced newline sequence: CR\n");
1540            break;
1541    
1542            case PCRE_NEWLINE_LF:
1543            fprintf(outfile, "Forced newline sequence: LF\n");
1544            break;
1545    
1546            case PCRE_NEWLINE_CRLF:
1547            fprintf(outfile, "Forced newline sequence: CRLF\n");
1548            break;
1549    
1550            case PCRE_NEWLINE_ANYCRLF:
1551            fprintf(outfile, "Forced newline sequence: ANYCRLF\n");
1552            break;
1553    
1554        if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0)          case PCRE_NEWLINE_ANY:
1555          fprintf(outfile, "Case state changes\n");          fprintf(outfile, "Forced newline sequence: ANY\n");
1556            break;
1557    
1558            default:
1559            break;
1560            }
1561    
1562        if (first_char == -1)        if (first_char == -1)
1563          {          {
1564          fprintf(outfile, "First char at start or follows \\n\n");          fprintf(outfile, "First char at start or follows newline\n");
1565          }          }
1566        else if (first_char < 0)        else if (first_char < 0)
1567          {          {
# Line 1108  while (!done) Line 1572  while (!done)
1572          int ch = first_char & 255;          int ch = first_char & 255;
1573          const char *caseless = ((first_char & REQ_CASELESS) == 0)?          const char *caseless = ((first_char & REQ_CASELESS) == 0)?
1574            "" : " (caseless)";            "" : " (caseless)";
1575          if (isprint(ch))          if (PRINTHEX(ch))
1576            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);
1577          else          else
1578            fprintf(outfile, "First char = %d%s\n", ch, caseless);            fprintf(outfile, "First char = %d%s\n", ch, caseless);
# Line 1123  while (!done) Line 1587  while (!done)
1587          int ch = need_char & 255;          int ch = need_char & 255;
1588          const char *caseless = ((need_char & REQ_CASELESS) == 0)?          const char *caseless = ((need_char & REQ_CASELESS) == 0)?
1589            "" : " (caseless)";            "" : " (caseless)";
1590          if (isprint(ch))          if (PRINTHEX(ch))
1591            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);
1592          else          else
1593            fprintf(outfile, "Need char = %d%s\n", ch, caseless);            fprintf(outfile, "Need char = %d%s\n", ch, caseless);
# Line 1159  while (!done) Line 1623  while (!done)
1623                    fprintf(outfile, "\n  ");                    fprintf(outfile, "\n  ");
1624                    c = 2;                    c = 2;
1625                    }                    }
1626                  if (isprint(i) && i != ' ')                  if (PRINTHEX(i) && i != ' ')
1627                    {                    {
1628                    fprintf(outfile, "%c ", i);                    fprintf(outfile, "%c ", i);
1629                    c += 2;                    c += 2;
# Line 1191  while (!done) Line 1655  while (!done)
1655        else        else
1656          {          {
1657          uschar sbuf[8];          uschar sbuf[8];
1658          sbuf[0] = (true_size >> 24)  & 255;          sbuf[0] = (uschar)((true_size >> 24) & 255);
1659          sbuf[1] = (true_size >> 16)  & 255;          sbuf[1] = (uschar)((true_size >> 16) & 255);
1660          sbuf[2] = (true_size >>  8)  & 255;          sbuf[2] = (uschar)((true_size >>  8) & 255);
1661          sbuf[3] = (true_size)  & 255;          sbuf[3] = (uschar)((true_size) & 255);
1662    
1663          sbuf[4] = (true_study_size >> 24)  & 255;          sbuf[4] = (uschar)((true_study_size >> 24) & 255);
1664          sbuf[5] = (true_study_size >> 16)  & 255;          sbuf[5] = (uschar)((true_study_size >> 16) & 255);
1665          sbuf[6] = (true_study_size >>  8)  & 255;          sbuf[6] = (uschar)((true_study_size >>  8) & 255);
1666          sbuf[7] = (true_study_size)  & 255;          sbuf[7] = (uschar)((true_study_size) & 255);
1667    
1668          if (fwrite(sbuf, 1, 8, f) < 8 ||          if (fwrite(sbuf, 1, 8, f) < 8 ||
1669              fwrite(re, 1, true_size, f) < true_size)              fwrite(re, 1, true_size, f) < true_size)
# Line 1218  while (!done) Line 1682  while (!done)
1682                  strerror(errno));                  strerror(errno));
1683                }                }
1684              else fprintf(outfile, "Study data written to %s\n", to_file);              else fprintf(outfile, "Study data written to %s\n", to_file);
1685    
1686              }              }
1687            }            }
1688          fclose(f);          fclose(f);
1689          }          }
1690    
1691          new_free(re);
1692          if (extra != NULL) new_free(extra);
1693          if (tables != NULL) new_free((void *)tables);
1694        continue;  /* With next regex */        continue;  /* With next regex */
1695        }        }
1696      }        /* End of non-POSIX compile */      }        /* End of non-POSIX compile */
# Line 1230  while (!done) Line 1699  while (!done)
1699    
1700    for (;;)    for (;;)
1701      {      {
1702      unsigned char *q;      uschar *q;
1703      unsigned char *bptr = dbuffer;      uschar *bptr;
1704      int *use_offsets = offsets;      int *use_offsets = offsets;
1705      int use_size_offsets = size_offsets;      int use_size_offsets = size_offsets;
1706      int callout_data = 0;      int callout_data = 0;
# Line 1244  while (!done) Line 1713  while (!done)
1713      int gmatched = 0;      int gmatched = 0;
1714      int start_offset = 0;      int start_offset = 0;
1715      int g_notempty = 0;      int g_notempty = 0;
1716        int use_dfa = 0;
1717    
1718      options = 0;      options = 0;
1719    
1720        *copynames = 0;
1721        *getnames = 0;
1722    
1723        copynamesptr = copynames;
1724        getnamesptr = getnames;
1725    
1726      pcre_callout = callout;      pcre_callout = callout;
1727      first_callout = 1;      first_callout = 1;
1728      callout_extra = 0;      callout_extra = 0;
# Line 1255  while (!done) Line 1731  while (!done)
1731      callout_fail_id = -1;      callout_fail_id = -1;
1732      show_malloc = 0;      show_malloc = 0;
1733    
1734      if (infile == stdin) printf("data> ");      if (extra != NULL) extra->flags &=
1735      if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)        ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
1736    
1737        len = 0;
1738        for (;;)
1739        {        {
1740        done = 1;        if (extend_inputline(infile, buffer + len, "data> ") == NULL)
1741        goto CONTINUE;          {
1742            if (len > 0) break;
1743            done = 1;
1744            goto CONTINUE;
1745            }
1746          if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1747          len = (int)strlen((char *)buffer);
1748          if (buffer[len-1] == '\n') break;
1749        }        }
     if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);  
1750    
     len = (int)strlen((char *)buffer);  
1751      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
1752      buffer[len] = 0;      buffer[len] = 0;
1753      if (len == 0) break;      if (len == 0) break;
# Line 1271  while (!done) Line 1755  while (!done)
1755      p = buffer;      p = buffer;
1756      while (isspace(*p)) p++;      while (isspace(*p)) p++;
1757    
1758      q = dbuffer;      bptr = q = dbuffer;
1759      while ((c = *p++) != 0)      while ((c = *p++) != 0)
1760        {        {
1761        int i = 0;        int i = 0;
# Line 1293  while (!done) Line 1777  while (!done)
1777          c -= '0';          c -= '0';
1778          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')
1779            c = c * 8 + *p++ - '0';            c = c * 8 + *p++ - '0';
1780    
1781    #if !defined NOUTF8
1782            if (use_utf8 && c > 255)
1783              {
1784              unsigned char buff8[8];
1785              int ii, utn;
1786              utn = ord2utf8(c, buff8);
1787              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1788              c = buff8[ii];   /* Last byte */
1789              }
1790    #endif
1791          break;          break;
1792    
1793          case 'x':          case 'x':
1794    
1795          /* Handle \x{..} specially - new Perl thing for utf8 */          /* Handle \x{..} specially - new Perl thing for utf8 */
1796    
1797    #if !defined NOUTF8
1798          if (*p == '{')          if (*p == '{')
1799            {            {
1800            unsigned char *pt = p;            unsigned char *pt = p;
# Line 1317  while (!done) Line 1813  while (!done)
1813              }              }
1814            /* Not correct form; fall through */            /* Not correct form; fall through */
1815            }            }
1816    #endif
1817    
1818          /* Ordinary \x */          /* Ordinary \x */
1819    
# Line 1352  while (!done) Line 1849  while (!done)
1849            }            }
1850          else if (isalnum(*p))          else if (isalnum(*p))
1851            {            {
1852            uschar name[256];            uschar *npp = copynamesptr;
           uschar *npp = name;  
1853            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1854              *npp++ = 0;
1855            *npp = 0;            *npp = 0;
1856            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)copynamesptr);
1857            if (n < 0)            if (n < 0)
1858              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", copynamesptr);
1859            else copystrings |= 1 << n;            copynamesptr = npp;
1860            }            }
1861          else if (*p == '+')          else if (*p == '+')
1862            {            {
# Line 1397  while (!done) Line 1894  while (!done)
1894            }            }
1895          continue;          continue;
1896    
1897    #if !defined NODFA
1898            case 'D':
1899    #if !defined NOPOSIX
1900            if (posix || do_posix)
1901              printf("** Can't use dfa matching in POSIX mode: \\D ignored\n");
1902            else
1903    #endif
1904              use_dfa = 1;
1905            continue;
1906    
1907            case 'F':
1908            options |= PCRE_DFA_SHORTEST;
1909            continue;
1910    #endif
1911    
1912          case 'G':          case 'G':
1913          if (isdigit(*p))          if (isdigit(*p))
1914            {            {
# Line 1405  while (!done) Line 1917  while (!done)
1917            }            }
1918          else if (isalnum(*p))          else if (isalnum(*p))
1919            {            {
1920            uschar name[256];            uschar *npp = getnamesptr;
           uschar *npp = name;  
1921            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1922              *npp++ = 0;
1923            *npp = 0;            *npp = 0;
1924            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)getnamesptr);
1925            if (n < 0)            if (n < 0)
1926              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", getnamesptr);
1927            else getstrings |= 1 << n;            getnamesptr = npp;
1928            }            }
1929          continue;          continue;
1930    
# Line 1438  while (!done) Line 1950  while (!done)
1950            if (offsets == NULL)            if (offsets == NULL)
1951              {              {
1952              printf("** Failed to get %d bytes of memory for offsets vector\n",              printf("** Failed to get %d bytes of memory for offsets vector\n",
1953                size_offsets_max * sizeof(int));                (int)(size_offsets_max * sizeof(int)));
1954              return 1;              yield = 1;
1955                goto EXIT;
1956              }              }
1957            }            }
1958          use_size_offsets = n;          use_size_offsets = n;
# Line 1450  while (!done) Line 1963  while (!done)
1963          options |= PCRE_PARTIAL;          options |= PCRE_PARTIAL;
1964          continue;          continue;
1965    
1966            case 'Q':
1967            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1968            if (extra == NULL)
1969              {
1970              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1971              extra->flags = 0;
1972              }
1973            extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
1974            extra->match_limit_recursion = n;
1975            continue;
1976    
1977            case 'q':
1978            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1979            if (extra == NULL)
1980              {
1981              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1982              extra->flags = 0;
1983              }
1984            extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
1985            extra->match_limit = n;
1986            continue;
1987    
1988    #if !defined NODFA
1989            case 'R':
1990            options |= PCRE_DFA_RESTART;
1991            continue;
1992    #endif
1993    
1994          case 'S':          case 'S':
1995          show_malloc = 1;          show_malloc = 1;
1996          continue;          continue;
# Line 1461  while (!done) Line 2002  while (!done)
2002          case '?':          case '?':
2003          options |= PCRE_NO_UTF8_CHECK;          options |= PCRE_NO_UTF8_CHECK;
2004          continue;          continue;
2005    
2006            case '<':
2007              {
2008              int x = check_newline(p, outfile);
2009              if (x == 0) goto NEXT_DATA;
2010              options |= x;
2011              while (*p++ != '>');
2012              }
2013            continue;
2014          }          }
2015        *q++ = c;        *q++ = c;
2016        }        }
2017      *q = 0;      *q = 0;
2018      len = q - dbuffer;      len = q - dbuffer;
2019    
2020        if ((all_use_dfa || use_dfa) && find_match_limit)
2021          {
2022          printf("**Match limit not relevant for DFA matching: ignored\n");
2023          find_match_limit = 0;
2024          }
2025    
2026      /* Handle matching via the POSIX interface, which does not      /* Handle matching via the POSIX interface, which does not
2027      support timing or playing with the match limit or callout data. */      support timing or playing with the match limit or callout data. */
2028    
# Line 1485  while (!done) Line 2041  while (!done)
2041    
2042        if (rc != 0)        if (rc != 0)
2043          {          {
2044          (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);          (void)regerror(rc, &preg, (char *)buffer, buffer_size);
2045          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
2046          }          }
2047          else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
2048                  != 0)
2049            {
2050            fprintf(outfile, "Matched with REG_NOSUB\n");
2051            }
2052        else        else
2053          {          {
2054          size_t i;          size_t i;
# Line 1519  while (!done) Line 2080  while (!done)
2080    
2081      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
2082        {        {
2083        if (timeit)        if (timeitm > 0)
2084          {          {
2085          register int i;          register int i;
2086          clock_t time_taken;          clock_t time_taken;
2087          clock_t start_time = clock();          clock_t start_time = clock();
2088          for (i = 0; i < LOOPREPEAT; i++)  
2089    #if !defined NODFA
2090            if (all_use_dfa || use_dfa)
2091              {
2092              int workspace[1000];
2093              for (i = 0; i < timeitm; i++)
2094                count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
2095                  options | g_notempty, use_offsets, use_size_offsets, workspace,
2096                  sizeof(workspace)/sizeof(int));
2097              }
2098            else
2099    #endif
2100    
2101            for (i = 0; i < timeitm; i++)
2102            count = pcre_exec(re, extra, (char *)bptr, len,            count = pcre_exec(re, extra, (char *)bptr, len,
2103              start_offset, options | g_notempty, use_offsets, use_size_offsets);              start_offset, options | g_notempty, use_offsets, use_size_offsets);
2104    
2105          time_taken = clock() - start_time;          time_taken = clock() - start_time;
2106          fprintf(outfile, "Execute time %.3f milliseconds\n",          fprintf(outfile, "Execute time %.4f milliseconds\n",
2107            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeitm) /
2108              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
2109          }          }
2110    
2111        /* 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
2112        varying limits in order to find the minimum value. */        varying limits in order to find the minimum value for the match limit and
2113          for the recursion limit. */
2114    
2115        if (find_match_limit)        if (find_match_limit)
2116          {          {
         int min = 0;  
         int mid = 64;  
         int max = -1;  
   
2117          if (extra == NULL)          if (extra == NULL)
2118            {            {
2119            extra = (pcre_extra *)malloc(sizeof(pcre_extra));            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2120            extra->flags = 0;            extra->flags = 0;
2121            }            }
         extra->flags |= PCRE_EXTRA_MATCH_LIMIT;  
2122    
2123          for (;;)          (void)check_match_limit(re, extra, bptr, len, start_offset,
2124            {            options|g_notempty, use_offsets, use_size_offsets,
2125            extra->match_limit = mid;            PCRE_EXTRA_MATCH_LIMIT, &(extra->match_limit),
2126            count = pcre_exec(re, extra, (char *)bptr, len, start_offset,            PCRE_ERROR_MATCHLIMIT, "match()");
2127              options | g_notempty, use_offsets, use_size_offsets);  
2128            if (count == PCRE_ERROR_MATCHLIMIT)          count = check_match_limit(re, extra, bptr, len, start_offset,
2129              {            options|g_notempty, use_offsets, use_size_offsets,
2130              /* fprintf(outfile, "Testing match limit = %d\n", mid); */            PCRE_EXTRA_MATCH_LIMIT_RECURSION, &(extra->match_limit_recursion),
2131              min = mid;            PCRE_ERROR_RECURSIONLIMIT, "match() recursion");
             mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;  
             }  
           else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||  
                                  count == PCRE_ERROR_PARTIAL)  
             {  
             if (mid == min + 1)  
               {  
               fprintf(outfile, "Minimum match limit = %d\n", mid);  
               break;  
               }  
             /* fprintf(outfile, "Testing match limit = %d\n", mid); */  
             max = mid;  
             mid = (min + mid)/2;  
             }  
           else break;    /* Some other error */  
           }  
   
         extra->flags &= ~PCRE_EXTRA_MATCH_LIMIT;  
2132          }          }
2133    
2134        /* If callout_data is set, use the interface with additional data */        /* If callout_data is set, use the interface with additional data */
# Line 1597  while (!done) Line 2150  while (!done)
2150        /* 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
2151        value of match_limit. */        value of match_limit. */
2152    
2153        else  #if !defined NODFA
2154          else if (all_use_dfa || use_dfa)
2155          {          {
2156          count = pcre_exec(re, extra, (char *)bptr, len,          int workspace[1000];
2157            start_offset, options | g_notempty, use_offsets, use_size_offsets);          count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
2158              options | g_notempty, use_offsets, use_size_offsets, workspace,
2159              sizeof(workspace)/sizeof(int));
2160            if (count == 0)
2161              {
2162              fprintf(outfile, "Matched, but too many subsidiary matches\n");
2163              count = use_size_offsets/2;
2164              }
2165          }          }
2166    #endif
2167    
2168        if (count == 0)        else
2169          {          {
2170          fprintf(outfile, "Matched, but too many substrings\n");          count = pcre_exec(re, extra, (char *)bptr, len,
2171          count = use_size_offsets/3;            start_offset, options | g_notempty, use_offsets, use_size_offsets);
2172            if (count == 0)
2173              {
2174              fprintf(outfile, "Matched, but too many substrings\n");
2175              count = use_size_offsets/3;
2176              }
2177          }          }
2178    
2179        /* Matched */        /* Matched */
2180    
2181        if (count >= 0)        if (count >= 0)
2182          {          {
2183          int i;          int i, maxcount;
2184    
2185    #if !defined NODFA
2186            if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else
2187    #endif
2188              maxcount = use_size_offsets/3;
2189    
2190            /* This is a check against a lunatic return value. */
2191    
2192            if (count > maxcount)
2193              {
2194              fprintf(outfile,
2195                "** PCRE error: returned count %d is too big for offset size %d\n",
2196                count, use_size_offsets);
2197              count = use_size_offsets/3;
2198              if (do_g || do_G)
2199                {
2200                fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G');
2201                do_g = do_G = FALSE;        /* Break g/G loop */
2202                }
2203              }
2204    
2205          for (i = 0; i < count * 2; i += 2)          for (i = 0; i < count * 2; i += 2)
2206            {            {
2207            if (use_offsets[i] < 0)            if (use_offsets[i] < 0)
# Line 1641  while (!done) Line 2229  while (!done)
2229            {            {
2230            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
2231              {              {
2232              char copybuffer[16];              char copybuffer[256];
2233              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,
2234                i, copybuffer, sizeof(copybuffer));                i, copybuffer, sizeof(copybuffer));
2235              if (rc < 0)              if (rc < 0)
# Line 1651  while (!done) Line 2239  while (!done)
2239              }              }
2240            }            }
2241    
2242            for (copynamesptr = copynames;
2243                 *copynamesptr != 0;
2244                 copynamesptr += (int)strlen((char*)copynamesptr) + 1)
2245              {
2246              char copybuffer[256];
2247              int rc = pcre_copy_named_substring(re, (char *)bptr, use_offsets,
2248                count, (char *)copynamesptr, copybuffer, sizeof(copybuffer));
2249              if (rc < 0)
2250                fprintf(outfile, "copy substring %s failed %d\n", copynamesptr, rc);
2251              else
2252                fprintf(outfile, "  C %s (%d) %s\n", copybuffer, rc, copynamesptr);
2253              }
2254    
2255          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2256            {            {
2257            if ((getstrings & (1 << i)) != 0)            if ((getstrings & (1 << i)) != 0)
# Line 1663  while (!done) Line 2264  while (!done)
2264              else              else
2265                {                {
2266                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
               /* free((void *)substring); */  
2267                pcre_free_substring(substring);                pcre_free_substring(substring);
2268                }                }
2269              }              }
2270            }            }
2271    
2272            for (getnamesptr = getnames;
2273                 *getnamesptr != 0;
2274                 getnamesptr += (int)strlen((char*)getnamesptr) + 1)
2275              {
2276              const char *substring;
2277              int rc = pcre_get_named_substring(re, (char *)bptr, use_offsets,
2278                count, (char *)getnamesptr, &substring);
2279              if (rc < 0)
2280                fprintf(outfile, "copy substring %s failed %d\n", getnamesptr, rc);
2281              else
2282                {
2283                fprintf(outfile, "  G %s (%d) %s\n", substring, rc, getnamesptr);
2284                pcre_free_substring(substring);
2285                }
2286              }
2287    
2288          if (getlist)          if (getlist)
2289            {            {
2290            const char **stringlist;            const char **stringlist;
# Line 1692  while (!done) Line 2308  while (!done)
2308    
2309        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2310          {          {
2311          fprintf(outfile, "Partial match\n");          fprintf(outfile, "Partial match");
2312    #if !defined NODFA
2313            if ((all_use_dfa || use_dfa) && use_size_offsets > 2)
2314              fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],
2315                bptr + use_offsets[0]);
2316    #endif
2317            fprintf(outfile, "\n");
2318          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2319          }          }
2320    
2321        /* 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
2322        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
2323        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
2324        matching, the advance must be one character, not one byte. Fudge the        string - that was checked before setting g_notempty.
2325        offset values to achieve this. We won't be at the end of the string -  
2326        that was checked before setting g_notempty. */        Complication arises in the case when the newline option is "any" or
2327          "anycrlf". If the previous match was at the end of a line terminated by
2328          CRLF, an advance of one character just passes the \r, whereas we should
2329          prefer the longer newline sequence, as does the code in pcre_exec().
2330          Fudge the offset value to achieve this.
2331    
2332          Otherwise, in the case of UTF-8 matching, the advance must be one
2333          character, not one byte. */
2334    
2335        else        else
2336          {          {
2337          if (g_notempty != 0)          if (g_notempty != 0)
2338            {            {
2339            int onechar = 1;            int onechar = 1;
2340              unsigned int obits = ((real_pcre *)re)->options;
2341            use_offsets[0] = start_offset;            use_offsets[0] = start_offset;
2342            if (use_utf8)            if ((obits & PCRE_NEWLINE_BITS) == 0)
2343                {
2344                int d;
2345                (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);
2346                obits = (d == '\r')? PCRE_NEWLINE_CR :
2347                        (d == '\n')? PCRE_NEWLINE_LF :
2348                        (d == ('\r'<<8 | '\n'))? PCRE_NEWLINE_CRLF :
2349                        (d == -2)? PCRE_NEWLINE_ANYCRLF :
2350                        (d == -1)? PCRE_NEWLINE_ANY : 0;
2351                }
2352              if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2353                   (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2354                  &&
2355                  start_offset < len - 1 &&
2356                  bptr[start_offset] == '\r' &&
2357                  bptr[start_offset+1] == '\n')
2358                onechar++;
2359              else if (use_utf8)
2360              {              {
2361              while (start_offset + onechar < len)              while (start_offset + onechar < len)
2362                {                {
# Line 1744  while (!done) Line 2391  while (!done)
2391        character. */        character. */
2392    
2393        g_notempty = 0;        g_notempty = 0;
2394    
2395        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2396          {          {
2397          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
# Line 1762  while (!done) Line 2410  while (!done)
2410          len -= use_offsets[1];          len -= use_offsets[1];
2411          }          }
2412        }  /* End of loop for /g and /G */        }  /* End of loop for /g and /G */
2413    
2414        NEXT_DATA: continue;
2415      }    /* End of loop for data lines */      }    /* End of loop for data lines */
2416    
2417    CONTINUE:    CONTINUE:
# Line 1770  while (!done) Line 2420  while (!done)
2420    if (posix || do_posix) regfree(&preg);    if (posix || do_posix) regfree(&preg);
2421  #endif  #endif
2422    
2423    if (re != NULL) free(re);    if (re != NULL) new_free(re);
2424    if (extra != NULL) free(extra);    if (extra != NULL) new_free(extra);
2425    if (tables != NULL)    if (tables != NULL)
2426      {      {
2427      free((void *)tables);      new_free((void *)tables);
2428      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");
2429        locale_set = 0;
2430      }      }
2431    }    }
2432    
2433  if (infile == stdin) fprintf(outfile, "\n");  if (infile == stdin) fprintf(outfile, "\n");
2434  return 0;  
2435    EXIT:
2436    
2437    if (infile != NULL && infile != stdin) fclose(infile);
2438    if (outfile != NULL && outfile != stdout) fclose(outfile);
2439    
2440    free(buffer);
2441    free(dbuffer);
2442    free(pbuffer);
2443    free(offsets);
2444    
2445    return yield;
2446  }  }
2447    
2448  /* End */  /* End of pcretest.c */

Legend:
Removed from v.75  
changed lines
  Added in v.343

  ViewVC Help
Powered by ViewVC 1.1.5