/[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 75 by nigel, Sat Feb 24 21:40:37 2007 UTC revision 151 by ph10, Tue Apr 17 15:07:29 2007 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 44  POSSIBILITY OF SUCH DAMAGE. Line 44  POSSIBILITY OF SUCH DAMAGE.
44  #include <locale.h>  #include <locale.h>
45  #include <errno.h>  #include <errno.h>
46    
 /* We need the internal info for displaying the results of pcre_study(). Also  
 for getting the opcodes for showing compiled code. */  
47    
48  #define PCRE_SPY        /* For Win32 build, import data, not export */  /* A number of things vary for Windows builds. Originally, pcretest opened its
49  #include "internal.h"  input and output without "b"; then I was told that "b" was needed in some
50    environments, so it was added for release 5.0 to both the input and output. (It
51    makes no difference on Unix-like systems.) Later I was told that it is wrong
52    for the input on Windows. I've now abstracted the modes into two macros that
53    are set here, to make it easier to fiddle with them, and removed "b" from the
54    input mode under Windows. */
55    
56    #if defined(_WIN32) || defined(WIN32)
57    #include <io.h>                /* For _setmode() */
58    #include <fcntl.h>             /* For _O_BINARY */
59    #define INPUT_MODE   "r"
60    #define OUTPUT_MODE  "wb"
61    
62    #else
63    #include <sys/time.h>          /* These two includes are needed */
64    #include <sys/resource.h>      /* for setrlimit(). */
65    #define INPUT_MODE   "rb"
66    #define OUTPUT_MODE  "wb"
67    #endif
68    
69    
70    /* We have to include pcre_internal.h because we need the internal info for
71    displaying the results of pcre_study() and we also need to know about the
72    internal macros, structures, and other internal data values; pcretest has
73    "inside information" compared to a program that strictly follows the PCRE API.
74    
75    Although pcre_internal.h does itself include pcre.h, we explicitly include it
76    here before pcre_internal.h so that the PCRE_EXP_xxx macros get set
77    appropriately for an application, not for building PCRE. */
78    
79    #include "pcre.h"
80    #include "pcre_internal.h"
81    
82    /* We need access to the data tables that PCRE uses. So as not to have to keep
83    two copies, we include the source file here, changing the names of the external
84    symbols to prevent clashes. */
85    
86    #define _pcre_utf8_table1      utf8_table1
87    #define _pcre_utf8_table1_size utf8_table1_size
88    #define _pcre_utf8_table2      utf8_table2
89    #define _pcre_utf8_table3      utf8_table3
90    #define _pcre_utf8_table4      utf8_table4
91    #define _pcre_utt              utt
92    #define _pcre_utt_size         utt_size
93    #define _pcre_OP_lengths       OP_lengths
94    
95    #include "pcre_tables.c"
96    
97    /* We also need the pcre_printint() function for printing out compiled
98    patterns. This function is in a separate file so that it can be included in
99    pcre_compile.c when that module is compiled with debugging enabled.
100    
101    The definition of the macro PRINTABLE, which determines whether to print an
102    output character as-is or as a hex value when showing compiled patterns, is
103    contained in this file. We uses it here also, in cases when the locale has not
104    been explicitly changed, so as to get consistent output from systems that
105    differ in their output from isprint() even in the "C" locale. */
106    
107    #include "pcre_printint.src"
108    
109    #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
110    
111    
112  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
113  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 117  Makefile. */
117  #include "pcreposix.h"  #include "pcreposix.h"
118  #endif  #endif
119    
120    /* It is also possible, for the benefit of the version currently imported into
121    Exim, to build pcretest without support for UTF8 (define NOUTF8), without the
122    interface to the DFA matcher (NODFA), and without the doublecheck of the old
123    "info" function (define NOINFOCHECK). In fact, we automatically cut out the
124    UTF8 support if PCRE is built without it. */
125    
126    #ifndef SUPPORT_UTF8
127    #ifndef NOUTF8
128    #define NOUTF8
129    #endif
130    #endif
131    
132    
133    /* Other parameters */
134    
135  #ifndef CLOCKS_PER_SEC  #ifndef CLOCKS_PER_SEC
136  #ifdef CLK_TCK  #ifdef CLK_TCK
137  #define CLOCKS_PER_SEC CLK_TCK  #define CLOCKS_PER_SEC CLK_TCK
# Line 66  Makefile. */ Line 140  Makefile. */
140  #endif  #endif
141  #endif  #endif
142    
143  #define LOOPREPEAT 500000  /* This is the default loop count for timing. */
144    
145  #define BUFFER_SIZE 30000  #define LOOPREPEAT 500000
 #define PBUFFER_SIZE BUFFER_SIZE  
 #define DBUFFER_SIZE BUFFER_SIZE  
146    
147    /* Static variables */
148    
149  static FILE *outfile;  static FILE *outfile;
150  static int log_store = 0;  static int log_store = 0;
# Line 80  static int callout_extra; Line 153  static int callout_extra;
153  static int callout_fail_count;  static int callout_fail_count;
154  static int callout_fail_id;  static int callout_fail_id;
155  static int first_callout;  static int first_callout;
156    static int locale_set = 0;
157  static int show_malloc;  static int show_malloc;
158  static int use_utf8;  static int use_utf8;
159  static size_t gotten_store;  static size_t gotten_store;
160    
161    /* The buffers grow automatically if very long input lines are encountered. */
162    
163    static int buffer_size = 50000;
164    static uschar *buffer = NULL;
165    static uschar *dbuffer = NULL;
166  static uschar *pbuffer = NULL;  static uschar *pbuffer = NULL;
167    
168    
 static const int utf8_table1[] = {  
   0x0000007f, 0x000007ff, 0x0000ffff, 0x001fffff, 0x03ffffff, 0x7fffffff};  
169    
170  static const int utf8_table2[] = {  /*************************************************
171    0, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc};  *        Read or extend an input line            *
172    *************************************************/
173    
174    /* Input lines are read into buffer, but both patterns and data lines can be
175    continued over multiple input lines. In addition, if the buffer fills up, we
176    want to automatically expand it so as to be able to handle extremely large
177    lines that are needed for certain stress tests. When the input buffer is
178    expanded, the other two buffers must also be expanded likewise, and the
179    contents of pbuffer, which are a copy of the input for callouts, must be
180    preserved (for when expansion happens for a data line). This is not the most
181    optimal way of handling this, but hey, this is just a test program!
182    
183    Arguments:
184      f            the file to read
185      start        where in buffer to start (this *must* be within buffer)
186    
187    Returns:       pointer to the start of new data
188                   could be a copy of start, or could be moved
189                   NULL if no data read and EOF reached
190    */
191    
192  static const int utf8_table3[] = {  static uschar *
193    0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01};  extend_inputline(FILE *f, uschar *start)
194    {
195    uschar *here = start;
196    
197    for (;;)
198      {
199      int rlen = buffer_size - (here - buffer);
200    
201      if (rlen > 1000)
202        {
203        int dlen;
204        if (fgets((char *)here, rlen,  f) == NULL)
205          return (here == start)? NULL : start;
206        dlen = (int)strlen((char *)here);
207        if (dlen > 0 && here[dlen - 1] == '\n') return start;
208        here += dlen;
209        }
210    
211      else
212        {
213        int new_buffer_size = 2*buffer_size;
214        uschar *new_buffer = (unsigned char *)malloc(new_buffer_size);
215        uschar *new_dbuffer = (unsigned char *)malloc(new_buffer_size);
216        uschar *new_pbuffer = (unsigned char *)malloc(new_buffer_size);
217    
218        if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
219          {
220          fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
221          exit(1);
222          }
223    
224        memcpy(new_buffer, buffer, buffer_size);
225        memcpy(new_pbuffer, pbuffer, buffer_size);
226    
227        buffer_size = new_buffer_size;
228    
229        start = new_buffer + (start - buffer);
230        here = new_buffer + (here - buffer);
231    
232        free(buffer);
233        free(dbuffer);
234        free(pbuffer);
235    
236        buffer = new_buffer;
237        dbuffer = new_dbuffer;
238        pbuffer = new_pbuffer;
239        }
240      }
241    
242    return NULL;  /* Control never gets here */
243    }
244    
245    
246    
 /*************************************************  
 *         Print compiled regex                   *  
 *************************************************/  
247    
 /* 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"  
248    
249    
250    
# Line 122  static uschar OP_lengths[] = { OP_LENGTH Line 254  static uschar OP_lengths[] = { OP_LENGTH
254    
255  /* 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
256  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
257  for unpicking the -o argument, so just keep it simple.  for unpicking arguments, so just keep it simple.
258    
259  Arguments:  Arguments:
260    str           string to be converted    str           string to be converted
# Line 143  return(result); Line 275  return(result);
275    
276    
277    
 /*************************************************  
 *       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;  
 }  
   
278    
279  /*************************************************  /*************************************************
280  *            Convert UTF-8 string to value       *  *            Convert UTF-8 string to value       *
# Line 188  return i + 1; Line 284  return i + 1;
284  and returns the value of the character.  and returns the value of the character.
285    
286  Argument:  Argument:
287    buffer   a pointer to the byte vector    utf8bytes   a pointer to the byte vector
288    vptr     a pointer to an int to receive the value    vptr        a pointer to an int to receive the value
289    
290  Returns:   >  0 => the number of bytes consumed  Returns:      >  0 => the number of bytes consumed
291             -6 to 0 => malformed UTF-8 character at offset = (-return)                -6 to 0 => malformed UTF-8 character at offset = (-return)
292  */  */
293    
294    #if !defined NOUTF8
295    
296  static int  static int
297  utf82ord(unsigned char *buffer, int *vptr)  utf82ord(unsigned char *utf8bytes, int *vptr)
298  {  {
299  int c = *buffer++;  int c = *utf8bytes++;
300  int d = c;  int d = c;
301  int i, j, s;  int i, j, s;
302    
# Line 218  d = (c & utf8_table3[i]) << s; Line 316  d = (c & utf8_table3[i]) << s;
316    
317  for (j = 0; j < i; j++)  for (j = 0; j < i; j++)
318    {    {
319    c = *buffer++;    c = *utf8bytes++;
320    if ((c & 0xc0) != 0x80) return -(j+1);    if ((c & 0xc0) != 0x80) return -(j+1);
321    s -= 6;    s -= 6;
322    d |= (c & 0x3f) << s;    d |= (c & 0x3f) << s;
# Line 226  for (j = 0; j < i; j++) Line 324  for (j = 0; j < i; j++)
324    
325  /* Check that encoding was the correct unique one */  /* Check that encoding was the correct unique one */
326    
327  for (j = 0; j < sizeof(utf8_table1)/sizeof(int); j++)  for (j = 0; j < utf8_table1_size; j++)
328    if (d <= utf8_table1[j]) break;    if (d <= utf8_table1[j]) break;
329  if (j != i) return -(i+1);  if (j != i) return -(i+1);
330    
# Line 236  if (j != i) return -(i+1); Line 334  if (j != i) return -(i+1);
334  return i+1;  return i+1;
335  }  }
336    
337    #endif
338    
339    
340    
341    /*************************************************
342    *       Convert character value to UTF-8         *
343    *************************************************/
344    
345    /* This function takes an integer value in the range 0 - 0x7fffffff
346    and encodes it as a UTF-8 character in 0 to 6 bytes.
347    
348    Arguments:
349      cvalue     the character value
350      utf8bytes  pointer to buffer for result - at least 6 bytes long
351    
352    Returns:     number of characters placed in the buffer
353    */
354    
355    #if !defined NOUTF8
356    
357    static int
358    ord2utf8(int cvalue, uschar *utf8bytes)
359    {
360    register int i, j;
361    for (i = 0; i < utf8_table1_size; i++)
362      if (cvalue <= utf8_table1[i]) break;
363    utf8bytes += i;
364    for (j = i; j > 0; j--)
365     {
366     *utf8bytes-- = 0x80 | (cvalue & 0x3f);
367     cvalue >>= 6;
368     }
369    *utf8bytes = utf8_table2[i] | cvalue;
370    return i + 1;
371    }
372    
373    #endif
374    
375    
376    
377  /*************************************************  /*************************************************
# Line 248  chars without printing. */ Line 384  chars without printing. */
384    
385  static int pchars(unsigned char *p, int length, FILE *f)  static int pchars(unsigned char *p, int length, FILE *f)
386  {  {
387  int c;  int c = 0;
388  int yield = 0;  int yield = 0;
389    
390  while (length-- > 0)  while (length-- > 0)
391    {    {
392    #if !defined NOUTF8
393    if (use_utf8)    if (use_utf8)
394      {      {
395      int rc = utf82ord(p, &c);      int rc = utf82ord(p, &c);
# Line 261  while (length-- > 0) Line 398  while (length-- > 0)
398        {        {
399        length -= rc - 1;        length -= rc - 1;
400        p += rc;        p += rc;
401        if (c < 256 && isprint(c))        if (PRINTHEX(c))
402          {          {
403          if (f != NULL) fprintf(f, "%c", c);          if (f != NULL) fprintf(f, "%c", c);
404          yield++;          yield++;
405          }          }
406        else        else
407          {          {
408          int n;          int n = 4;
409          if (f != NULL) fprintf(f, "\\x{%02x}%n", c, &n);          if (f != NULL) fprintf(f, "\\x{%02x}", c);
410          yield += n;          yield += (n <= 0x000000ff)? 2 :
411                     (n <= 0x00000fff)? 3 :
412                     (n <= 0x0000ffff)? 4 :
413                     (n <= 0x000fffff)? 5 : 6;
414          }          }
415        continue;        continue;
416        }        }
417      }      }
418    #endif
419    
420     /* Not UTF-8, or malformed UTF-8  */     /* Not UTF-8, or malformed UTF-8  */
421    
422    if (isprint(c = *(p++)))    c = *p++;
423      if (PRINTHEX(c))
424      {      {
425      if (f != NULL) fprintf(f, "%c", c);      if (f != NULL) fprintf(f, "%c", c);
426      yield++;      yield++;
# Line 403  static void *new_malloc(size_t size) Line 545  static void *new_malloc(size_t size)
545  void *block = malloc(size);  void *block = malloc(size);
546  gotten_store = size;  gotten_store = size;
547  if (show_malloc)  if (show_malloc)
548    fprintf(outfile, "malloc       %3d %p\n", size, block);    fprintf(outfile, "malloc       %3d %p\n", (int)size, block);
549  return block;  return block;
550  }  }
551    
# Line 421  static void *stack_malloc(size_t size) Line 563  static void *stack_malloc(size_t size)
563  {  {
564  void *block = malloc(size);  void *block = malloc(size);
565  if (show_malloc)  if (show_malloc)
566    fprintf(outfile, "stack_malloc %3d %p\n", size, block);    fprintf(outfile, "stack_malloc %3d %p\n", (int)size, block);
567  return block;  return block;
568  }  }
569    
# Line 452  if ((rc = pcre_fullinfo(re, study, optio Line 594  if ((rc = pcre_fullinfo(re, study, optio
594  *         Byte flipping function                 *  *         Byte flipping function                 *
595  *************************************************/  *************************************************/
596    
597  static long int  static unsigned long int
598  byteflip(long int value, int n)  byteflip(unsigned long int value, int n)
599  {  {
600  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);
601  return ((value & 0x000000ff) << 24) |  return ((value & 0x000000ff) << 24) |
# Line 466  return ((value & 0x000000ff) << 24) | Line 608  return ((value & 0x000000ff) << 24) |
608    
609    
610  /*************************************************  /*************************************************
611    *        Check match or recursion limit          *
612    *************************************************/
613    
614    static int
615    check_match_limit(pcre *re, pcre_extra *extra, uschar *bptr, int len,
616      int start_offset, int options, int *use_offsets, int use_size_offsets,
617      int flag, unsigned long int *limit, int errnumber, const char *msg)
618    {
619    int count;
620    int min = 0;
621    int mid = 64;
622    int max = -1;
623    
624    extra->flags |= flag;
625    
626    for (;;)
627      {
628      *limit = mid;
629    
630      count = pcre_exec(re, extra, (char *)bptr, len, start_offset, options,
631        use_offsets, use_size_offsets);
632    
633      if (count == errnumber)
634        {
635        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
636        min = mid;
637        mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;
638        }
639    
640      else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||
641                             count == PCRE_ERROR_PARTIAL)
642        {
643        if (mid == min + 1)
644          {
645          fprintf(outfile, "Minimum %s limit = %d\n", msg, mid);
646          break;
647          }
648        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
649        max = mid;
650        mid = (min + mid)/2;
651        }
652      else break;    /* Some other error */
653      }
654    
655    extra->flags &= ~flag;
656    return count;
657    }
658    
659    
660    
661    /*************************************************
662    *         Check newline indicator                *
663    *************************************************/
664    
665    /* This is used both at compile and run-time to check for <xxx> escapes, where
666    xxx is LF, CR, CRLF, ANYCRLF, or ANY. Print a message and return 0 if there is
667    no match.
668    
669    Arguments:
670      p           points after the leading '<'
671      f           file for error message
672    
673    Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
674    */
675    
676    static int
677    check_newline(uschar *p, FILE *f)
678    {
679    if (strncmp((char *)p, "cr>", 3) == 0) return PCRE_NEWLINE_CR;
680    if (strncmp((char *)p, "lf>", 3) == 0) return PCRE_NEWLINE_LF;
681    if (strncmp((char *)p, "crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
682    if (strncmp((char *)p, "anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF;
683    if (strncmp((char *)p, "any>", 4) == 0) return PCRE_NEWLINE_ANY;
684    fprintf(f, "Unknown newline type at: <%s\n", p);
685    return 0;
686    }
687    
688    
689    
690    /*************************************************
691    *             Usage function                     *
692    *************************************************/
693    
694    static void
695    usage(void)
696    {
697    printf("Usage:     pcretest [options] [<input> [<output>]]\n");
698    printf("  -b       show compiled code (bytecode)\n");
699    printf("  -C       show PCRE compile-time options and exit\n");
700    printf("  -d       debug: show compiled code and information (-b and -i)\n");
701    #if !defined NODFA
702    printf("  -dfa     force DFA matching for all subjects\n");
703    #endif
704    printf("  -help    show usage information\n");
705    printf("  -i       show information about compiled patterns\n"
706           "  -m       output memory used information\n"
707           "  -o <n>   set size of offsets vector to <n>\n");
708    #if !defined NOPOSIX
709    printf("  -p       use POSIX interface\n");
710    #endif
711    printf("  -q       quiet: do not output PCRE version number at start\n");
712    printf("  -S <n>   set stack size to <n> megabytes\n");
713    printf("  -s       output store (memory) used information\n"
714           "  -t       time compilation and execution\n");
715    printf("  -t <n>   time compilation and execution, repeating <n> times\n");
716    printf("  -tm      time execution (matching) only\n");
717    printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
718    }
719    
720    
721    
722    /*************************************************
723  *                Main Program                    *  *                Main Program                    *
724  *************************************************/  *************************************************/
725    
# Line 480  int options = 0; Line 734  int options = 0;
734  int study_options = 0;  int study_options = 0;
735  int op = 1;  int op = 1;
736  int timeit = 0;  int timeit = 0;
737    int timeitm = 0;
738  int showinfo = 0;  int showinfo = 0;
739  int showstore = 0;  int showstore = 0;
740    int quiet = 0;
741  int size_offsets = 45;  int size_offsets = 45;
742  int size_offsets_max;  int size_offsets_max;
743  int *offsets;  int *offsets = NULL;
744  #if !defined NOPOSIX  #if !defined NOPOSIX
745  int posix = 0;  int posix = 0;
746  #endif  #endif
747  int debug = 0;  int debug = 0;
748  int done = 0;  int done = 0;
749    int all_use_dfa = 0;
750    int yield = 0;
751    int stack_size;
752    
753    /* These vectors store, end-to-end, a list of captured substring names. Assume
754    that 1024 is plenty long enough for the few names we'll be testing. */
755    
756    uschar copynames[1024];
757    uschar getnames[1024];
758    
759  unsigned char *buffer;  uschar *copynamesptr;
760  unsigned char *dbuffer;  uschar *getnamesptr;
761    
762  /* Get buffers from malloc() so that Electric Fence will check their misuse  /* Get buffers from malloc() so that Electric Fence will check their misuse
763  when I am debugging. */  when I am debugging. They grow automatically when very long lines are read. */
764    
765  buffer = (unsigned char *)malloc(BUFFER_SIZE);  buffer = (unsigned char *)malloc(buffer_size);
766  dbuffer = (unsigned char *)malloc(DBUFFER_SIZE);  dbuffer = (unsigned char *)malloc(buffer_size);
767  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.  */  
768    
769  #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) */  
770    
771  outfile = stdout;  outfile = stdout;
772    
773    /* The following  _setmode() stuff is some Windows magic that tells its runtime
774    library to translate CRLF into a single LF character. At least, that's what
775    I've been told: never having used Windows I take this all on trust. Originally
776    it set 0x8000, but then I was advised that _O_BINARY was better. */
777    
778    #if defined(_WIN32) || defined(WIN32)
779    _setmode( _fileno( stdout ), _O_BINARY );
780    #endif
781    
782  /* Scan options */  /* Scan options */
783    
784  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
# Line 519  while (argc > 1 && argv[op][0] == '-') Line 787  while (argc > 1 && argv[op][0] == '-')
787    
788    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
789      showstore = 1;      showstore = 1;
790    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-q") == 0) quiet = 1;
791      else if (strcmp(argv[op], "-b") == 0) debug = 1;
792    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
793    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
794    #if !defined NODFA
795      else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
796    #endif
797    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&
798        ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),        ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),
799          *endptr == 0))          *endptr == 0))
# Line 529  while (argc > 1 && argv[op][0] == '-') Line 801  while (argc > 1 && argv[op][0] == '-')
801      op++;      op++;
802      argc--;      argc--;
803      }      }
804      else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
805        {
806        int both = argv[op][2] == 0;
807        int temp;
808        if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr),
809                         *endptr == 0))
810          {
811          timeitm = temp;
812          op++;
813          argc--;
814          }
815        else timeitm = LOOPREPEAT;
816        if (both) timeit = timeitm;
817        }
818      else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
819          ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
820            *endptr == 0))
821        {
822    #if defined(_WIN32) || defined(WIN32)
823        printf("PCRE: -S not supported on this OS\n");
824        exit(1);
825    #else
826        int rc;
827        struct rlimit rlim;
828        getrlimit(RLIMIT_STACK, &rlim);
829        rlim.rlim_cur = stack_size * 1024 * 1024;
830        rc = setrlimit(RLIMIT_STACK, &rlim);
831        if (rc != 0)
832          {
833        printf("PCRE: setrlimit() failed with error %d\n", rc);
834        exit(1);
835          }
836        op++;
837        argc--;
838    #endif
839        }
840  #if !defined NOPOSIX  #if !defined NOPOSIX
841    else if (strcmp(argv[op], "-p") == 0) posix = 1;    else if (strcmp(argv[op], "-p") == 0) posix = 1;
842  #endif  #endif
# Line 542  while (argc > 1 && argv[op][0] == '-') Line 850  while (argc > 1 && argv[op][0] == '-')
850      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
851      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
852      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
853      printf("  Newline character is %s\n", (rc == '\r')? "CR" : "LF");      printf("  Newline sequence is %s\n", (rc == '\r')? "CR" :
854          (rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" :
855          (rc == -2)? "ANYCRLF" :
856          (rc == -1)? "ANY" : "???");
857      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);
858      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
859      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
860      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
861      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);
862      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %d\n", rc);
863        (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &rc);
864        printf("  Default recursion depth limit = %d\n", rc);
865      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
866      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
867      exit(0);      goto EXIT;
868        }
869      else if (strcmp(argv[op], "-help") == 0 ||
870               strcmp(argv[op], "--help") == 0)
871        {
872        usage();
873        goto EXIT;
874      }      }
875    else    else
876      {      {
877      printf("** Unknown or malformed option %s\n", argv[op]);      printf("** Unknown or malformed option %s\n", argv[op]);
878      printf("Usage:   pcretest [-d] [-i] [-o <n>] [-p] [-s] [-t] [<input> [<output>]]\n");      usage();
879      printf("  -C     show PCRE compile-time options and exit\n");      yield = 1;
880      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;  
881      }      }
882    op++;    op++;
883    argc--;    argc--;
# Line 580  offsets = (int *)malloc(size_offsets_max Line 890  offsets = (int *)malloc(size_offsets_max
890  if (offsets == NULL)  if (offsets == NULL)
891    {    {
892    printf("** Failed to get %d bytes of memory for offsets vector\n",    printf("** Failed to get %d bytes of memory for offsets vector\n",
893      size_offsets_max * sizeof(int));      (int)(size_offsets_max * sizeof(int)));
894    return 1;    yield = 1;
895      goto EXIT;
896    }    }
897    
898  /* Sort out the input and output files */  /* Sort out the input and output files */
899    
900  if (argc > 1)  if (argc > 1)
901    {    {
902    infile = fopen(argv[op], "rb");    infile = fopen(argv[op], INPUT_MODE);
903    if (infile == NULL)    if (infile == NULL)
904      {      {
905      printf("** Failed to open %s\n", argv[op]);      printf("** Failed to open %s\n", argv[op]);
906      return 1;      yield = 1;
907        goto EXIT;
908      }      }
909    }    }
910    
911  if (argc > 2)  if (argc > 2)
912    {    {
913    outfile = fopen(argv[op+1], "wb");    outfile = fopen(argv[op+1], OUTPUT_MODE);
914    if (outfile == NULL)    if (outfile == NULL)
915      {      {
916      printf("** Failed to open %s\n", argv[op+1]);      printf("** Failed to open %s\n", argv[op+1]);
917      return 1;      yield = 1;
918        goto EXIT;
919      }      }
920    }    }
921    
# Line 613  pcre_free = new_free; Line 926  pcre_free = new_free;
926  pcre_stack_malloc = stack_malloc;  pcre_stack_malloc = stack_malloc;
927  pcre_stack_free = stack_free;  pcre_stack_free = stack_free;
928    
929  /* Heading line, then prompt for first regex if stdin */  /* Heading line unless quiet, then prompt for first regex if stdin */
930    
931  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  if (!quiet) fprintf(outfile, "PCRE version %s\n\n", pcre_version());
932    
933  /* Main loop */  /* Main loop */
934    
# Line 637  while (!done) Line 950  while (!done)
950    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
951    int do_study = 0;    int do_study = 0;
952    int do_debug = debug;    int do_debug = debug;
953      int debug_lengths = 1;
954    int do_G = 0;    int do_G = 0;
955    int do_g = 0;    int do_g = 0;
956    int do_showinfo = showinfo;    int do_showinfo = showinfo;
957    int do_showrest = 0;    int do_showrest = 0;
958    int do_flip = 0;    int do_flip = 0;
959    int erroroffset, len, delimiter;    int erroroffset, len, delimiter, poffset;
960    
961    use_utf8 = 0;    use_utf8 = 0;
962    
963    if (infile == stdin) printf("  re> ");    if (infile == stdin) printf("  re> ");
964    if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break;    if (extend_inputline(infile, buffer) == NULL) break;
965    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
966    fflush(outfile);    fflush(outfile);
967    
# Line 659  while (!done) Line 973  while (!done)
973    
974    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
975      {      {
976      unsigned long int magic;      unsigned long int magic, get_options;
977      uschar sbuf[8];      uschar sbuf[8];
978      FILE *f;      FILE *f;
979    
# Line 707  while (!done) Line 1021  while (!done)
1021    
1022      /* Need to know if UTF-8 for printing data strings */      /* Need to know if UTF-8 for printing data strings */
1023    
1024      new_info(re, NULL, PCRE_INFO_OPTIONS, &options);      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1025      use_utf8 = (options & PCRE_UTF8) != 0;      use_utf8 = (get_options & PCRE_UTF8) != 0;
1026    
1027      /* Now see if there is any following study data */      /* Now see if there is any following study data */
1028    
# Line 752  while (!done) Line 1066  while (!done)
1066      }      }
1067    
1068    pp = p;    pp = p;
1069      poffset = p - buffer;
1070    
1071    for(;;)    for(;;)
1072      {      {
# Line 762  while (!done) Line 1077  while (!done)
1077        pp++;        pp++;
1078        }        }
1079      if (*pp != 0) break;      if (*pp != 0) break;
   
     len = BUFFER_SIZE - (pp - buffer);  
     if (len < 256)  
       {  
       fprintf(outfile, "** Expression too long - missing delimiter?\n");  
       goto SKIP_DATA;  
       }  
   
1080      if (infile == stdin) printf("    > ");      if (infile == stdin) printf("    > ");
1081      if (fgets((char *)pp, len, infile) == NULL)      if ((pp = extend_inputline(infile, pp)) == NULL)
1082        {        {
1083        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
1084        done = 1;        done = 1;
# Line 780  while (!done) Line 1087  while (!done)
1087      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
1088      }      }
1089    
1090      /* The buffer may have moved while being extended; reset the start of data
1091      pointer to the correct relative point in the buffer. */
1092    
1093      p = buffer + poffset;
1094    
1095    /* If the first character after the delimiter is backslash, make    /* If the first character after the delimiter is backslash, make
1096    the pattern end with backslash. This is purely to provide a way    the pattern end with backslash. This is purely to provide a way
1097    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 1114  while (!done)
1114      {      {
1115      switch (*pp++)      switch (*pp++)
1116        {        {
1117          case 'f': options |= PCRE_FIRSTLINE; break;
1118        case 'g': do_g = 1; break;        case 'g': do_g = 1; break;
1119        case 'i': options |= PCRE_CASELESS; break;        case 'i': options |= PCRE_CASELESS; break;
1120        case 'm': options |= PCRE_MULTILINE; break;        case 'm': options |= PCRE_MULTILINE; break;
# Line 810  while (!done) Line 1123  while (!done)
1123    
1124        case '+': do_showrest = 1; break;        case '+': do_showrest = 1; break;
1125        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
1126          case 'B': do_debug = 1; break;
1127        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
1128        case 'D': do_debug = do_showinfo = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
1129        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
1130        case 'F': do_flip = 1; break;        case 'F': do_flip = 1; break;
1131        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1132        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1133          case 'J': options |= PCRE_DUPNAMES; break;
1134        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1135        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1136    
# Line 826  while (!done) Line 1141  while (!done)
1141        case 'S': do_study = 1; break;        case 'S': do_study = 1; break;
1142        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1143        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
1144          case 'Z': debug_lengths = 0; break;
1145        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1146        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
1147    
1148        case 'L':        case 'L':
1149        ppp = pp;        ppp = pp;
1150        while (*ppp != '\n' && *ppp != ' ') ppp++;        /* The '\r' test here is so that it works on Windows. */
1151          /* The '0' test is just in case this is an unterminated line. */
1152          while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
1153        *ppp = 0;        *ppp = 0;
1154        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
1155          {          {
1156          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
1157          goto SKIP_DATA;          goto SKIP_DATA;
1158          }          }
1159          locale_set = 1;
1160        tables = pcre_maketables();        tables = pcre_maketables();
1161        pp = ppp;        pp = ppp;
1162        break;        break;
# Line 849  while (!done) Line 1168  while (!done)
1168        *pp = 0;        *pp = 0;
1169        break;        break;
1170    
1171        case '\n': case ' ': break;        case '<':
1172            {
1173            int x = check_newline(pp, outfile);
1174            if (x == 0) goto SKIP_DATA;
1175            options |= x;
1176            while (*pp++ != '>');
1177            }
1178          break;
1179    
1180          case '\r':                      /* So that it works in Windows */
1181          case '\n':
1182          case ' ':
1183          break;
1184    
1185        default:        default:
1186        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);
# Line 869  while (!done) Line 1200  while (!done)
1200    
1201      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
1202      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
1203        if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1204        if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1205        if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1206    
1207      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1208    
1209      /* 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 1211  while (!done)
1211    
1212      if (rc != 0)      if (rc != 0)
1213        {        {
1214        (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);        (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1215        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
1216        goto SKIP_DATA;        goto SKIP_DATA;
1217        }        }
# Line 888  while (!done) Line 1223  while (!done)
1223  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1224    
1225      {      {
1226      if (timeit)      if (timeit > 0)
1227        {        {
1228        register int i;        register int i;
1229        clock_t time_taken;        clock_t time_taken;
1230        clock_t start_time = clock();        clock_t start_time = clock();
1231        for (i = 0; i < LOOPREPEAT; i++)        for (i = 0; i < timeit; i++)
1232          {          {
1233          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1234          if (re != NULL) free(re);          if (re != NULL) free(re);
1235          }          }
1236        time_taken = clock() - start_time;        time_taken = clock() - start_time;
1237        fprintf(outfile, "Compile time %.3f milliseconds\n",        fprintf(outfile, "Compile time %.4f milliseconds\n",
1238          (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /          (((double)time_taken * 1000.0) / (double)timeit) /
1239            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
1240        }        }
1241    
# Line 917  while (!done) Line 1252  while (!done)
1252          {          {
1253          for (;;)          for (;;)
1254            {            {
1255            if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)            if (extend_inputline(infile, buffer) == NULL)
1256              {              {
1257              done = 1;              done = 1;
1258              goto CONTINUE;              goto CONTINUE;
# Line 952  while (!done) Line 1287  while (!done)
1287    
1288      if (do_study)      if (do_study)
1289        {        {
1290        if (timeit)        if (timeit > 0)
1291          {          {
1292          register int i;          register int i;
1293          clock_t time_taken;          clock_t time_taken;
1294          clock_t start_time = clock();          clock_t start_time = clock();
1295          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeit; i++)
1296            extra = pcre_study(re, study_options, &error);            extra = pcre_study(re, study_options, &error);
1297          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1298          if (extra != NULL) free(extra);          if (extra != NULL) free(extra);
1299          fprintf(outfile, "  Study time %.3f milliseconds\n",          fprintf(outfile, "  Study time %.4f milliseconds\n",
1300            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeit) /
1301              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
1302          }          }
1303        extra = pcre_study(re, study_options, &error);        extra = pcre_study(re, study_options, &error);
# Line 1005  while (!done) Line 1340  while (!done)
1340    
1341      SHOW_INFO:      SHOW_INFO:
1342    
1343        if (do_debug)
1344          {
1345          fprintf(outfile, "------------------------------------------------------------------\n");
1346          pcre_printint(re, outfile, debug_lengths);
1347          }
1348    
1349      if (do_showinfo)      if (do_showinfo)
1350        {        {
1351        unsigned long int get_options, all_options;        unsigned long int get_options, all_options;
1352    #if !defined NOINFOCHECK
1353        int old_first_char, old_options, old_count;        int old_first_char, old_options, old_count;
1354    #endif
1355        int count, backrefmax, first_char, need_char;        int count, backrefmax, first_char, need_char;
1356        int nameentrysize, namecount;        int nameentrysize, namecount;
1357        const uschar *nametable;        const uschar *nametable;
1358    
       if (do_debug)  
         {  
         fprintf(outfile, "------------------------------------------------------------------\n");  
         print_internals(re, outfile);  
         }  
   
1359        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1360        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1361        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
# Line 1029  while (!done) Line 1366  while (!done)
1366        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);
1367        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);
1368    
1369    #if !defined NOINFOCHECK
1370        old_count = pcre_info(re, &old_options, &old_first_char);        old_count = pcre_info(re, &old_options, &old_first_char);
1371        if (count < 0) fprintf(outfile,        if (count < 0) fprintf(outfile,
1372          "Error %d from pcre_info()\n", count);          "Error %d from pcre_info()\n", count);
# Line 1046  while (!done) Line 1384  while (!done)
1384            "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",            "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",
1385              get_options, old_options);              get_options, old_options);
1386          }          }
1387    #endif
1388    
1389        if (size != regex_gotten_store) fprintf(outfile,        if (size != regex_gotten_store) fprintf(outfile,
1390          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",
1391          size, regex_gotten_store);          (int)size, (int)regex_gotten_store);
1392    
1393        fprintf(outfile, "Capturing subpattern count = %d\n", count);        fprintf(outfile, "Capturing subpattern count = %d\n", count);
1394        if (backrefmax > 0)        if (backrefmax > 0)
# Line 1074  while (!done) Line 1413  while (!done)
1413        if (do_flip)        if (do_flip)
1414          {          {
1415          all_options = byteflip(all_options, sizeof(all_options));          all_options = byteflip(all_options, sizeof(all_options));
1416          }           }
1417    
1418        if ((all_options & PCRE_NOPARTIAL) != 0)        if ((all_options & PCRE_NOPARTIAL) != 0)
1419          fprintf(outfile, "Partial matching not supported\n");          fprintf(outfile, "Partial matching not supported\n");
1420    
1421        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1422          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\n",
1423            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1424            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1425            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
1426            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",
1427              ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",
1428            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",
1429            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
1430            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
1431            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1432              ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1433            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1434            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "");            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1435              ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1436    
1437        if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0)        switch (get_options & PCRE_NEWLINE_BITS)
1438          fprintf(outfile, "Case state changes\n");          {
1439            case PCRE_NEWLINE_CR:
1440            fprintf(outfile, "Forced newline sequence: CR\n");
1441            break;
1442    
1443            case PCRE_NEWLINE_LF:
1444            fprintf(outfile, "Forced newline sequence: LF\n");
1445            break;
1446    
1447            case PCRE_NEWLINE_CRLF:
1448            fprintf(outfile, "Forced newline sequence: CRLF\n");
1449            break;
1450    
1451            case PCRE_NEWLINE_ANYCRLF:
1452            fprintf(outfile, "Forced newline sequence: ANYCRLF\n");
1453            break;
1454    
1455            case PCRE_NEWLINE_ANY:
1456            fprintf(outfile, "Forced newline sequence: ANY\n");
1457            break;
1458    
1459            default:
1460            break;
1461            }
1462    
1463        if (first_char == -1)        if (first_char == -1)
1464          {          {
1465          fprintf(outfile, "First char at start or follows \\n\n");          fprintf(outfile, "First char at start or follows newline\n");
1466          }          }
1467        else if (first_char < 0)        else if (first_char < 0)
1468          {          {
# Line 1108  while (!done) Line 1473  while (!done)
1473          int ch = first_char & 255;          int ch = first_char & 255;
1474          const char *caseless = ((first_char & REQ_CASELESS) == 0)?          const char *caseless = ((first_char & REQ_CASELESS) == 0)?
1475            "" : " (caseless)";            "" : " (caseless)";
1476          if (isprint(ch))          if (PRINTHEX(ch))
1477            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);
1478          else          else
1479            fprintf(outfile, "First char = %d%s\n", ch, caseless);            fprintf(outfile, "First char = %d%s\n", ch, caseless);
# Line 1123  while (!done) Line 1488  while (!done)
1488          int ch = need_char & 255;          int ch = need_char & 255;
1489          const char *caseless = ((need_char & REQ_CASELESS) == 0)?          const char *caseless = ((need_char & REQ_CASELESS) == 0)?
1490            "" : " (caseless)";            "" : " (caseless)";
1491          if (isprint(ch))          if (PRINTHEX(ch))
1492            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);
1493          else          else
1494            fprintf(outfile, "Need char = %d%s\n", ch, caseless);            fprintf(outfile, "Need char = %d%s\n", ch, caseless);
# Line 1159  while (!done) Line 1524  while (!done)
1524                    fprintf(outfile, "\n  ");                    fprintf(outfile, "\n  ");
1525                    c = 2;                    c = 2;
1526                    }                    }
1527                  if (isprint(i) && i != ' ')                  if (PRINTHEX(i) && i != ' ')
1528                    {                    {
1529                    fprintf(outfile, "%c ", i);                    fprintf(outfile, "%c ", i);
1530                    c += 2;                    c += 2;
# Line 1218  while (!done) Line 1583  while (!done)
1583                  strerror(errno));                  strerror(errno));
1584                }                }
1585              else fprintf(outfile, "Study data written to %s\n", to_file);              else fprintf(outfile, "Study data written to %s\n", to_file);
1586    
1587              }              }
1588            }            }
1589          fclose(f);          fclose(f);
1590          }          }
1591    
1592          new_free(re);
1593          if (extra != NULL) new_free(extra);
1594          if (tables != NULL) new_free((void *)tables);
1595        continue;  /* With next regex */        continue;  /* With next regex */
1596        }        }
1597      }        /* End of non-POSIX compile */      }        /* End of non-POSIX compile */
# Line 1230  while (!done) Line 1600  while (!done)
1600    
1601    for (;;)    for (;;)
1602      {      {
1603      unsigned char *q;      uschar *q;
1604      unsigned char *bptr = dbuffer;      uschar *bptr;
1605      int *use_offsets = offsets;      int *use_offsets = offsets;
1606      int use_size_offsets = size_offsets;      int use_size_offsets = size_offsets;
1607      int callout_data = 0;      int callout_data = 0;
# Line 1244  while (!done) Line 1614  while (!done)
1614      int gmatched = 0;      int gmatched = 0;
1615      int start_offset = 0;      int start_offset = 0;
1616      int g_notempty = 0;      int g_notempty = 0;
1617        int use_dfa = 0;
1618    
1619      options = 0;      options = 0;
1620    
1621        *copynames = 0;
1622        *getnames = 0;
1623    
1624        copynamesptr = copynames;
1625        getnamesptr = getnames;
1626    
1627      pcre_callout = callout;      pcre_callout = callout;
1628      first_callout = 1;      first_callout = 1;
1629      callout_extra = 0;      callout_extra = 0;
# Line 1255  while (!done) Line 1632  while (!done)
1632      callout_fail_id = -1;      callout_fail_id = -1;
1633      show_malloc = 0;      show_malloc = 0;
1634    
1635      if (infile == stdin) printf("data> ");      if (extra != NULL) extra->flags &=
1636      if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)        ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
1637    
1638        len = 0;
1639        for (;;)
1640        {        {
1641        done = 1;        if (infile == stdin) printf("data> ");
1642        goto CONTINUE;        if (extend_inputline(infile, buffer + len) == NULL)
1643            {
1644            if (len > 0) break;
1645            done = 1;
1646            goto CONTINUE;
1647            }
1648          if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1649          len = (int)strlen((char *)buffer);
1650          if (buffer[len-1] == '\n') break;
1651        }        }
     if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);  
1652    
     len = (int)strlen((char *)buffer);  
1653      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
1654      buffer[len] = 0;      buffer[len] = 0;
1655      if (len == 0) break;      if (len == 0) break;
# Line 1271  while (!done) Line 1657  while (!done)
1657      p = buffer;      p = buffer;
1658      while (isspace(*p)) p++;      while (isspace(*p)) p++;
1659    
1660      q = dbuffer;      bptr = q = dbuffer;
1661      while ((c = *p++) != 0)      while ((c = *p++) != 0)
1662        {        {
1663        int i = 0;        int i = 0;
# Line 1293  while (!done) Line 1679  while (!done)
1679          c -= '0';          c -= '0';
1680          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')
1681            c = c * 8 + *p++ - '0';            c = c * 8 + *p++ - '0';
1682    
1683    #if !defined NOUTF8
1684            if (use_utf8 && c > 255)
1685              {
1686              unsigned char buff8[8];
1687              int ii, utn;
1688              utn = ord2utf8(c, buff8);
1689              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1690              c = buff8[ii];   /* Last byte */
1691              }
1692    #endif
1693          break;          break;
1694    
1695          case 'x':          case 'x':
1696    
1697          /* Handle \x{..} specially - new Perl thing for utf8 */          /* Handle \x{..} specially - new Perl thing for utf8 */
1698    
1699    #if !defined NOUTF8
1700          if (*p == '{')          if (*p == '{')
1701            {            {
1702            unsigned char *pt = p;            unsigned char *pt = p;
# Line 1317  while (!done) Line 1715  while (!done)
1715              }              }
1716            /* Not correct form; fall through */            /* Not correct form; fall through */
1717            }            }
1718    #endif
1719    
1720          /* Ordinary \x */          /* Ordinary \x */
1721    
# Line 1352  while (!done) Line 1751  while (!done)
1751            }            }
1752          else if (isalnum(*p))          else if (isalnum(*p))
1753            {            {
1754            uschar name[256];            uschar *npp = copynamesptr;
           uschar *npp = name;  
1755            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1756              *npp++ = 0;
1757            *npp = 0;            *npp = 0;
1758            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)copynamesptr);
1759            if (n < 0)            if (n < 0)
1760              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", copynamesptr);
1761            else copystrings |= 1 << n;            copynamesptr = npp;
1762            }            }
1763          else if (*p == '+')          else if (*p == '+')
1764            {            {
# Line 1397  while (!done) Line 1796  while (!done)
1796            }            }
1797          continue;          continue;
1798    
1799    #if !defined NODFA
1800            case 'D':
1801    #if !defined NOPOSIX
1802            if (posix || do_posix)
1803              printf("** Can't use dfa matching in POSIX mode: \\D ignored\n");
1804            else
1805    #endif
1806              use_dfa = 1;
1807            continue;
1808    
1809            case 'F':
1810            options |= PCRE_DFA_SHORTEST;
1811            continue;
1812    #endif
1813    
1814          case 'G':          case 'G':
1815          if (isdigit(*p))          if (isdigit(*p))
1816            {            {
# Line 1405  while (!done) Line 1819  while (!done)
1819            }            }
1820          else if (isalnum(*p))          else if (isalnum(*p))
1821            {            {
1822            uschar name[256];            uschar *npp = getnamesptr;
           uschar *npp = name;  
1823            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1824              *npp++ = 0;
1825            *npp = 0;            *npp = 0;
1826            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)getnamesptr);
1827            if (n < 0)            if (n < 0)
1828              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", getnamesptr);
1829            else getstrings |= 1 << n;            getnamesptr = npp;
1830            }            }
1831          continue;          continue;
1832    
# Line 1438  while (!done) Line 1852  while (!done)
1852            if (offsets == NULL)            if (offsets == NULL)
1853              {              {
1854              printf("** Failed to get %d bytes of memory for offsets vector\n",              printf("** Failed to get %d bytes of memory for offsets vector\n",
1855                size_offsets_max * sizeof(int));                (int)(size_offsets_max * sizeof(int)));
1856              return 1;              yield = 1;
1857                goto EXIT;
1858              }              }
1859            }            }
1860          use_size_offsets = n;          use_size_offsets = n;
# Line 1450  while (!done) Line 1865  while (!done)
1865          options |= PCRE_PARTIAL;          options |= PCRE_PARTIAL;
1866          continue;          continue;
1867    
1868            case 'Q':
1869            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1870            if (extra == NULL)
1871              {
1872              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1873              extra->flags = 0;
1874              }
1875            extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
1876            extra->match_limit_recursion = n;
1877            continue;
1878    
1879            case 'q':
1880            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1881            if (extra == NULL)
1882              {
1883              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1884              extra->flags = 0;
1885              }
1886            extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
1887            extra->match_limit = n;
1888            continue;
1889    
1890    #if !defined NODFA
1891            case 'R':
1892            options |= PCRE_DFA_RESTART;
1893            continue;
1894    #endif
1895    
1896          case 'S':          case 'S':
1897          show_malloc = 1;          show_malloc = 1;
1898          continue;          continue;
# Line 1461  while (!done) Line 1904  while (!done)
1904          case '?':          case '?':
1905          options |= PCRE_NO_UTF8_CHECK;          options |= PCRE_NO_UTF8_CHECK;
1906          continue;          continue;
1907    
1908            case '<':
1909              {
1910              int x = check_newline(p, outfile);
1911              if (x == 0) goto NEXT_DATA;
1912              options |= x;
1913              while (*p++ != '>');
1914              }
1915            continue;
1916          }          }
1917        *q++ = c;        *q++ = c;
1918        }        }
1919      *q = 0;      *q = 0;
1920      len = q - dbuffer;      len = q - dbuffer;
1921    
1922        if ((all_use_dfa || use_dfa) && find_match_limit)
1923          {
1924          printf("**Match limit not relevant for DFA matching: ignored\n");
1925          find_match_limit = 0;
1926          }
1927    
1928      /* Handle matching via the POSIX interface, which does not      /* Handle matching via the POSIX interface, which does not
1929      support timing or playing with the match limit or callout data. */      support timing or playing with the match limit or callout data. */
1930    
# Line 1485  while (!done) Line 1943  while (!done)
1943    
1944        if (rc != 0)        if (rc != 0)
1945          {          {
1946          (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);          (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1947          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
1948          }          }
1949          else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
1950                  != 0)
1951            {
1952            fprintf(outfile, "Matched with REG_NOSUB\n");
1953            }
1954        else        else
1955          {          {
1956          size_t i;          size_t i;
# Line 1519  while (!done) Line 1982  while (!done)
1982    
1983      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
1984        {        {
1985        if (timeit)        if (timeitm > 0)
1986          {          {
1987          register int i;          register int i;
1988          clock_t time_taken;          clock_t time_taken;
1989          clock_t start_time = clock();          clock_t start_time = clock();
1990          for (i = 0; i < LOOPREPEAT; i++)  
1991    #if !defined NODFA
1992            if (all_use_dfa || use_dfa)
1993              {
1994              int workspace[1000];
1995              for (i = 0; i < timeitm; i++)
1996                count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
1997                  options | g_notempty, use_offsets, use_size_offsets, workspace,
1998                  sizeof(workspace)/sizeof(int));
1999              }
2000            else
2001    #endif
2002    
2003            for (i = 0; i < timeitm; i++)
2004            count = pcre_exec(re, extra, (char *)bptr, len,            count = pcre_exec(re, extra, (char *)bptr, len,
2005              start_offset, options | g_notempty, use_offsets, use_size_offsets);              start_offset, options | g_notempty, use_offsets, use_size_offsets);
2006    
2007          time_taken = clock() - start_time;          time_taken = clock() - start_time;
2008          fprintf(outfile, "Execute time %.3f milliseconds\n",          fprintf(outfile, "Execute time %.4f milliseconds\n",
2009            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeitm) /
2010              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
2011          }          }
2012    
2013        /* 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
2014        varying limits in order to find the minimum value. */        varying limits in order to find the minimum value for the match limit and
2015          for the recursion limit. */
2016    
2017        if (find_match_limit)        if (find_match_limit)
2018          {          {
         int min = 0;  
         int mid = 64;  
         int max = -1;  
   
2019          if (extra == NULL)          if (extra == NULL)
2020            {            {
2021            extra = (pcre_extra *)malloc(sizeof(pcre_extra));            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2022            extra->flags = 0;            extra->flags = 0;
2023            }            }
         extra->flags |= PCRE_EXTRA_MATCH_LIMIT;  
2024    
2025          for (;;)          (void)check_match_limit(re, extra, bptr, len, start_offset,
2026            {            options|g_notempty, use_offsets, use_size_offsets,
2027            extra->match_limit = mid;            PCRE_EXTRA_MATCH_LIMIT, &(extra->match_limit),
2028            count = pcre_exec(re, extra, (char *)bptr, len, start_offset,            PCRE_ERROR_MATCHLIMIT, "match()");
2029              options | g_notempty, use_offsets, use_size_offsets);  
2030            if (count == PCRE_ERROR_MATCHLIMIT)          count = check_match_limit(re, extra, bptr, len, start_offset,
2031              {            options|g_notempty, use_offsets, use_size_offsets,
2032              /* fprintf(outfile, "Testing match limit = %d\n", mid); */            PCRE_EXTRA_MATCH_LIMIT_RECURSION, &(extra->match_limit_recursion),
2033              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;  
2034          }          }
2035    
2036        /* 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 2052  while (!done)
2052        /* 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
2053        value of match_limit. */        value of match_limit. */
2054    
2055        else  #if !defined NODFA
2056          else if (all_use_dfa || use_dfa)
2057          {          {
2058          count = pcre_exec(re, extra, (char *)bptr, len,          int workspace[1000];
2059            start_offset, options | g_notempty, use_offsets, use_size_offsets);          count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
2060              options | g_notempty, use_offsets, use_size_offsets, workspace,
2061              sizeof(workspace)/sizeof(int));
2062            if (count == 0)
2063              {
2064              fprintf(outfile, "Matched, but too many subsidiary matches\n");
2065              count = use_size_offsets/2;
2066              }
2067          }          }
2068    #endif
2069    
2070        if (count == 0)        else
2071          {          {
2072          fprintf(outfile, "Matched, but too many substrings\n");          count = pcre_exec(re, extra, (char *)bptr, len,
2073          count = use_size_offsets/3;            start_offset, options | g_notempty, use_offsets, use_size_offsets);
2074            if (count == 0)
2075              {
2076              fprintf(outfile, "Matched, but too many substrings\n");
2077              count = use_size_offsets/3;
2078              }
2079          }          }
2080    
2081        /* Matched */        /* Matched */
2082    
2083        if (count >= 0)        if (count >= 0)
2084          {          {
2085          int i;          int i, maxcount;
2086    
2087    #if !defined NODFA
2088            if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else
2089    #endif
2090              maxcount = use_size_offsets/3;
2091    
2092            /* This is a check against a lunatic return value. */
2093    
2094            if (count > maxcount)
2095              {
2096              fprintf(outfile,
2097                "** PCRE error: returned count %d is too big for offset size %d\n",
2098                count, use_size_offsets);
2099              count = use_size_offsets/3;
2100              if (do_g || do_G)
2101                {
2102                fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G');
2103                do_g = do_G = FALSE;        /* Break g/G loop */
2104                }
2105              }
2106    
2107          for (i = 0; i < count * 2; i += 2)          for (i = 0; i < count * 2; i += 2)
2108            {            {
2109            if (use_offsets[i] < 0)            if (use_offsets[i] < 0)
# Line 1641  while (!done) Line 2131  while (!done)
2131            {            {
2132            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
2133              {              {
2134              char copybuffer[16];              char copybuffer[256];
2135              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,
2136                i, copybuffer, sizeof(copybuffer));                i, copybuffer, sizeof(copybuffer));
2137              if (rc < 0)              if (rc < 0)
# Line 1651  while (!done) Line 2141  while (!done)
2141              }              }
2142            }            }
2143    
2144            for (copynamesptr = copynames;
2145                 *copynamesptr != 0;
2146                 copynamesptr += (int)strlen((char*)copynamesptr) + 1)
2147              {
2148              char copybuffer[256];
2149              int rc = pcre_copy_named_substring(re, (char *)bptr, use_offsets,
2150                count, (char *)copynamesptr, copybuffer, sizeof(copybuffer));
2151              if (rc < 0)
2152                fprintf(outfile, "copy substring %s failed %d\n", copynamesptr, rc);
2153              else
2154                fprintf(outfile, "  C %s (%d) %s\n", copybuffer, rc, copynamesptr);
2155              }
2156    
2157          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2158            {            {
2159            if ((getstrings & (1 << i)) != 0)            if ((getstrings & (1 << i)) != 0)
# Line 1663  while (!done) Line 2166  while (!done)
2166              else              else
2167                {                {
2168                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
               /* free((void *)substring); */  
2169                pcre_free_substring(substring);                pcre_free_substring(substring);
2170                }                }
2171              }              }
2172            }            }
2173    
2174            for (getnamesptr = getnames;
2175                 *getnamesptr != 0;
2176                 getnamesptr += (int)strlen((char*)getnamesptr) + 1)
2177              {
2178              const char *substring;
2179              int rc = pcre_get_named_substring(re, (char *)bptr, use_offsets,
2180                count, (char *)getnamesptr, &substring);
2181              if (rc < 0)
2182                fprintf(outfile, "copy substring %s failed %d\n", getnamesptr, rc);
2183              else
2184                {
2185                fprintf(outfile, "  G %s (%d) %s\n", substring, rc, getnamesptr);
2186                pcre_free_substring(substring);
2187                }
2188              }
2189    
2190          if (getlist)          if (getlist)
2191            {            {
2192            const char **stringlist;            const char **stringlist;
# Line 1692  while (!done) Line 2210  while (!done)
2210    
2211        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2212          {          {
2213          fprintf(outfile, "Partial match\n");          fprintf(outfile, "Partial match");
2214    #if !defined NODFA
2215            if ((all_use_dfa || use_dfa) && use_size_offsets > 2)
2216              fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],
2217                bptr + use_offsets[0]);
2218    #endif
2219            fprintf(outfile, "\n");
2220          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2221          }          }
2222    
2223        /* 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
2224        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
2225        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
2226        matching, the advance must be one character, not one byte. Fudge the        string - that was checked before setting g_notempty.
2227        offset values to achieve this. We won't be at the end of the string -  
2228        that was checked before setting g_notempty. */        Complication arises in the case when the newline option is "any" or
2229          "anycrlf". If the previous match was at the end of a line terminated by
2230          CRLF, an advance of one character just passes the \r, whereas we should
2231          prefer the longer newline sequence, as does the code in pcre_exec().
2232          Fudge the offset value to achieve this.
2233    
2234          Otherwise, in the case of UTF-8 matching, the advance must be one
2235          character, not one byte. */
2236    
2237        else        else
2238          {          {
2239          if (g_notempty != 0)          if (g_notempty != 0)
2240            {            {
2241            int onechar = 1;            int onechar = 1;
2242              unsigned int obits = ((real_pcre *)re)->options;
2243            use_offsets[0] = start_offset;            use_offsets[0] = start_offset;
2244            if (use_utf8)            if ((obits & PCRE_NEWLINE_BITS) == 0)
2245                {
2246                int d;
2247                (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);
2248                obits = (d == '\r')? PCRE_NEWLINE_CR :
2249                        (d == '\n')? PCRE_NEWLINE_LF :
2250                        (d == ('\r'<<8 | '\n'))? PCRE_NEWLINE_CRLF :
2251                        (d == -2)? PCRE_NEWLINE_ANYCRLF :
2252                        (d == -1)? PCRE_NEWLINE_ANY : 0;
2253                }
2254              if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2255                   (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2256                  &&
2257                  start_offset < len - 1 &&
2258                  bptr[start_offset] == '\r' &&
2259                  bptr[start_offset+1] == '\n')
2260                onechar++;
2261              else if (use_utf8)
2262              {              {
2263              while (start_offset + onechar < len)              while (start_offset + onechar < len)
2264                {                {
# Line 1744  while (!done) Line 2293  while (!done)
2293        character. */        character. */
2294    
2295        g_notempty = 0;        g_notempty = 0;
2296    
2297        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2298          {          {
2299          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
# Line 1762  while (!done) Line 2312  while (!done)
2312          len -= use_offsets[1];          len -= use_offsets[1];
2313          }          }
2314        }  /* End of loop for /g and /G */        }  /* End of loop for /g and /G */
2315    
2316        NEXT_DATA: continue;
2317      }    /* End of loop for data lines */      }    /* End of loop for data lines */
2318    
2319    CONTINUE:    CONTINUE:
# Line 1770  while (!done) Line 2322  while (!done)
2322    if (posix || do_posix) regfree(&preg);    if (posix || do_posix) regfree(&preg);
2323  #endif  #endif
2324    
2325    if (re != NULL) free(re);    if (re != NULL) new_free(re);
2326    if (extra != NULL) free(extra);    if (extra != NULL) new_free(extra);
2327    if (tables != NULL)    if (tables != NULL)
2328      {      {
2329      free((void *)tables);      new_free((void *)tables);
2330      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");
2331        locale_set = 0;
2332      }      }
2333    }    }
2334    
2335  if (infile == stdin) fprintf(outfile, "\n");  if (infile == stdin) fprintf(outfile, "\n");
2336  return 0;  
2337    EXIT:
2338    
2339    if (infile != NULL && infile != stdin) fclose(infile);
2340    if (outfile != NULL && outfile != stdout) fclose(outfile);
2341    
2342    free(buffer);
2343    free(dbuffer);
2344    free(pbuffer);
2345    free(offsets);
2346    
2347    return yield;
2348  }  }
2349    
2350  /* End */  /* End of pcretest.c */

Legend:
Removed from v.75  
changed lines
  Added in v.151

  ViewVC Help
Powered by ViewVC 1.1.5