/[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 227 by ph10, Tue Aug 21 15:00:15 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 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    
 /* We need the internal info for displaying the results of pcre_study(). Also  
 for getting the opcodes for showing compiled code. */  
51    
52  #define PCRE_SPY        /* For Win32 build, import data, not export */  /* A number of things vary for Windows builds. Originally, pcretest opened its
53  #include "internal.h"  input and output without "b"; then I was told that "b" was needed in some
54    environments, so it was added for release 5.0 to both the input and output. (It
55    makes no difference on Unix-like systems.) Later I was told that it is wrong
56    for the input on Windows. I've now abstracted the modes into two macros that
57    are set here, to make it easier to fiddle with them, and removed "b" from the
58    input mode under Windows. */
59    
60    #if defined(_WIN32) || defined(WIN32)
61    #include <io.h>                /* For _setmode() */
62    #include <fcntl.h>             /* For _O_BINARY */
63    #define INPUT_MODE   "r"
64    #define OUTPUT_MODE  "wb"
65    
66    #else
67    #include <sys/time.h>          /* These two includes are needed */
68    #include <sys/resource.h>      /* for setrlimit(). */
69    #define INPUT_MODE   "rb"
70    #define OUTPUT_MODE  "wb"
71    #endif
72    
73    
74    /* We have to include pcre_internal.h because we need the internal info for
75    displaying the results of pcre_study() and we also need to know about the
76    internal macros, structures, and other internal data values; pcretest has
77    "inside information" compared to a program that strictly follows the PCRE API.
78    
79    Although pcre_internal.h does itself include pcre.h, we explicitly include it
80    here before pcre_internal.h so that the PCRE_EXP_xxx macros get set
81    appropriately for an application, not for building PCRE. */
82    
83    #include "pcre.h"
84    #include "pcre_internal.h"
85    
86    /* We need access to the data tables that PCRE uses. So as not to have to keep
87    two copies, we include the source file here, changing the names of the external
88    symbols to prevent clashes. */
89    
90    #define _pcre_utf8_table1      utf8_table1
91    #define _pcre_utf8_table1_size utf8_table1_size
92    #define _pcre_utf8_table2      utf8_table2
93    #define _pcre_utf8_table3      utf8_table3
94    #define _pcre_utf8_table4      utf8_table4
95    #define _pcre_utt              utt
96    #define _pcre_utt_size         utt_size
97    #define _pcre_OP_lengths       OP_lengths
98    
99    #include "pcre_tables.c"
100    
101    /* We also need the pcre_printint() function for printing out compiled
102    patterns. This function is in a separate file so that it can be included in
103    pcre_compile.c when that module is compiled with debugging enabled.
104    
105    The definition of the macro PRINTABLE, which determines whether to print an
106    output character as-is or as a hex value when showing compiled patterns, is
107    contained in this file. We uses it here also, in cases when the locale has not
108    been explicitly changed, so as to get consistent output from systems that
109    differ in their output from isprint() even in the "C" locale. */
110    
111    #include "pcre_printint.src"
112    
113    #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
114    
115    
116  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
117  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 121  Makefile. */
121  #include "pcreposix.h"  #include "pcreposix.h"
122  #endif  #endif
123    
124    /* It is also possible, for the benefit of the version currently imported into
125    Exim, to build pcretest without support for UTF8 (define NOUTF8), without the
126    interface to the DFA matcher (NODFA), and without the doublecheck of the old
127    "info" function (define NOINFOCHECK). In fact, we automatically cut out the
128    UTF8 support if PCRE is built without it. */
129    
130    #ifndef SUPPORT_UTF8
131    #ifndef NOUTF8
132    #define NOUTF8
133    #endif
134    #endif
135    
136    
137    /* Other parameters */
138    
139  #ifndef CLOCKS_PER_SEC  #ifndef CLOCKS_PER_SEC
140  #ifdef CLK_TCK  #ifdef CLK_TCK
141  #define CLOCKS_PER_SEC CLK_TCK  #define CLOCKS_PER_SEC CLK_TCK
# Line 66  Makefile. */ Line 144  Makefile. */
144  #endif  #endif
145  #endif  #endif
146    
147  #define LOOPREPEAT 500000  /* This is the default loop count for timing. */
148    
149  #define BUFFER_SIZE 30000  #define LOOPREPEAT 500000
 #define PBUFFER_SIZE BUFFER_SIZE  
 #define DBUFFER_SIZE BUFFER_SIZE  
150    
151    /* Static variables */
152    
153  static FILE *outfile;  static FILE *outfile;
154  static int log_store = 0;  static int log_store = 0;
# Line 79  static int callout_count; Line 156  static int callout_count;
156  static int callout_extra;  static int callout_extra;
157  static int callout_fail_count;  static int callout_fail_count;
158  static int callout_fail_id;  static int callout_fail_id;
159    static int debug_lengths;
160  static int first_callout;  static int first_callout;
161    static int locale_set = 0;
162  static int show_malloc;  static int show_malloc;
163  static int use_utf8;  static int use_utf8;
164  static size_t gotten_store;  static size_t gotten_store;
165    
166    /* The buffers grow automatically if very long input lines are encountered. */
167    
168    static int buffer_size = 50000;
169    static uschar *buffer = NULL;
170    static uschar *dbuffer = NULL;
171  static uschar *pbuffer = NULL;  static uschar *pbuffer = NULL;
172    
173    
 static const int utf8_table1[] = {  
   0x0000007f, 0x000007ff, 0x0000ffff, 0x001fffff, 0x03ffffff, 0x7fffffff};  
174    
175  static const int utf8_table2[] = {  /*************************************************
176    0, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc};  *        Read or extend an input line            *
177    *************************************************/
178    
179  static const int utf8_table3[] = {  /* Input lines are read into buffer, but both patterns and data lines can be
180    0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01};  continued over multiple input lines. In addition, if the buffer fills up, we
181    want to automatically expand it so as to be able to handle extremely large
182    lines that are needed for certain stress tests. When the input buffer is
183    expanded, the other two buffers must also be expanded likewise, and the
184    contents of pbuffer, which are a copy of the input for callouts, must be
185    preserved (for when expansion happens for a data line). This is not the most
186    optimal way of handling this, but hey, this is just a test program!
187    
188    Arguments:
189      f            the file to read
190      start        where in buffer to start (this *must* be within buffer)
191    
192    Returns:       pointer to the start of new data
193                   could be a copy of start, or could be moved
194                   NULL if no data read and EOF reached
195    */
196    
197    static uschar *
198    extend_inputline(FILE *f, uschar *start)
199    {
200    uschar *here = start;
201    
202    for (;;)
203      {
204      int rlen = buffer_size - (here - buffer);
205    
206      if (rlen > 1000)
207        {
208        int dlen;
209        if (fgets((char *)here, rlen,  f) == NULL)
210          return (here == start)? NULL : start;
211        dlen = (int)strlen((char *)here);
212        if (dlen > 0 && here[dlen - 1] == '\n') return start;
213        here += dlen;
214        }
215    
216      else
217        {
218        int new_buffer_size = 2*buffer_size;
219        uschar *new_buffer = (unsigned char *)malloc(new_buffer_size);
220        uschar *new_dbuffer = (unsigned char *)malloc(new_buffer_size);
221        uschar *new_pbuffer = (unsigned char *)malloc(new_buffer_size);
222    
223        if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
224          {
225          fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
226          exit(1);
227          }
228    
229        memcpy(new_buffer, buffer, buffer_size);
230        memcpy(new_pbuffer, pbuffer, buffer_size);
231    
232        buffer_size = new_buffer_size;
233    
234        start = new_buffer + (start - buffer);
235        here = new_buffer + (here - buffer);
236    
237        free(buffer);
238        free(dbuffer);
239        free(pbuffer);
240    
241        buffer = new_buffer;
242        dbuffer = new_dbuffer;
243        pbuffer = new_pbuffer;
244        }
245      }
246    
247    return NULL;  /* Control never gets here */
248    }
249    
250    
251    
 /*************************************************  
 *         Print compiled regex                   *  
 *************************************************/  
252    
 /* 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"  
253    
254    
255    
# Line 122  static uschar OP_lengths[] = { OP_LENGTH Line 259  static uschar OP_lengths[] = { OP_LENGTH
259    
260  /* 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
261  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
262  for unpicking the -o argument, so just keep it simple.  for unpicking arguments, so just keep it simple.
263    
264  Arguments:  Arguments:
265    str           string to be converted    str           string to be converted
# Line 143  return(result); Line 280  return(result);
280    
281    
282    
 /*************************************************  
 *       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;  
 }  
   
283    
284  /*************************************************  /*************************************************
285  *            Convert UTF-8 string to value       *  *            Convert UTF-8 string to value       *
# Line 188  return i + 1; Line 289  return i + 1;
289  and returns the value of the character.  and returns the value of the character.
290    
291  Argument:  Argument:
292    buffer   a pointer to the byte vector    utf8bytes   a pointer to the byte vector
293    vptr     a pointer to an int to receive the value    vptr        a pointer to an int to receive the value
294    
295  Returns:   >  0 => the number of bytes consumed  Returns:      >  0 => the number of bytes consumed
296             -6 to 0 => malformed UTF-8 character at offset = (-return)                -6 to 0 => malformed UTF-8 character at offset = (-return)
297  */  */
298    
299    #if !defined NOUTF8
300    
301  static int  static int
302  utf82ord(unsigned char *buffer, int *vptr)  utf82ord(unsigned char *utf8bytes, int *vptr)
303  {  {
304  int c = *buffer++;  int c = *utf8bytes++;
305  int d = c;  int d = c;
306  int i, j, s;  int i, j, s;
307    
# Line 218  d = (c & utf8_table3[i]) << s; Line 321  d = (c & utf8_table3[i]) << s;
321    
322  for (j = 0; j < i; j++)  for (j = 0; j < i; j++)
323    {    {
324    c = *buffer++;    c = *utf8bytes++;
325    if ((c & 0xc0) != 0x80) return -(j+1);    if ((c & 0xc0) != 0x80) return -(j+1);
326    s -= 6;    s -= 6;
327    d |= (c & 0x3f) << s;    d |= (c & 0x3f) << s;
# Line 226  for (j = 0; j < i; j++) Line 329  for (j = 0; j < i; j++)
329    
330  /* Check that encoding was the correct unique one */  /* Check that encoding was the correct unique one */
331    
332  for (j = 0; j < sizeof(utf8_table1)/sizeof(int); j++)  for (j = 0; j < utf8_table1_size; j++)
333    if (d <= utf8_table1[j]) break;    if (d <= utf8_table1[j]) break;
334  if (j != i) return -(i+1);  if (j != i) return -(i+1);
335    
# Line 236  if (j != i) return -(i+1); Line 339  if (j != i) return -(i+1);
339  return i+1;  return i+1;
340  }  }
341    
342    #endif
343    
344    
345    
346    /*************************************************
347    *       Convert character value to UTF-8         *
348    *************************************************/
349    
350    /* This function takes an integer value in the range 0 - 0x7fffffff
351    and encodes it as a UTF-8 character in 0 to 6 bytes.
352    
353    Arguments:
354      cvalue     the character value
355      utf8bytes  pointer to buffer for result - at least 6 bytes long
356    
357    Returns:     number of characters placed in the buffer
358    */
359    
360    #if !defined NOUTF8
361    
362    static int
363    ord2utf8(int cvalue, uschar *utf8bytes)
364    {
365    register int i, j;
366    for (i = 0; i < utf8_table1_size; i++)
367      if (cvalue <= utf8_table1[i]) break;
368    utf8bytes += i;
369    for (j = i; j > 0; j--)
370     {
371     *utf8bytes-- = 0x80 | (cvalue & 0x3f);
372     cvalue >>= 6;
373     }
374    *utf8bytes = utf8_table2[i] | cvalue;
375    return i + 1;
376    }
377    
378    #endif
379    
380    
381    
382  /*************************************************  /*************************************************
# Line 248  chars without printing. */ Line 389  chars without printing. */
389    
390  static int pchars(unsigned char *p, int length, FILE *f)  static int pchars(unsigned char *p, int length, FILE *f)
391  {  {
392  int c;  int c = 0;
393  int yield = 0;  int yield = 0;
394    
395  while (length-- > 0)  while (length-- > 0)
396    {    {
397    #if !defined NOUTF8
398    if (use_utf8)    if (use_utf8)
399      {      {
400      int rc = utf82ord(p, &c);      int rc = utf82ord(p, &c);
# Line 261  while (length-- > 0) Line 403  while (length-- > 0)
403        {        {
404        length -= rc - 1;        length -= rc - 1;
405        p += rc;        p += rc;
406        if (c < 256 && isprint(c))        if (PRINTHEX(c))
407          {          {
408          if (f != NULL) fprintf(f, "%c", c);          if (f != NULL) fprintf(f, "%c", c);
409          yield++;          yield++;
410          }          }
411        else        else
412          {          {
413          int n;          int n = 4;
414          if (f != NULL) fprintf(f, "\\x{%02x}%n", c, &n);          if (f != NULL) fprintf(f, "\\x{%02x}", c);
415          yield += n;          yield += (n <= 0x000000ff)? 2 :
416                     (n <= 0x00000fff)? 3 :
417                     (n <= 0x0000ffff)? 4 :
418                     (n <= 0x000fffff)? 5 : 6;
419          }          }
420        continue;        continue;
421        }        }
422      }      }
423    #endif
424    
425     /* Not UTF-8, or malformed UTF-8  */     /* Not UTF-8, or malformed UTF-8  */
426    
427    if (isprint(c = *(p++)))    c = *p++;
428      if (PRINTHEX(c))
429      {      {
430      if (f != NULL) fprintf(f, "%c", c);      if (f != NULL) fprintf(f, "%c", c);
431      yield++;      yield++;
# Line 403  static void *new_malloc(size_t size) Line 550  static void *new_malloc(size_t size)
550  void *block = malloc(size);  void *block = malloc(size);
551  gotten_store = size;  gotten_store = size;
552  if (show_malloc)  if (show_malloc)
553    fprintf(outfile, "malloc       %3d %p\n", size, block);    fprintf(outfile, "malloc       %3d %p\n", (int)size, block);
554  return block;  return block;
555  }  }
556    
# Line 421  static void *stack_malloc(size_t size) Line 568  static void *stack_malloc(size_t size)
568  {  {
569  void *block = malloc(size);  void *block = malloc(size);
570  if (show_malloc)  if (show_malloc)
571    fprintf(outfile, "stack_malloc %3d %p\n", size, block);    fprintf(outfile, "stack_malloc %3d %p\n", (int)size, block);
572  return block;  return block;
573  }  }
574    
# Line 452  if ((rc = pcre_fullinfo(re, study, optio Line 599  if ((rc = pcre_fullinfo(re, study, optio
599  *         Byte flipping function                 *  *         Byte flipping function                 *
600  *************************************************/  *************************************************/
601    
602  static long int  static unsigned long int
603  byteflip(long int value, int n)  byteflip(unsigned long int value, int n)
604  {  {
605  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);
606  return ((value & 0x000000ff) << 24) |  return ((value & 0x000000ff) << 24) |
# Line 466  return ((value & 0x000000ff) << 24) | Line 613  return ((value & 0x000000ff) << 24) |
613    
614    
615  /*************************************************  /*************************************************
616    *        Check match or recursion limit          *
617    *************************************************/
618    
619    static int
620    check_match_limit(pcre *re, pcre_extra *extra, uschar *bptr, int len,
621      int start_offset, int options, int *use_offsets, int use_size_offsets,
622      int flag, unsigned long int *limit, int errnumber, const char *msg)
623    {
624    int count;
625    int min = 0;
626    int mid = 64;
627    int max = -1;
628    
629    extra->flags |= flag;
630    
631    for (;;)
632      {
633      *limit = mid;
634    
635      count = pcre_exec(re, extra, (char *)bptr, len, start_offset, options,
636        use_offsets, use_size_offsets);
637    
638      if (count == errnumber)
639        {
640        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
641        min = mid;
642        mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;
643        }
644    
645      else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||
646                             count == PCRE_ERROR_PARTIAL)
647        {
648        if (mid == min + 1)
649          {
650          fprintf(outfile, "Minimum %s limit = %d\n", msg, mid);
651          break;
652          }
653        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
654        max = mid;
655        mid = (min + mid)/2;
656        }
657      else break;    /* Some other error */
658      }
659    
660    extra->flags &= ~flag;
661    return count;
662    }
663    
664    
665    
666    /*************************************************
667    *         Case-independent strncmp() function    *
668    *************************************************/
669    
670    /*
671    Arguments:
672      s         first string
673      t         second string
674      n         number of characters to compare
675    
676    Returns:    < 0, = 0, or > 0, according to the comparison
677    */
678    
679    static int
680    strncmpic(uschar *s, uschar *t, int n)
681    {
682    while (n--)
683      {
684      int c = tolower(*s++) - tolower(*t++);
685      if (c) return c;
686      }
687    return 0;
688    }
689    
690    
691    
692    /*************************************************
693    *         Check newline indicator                *
694    *************************************************/
695    
696    /* This is used both at compile and run-time to check for <xxx> escapes, where
697    xxx is LF, CR, CRLF, ANYCRLF, or ANY. Print a message and return 0 if there is
698    no match.
699    
700    Arguments:
701      p           points after the leading '<'
702      f           file for error message
703    
704    Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
705    */
706    
707    static int
708    check_newline(uschar *p, FILE *f)
709    {
710    if (strncmpic(p, (uschar *)"cr>", 3) == 0) return PCRE_NEWLINE_CR;
711    if (strncmpic(p, (uschar *)"lf>", 3) == 0) return PCRE_NEWLINE_LF;
712    if (strncmpic(p, (uschar *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
713    if (strncmpic(p, (uschar *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF;
714    if (strncmpic(p, (uschar *)"any>", 4) == 0) return PCRE_NEWLINE_ANY;
715    fprintf(f, "Unknown newline type at: <%s\n", p);
716    return 0;
717    }
718    
719    
720    
721    /*************************************************
722    *             Usage function                     *
723    *************************************************/
724    
725    static void
726    usage(void)
727    {
728    printf("Usage:     pcretest [options] [<input> [<output>]]\n");
729    printf("  -b       show compiled code (bytecode)\n");
730    printf("  -C       show PCRE compile-time options and exit\n");
731    printf("  -d       debug: show compiled code and information (-b and -i)\n");
732    #if !defined NODFA
733    printf("  -dfa     force DFA matching for all subjects\n");
734    #endif
735    printf("  -help    show usage information\n");
736    printf("  -i       show information about compiled patterns\n"
737           "  -m       output memory used information\n"
738           "  -o <n>   set size of offsets vector to <n>\n");
739    #if !defined NOPOSIX
740    printf("  -p       use POSIX interface\n");
741    #endif
742    printf("  -q       quiet: do not output PCRE version number at start\n");
743    printf("  -S <n>   set stack size to <n> megabytes\n");
744    printf("  -s       output store (memory) used information\n"
745           "  -t       time compilation and execution\n");
746    printf("  -t <n>   time compilation and execution, repeating <n> times\n");
747    printf("  -tm      time execution (matching) only\n");
748    printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
749    }
750    
751    
752    
753    /*************************************************
754  *                Main Program                    *  *                Main Program                    *
755  *************************************************/  *************************************************/
756    
# Line 480  int options = 0; Line 765  int options = 0;
765  int study_options = 0;  int study_options = 0;
766  int op = 1;  int op = 1;
767  int timeit = 0;  int timeit = 0;
768    int timeitm = 0;
769  int showinfo = 0;  int showinfo = 0;
770  int showstore = 0;  int showstore = 0;
771    int quiet = 0;
772  int size_offsets = 45;  int size_offsets = 45;
773  int size_offsets_max;  int size_offsets_max;
774  int *offsets;  int *offsets = NULL;
775  #if !defined NOPOSIX  #if !defined NOPOSIX
776  int posix = 0;  int posix = 0;
777  #endif  #endif
778  int debug = 0;  int debug = 0;
779  int done = 0;  int done = 0;
780    int all_use_dfa = 0;
781    int yield = 0;
782    int stack_size;
783    
784    /* These vectors store, end-to-end, a list of captured substring names. Assume
785    that 1024 is plenty long enough for the few names we'll be testing. */
786    
787    uschar copynames[1024];
788    uschar getnames[1024];
789    
790  unsigned char *buffer;  uschar *copynamesptr;
791  unsigned char *dbuffer;  uschar *getnamesptr;
792    
793  /* Get buffers from malloc() so that Electric Fence will check their misuse  /* Get buffers from malloc() so that Electric Fence will check their misuse
794  when I am debugging. */  when I am debugging. They grow automatically when very long lines are read. */
795    
796  buffer = (unsigned char *)malloc(BUFFER_SIZE);  buffer = (unsigned char *)malloc(buffer_size);
797  dbuffer = (unsigned char *)malloc(DBUFFER_SIZE);  dbuffer = (unsigned char *)malloc(buffer_size);
798  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.  */  
799    
800  #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) */  
801    
802  outfile = stdout;  outfile = stdout;
803    
804    /* The following  _setmode() stuff is some Windows magic that tells its runtime
805    library to translate CRLF into a single LF character. At least, that's what
806    I've been told: never having used Windows I take this all on trust. Originally
807    it set 0x8000, but then I was advised that _O_BINARY was better. */
808    
809    #if defined(_WIN32) || defined(WIN32)
810    _setmode( _fileno( stdout ), _O_BINARY );
811    #endif
812    
813  /* Scan options */  /* Scan options */
814    
815  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
# Line 519  while (argc > 1 && argv[op][0] == '-') Line 818  while (argc > 1 && argv[op][0] == '-')
818    
819    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
820      showstore = 1;      showstore = 1;
821    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-q") == 0) quiet = 1;
822      else if (strcmp(argv[op], "-b") == 0) debug = 1;
823    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
824    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
825    #if !defined NODFA
826      else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
827    #endif
828    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&
829        ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),        ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),
830          *endptr == 0))          *endptr == 0))
# Line 529  while (argc > 1 && argv[op][0] == '-') Line 832  while (argc > 1 && argv[op][0] == '-')
832      op++;      op++;
833      argc--;      argc--;
834      }      }
835      else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
836        {
837        int both = argv[op][2] == 0;
838        int temp;
839        if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr),
840                         *endptr == 0))
841          {
842          timeitm = temp;
843          op++;
844          argc--;
845          }
846        else timeitm = LOOPREPEAT;
847        if (both) timeit = timeitm;
848        }
849      else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
850          ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
851            *endptr == 0))
852        {
853    #if defined(_WIN32) || defined(WIN32)
854        printf("PCRE: -S not supported on this OS\n");
855        exit(1);
856    #else
857        int rc;
858        struct rlimit rlim;
859        getrlimit(RLIMIT_STACK, &rlim);
860        rlim.rlim_cur = stack_size * 1024 * 1024;
861        rc = setrlimit(RLIMIT_STACK, &rlim);
862        if (rc != 0)
863          {
864        printf("PCRE: setrlimit() failed with error %d\n", rc);
865        exit(1);
866          }
867        op++;
868        argc--;
869    #endif
870        }
871  #if !defined NOPOSIX  #if !defined NOPOSIX
872    else if (strcmp(argv[op], "-p") == 0) posix = 1;    else if (strcmp(argv[op], "-p") == 0) posix = 1;
873  #endif  #endif
# Line 542  while (argc > 1 && argv[op][0] == '-') Line 881  while (argc > 1 && argv[op][0] == '-')
881      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
882      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
883      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
884      printf("  Newline character is %s\n", (rc == '\r')? "CR" : "LF");      printf("  Newline sequence is %s\n", (rc == '\r')? "CR" :
885          (rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" :
886          (rc == -2)? "ANYCRLF" :
887          (rc == -1)? "ANY" : "???");
888      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);
889      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
890      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
891      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
892      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);
893      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %d\n", rc);
894        (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &rc);
895        printf("  Default recursion depth limit = %d\n", rc);
896      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
897      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
898      exit(0);      goto EXIT;
899        }
900      else if (strcmp(argv[op], "-help") == 0 ||
901               strcmp(argv[op], "--help") == 0)
902        {
903        usage();
904        goto EXIT;
905      }      }
906    else    else
907      {      {
908      printf("** Unknown or malformed option %s\n", argv[op]);      printf("** Unknown or malformed option %s\n", argv[op]);
909      printf("Usage:   pcretest [-d] [-i] [-o <n>] [-p] [-s] [-t] [<input> [<output>]]\n");      usage();
910      printf("  -C     show PCRE compile-time options and exit\n");      yield = 1;
911      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;  
912      }      }
913    op++;    op++;
914    argc--;    argc--;
# Line 580  offsets = (int *)malloc(size_offsets_max Line 921  offsets = (int *)malloc(size_offsets_max
921  if (offsets == NULL)  if (offsets == NULL)
922    {    {
923    printf("** Failed to get %d bytes of memory for offsets vector\n",    printf("** Failed to get %d bytes of memory for offsets vector\n",
924      size_offsets_max * sizeof(int));      (int)(size_offsets_max * sizeof(int)));
925    return 1;    yield = 1;
926      goto EXIT;
927    }    }
928    
929  /* Sort out the input and output files */  /* Sort out the input and output files */
930    
931  if (argc > 1)  if (argc > 1)
932    {    {
933    infile = fopen(argv[op], "rb");    infile = fopen(argv[op], INPUT_MODE);
934    if (infile == NULL)    if (infile == NULL)
935      {      {
936      printf("** Failed to open %s\n", argv[op]);      printf("** Failed to open %s\n", argv[op]);
937      return 1;      yield = 1;
938        goto EXIT;
939      }      }
940    }    }
941    
942  if (argc > 2)  if (argc > 2)
943    {    {
944    outfile = fopen(argv[op+1], "wb");    outfile = fopen(argv[op+1], OUTPUT_MODE);
945    if (outfile == NULL)    if (outfile == NULL)
946      {      {
947      printf("** Failed to open %s\n", argv[op+1]);      printf("** Failed to open %s\n", argv[op+1]);
948      return 1;      yield = 1;
949        goto EXIT;
950      }      }
951    }    }
952    
# Line 613  pcre_free = new_free; Line 957  pcre_free = new_free;
957  pcre_stack_malloc = stack_malloc;  pcre_stack_malloc = stack_malloc;
958  pcre_stack_free = stack_free;  pcre_stack_free = stack_free;
959    
960  /* Heading line, then prompt for first regex if stdin */  /* Heading line unless quiet, then prompt for first regex if stdin */
961    
962  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  if (!quiet) fprintf(outfile, "PCRE version %s\n\n", pcre_version());
963    
964  /* Main loop */  /* Main loop */
965    
# Line 642  while (!done) Line 986  while (!done)
986    int do_showinfo = showinfo;    int do_showinfo = showinfo;
987    int do_showrest = 0;    int do_showrest = 0;
988    int do_flip = 0;    int do_flip = 0;
989    int erroroffset, len, delimiter;    int erroroffset, len, delimiter, poffset;
990    
991    use_utf8 = 0;    use_utf8 = 0;
992      debug_lengths = 1;
993    
994    if (infile == stdin) printf("  re> ");    if (infile == stdin) printf("  re> ");
995    if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break;    if (extend_inputline(infile, buffer) == NULL) break;
996    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
997    fflush(outfile);    fflush(outfile);
998    
# Line 659  while (!done) Line 1004  while (!done)
1004    
1005    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
1006      {      {
1007      unsigned long int magic;      unsigned long int magic, get_options;
1008      uschar sbuf[8];      uschar sbuf[8];
1009      FILE *f;      FILE *f;
1010    
# Line 707  while (!done) Line 1052  while (!done)
1052    
1053      /* Need to know if UTF-8 for printing data strings */      /* Need to know if UTF-8 for printing data strings */
1054    
1055      new_info(re, NULL, PCRE_INFO_OPTIONS, &options);      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1056      use_utf8 = (options & PCRE_UTF8) != 0;      use_utf8 = (get_options & PCRE_UTF8) != 0;
1057    
1058      /* Now see if there is any following study data */      /* Now see if there is any following study data */
1059    
# Line 752  while (!done) Line 1097  while (!done)
1097      }      }
1098    
1099    pp = p;    pp = p;
1100      poffset = p - buffer;
1101    
1102    for(;;)    for(;;)
1103      {      {
# Line 762  while (!done) Line 1108  while (!done)
1108        pp++;        pp++;
1109        }        }
1110      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;  
       }  
   
1111      if (infile == stdin) printf("    > ");      if (infile == stdin) printf("    > ");
1112      if (fgets((char *)pp, len, infile) == NULL)      if ((pp = extend_inputline(infile, pp)) == NULL)
1113        {        {
1114        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
1115        done = 1;        done = 1;
# Line 780  while (!done) Line 1118  while (!done)
1118      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
1119      }      }
1120    
1121      /* The buffer may have moved while being extended; reset the start of data
1122      pointer to the correct relative point in the buffer. */
1123    
1124      p = buffer + poffset;
1125    
1126    /* If the first character after the delimiter is backslash, make    /* If the first character after the delimiter is backslash, make
1127    the pattern end with backslash. This is purely to provide a way    the pattern end with backslash. This is purely to provide a way
1128    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 1145  while (!done)
1145      {      {
1146      switch (*pp++)      switch (*pp++)
1147        {        {
1148          case 'f': options |= PCRE_FIRSTLINE; break;
1149        case 'g': do_g = 1; break;        case 'g': do_g = 1; break;
1150        case 'i': options |= PCRE_CASELESS; break;        case 'i': options |= PCRE_CASELESS; break;
1151        case 'm': options |= PCRE_MULTILINE; break;        case 'm': options |= PCRE_MULTILINE; break;
# Line 810  while (!done) Line 1154  while (!done)
1154    
1155        case '+': do_showrest = 1; break;        case '+': do_showrest = 1; break;
1156        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
1157          case 'B': do_debug = 1; break;
1158        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
1159        case 'D': do_debug = do_showinfo = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
1160        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
1161        case 'F': do_flip = 1; break;        case 'F': do_flip = 1; break;
1162        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1163        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1164          case 'J': options |= PCRE_DUPNAMES; break;
1165        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1166        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1167    
# Line 826  while (!done) Line 1172  while (!done)
1172        case 'S': do_study = 1; break;        case 'S': do_study = 1; break;
1173        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1174        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
1175          case 'Z': debug_lengths = 0; break;
1176        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1177        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
1178    
1179        case 'L':        case 'L':
1180        ppp = pp;        ppp = pp;
1181        while (*ppp != '\n' && *ppp != ' ') ppp++;        /* The '\r' test here is so that it works on Windows. */
1182          /* The '0' test is just in case this is an unterminated line. */
1183          while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
1184        *ppp = 0;        *ppp = 0;
1185        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
1186          {          {
1187          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
1188          goto SKIP_DATA;          goto SKIP_DATA;
1189          }          }
1190          locale_set = 1;
1191        tables = pcre_maketables();        tables = pcre_maketables();
1192        pp = ppp;        pp = ppp;
1193        break;        break;
# Line 849  while (!done) Line 1199  while (!done)
1199        *pp = 0;        *pp = 0;
1200        break;        break;
1201    
1202        case '\n': case ' ': break;        case '<':
1203            {
1204            int x = check_newline(pp, outfile);
1205            if (x == 0) goto SKIP_DATA;
1206            options |= x;
1207            while (*pp++ != '>');
1208            }
1209          break;
1210    
1211          case '\r':                      /* So that it works in Windows */
1212          case '\n':
1213          case ' ':
1214          break;
1215    
1216        default:        default:
1217        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);
# Line 869  while (!done) Line 1231  while (!done)
1231    
1232      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
1233      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
1234        if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1235        if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1236        if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1237    
1238      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1239    
1240      /* 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 1242  while (!done)
1242    
1243      if (rc != 0)      if (rc != 0)
1244        {        {
1245        (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);        (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1246        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
1247        goto SKIP_DATA;        goto SKIP_DATA;
1248        }        }
# Line 888  while (!done) Line 1254  while (!done)
1254  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1255    
1256      {      {
1257      if (timeit)      if (timeit > 0)
1258        {        {
1259        register int i;        register int i;
1260        clock_t time_taken;        clock_t time_taken;
1261        clock_t start_time = clock();        clock_t start_time = clock();
1262        for (i = 0; i < LOOPREPEAT; i++)        for (i = 0; i < timeit; i++)
1263          {          {
1264          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1265          if (re != NULL) free(re);          if (re != NULL) free(re);
1266          }          }
1267        time_taken = clock() - start_time;        time_taken = clock() - start_time;
1268        fprintf(outfile, "Compile time %.3f milliseconds\n",        fprintf(outfile, "Compile time %.4f milliseconds\n",
1269          (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /          (((double)time_taken * 1000.0) / (double)timeit) /
1270            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
1271        }        }
1272    
# Line 917  while (!done) Line 1283  while (!done)
1283          {          {
1284          for (;;)          for (;;)
1285            {            {
1286            if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)            if (extend_inputline(infile, buffer) == NULL)
1287              {              {
1288              done = 1;              done = 1;
1289              goto CONTINUE;              goto CONTINUE;
# Line 952  while (!done) Line 1318  while (!done)
1318    
1319      if (do_study)      if (do_study)
1320        {        {
1321        if (timeit)        if (timeit > 0)
1322          {          {
1323          register int i;          register int i;
1324          clock_t time_taken;          clock_t time_taken;
1325          clock_t start_time = clock();          clock_t start_time = clock();
1326          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeit; i++)
1327            extra = pcre_study(re, study_options, &error);            extra = pcre_study(re, study_options, &error);
1328          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1329          if (extra != NULL) free(extra);          if (extra != NULL) free(extra);
1330          fprintf(outfile, "  Study time %.3f milliseconds\n",          fprintf(outfile, "  Study time %.4f milliseconds\n",
1331            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeit) /
1332              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
1333          }          }
1334        extra = pcre_study(re, study_options, &error);        extra = pcre_study(re, study_options, &error);
# Line 1005  while (!done) Line 1371  while (!done)
1371    
1372      SHOW_INFO:      SHOW_INFO:
1373    
1374        if (do_debug)
1375          {
1376          fprintf(outfile, "------------------------------------------------------------------\n");
1377          pcre_printint(re, outfile, debug_lengths);
1378          }
1379    
1380      if (do_showinfo)      if (do_showinfo)
1381        {        {
1382        unsigned long int get_options, all_options;        unsigned long int get_options, all_options;
1383    #if !defined NOINFOCHECK
1384        int old_first_char, old_options, old_count;        int old_first_char, old_options, old_count;
1385        int count, backrefmax, first_char, need_char;  #endif
1386          int count, backrefmax, first_char, need_char, okpartial, jchanged,
1387            hascrorlf;
1388        int nameentrysize, namecount;        int nameentrysize, namecount;
1389        const uschar *nametable;        const uschar *nametable;
1390    
       if (do_debug)  
         {  
         fprintf(outfile, "------------------------------------------------------------------\n");  
         print_internals(re, outfile);  
         }  
   
1391        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1392        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1393        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
# Line 1028  while (!done) Line 1397  while (!done)
1397        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);
1398        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);
1399        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);
1400          new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial);
1401          new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged);
1402          new_info(re, NULL, PCRE_INFO_HASCRORLF, &hascrorlf);
1403    
1404    #if !defined NOINFOCHECK
1405        old_count = pcre_info(re, &old_options, &old_first_char);        old_count = pcre_info(re, &old_options, &old_first_char);
1406        if (count < 0) fprintf(outfile,        if (count < 0) fprintf(outfile,
1407          "Error %d from pcre_info()\n", count);          "Error %d from pcre_info()\n", count);
# Line 1046  while (!done) Line 1419  while (!done)
1419            "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",            "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",
1420              get_options, old_options);              get_options, old_options);
1421          }          }
1422    #endif
1423    
1424        if (size != regex_gotten_store) fprintf(outfile,        if (size != regex_gotten_store) fprintf(outfile,
1425          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",
1426          size, regex_gotten_store);          (int)size, (int)regex_gotten_store);
1427    
1428        fprintf(outfile, "Capturing subpattern count = %d\n", count);        fprintf(outfile, "Capturing subpattern count = %d\n", count);
1429        if (backrefmax > 0)        if (backrefmax > 0)
# Line 1067  while (!done) Line 1441  while (!done)
1441            }            }
1442          }          }
1443    
1444        /* The NOPARTIAL bit is a private bit in the options, so we have        if (!okpartial) fprintf(outfile, "Partial matching not supported\n");
1445        to fish it out via out back door */        if (hascrorlf) fprintf(outfile, "Contains explicit CR or LF match\n");
1446    
1447        all_options = ((real_pcre *)re)->options;        all_options = ((real_pcre *)re)->options;
1448        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");  
1449    
1450        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1451          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",
1452            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1453            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1454            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
1455            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",
1456              ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",
1457            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",
1458            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
1459            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
1460            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1461              ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1462            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1463            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "");            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1464              ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1465    
1466          if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
1467    
1468          switch (get_options & PCRE_NEWLINE_BITS)
1469            {
1470            case PCRE_NEWLINE_CR:
1471            fprintf(outfile, "Forced newline sequence: CR\n");
1472            break;
1473    
1474            case PCRE_NEWLINE_LF:
1475            fprintf(outfile, "Forced newline sequence: LF\n");
1476            break;
1477    
1478            case PCRE_NEWLINE_CRLF:
1479            fprintf(outfile, "Forced newline sequence: CRLF\n");
1480            break;
1481    
1482        if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0)          case PCRE_NEWLINE_ANYCRLF:
1483          fprintf(outfile, "Case state changes\n");          fprintf(outfile, "Forced newline sequence: ANYCRLF\n");
1484            break;
1485    
1486            case PCRE_NEWLINE_ANY:
1487            fprintf(outfile, "Forced newline sequence: ANY\n");
1488            break;
1489    
1490            default:
1491            break;
1492            }
1493    
1494        if (first_char == -1)        if (first_char == -1)
1495          {          {
1496          fprintf(outfile, "First char at start or follows \\n\n");          fprintf(outfile, "First char at start or follows newline\n");
1497          }          }
1498        else if (first_char < 0)        else if (first_char < 0)
1499          {          {
# Line 1108  while (!done) Line 1504  while (!done)
1504          int ch = first_char & 255;          int ch = first_char & 255;
1505          const char *caseless = ((first_char & REQ_CASELESS) == 0)?          const char *caseless = ((first_char & REQ_CASELESS) == 0)?
1506            "" : " (caseless)";            "" : " (caseless)";
1507          if (isprint(ch))          if (PRINTHEX(ch))
1508            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);
1509          else          else
1510            fprintf(outfile, "First char = %d%s\n", ch, caseless);            fprintf(outfile, "First char = %d%s\n", ch, caseless);
# Line 1123  while (!done) Line 1519  while (!done)
1519          int ch = need_char & 255;          int ch = need_char & 255;
1520          const char *caseless = ((need_char & REQ_CASELESS) == 0)?          const char *caseless = ((need_char & REQ_CASELESS) == 0)?
1521            "" : " (caseless)";            "" : " (caseless)";
1522          if (isprint(ch))          if (PRINTHEX(ch))
1523            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);
1524          else          else
1525            fprintf(outfile, "Need char = %d%s\n", ch, caseless);            fprintf(outfile, "Need char = %d%s\n", ch, caseless);
# Line 1159  while (!done) Line 1555  while (!done)
1555                    fprintf(outfile, "\n  ");                    fprintf(outfile, "\n  ");
1556                    c = 2;                    c = 2;
1557                    }                    }
1558                  if (isprint(i) && i != ' ')                  if (PRINTHEX(i) && i != ' ')
1559                    {                    {
1560                    fprintf(outfile, "%c ", i);                    fprintf(outfile, "%c ", i);
1561                    c += 2;                    c += 2;
# Line 1218  while (!done) Line 1614  while (!done)
1614                  strerror(errno));                  strerror(errno));
1615                }                }
1616              else fprintf(outfile, "Study data written to %s\n", to_file);              else fprintf(outfile, "Study data written to %s\n", to_file);
1617    
1618              }              }
1619            }            }
1620          fclose(f);          fclose(f);
1621          }          }
1622    
1623          new_free(re);
1624          if (extra != NULL) new_free(extra);
1625          if (tables != NULL) new_free((void *)tables);
1626        continue;  /* With next regex */        continue;  /* With next regex */
1627        }        }
1628      }        /* End of non-POSIX compile */      }        /* End of non-POSIX compile */
# Line 1230  while (!done) Line 1631  while (!done)
1631    
1632    for (;;)    for (;;)
1633      {      {
1634      unsigned char *q;      uschar *q;
1635      unsigned char *bptr = dbuffer;      uschar *bptr;
1636      int *use_offsets = offsets;      int *use_offsets = offsets;
1637      int use_size_offsets = size_offsets;      int use_size_offsets = size_offsets;
1638      int callout_data = 0;      int callout_data = 0;
# Line 1244  while (!done) Line 1645  while (!done)
1645      int gmatched = 0;      int gmatched = 0;
1646      int start_offset = 0;      int start_offset = 0;
1647      int g_notempty = 0;      int g_notempty = 0;
1648        int use_dfa = 0;
1649    
1650      options = 0;      options = 0;
1651    
1652        *copynames = 0;
1653        *getnames = 0;
1654    
1655        copynamesptr = copynames;
1656        getnamesptr = getnames;
1657    
1658      pcre_callout = callout;      pcre_callout = callout;
1659      first_callout = 1;      first_callout = 1;
1660      callout_extra = 0;      callout_extra = 0;
# Line 1255  while (!done) Line 1663  while (!done)
1663      callout_fail_id = -1;      callout_fail_id = -1;
1664      show_malloc = 0;      show_malloc = 0;
1665    
1666      if (infile == stdin) printf("data> ");      if (extra != NULL) extra->flags &=
1667      if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)        ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
1668    
1669        len = 0;
1670        for (;;)
1671        {        {
1672        done = 1;        if (infile == stdin) printf("data> ");
1673        goto CONTINUE;        if (extend_inputline(infile, buffer + len) == NULL)
1674            {
1675            if (len > 0) break;
1676            done = 1;
1677            goto CONTINUE;
1678            }
1679          if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1680          len = (int)strlen((char *)buffer);
1681          if (buffer[len-1] == '\n') break;
1682        }        }
     if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);  
1683    
     len = (int)strlen((char *)buffer);  
1684      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
1685      buffer[len] = 0;      buffer[len] = 0;
1686      if (len == 0) break;      if (len == 0) break;
# Line 1271  while (!done) Line 1688  while (!done)
1688      p = buffer;      p = buffer;
1689      while (isspace(*p)) p++;      while (isspace(*p)) p++;
1690    
1691      q = dbuffer;      bptr = q = dbuffer;
1692      while ((c = *p++) != 0)      while ((c = *p++) != 0)
1693        {        {
1694        int i = 0;        int i = 0;
# Line 1293  while (!done) Line 1710  while (!done)
1710          c -= '0';          c -= '0';
1711          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')
1712            c = c * 8 + *p++ - '0';            c = c * 8 + *p++ - '0';
1713    
1714    #if !defined NOUTF8
1715            if (use_utf8 && c > 255)
1716              {
1717              unsigned char buff8[8];
1718              int ii, utn;
1719              utn = ord2utf8(c, buff8);
1720              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1721              c = buff8[ii];   /* Last byte */
1722              }
1723    #endif
1724          break;          break;
1725    
1726          case 'x':          case 'x':
1727    
1728          /* Handle \x{..} specially - new Perl thing for utf8 */          /* Handle \x{..} specially - new Perl thing for utf8 */
1729    
1730    #if !defined NOUTF8
1731          if (*p == '{')          if (*p == '{')
1732            {            {
1733            unsigned char *pt = p;            unsigned char *pt = p;
# Line 1317  while (!done) Line 1746  while (!done)
1746              }              }
1747            /* Not correct form; fall through */            /* Not correct form; fall through */
1748            }            }
1749    #endif
1750    
1751          /* Ordinary \x */          /* Ordinary \x */
1752    
# Line 1352  while (!done) Line 1782  while (!done)
1782            }            }
1783          else if (isalnum(*p))          else if (isalnum(*p))
1784            {            {
1785            uschar name[256];            uschar *npp = copynamesptr;
           uschar *npp = name;  
1786            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1787              *npp++ = 0;
1788            *npp = 0;            *npp = 0;
1789            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)copynamesptr);
1790            if (n < 0)            if (n < 0)
1791              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", copynamesptr);
1792            else copystrings |= 1 << n;            copynamesptr = npp;
1793            }            }
1794          else if (*p == '+')          else if (*p == '+')
1795            {            {
# Line 1397  while (!done) Line 1827  while (!done)
1827            }            }
1828          continue;          continue;
1829    
1830    #if !defined NODFA
1831            case 'D':
1832    #if !defined NOPOSIX
1833            if (posix || do_posix)
1834              printf("** Can't use dfa matching in POSIX mode: \\D ignored\n");
1835            else
1836    #endif
1837              use_dfa = 1;
1838            continue;
1839    
1840            case 'F':
1841            options |= PCRE_DFA_SHORTEST;
1842            continue;
1843    #endif
1844    
1845          case 'G':          case 'G':
1846          if (isdigit(*p))          if (isdigit(*p))
1847            {            {
# Line 1405  while (!done) Line 1850  while (!done)
1850            }            }
1851          else if (isalnum(*p))          else if (isalnum(*p))
1852            {            {
1853            uschar name[256];            uschar *npp = getnamesptr;
           uschar *npp = name;  
1854            while (isalnum(*p)) *npp++ = *p++;            while (isalnum(*p)) *npp++ = *p++;
1855              *npp++ = 0;
1856            *npp = 0;            *npp = 0;
1857            n = pcre_get_stringnumber(re, (char *)name);            n = pcre_get_stringnumber(re, (char *)getnamesptr);
1858            if (n < 0)            if (n < 0)
1859              fprintf(outfile, "no parentheses with name \"%s\"\n", name);              fprintf(outfile, "no parentheses with name \"%s\"\n", getnamesptr);
1860            else getstrings |= 1 << n;            getnamesptr = npp;
1861            }            }
1862          continue;          continue;
1863    
# Line 1438  while (!done) Line 1883  while (!done)
1883            if (offsets == NULL)            if (offsets == NULL)
1884              {              {
1885              printf("** Failed to get %d bytes of memory for offsets vector\n",              printf("** Failed to get %d bytes of memory for offsets vector\n",
1886                size_offsets_max * sizeof(int));                (int)(size_offsets_max * sizeof(int)));
1887              return 1;              yield = 1;
1888                goto EXIT;
1889              }              }
1890            }            }
1891          use_size_offsets = n;          use_size_offsets = n;
# Line 1450  while (!done) Line 1896  while (!done)
1896          options |= PCRE_PARTIAL;          options |= PCRE_PARTIAL;
1897          continue;          continue;
1898    
1899            case 'Q':
1900            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1901            if (extra == NULL)
1902              {
1903              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1904              extra->flags = 0;
1905              }
1906            extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
1907            extra->match_limit_recursion = n;
1908            continue;
1909    
1910            case 'q':
1911            while(isdigit(*p)) n = n * 10 + *p++ - '0';
1912            if (extra == NULL)
1913              {
1914              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1915              extra->flags = 0;
1916              }
1917            extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
1918            extra->match_limit = n;
1919            continue;
1920    
1921    #if !defined NODFA
1922            case 'R':
1923            options |= PCRE_DFA_RESTART;
1924            continue;
1925    #endif
1926    
1927          case 'S':          case 'S':
1928          show_malloc = 1;          show_malloc = 1;
1929          continue;          continue;
# Line 1461  while (!done) Line 1935  while (!done)
1935          case '?':          case '?':
1936          options |= PCRE_NO_UTF8_CHECK;          options |= PCRE_NO_UTF8_CHECK;
1937          continue;          continue;
1938    
1939            case '<':
1940              {
1941              int x = check_newline(p, outfile);
1942              if (x == 0) goto NEXT_DATA;
1943              options |= x;
1944              while (*p++ != '>');
1945              }
1946            continue;
1947          }          }
1948        *q++ = c;        *q++ = c;
1949        }        }
1950      *q = 0;      *q = 0;
1951      len = q - dbuffer;      len = q - dbuffer;
1952    
1953        if ((all_use_dfa || use_dfa) && find_match_limit)
1954          {
1955          printf("**Match limit not relevant for DFA matching: ignored\n");
1956          find_match_limit = 0;
1957          }
1958    
1959      /* Handle matching via the POSIX interface, which does not      /* Handle matching via the POSIX interface, which does not
1960      support timing or playing with the match limit or callout data. */      support timing or playing with the match limit or callout data. */
1961    
# Line 1485  while (!done) Line 1974  while (!done)
1974    
1975        if (rc != 0)        if (rc != 0)
1976          {          {
1977          (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);          (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1978          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
1979          }          }
1980          else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
1981                  != 0)
1982            {
1983            fprintf(outfile, "Matched with REG_NOSUB\n");
1984            }
1985        else        else
1986          {          {
1987          size_t i;          size_t i;
# Line 1519  while (!done) Line 2013  while (!done)
2013    
2014      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
2015        {        {
2016        if (timeit)        if (timeitm > 0)
2017          {          {
2018          register int i;          register int i;
2019          clock_t time_taken;          clock_t time_taken;
2020          clock_t start_time = clock();          clock_t start_time = clock();
2021          for (i = 0; i < LOOPREPEAT; i++)  
2022    #if !defined NODFA
2023            if (all_use_dfa || use_dfa)
2024              {
2025              int workspace[1000];
2026              for (i = 0; i < timeitm; i++)
2027                count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
2028                  options | g_notempty, use_offsets, use_size_offsets, workspace,
2029                  sizeof(workspace)/sizeof(int));
2030              }
2031            else
2032    #endif
2033    
2034            for (i = 0; i < timeitm; i++)
2035            count = pcre_exec(re, extra, (char *)bptr, len,            count = pcre_exec(re, extra, (char *)bptr, len,
2036              start_offset, options | g_notempty, use_offsets, use_size_offsets);              start_offset, options | g_notempty, use_offsets, use_size_offsets);
2037    
2038          time_taken = clock() - start_time;          time_taken = clock() - start_time;
2039          fprintf(outfile, "Execute time %.3f milliseconds\n",          fprintf(outfile, "Execute time %.4f milliseconds\n",
2040            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            (((double)time_taken * 1000.0) / (double)timeitm) /
2041              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
2042          }          }
2043    
2044        /* 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
2045        varying limits in order to find the minimum value. */        varying limits in order to find the minimum value for the match limit and
2046          for the recursion limit. */
2047    
2048        if (find_match_limit)        if (find_match_limit)
2049          {          {
         int min = 0;  
         int mid = 64;  
         int max = -1;  
   
2050          if (extra == NULL)          if (extra == NULL)
2051            {            {
2052            extra = (pcre_extra *)malloc(sizeof(pcre_extra));            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2053            extra->flags = 0;            extra->flags = 0;
2054            }            }
         extra->flags |= PCRE_EXTRA_MATCH_LIMIT;  
   
         for (;;)  
           {  
           extra->match_limit = mid;  
           count = pcre_exec(re, extra, (char *)bptr, len, start_offset,  
             options | g_notempty, use_offsets, use_size_offsets);  
           if (count == PCRE_ERROR_MATCHLIMIT)  
             {  
             /* fprintf(outfile, "Testing match limit = %d\n", mid); */  
             min = mid;  
             mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;  
             }  
           else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||  
                                  count == PCRE_ERROR_PARTIAL)  
             {  
             if (mid == min + 1)  
               {  
               fprintf(outfile, "Minimum match limit = %d\n", mid);  
               break;  
               }  
             /* fprintf(outfile, "Testing match limit = %d\n", mid); */  
             max = mid;  
             mid = (min + mid)/2;  
             }  
           else break;    /* Some other error */  
           }  
2055    
2056          extra->flags &= ~PCRE_EXTRA_MATCH_LIMIT;          (void)check_match_limit(re, extra, bptr, len, start_offset,
2057              options|g_notempty, use_offsets, use_size_offsets,
2058              PCRE_EXTRA_MATCH_LIMIT, &(extra->match_limit),
2059              PCRE_ERROR_MATCHLIMIT, "match()");
2060    
2061            count = check_match_limit(re, extra, bptr, len, start_offset,
2062              options|g_notempty, use_offsets, use_size_offsets,
2063              PCRE_EXTRA_MATCH_LIMIT_RECURSION, &(extra->match_limit_recursion),
2064              PCRE_ERROR_RECURSIONLIMIT, "match() recursion");
2065          }          }
2066    
2067        /* 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 2083  while (!done)
2083        /* 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
2084        value of match_limit. */        value of match_limit. */
2085    
2086        else  #if !defined NODFA
2087          else if (all_use_dfa || use_dfa)
2088          {          {
2089          count = pcre_exec(re, extra, (char *)bptr, len,          int workspace[1000];
2090            start_offset, options | g_notempty, use_offsets, use_size_offsets);          count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
2091              options | g_notempty, use_offsets, use_size_offsets, workspace,
2092              sizeof(workspace)/sizeof(int));
2093            if (count == 0)
2094              {
2095              fprintf(outfile, "Matched, but too many subsidiary matches\n");
2096              count = use_size_offsets/2;
2097              }
2098          }          }
2099    #endif
2100    
2101        if (count == 0)        else
2102          {          {
2103          fprintf(outfile, "Matched, but too many substrings\n");          count = pcre_exec(re, extra, (char *)bptr, len,
2104          count = use_size_offsets/3;            start_offset, options | g_notempty, use_offsets, use_size_offsets);
2105            if (count == 0)
2106              {
2107              fprintf(outfile, "Matched, but too many substrings\n");
2108              count = use_size_offsets/3;
2109              }
2110          }          }
2111    
2112        /* Matched */        /* Matched */
2113    
2114        if (count >= 0)        if (count >= 0)
2115          {          {
2116          int i;          int i, maxcount;
2117    
2118    #if !defined NODFA
2119            if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else
2120    #endif
2121              maxcount = use_size_offsets/3;
2122    
2123            /* This is a check against a lunatic return value. */
2124    
2125            if (count > maxcount)
2126              {
2127              fprintf(outfile,
2128                "** PCRE error: returned count %d is too big for offset size %d\n",
2129                count, use_size_offsets);
2130              count = use_size_offsets/3;
2131              if (do_g || do_G)
2132                {
2133                fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G');
2134                do_g = do_G = FALSE;        /* Break g/G loop */
2135                }
2136              }
2137    
2138          for (i = 0; i < count * 2; i += 2)          for (i = 0; i < count * 2; i += 2)
2139            {            {
2140            if (use_offsets[i] < 0)            if (use_offsets[i] < 0)
# Line 1641  while (!done) Line 2162  while (!done)
2162            {            {
2163            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
2164              {              {
2165              char copybuffer[16];              char copybuffer[256];
2166              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,              int rc = pcre_copy_substring((char *)bptr, use_offsets, count,
2167                i, copybuffer, sizeof(copybuffer));                i, copybuffer, sizeof(copybuffer));
2168              if (rc < 0)              if (rc < 0)
# Line 1651  while (!done) Line 2172  while (!done)
2172              }              }
2173            }            }
2174    
2175            for (copynamesptr = copynames;
2176                 *copynamesptr != 0;
2177                 copynamesptr += (int)strlen((char*)copynamesptr) + 1)
2178              {
2179              char copybuffer[256];
2180              int rc = pcre_copy_named_substring(re, (char *)bptr, use_offsets,
2181                count, (char *)copynamesptr, copybuffer, sizeof(copybuffer));
2182              if (rc < 0)
2183                fprintf(outfile, "copy substring %s failed %d\n", copynamesptr, rc);
2184              else
2185                fprintf(outfile, "  C %s (%d) %s\n", copybuffer, rc, copynamesptr);
2186              }
2187    
2188          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2189            {            {
2190            if ((getstrings & (1 << i)) != 0)            if ((getstrings & (1 << i)) != 0)
# Line 1663  while (!done) Line 2197  while (!done)
2197              else              else
2198                {                {
2199                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);                fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
               /* free((void *)substring); */  
2200                pcre_free_substring(substring);                pcre_free_substring(substring);
2201                }                }
2202              }              }
2203            }            }
2204    
2205            for (getnamesptr = getnames;
2206                 *getnamesptr != 0;
2207                 getnamesptr += (int)strlen((char*)getnamesptr) + 1)
2208              {
2209              const char *substring;
2210              int rc = pcre_get_named_substring(re, (char *)bptr, use_offsets,
2211                count, (char *)getnamesptr, &substring);
2212              if (rc < 0)
2213                fprintf(outfile, "copy substring %s failed %d\n", getnamesptr, rc);
2214              else
2215                {
2216                fprintf(outfile, "  G %s (%d) %s\n", substring, rc, getnamesptr);
2217                pcre_free_substring(substring);
2218                }
2219              }
2220    
2221          if (getlist)          if (getlist)
2222            {            {
2223            const char **stringlist;            const char **stringlist;
# Line 1692  while (!done) Line 2241  while (!done)
2241    
2242        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2243          {          {
2244          fprintf(outfile, "Partial match\n");          fprintf(outfile, "Partial match");
2245    #if !defined NODFA
2246            if ((all_use_dfa || use_dfa) && use_size_offsets > 2)
2247              fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],
2248                bptr + use_offsets[0]);
2249    #endif
2250            fprintf(outfile, "\n");
2251          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2252          }          }
2253    
2254        /* 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
2255        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
2256        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
2257        matching, the advance must be one character, not one byte. Fudge the        string - that was checked before setting g_notempty.
2258        offset values to achieve this. We won't be at the end of the string -  
2259        that was checked before setting g_notempty. */        Complication arises in the case when the newline option is "any" or
2260          "anycrlf". If the previous match was at the end of a line terminated by
2261          CRLF, an advance of one character just passes the \r, whereas we should
2262          prefer the longer newline sequence, as does the code in pcre_exec().
2263          Fudge the offset value to achieve this.
2264    
2265          Otherwise, in the case of UTF-8 matching, the advance must be one
2266          character, not one byte. */
2267    
2268        else        else
2269          {          {
2270          if (g_notempty != 0)          if (g_notempty != 0)
2271            {            {
2272            int onechar = 1;            int onechar = 1;
2273              unsigned int obits = ((real_pcre *)re)->options;
2274            use_offsets[0] = start_offset;            use_offsets[0] = start_offset;
2275            if (use_utf8)            if ((obits & PCRE_NEWLINE_BITS) == 0)
2276                {
2277                int d;
2278                (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);
2279                obits = (d == '\r')? PCRE_NEWLINE_CR :
2280                        (d == '\n')? PCRE_NEWLINE_LF :
2281                        (d == ('\r'<<8 | '\n'))? PCRE_NEWLINE_CRLF :
2282                        (d == -2)? PCRE_NEWLINE_ANYCRLF :
2283                        (d == -1)? PCRE_NEWLINE_ANY : 0;
2284                }
2285              if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2286                   (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2287                  &&
2288                  start_offset < len - 1 &&
2289                  bptr[start_offset] == '\r' &&
2290                  bptr[start_offset+1] == '\n')
2291                onechar++;
2292              else if (use_utf8)
2293              {              {
2294              while (start_offset + onechar < len)              while (start_offset + onechar < len)
2295                {                {
# Line 1744  while (!done) Line 2324  while (!done)
2324        character. */        character. */
2325    
2326        g_notempty = 0;        g_notempty = 0;
2327    
2328        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2329          {          {
2330          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
# Line 1762  while (!done) Line 2343  while (!done)
2343          len -= use_offsets[1];          len -= use_offsets[1];
2344          }          }
2345        }  /* End of loop for /g and /G */        }  /* End of loop for /g and /G */
2346    
2347        NEXT_DATA: continue;
2348      }    /* End of loop for data lines */      }    /* End of loop for data lines */
2349    
2350    CONTINUE:    CONTINUE:
# Line 1770  while (!done) Line 2353  while (!done)
2353    if (posix || do_posix) regfree(&preg);    if (posix || do_posix) regfree(&preg);
2354  #endif  #endif
2355    
2356    if (re != NULL) free(re);    if (re != NULL) new_free(re);
2357    if (extra != NULL) free(extra);    if (extra != NULL) new_free(extra);
2358    if (tables != NULL)    if (tables != NULL)
2359      {      {
2360      free((void *)tables);      new_free((void *)tables);
2361      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");
2362        locale_set = 0;
2363      }      }
2364    }    }
2365    
2366  if (infile == stdin) fprintf(outfile, "\n");  if (infile == stdin) fprintf(outfile, "\n");
2367  return 0;  
2368    EXIT:
2369    
2370    if (infile != NULL && infile != stdin) fclose(infile);
2371    if (outfile != NULL && outfile != stdout) fclose(outfile);
2372    
2373    free(buffer);
2374    free(dbuffer);
2375    free(pbuffer);
2376    free(offsets);
2377    
2378    return yield;
2379  }  }
2380    
2381  /* End */  /* End of pcretest.c */

Legend:
Removed from v.75  
changed lines
  Added in v.227

  ViewVC Help
Powered by ViewVC 1.1.5