/[pcre]/code/trunk/pcretest.c

Diff of /code/trunk/pcretest.c

Parent Directory | Revision Log | View Patch Patch

-revision 9 by nigel,
Sat Feb 24 21:38:13 2007 UTC
+revision 107 by ph10,
Wed Mar  7 11:02:28 2007 UTC
 Line 2
  *             PCRE testing program               *
  *************************************************/
+ /* This program was hacked up as a tester for PCRE. I really should have
+ written it more tidily in the first place. Will I ever learn? It has grown and
+ been extended and consequently is now rather, er, *very* untidy in places.
+ -----------------------------------------------------------------------------
+ Redistribution and use in source and binary forms, with or without
+ modification, are permitted provided that the following conditions are met:
+     * Redistributions of source code must retain the above copyright notice,
+       this list of conditions and the following disclaimer.
+     * Redistributions in binary form must reproduce the above copyright
+       notice, this list of conditions and the following disclaimer in the
+       documentation and/or other materials provided with the distribution.
+     * Neither the name of the University of Cambridge nor the names of its
+       contributors may be used to endorse or promote products derived from
+       this software without specific prior written permission.
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+ LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ POSSIBILITY OF SUCH DAMAGE.
+ -----------------------------------------------------------------------------
+ */
  #include <ctype.h>
  #include <stdio.h>
  #include <string.h>
  #include <stdlib.h>
  #include <time.h>
+ #include <locale.h>
+ #include <errno.h>
+ /* A number of things vary for Windows builds. Originally, pcretest opened its
+ input and output without "b"; then I was told that "b" was needed in some
+ environments, so it was added for release 5.0 to both the input and output. (It
+ makes no difference on Unix-like systems.) Later I was told that it is wrong
+ for the input on Windows. I've now abstracted the modes into two macros that
+ are set here, to make it easier to fiddle with them, and removed "b" from the
+ input mode under Windows. */
+ #if defined(_WIN32) || defined(WIN32)
+ #include <io.h>                /* For _setmode() */
+ #include <fcntl.h>             /* For _O_BINARY */
+ #define INPUT_MODE   "r"
+ #define OUTPUT_MODE  "wb"
+ #else
+ #include <sys/time.h>          /* These two includes are needed */
+ #include <sys/resource.h>      /* for setrlimit(). */
+ #define INPUT_MODE   "rb"
+ #define OUTPUT_MODE  "wb"
+ #endif
+ #define PCRE_SPY        /* For Win32 build, import data, not export */
+ /* We include pcre_internal.h because we need the internal info for displaying
+ the results of pcre_study() and we also need to know about the internal
+ macros, structures, and other internal data values; pcretest has "inside
+ information" compared to a program that strictly follows the PCRE API. */
+ #include "pcre_internal.h"
+ /* We need access to the data tables that PCRE uses. So as not to have to keep
+ two copies, we include the source file here, changing the names of the external
+ symbols to prevent clashes. */
+ #define _pcre_utf8_table1      utf8_table1
+ #define _pcre_utf8_table1_size utf8_table1_size
+ #define _pcre_utf8_table2      utf8_table2
+ #define _pcre_utf8_table3      utf8_table3
+ #define _pcre_utf8_table4      utf8_table4
+ #define _pcre_utt              utt
+ #define _pcre_utt_size         utt_size
+ #define _pcre_OP_lengths       OP_lengths
+ #include "pcre_tables.c"
- /* Use the internal info for displaying the results of pcre_study(). */
+ /* We also need the pcre_printint() function for printing out compiled
+ patterns. This function is in a separate file so that it can be included in
+ pcre_compile.c when that module is compiled with debugging enabled.
- #include "internal.h"
+ The definition of the macro PRINTABLE, which determines whether to print an
+ output character as-is or as a hex value when showing compiled patterns, is
+ contained in this file. We uses it here also, in cases when the locale has not
+ been explicitly changed, so as to get consistent output from systems that
+ differ in their output from isprint() even in the "C" locale. */
+ #include "pcre_printint.src"
+ #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
+ /* It is possible to compile this test program without including support for
+ testing the POSIX interface, though this is not available via the standard
+ Makefile. */
+ #if !defined NOPOSIX
  #include "pcreposix.h"
+ #endif
+ /* It is also possible, for the benefit of the version currently imported into
+ Exim, to build pcretest without support for UTF8 (define NOUTF8), without the
+ interface to the DFA matcher (NODFA), and without the doublecheck of the old
+ "info" function (define NOINFOCHECK). In fact, we automatically cut out the
+ UTF8 support if PCRE is built without it. */
+ #ifndef SUPPORT_UTF8
+ #ifndef NOUTF8
+ #define NOUTF8
+ #endif
+ #endif
+ /* Other parameters */
  #ifndef CLOCKS_PER_SEC
  #ifdef CLK_TCK
-Line 21
+Line 137
  #endif
  #endif
+ /* This is the default loop count for timing. */
+ #define LOOPREPEAT 500000
+ /* Static variables */
  static FILE *outfile;
  static int log_store = 0;
+ static int callout_count;
+ static int callout_extra;
+ static int callout_fail_count;
+ static int callout_fail_id;
+ static int first_callout;
+ static int locale_set = 0;
+ static int show_malloc;
+ static int use_utf8;
+ static size_t gotten_store;
+ /* The buffers grow automatically if very long input lines are encountered. */
+ static int buffer_size = 50000;
+ static uschar *buffer = NULL;
+ static uschar *dbuffer = NULL;
+ static uschar *pbuffer = NULL;
- /* Debugging function to print the internal form of the regex. This is the same
+ /*************************************************
- code as contained in pcre.c under the DEBUG macro. */
+ *        Read or extend an input line            *
+ *************************************************/
- static const char *OP_names[] = {
+ /* Input lines are read into buffer, but both patterns and data lines can be
-   "End", "\\A", "\\B", "\\b", "\\D", "\\d",
+ continued over multiple input lines. In addition, if the buffer fills up, we
-   "\\S", "\\s", "\\W", "\\w", "Cut", "\\Z", "^", "$", "Any", "chars",
+ want to automatically expand it so as to be able to handle extremely large
-   "not",
+ lines that are needed for certain stress tests. When the input buffer is
-   "*", "*?", "+", "+?", "?", "??", "{", "{", "{",
+ expanded, the other two buffers must also be expanded likewise, and the
-   "*", "*?", "+", "+?", "?", "??", "{", "{", "{",
+ contents of pbuffer, which are a copy of the input for callouts, must be
-   "*", "*?", "+", "+?", "?", "??", "{", "{", "{",
+ preserved (for when expansion happens for a data line). This is not the most
-   "*", "*?", "+", "+?", "?", "??", "{", "{",
+ optimal way of handling this, but hey, this is just a test program!
-   "class", "Ref",
-   "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", "Once",
+ Arguments:
-   "Brazero", "Braminzero", "Bra"
+   f            the file to read
- };
+   start        where in buffer to start (this *must* be within buffer)
+ Returns:       pointer to the start of new data
- static void print_internals(pcre *re)
+                could be a copy of start, or could be moved
- {
+                NULL if no data read and EOF reached
- unsigned char *code = ((real_pcre *)re)->code;
+ */
- printf("------------------------------------------------------------------\n");
- for(;;)
-   {
-   int c;
-   int charlength;
-   printf("%3d ", code - ((real_pcre *)re)->code);
-   if (*code >= OP_BRA)
-     {
-     printf("%3d Bra %d", (code[1] << 8) + code[2], *code - OP_BRA);
-     code += 2;
-     }
-   else switch(*code)
-     {
-     case OP_END:
-     printf("    %s\n", OP_names[*code]);
-     printf("------------------------------------------------------------------\n");
-     return;
-     case OP_CHARS:
-     charlength = *(++code);
-     printf("%3d ", charlength);
-     while (charlength-- > 0)
-       if (isprint(c = *(++code))) printf("%c", c); else printf("\\x%02x", c);
-     break;
-     case OP_KETRMAX:
-     case OP_KETRMIN:
-     case OP_ALT:
-     case OP_KET:
-     case OP_ASSERT:
-     case OP_ASSERT_NOT:
-     case OP_ONCE:
-     printf("%3d %s", (code[1] << 8) + code[2], OP_names[*code]);
-     code += 2;
-     break;
-     case OP_STAR:
-     case OP_MINSTAR:
-     case OP_PLUS:
-     case OP_MINPLUS:
-     case OP_QUERY:
-     case OP_MINQUERY:
-     case OP_TYPESTAR:
-     case OP_TYPEMINSTAR:
-     case OP_TYPEPLUS:
-     case OP_TYPEMINPLUS:
-     case OP_TYPEQUERY:
-     case OP_TYPEMINQUERY:
-     if (*code >= OP_TYPESTAR)
-       printf("    %s", OP_names[code[1]]);
-     else if (isprint(c = code[1])) printf("    %c", c);
-       else printf("    \\x%02x", c);
-     printf("%s", OP_names[*code++]);
-     break;
-     case OP_EXACT:
-     case OP_UPTO:
-     case OP_MINUPTO:
-     if (isprint(c = code[3])) printf("    %c{", c);
-       else printf("    \\x%02x{", c);
-     if (*code != OP_EXACT) printf(",");
-     printf("%d}", (code[1] << 8) + code[2]);
-     if (*code == OP_MINUPTO) printf("?");
-     code += 3;
-     break;
-     case OP_TYPEEXACT:
-     case OP_TYPEUPTO:
-     case OP_TYPEMINUPTO:
-     printf("    %s{", OP_names[code[3]]);
-     if (*code != OP_TYPEEXACT) printf(",");
-     printf("%d}", (code[1] << 8) + code[2]);
-     if (*code == OP_TYPEMINUPTO) printf("?");
-     code += 3;
-     break;
-     case OP_NOT:
-     if (isprint(c = *(++code))) printf("    [^%c]", c);
-       else printf("    [^\\x%02x]", c);
-     break;
-     case OP_NOTSTAR:
-     case OP_NOTMINSTAR:
-     case OP_NOTPLUS:
-     case OP_NOTMINPLUS:
-     case OP_NOTQUERY:
-     case OP_NOTMINQUERY:
-     if (isprint(c = code[1])) printf("    [^%c]", c);
-       else printf("    [^\\x%02x]", c);
-     printf("%s", OP_names[*code++]);
-     break;
-     case OP_NOTEXACT:
-     case OP_NOTUPTO:
-     case OP_NOTMINUPTO:
-     if (isprint(c = code[3])) printf("    [^%c]{", c);
-       else printf("    [^\\x%02x]{", c);
-     if (*code != OP_NOTEXACT) printf(",");
-     printf("%d}", (code[1] << 8) + code[2]);
-     if (*code == OP_NOTMINUPTO) printf("?");
-     code += 3;
-     break;
-     case OP_REF:
-     printf("    \\%d", *(++code));
-     code++;
-     goto CLASS_REF_REPEAT;
-     case OP_CLASS:
-       {
-       int i, min, max;
-       code++;
-       printf("    [");
-       for (i = 0; i < 256; i++)
-         {
-         if ((code[i/8] & (1 << (i&7))) != 0)
-           {
-           int j;
-           for (j = i+1; j < 256; j++)
-             if ((code[j/8] & (1 << (j&7))) == 0) break;
-           if (i == '-' || i == ']') printf("\\");
-           if (isprint(i)) printf("%c", i); else printf("\\x%02x", i);
-           if (--j > i)
-             {
-             printf("-");
-             if (j == '-' || j == ']') printf("\\");
-             if (isprint(j)) printf("%c", j); else printf("\\x%02x", j);
-             }
-           i = j;
-           }
-         }
-       printf("]");
-       code += 32;
-       CLASS_REF_REPEAT:
-       switch(*code)
-         {
-         case OP_CRSTAR:
-         case OP_CRMINSTAR:
-         case OP_CRPLUS:
-         case OP_CRMINPLUS:
-         case OP_CRQUERY:
-         case OP_CRMINQUERY:
-         printf("%s", OP_names[*code]);
-         break;
-         case OP_CRRANGE:
+ static uschar *
-         case OP_CRMINRANGE:
+ extend_inputline(FILE *f, uschar *start)
-         min = (code[1] << 8) + code[2];
+ {
-         max = (code[3] << 8) + code[4];
+ uschar *here = start;
-         if (max == 0) printf("{%d,}", min);
-         else printf("{%d,%d}", min, max);
-         if (*code == OP_CRMINRANGE) printf("?");
-         code += 4;
-         break;
-         default:
+ for (;;)
-         code--;
+   {
-         }
+   int rlen = buffer_size - (here - buffer);
+   if (rlen > 1000)
+     {
+     int dlen;
+     if (fgets((char *)here, rlen,  f) == NULL)
+       return (here == start)? NULL : start;
+     dlen = (int)strlen((char *)here);
+     if (dlen > 0 && here[dlen - 1] == '\n') return start;
+     here += dlen;
+     }
+   else
+     {
+     int new_buffer_size = 2*buffer_size;
+     uschar *new_buffer = (unsigned char *)malloc(new_buffer_size);
+     uschar *new_dbuffer = (unsigned char *)malloc(new_buffer_size);
+     uschar *new_pbuffer = (unsigned char *)malloc(new_buffer_size);
+     if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
+       {
+       fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
+       exit(1);
        }
-     break;
-     /* Anything else is just a one-node item */
+     memcpy(new_buffer, buffer, buffer_size);
+     memcpy(new_pbuffer, pbuffer, buffer_size);
+     buffer_size = new_buffer_size;
-     default:
+     start = new_buffer + (start - buffer);
-     printf("    %s", OP_names[*code]);
+     here = new_buffer + (here - buffer);
-     break;
+     free(buffer);
+     free(dbuffer);
+     free(pbuffer);
+     buffer = new_buffer;
+     dbuffer = new_dbuffer;
+     pbuffer = new_pbuffer;
      }
+   }
+ return NULL;  /* Control never gets here */
+ }
-   code++;
-   printf("\n");
+ /*************************************************
+ *          Read number from string               *
+ *************************************************/
+ /* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess
+ around with conditional compilation, just do the job by hand. It is only used
+ for unpicking arguments, so just keep it simple.
+ Arguments:
+   str           string to be converted
+   endptr        where to put the end pointer
+ Returns:        the unsigned long
+ */
+ static int
+ get_value(unsigned char *str, unsigned char **endptr)
+ {
+ int result = 0;
+ while(*str != 0 && isspace(*str)) str++;
+ while (isdigit(*str)) result = result * 10 + (int)(*str++ - '0');
+ *endptr = str;
+ return(result);
+ }
+ /*************************************************
+ *            Convert UTF-8 string to value       *
+ *************************************************/
+ /* This function takes one or more bytes that represents a UTF-8 character,
+ and returns the value of the character.
+ Argument:
+   utf8bytes   a pointer to the byte vector
+   vptr        a pointer to an int to receive the value
+ Returns:      >  0 => the number of bytes consumed
+               -6 to 0 => malformed UTF-8 character at offset = (-return)
+ */
+ #if !defined NOUTF8
+ static int
+ utf82ord(unsigned char *utf8bytes, int *vptr)
+ {
+ int c = *utf8bytes++;
+ int d = c;
+ int i, j, s;
+ for (i = -1; i < 6; i++)               /* i is number of additional bytes */
+   {
+   if ((d & 0x80) == 0) break;
+   d <<= 1;
+   }
+ if (i == -1) { *vptr = c; return 1; }  /* ascii character */
+ if (i == 0 || i == 6) return 0;        /* invalid UTF-8 */
+ /* i now has a value in the range 1-5 */
+ s = 6*i;
+ d = (c & utf8_table3[i]) << s;
+ for (j = 0; j < i; j++)
+   {
+   c = *utf8bytes++;
+   if ((c & 0xc0) != 0x80) return -(j+1);
+   s -= 6;
+   d |= (c & 0x3f) << s;
    }
+ /* Check that encoding was the correct unique one */
+ for (j = 0; j < utf8_table1_size; j++)
+   if (d <= utf8_table1[j]) break;
+ if (j != i) return -(i+1);
+ /* Valid value */
+ *vptr = d;
+ return i+1;
+ }
+ #endif
+ /*************************************************
+ *       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
+   utf8bytes  pointer to buffer for result - at least 6 bytes long
+ Returns:     number of characters placed in the buffer
+ */
+ #if !defined NOUTF8
+ static int
+ ord2utf8(int cvalue, uschar *utf8bytes)
+ {
+ register int i, j;
+ for (i = 0; i < utf8_table1_size; i++)
+   if (cvalue <= utf8_table1[i]) break;
+ utf8bytes += i;
+ for (j = i; j > 0; j--)
+  {
+  *utf8bytes-- = 0x80 | (cvalue & 0x3f);
+  cvalue >>= 6;
+  }
+ *utf8bytes = utf8_table2[i] | cvalue;
+ return i + 1;
  }
+ #endif
- /* Character string printing function. */
+ /*************************************************
+ *             Print character string             *
+ *************************************************/
- static void pchars(unsigned char *p, int length)
+ /* Character string printing function. Must handle UTF-8 strings in utf8
+ mode. Yields number of characters printed. If handed a NULL file, just counts
+ chars without printing. */
+ static int pchars(unsigned char *p, int length, FILE *f)
  {
- int c;
+ int c = 0;
+ int yield = 0;
  while (length-- > 0)
-   if (isprint(c = *(p++))) fprintf(outfile, "%c", c);
+   {
-     else fprintf(outfile, "\\x%02x", c);
+ #if !defined NOUTF8
+   if (use_utf8)
+     {
+     int rc = utf82ord(p, &c);
+     if (rc > 0 && rc <= length + 1)   /* Mustn't run over the end */
+       {
+       length -= rc - 1;
+       p += rc;
+       if (PRINTHEX(c))
+         {
+         if (f != NULL) fprintf(f, "%c", c);
+         yield++;
+         }
+       else
+         {
+         int n = 4;
+         if (f != NULL) fprintf(f, "\\x{%02x}", c);
+         yield += (n <= 0x000000ff)? 2 :
+                  (n <= 0x00000fff)? 3 :
+                  (n <= 0x0000ffff)? 4 :
+                  (n <= 0x000fffff)? 5 : 6;
+         }
+       continue;
+       }
+     }
+ #endif
+    /* Not UTF-8, or malformed UTF-8  */
+   c = *p++;
+   if (PRINTHEX(c))
+     {
+     if (f != NULL) fprintf(f, "%c", c);
+     yield++;
+     }
+   else
+     {
+     if (f != NULL) fprintf(f, "\\x%02x", c);
+     yield += 4;
+     }
+   }
+ return yield;
+ }
+ /*************************************************
+ *              Callout function                  *
+ *************************************************/
+ /* Called from PCRE as a result of the (?C) item. We print out where we are in
+ the match. Yield zero unless more callouts than the fail count, or the callout
+ data is not zero. */
+ static int callout(pcre_callout_block *cb)
+ {
+ FILE *f = (first_callout | callout_extra)? outfile : NULL;
+ int i, pre_start, post_start, subject_length;
+ if (callout_extra)
+   {
+   fprintf(f, "Callout %d: last capture = %d\n",
+     cb->callout_number, cb->capture_last);
+   for (i = 0; i < cb->capture_top * 2; i += 2)
+     {
+     if (cb->offset_vector[i] < 0)
+       fprintf(f, "%2d: <unset>\n", i/2);
+     else
+       {
+       fprintf(f, "%2d: ", i/2);
+       (void)pchars((unsigned char *)cb->subject + cb->offset_vector[i],
+         cb->offset_vector[i+1] - cb->offset_vector[i], f);
+       fprintf(f, "\n");
+       }
+     }
+   }
+ /* Re-print the subject in canonical form, the first time or if giving full
+ datails. On subsequent calls in the same match, we use pchars just to find the
+ printed lengths of the substrings. */
+ if (f != NULL) fprintf(f, "--->");
+ pre_start = pchars((unsigned char *)cb->subject, cb->start_match, f);
+ post_start = pchars((unsigned char *)(cb->subject + cb->start_match),
+   cb->current_position - cb->start_match, f);
+ subject_length = pchars((unsigned char *)cb->subject, cb->subject_length, NULL);
+ (void)pchars((unsigned char *)(cb->subject + cb->current_position),
+   cb->subject_length - cb->current_position, f);
+ if (f != NULL) fprintf(f, "\n");
+ /* Always print appropriate indicators, with callout number if not already
+ shown. For automatic callouts, show the pattern offset. */
+ if (cb->callout_number == 255)
+   {
+   fprintf(outfile, "%+3d ", cb->pattern_position);
+   if (cb->pattern_position > 99) fprintf(outfile, "\n    ");
+   }
+ else
+   {
+   if (callout_extra) fprintf(outfile, "    ");
+     else fprintf(outfile, "%3d ", cb->callout_number);
+   }
+ for (i = 0; i < pre_start; i++) fprintf(outfile, " ");
+ fprintf(outfile, "^");
+ if (post_start > 0)
+   {
+   for (i = 0; i < post_start - 1; i++) fprintf(outfile, " ");
+   fprintf(outfile, "^");
+   }
+ for (i = 0; i < subject_length - pre_start - post_start + 4; i++)
+   fprintf(outfile, " ");
+ fprintf(outfile, "%.*s", (cb->next_item_length == 0)? 1 : cb->next_item_length,
+   pbuffer + cb->pattern_position);
+ fprintf(outfile, "\n");
+ first_callout = 0;
+ if (cb->callout_data != NULL)
+   {
+   int callout_data = *((int *)(cb->callout_data));
+   if (callout_data != 0)
+     {
+     fprintf(outfile, "Callout data = %d\n", callout_data);
+     return callout_data;
+     }
+   }
+ return (cb->callout_number != callout_fail_id)? 0 :
+        (++callout_count >= callout_fail_count)? 1 : 0;
  }
+ /*************************************************
+ *            Local malloc functions              *
+ *************************************************/
  /* Alternative malloc function, to test functionality and show the size of the
  compiled re. */
  static void *new_malloc(size_t size)
  {
- if (log_store) fprintf(outfile, "Store size request: %d\n", (int)size);
+ void *block = malloc(size);
- return malloc(size);
+ gotten_store = size;
+ if (show_malloc)
+   fprintf(outfile, "malloc       %3d %p\n", (int)size, block);
+ return block;
+ }
+ static void new_free(void *block)
+ {
+ if (show_malloc)
+   fprintf(outfile, "free             %p\n", block);
+ free(block);
+ }
+ /* For recursion malloc/free, to test stacking calls */
+ static void *stack_malloc(size_t size)
+ {
+ void *block = malloc(size);
+ if (show_malloc)
+   fprintf(outfile, "stack_malloc %3d %p\n", (int)size, block);
+ return block;
+ }
+ static void stack_free(void *block)
+ {
+ if (show_malloc)
+   fprintf(outfile, "stack_free       %p\n", block);
+ free(block);
+ }
+ /*************************************************
+ *          Call pcre_fullinfo()                  *
+ *************************************************/
+ /* Get one piece of information from the pcre_fullinfo() function */
+ static void new_info(pcre *re, pcre_extra *study, int option, void *ptr)
+ {
+ int rc;
+ if ((rc = pcre_fullinfo(re, study, option, ptr)) < 0)
+   fprintf(outfile, "Error %d from pcre_fullinfo(%d)\n", rc, option);
+ }
+ /*************************************************
+ *         Byte flipping function                 *
+ *************************************************/
+ static unsigned long int
+ byteflip(unsigned long int value, int n)
+ {
+ if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);
+ return ((value & 0x000000ff) << 24) |
+        ((value & 0x0000ff00) <<  8) |
+        ((value & 0x00ff0000) >>  8) |
+        ((value & 0xff000000) >> 24);
+ }
+ /*************************************************
+ *        Check match or recursion limit          *
+ *************************************************/
+ static int
+ check_match_limit(pcre *re, pcre_extra *extra, uschar *bptr, int len,
+   int start_offset, int options, int *use_offsets, int use_size_offsets,
+   int flag, unsigned long int *limit, int errnumber, const char *msg)
+ {
+ int count;
+ int min = 0;
+ int mid = 64;
+ int max = -1;
+ extra->flags |= flag;
+ for (;;)
+   {
+   *limit = mid;
+   count = pcre_exec(re, extra, (char *)bptr, len, start_offset, options,
+     use_offsets, use_size_offsets);
+   if (count == errnumber)
+     {
+     /* fprintf(outfile, "Testing %s limit = %d\n", msg, 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 %s limit = %d\n", msg, mid);
+       break;
+       }
+     /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
+     max = mid;
+     mid = (min + mid)/2;
+     }
+   else break;    /* Some other error */
+   }
+ extra->flags &= ~flag;
+ return count;
+ }
+ /*************************************************
+ *         Check newline indicator                *
+ *************************************************/
+ /* This is used both at compile and run-time to check for <xxx> escapes, where
+ xxx is LF, CR, CRLF, or ANY. Print a message and return 0 if there is no match.
+ Arguments:
+   p           points after the leading '<'
+   f           file for error message
+ Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
+ */
+ static int
+ check_newline(uschar *p, FILE *f)
+ {
+ if (strncmp((char *)p, "cr>", 3) == 0) return PCRE_NEWLINE_CR;
+ if (strncmp((char *)p, "lf>", 3) == 0) return PCRE_NEWLINE_LF;
+ if (strncmp((char *)p, "crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
+ if (strncmp((char *)p, "any>", 4) == 0) return PCRE_NEWLINE_ANY;
+ fprintf(f, "Unknown newline type at: <%s\n", p);
+ return 0;
  }
+ /*************************************************
+ *             Usage function                     *
+ *************************************************/
+ static void
+ usage(void)
+ {
+ printf("Usage:     pcretest [options] [<input> [<output>]]\n");
+ printf("  -b       show compiled code (bytecode)\n");
+ printf("  -C       show PCRE compile-time options and exit\n");
+ printf("  -d       debug: show compiled code and information (-b and -i)\n");
+ #if !defined NODFA
+ printf("  -dfa     force DFA matching for all subjects\n");
+ #endif
+ printf("  -help    show usage information\n");
+ printf("  -i       show information about compiled patterns\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("  -q       quiet: do not output PCRE version number at start\n");
+ printf("  -S <n>   set stack size to <n> megabytes\n");
+ printf("  -s       output store (memory) used information\n"
+        "  -t       time compilation and execution\n");
+ printf("  -t <n>   time compilation and execution, repeating <n> times\n");
+ printf("  -tm      time execution (matching) only\n");
+ printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
+ }
+ /*************************************************
+ *                Main Program                    *
+ *************************************************/
  /* Read lines from named file or stdin and write to named file or stdout; lines
  consist of a regular expression, in delimiters and optionally followed by
  options, followed by a set of test data, terminated by an empty line. */
-Line 265 
 int options = 0;
+Line 729 
 int options = 0;
  int study_options = 0;
  int op = 1;
  int timeit = 0;
+ int timeitm = 0;
  int showinfo = 0;
+ int showstore = 0;
+ int quiet = 0;
+ int size_offsets = 45;
+ int size_offsets_max;
+ int *offsets = NULL;
+ #if !defined NOPOSIX
  int posix = 0;
+ #endif
  int debug = 0;
- unsigned char buffer[30000];
+ int done = 0;
- unsigned char dbuffer[1024];
+ int all_use_dfa = 0;
+ int yield = 0;
+ int stack_size;
- /* Static so that new_malloc can use it. */
+ /* These vectors store, end-to-end, a list of captured substring names. Assume
+ that 1024 is plenty long enough for the few names we'll be testing. */
+ uschar copynames[1024];
+ uschar getnames[1024];
+ uschar *copynamesptr;
+ uschar *getnamesptr;
+ /* Get buffers from malloc() so that Electric Fence will check their misuse
+ when I am debugging. They grow automatically when very long lines are read. */
+ buffer = (unsigned char *)malloc(buffer_size);
+ dbuffer = (unsigned char *)malloc(buffer_size);
+ pbuffer = (unsigned char *)malloc(buffer_size);
+ /* The outfile variable is static so that new_malloc can use it. */
  outfile = stdout;
+ /* The following  _setmode() stuff is some Windows magic that tells its runtime
+ library to translate CRLF into a single LF character. At least, that's what
+ I've been told: never having used Windows I take this all on trust. Originally
+ it set 0x8000, but then I was advised that _O_BINARY was better. */
+ #if defined(_WIN32) || defined(WIN32)
+ _setmode( _fileno( stdout ), _O_BINARY );
+ #endif
  /* Scan options */
  while (argc > 1 && argv[op][0] == '-')
    {
-   if (strcmp(argv[op], "-s") == 0) log_store = 1;
+   unsigned char *endptr;
-   else if (strcmp(argv[op], "-t") == 0) timeit = 1;
+   if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
+     showstore = 1;
+   else if (strcmp(argv[op], "-q") == 0) quiet = 1;
+   else if (strcmp(argv[op], "-b") == 0) debug = 1;
    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
+ #if !defined NODFA
+   else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
+ #endif
+   else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&
+       ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),
+         *endptr == 0))
+     {
+     op++;
+     argc--;
+     }
+   else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
+     {
+     int both = argv[op][2] == 0;
+     int temp;
+     if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr),
+                      *endptr == 0))
+       {
+       timeitm = temp;
+       op++;
+       argc--;
+       }
+     else timeitm = LOOPREPEAT;
+     if (both) timeit = timeitm;
+     }
+   else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
+       ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
+         *endptr == 0))
+     {
+ #if defined(_WIN32) || defined(WIN32)
+     printf("PCRE: -S not supported on this OS\n");
+     exit(1);
+ #else
+     int rc;
+     struct rlimit rlim;
+     getrlimit(RLIMIT_STACK, &rlim);
+     rlim.rlim_cur = stack_size * 1024 * 1024;
+     rc = setrlimit(RLIMIT_STACK, &rlim);
+     if (rc != 0)
+       {
+     printf("PCRE: setrlimit() failed with error %d\n", rc);
+     exit(1);
+       }
+     op++;
+     argc--;
+ #endif
+     }
+ #if !defined NOPOSIX
    else if (strcmp(argv[op], "-p") == 0) posix = 1;
+ #endif
+   else if (strcmp(argv[op], "-C") == 0)
+     {
+     int rc;
+     printf("PCRE version %s\n", pcre_version());
+     printf("Compiled with\n");
+     (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
+     printf("  %sUTF-8 support\n", rc? "" : "No ");
+     (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
+     printf("  %sUnicode properties support\n", rc? "" : "No ");
+     (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
+     printf("  Newline sequence is %s\n", (rc == '\r')? "CR" :
+       (rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" :
+       (rc == -1)? "ANY" : "???");
+     (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);
+     printf("  Internal link size = %d\n", rc);
+     (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
+     printf("  POSIX malloc threshold = %d\n", rc);
+     (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);
+     printf("  Default match limit = %d\n", rc);
+     (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &rc);
+     printf("  Default recursion depth limit = %d\n", rc);
+     (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
+     printf("  Match recursion uses %s\n", rc? "stack" : "heap");
+     exit(0);
+     }
+   else if (strcmp(argv[op], "-help") == 0 ||
+            strcmp(argv[op], "--help") == 0)
+     {
+     usage();
+     goto EXIT;
+     }
    else
      {
-     printf("*** Unknown option %s\n", argv[op]);
+     printf("** Unknown or malformed option %s\n", argv[op]);
-     return 1;
+     usage();
+     yield = 1;
+     goto EXIT;
      }
    op++;
    argc--;
    }
+ /* Get the store for the offsets vector, and remember what it was */
+ size_offsets_max = size_offsets;
+ offsets = (int *)malloc(size_offsets_max * sizeof(int));
+ if (offsets == NULL)
+   {
+   printf("** Failed to get %d bytes of memory for offsets vector\n",
+     size_offsets_max * sizeof(int));
+   yield = 1;
+   goto EXIT;
+   }
  /* Sort out the input and output files */
  if (argc > 1)
    {
-   infile = fopen(argv[op], "r");
+   infile = fopen(argv[op], INPUT_MODE);
    if (infile == NULL)
      {
      printf("** Failed to open %s\n", argv[op]);
-     return 1;
+     yield = 1;
+     goto EXIT;
      }
    }
  if (argc > 2)
    {
-   outfile = fopen(argv[op+1], "w");
+   outfile = fopen(argv[op+1], OUTPUT_MODE);
    if (outfile == NULL)
      {
      printf("** Failed to open %s\n", argv[op+1]);
-     return 1;
+     yield = 1;
+     goto EXIT;
      }
    }
  /* Set alternative malloc function */
  pcre_malloc = new_malloc;
+ pcre_free = new_free;
+ pcre_stack_malloc = stack_malloc;
+ pcre_stack_free = stack_free;
- /* Heading line, then prompt for first re if stdin */
+ /* Heading line unless quiet, then prompt for first regex if stdin */
- fprintf(outfile, "Testing Perl-Compatible Regular Expressions\n");
+ if (!quiet) fprintf(outfile, "PCRE version %s\n\n", pcre_version());
- fprintf(outfile, "PCRE version %s\n\n", pcre_version());
  /* Main loop */
- for (;;)
+ while (!done)
    {
    pcre *re = NULL;
    pcre_extra *extra = NULL;
+ #if !defined NOPOSIX  /* There are still compilers that require no indent */
    regex_t preg;
+   int do_posix = 0;
+ #endif
    const char *error;
-   unsigned char *p, *pp;
+   unsigned char *p, *pp, *ppp;
+   unsigned char *to_file = NULL;
+   const unsigned char *tables = NULL;
+   unsigned long int true_size, true_study_size = 0;
+   size_t size, regex_gotten_store;
    int do_study = 0;
-   int do_debug = 0;
+   int do_debug = debug;
-   int do_posix = 0;
+   int do_G = 0;
-   int erroroffset, len, delimiter;
+   int do_g = 0;
+   int do_showinfo = showinfo;
+   int do_showrest = 0;
+   int do_flip = 0;
+   int erroroffset, len, delimiter, poffset;
+   use_utf8 = 0;
    if (infile == stdin) printf("  re> ");
-   if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) break;
+   if (extend_inputline(infile, buffer) == NULL) break;
-   if (infile != stdin) fprintf(outfile, (char *)buffer);
+   if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
+   fflush(outfile);
    p = buffer;
    while (isspace(*p)) p++;
    if (*p == 0) continue;
-   /* Get the delimiter and seek the end of the pattern; if is isn't
+   /* See if the pattern is to be loaded pre-compiled from a file. */
-   complete, read more. */
+   if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
+     {
+     unsigned long int magic, get_options;
+     uschar sbuf[8];
+     FILE *f;
+     p++;
+     pp = p + (int)strlen((char *)p);
+     while (isspace(pp[-1])) pp--;
+     *pp = 0;
+     f = fopen((char *)p, "rb");
+     if (f == NULL)
+       {
+       fprintf(outfile, "Failed to open %s: %s\n", p, strerror(errno));
+       continue;
+       }
+     if (fread(sbuf, 1, 8, f) != 8) goto FAIL_READ;
+     true_size =
+       (sbuf[0] << 24) | (sbuf[1] << 16) | (sbuf[2] << 8) | sbuf[3];
+     true_study_size =
+       (sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7];
+     re = (real_pcre *)new_malloc(true_size);
+     regex_gotten_store = gotten_store;
+     if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ;
+     magic = ((real_pcre *)re)->magic_number;
+     if (magic != MAGIC_NUMBER)
+       {
+       if (byteflip(magic, sizeof(magic)) == MAGIC_NUMBER)
+         {
+         do_flip = 1;
+         }
+       else
+         {
+         fprintf(outfile, "Data in %s is not a compiled PCRE regex\n", p);
+         fclose(f);
+         continue;
+         }
+       }
+     fprintf(outfile, "Compiled regex%s loaded from %s\n",
+       do_flip? " (byte-inverted)" : "", p);
+     /* Need to know if UTF-8 for printing data strings */
+     new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
+     use_utf8 = (get_options & PCRE_UTF8) != 0;
+     /* Now see if there is any following study data */
+     if (true_study_size != 0)
+       {
+       pcre_study_data *psd;
+       extra = (pcre_extra *)new_malloc(sizeof(pcre_extra) + true_study_size);
+       extra->flags = PCRE_EXTRA_STUDY_DATA;
+       psd = (pcre_study_data *)(((char *)extra) + sizeof(pcre_extra));
+       extra->study_data = psd;
+       if (fread(psd, 1, true_study_size, f) != true_study_size)
+         {
+         FAIL_READ:
+         fprintf(outfile, "Failed to read data from %s\n", p);
+         if (extra != NULL) new_free(extra);
+         if (re != NULL) new_free(re);
+         fclose(f);
+         continue;
+         }
+       fprintf(outfile, "Study data loaded from %s\n", p);
+       do_study = 1;     /* To get the data output if requested */
+       }
+     else fprintf(outfile, "No study data\n");
+     fclose(f);
+     goto SHOW_INFO;
+     }
+   /* In-line pattern (the usual case). Get the delimiter and seek the end of
+   the pattern; if is isn't complete, read more. */
    delimiter = *p++;
-   if (isalnum(delimiter))
+   if (isalnum(delimiter) || delimiter == '\\')
      {
-     fprintf(outfile, "** Delimiter must not be alphameric\n");
+     fprintf(outfile, "** Delimiter must not be alphameric or \\\n");
      goto SKIP_DATA;
      }
    pp = p;
+   poffset = p - buffer;
    for(;;)
      {
-     while (*pp != 0 && *pp != delimiter) pp++;
+     while (*pp != 0)
-     if (*pp != 0) break;
-     len = sizeof(buffer) - (pp - buffer);
-     if (len < 256)
        {
-       fprintf(outfile, "** Expression too long - missing delimiter?\n");
+       if (*pp == '\\' && pp[1] != 0) pp++;
-       goto SKIP_DATA;
+         else if (*pp == delimiter) break;
+       pp++;
        }
+     if (*pp != 0) break;
      if (infile == stdin) printf("    > ");
-     if (fgets((char *)pp, len, infile) == NULL)
+     if ((pp = extend_inputline(infile, pp)) == NULL)
        {
        fprintf(outfile, "** Unexpected EOF\n");
-       goto END_OFF;
+       done = 1;
+       goto CONTINUE;
        }
-     if (infile != stdin) fprintf(outfile, (char *)pp);
+     if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
      }
-   /* Terminate the pattern at the delimiter */
+   /* The buffer may have moved while being extended; reset the start of data
+   pointer to the correct relative point in the buffer. */
+   p = buffer + poffset;
+   /* If the first character after the delimiter is backslash, make
+   the pattern end with backslash. This is purely to provide a way
+   of testing for the error message when a pattern ends with backslash. */
+   if (pp[1] == '\\') *pp++ = '\\';
+   /* Terminate the pattern at the delimiter, and save a copy of the pattern
+   for callouts. */
    *pp++ = 0;
+   strcpy((char *)pbuffer, (char *)p);
    /* Look for options after final delimiter */
    options = 0;
    study_options = 0;
+   log_store = showstore;  /* default from command line */
    while (*pp != 0)
      {
      switch (*pp++)
        {
+       case 'f': options |= PCRE_FIRSTLINE; break;
+       case 'g': do_g = 1; break;
        case 'i': options |= PCRE_CASELESS; break;
        case 'm': options |= PCRE_MULTILINE; break;
        case 's': options |= PCRE_DOTALL; break;
        case 'x': options |= PCRE_EXTENDED; break;
+       case '+': do_showrest = 1; break;
        case 'A': options |= PCRE_ANCHORED; break;
-       case 'D': do_debug = 1; break;
+       case 'B': do_debug = 1; break;
+       case 'C': options |= PCRE_AUTO_CALLOUT; break;
+       case 'D': do_debug = do_showinfo = 1; break;
        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
+       case 'F': do_flip = 1; break;
+       case 'G': do_G = 1; break;
+       case 'I': do_showinfo = 1; break;
+       case 'J': options |= PCRE_DUPNAMES; break;
+       case 'M': log_store = 1; break;
+       case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
+ #if !defined NOPOSIX
        case 'P': do_posix = 1; break;
+ #endif
        case 'S': do_study = 1; break;
-       case 'I': study_options |= PCRE_CASELESS; break;
+       case 'U': options |= PCRE_UNGREEDY; break;
        case 'X': options |= PCRE_EXTRA; break;
-       case '\n': case ' ': break;
+       case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
+       case '?': options |= PCRE_NO_UTF8_CHECK; break;
+       case 'L':
+       ppp = pp;
+       /* The '\r' test here is so that it works on Windows. */
+       /* The '0' test is just in case this is an unterminated line. */
+       while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
+       *ppp = 0;
+       if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
+         {
+         fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
+         goto SKIP_DATA;
+         }
+       locale_set = 1;
+       tables = pcre_maketables();
+       pp = ppp;
+       break;
+       case '>':
+       to_file = pp;
+       while (*pp != 0) pp++;
+       while (isspace(pp[-1])) pp--;
+       *pp = 0;
+       break;
+       case '<':
+         {
+         int x = check_newline(pp, outfile);
+         if (x == 0) goto SKIP_DATA;
+         options |= x;
+         while (*pp++ != '>');
+         }
+       break;
+       case '\r':                      /* So that it works in Windows */
+       case '\n':
+       case ' ':
+       break;
        default:
        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);
        goto SKIP_DATA;
        }
      }
-   /* Handle compiing via the POSIX interface, which doesn't support the
+   /* Handle compiling via the POSIX interface, which doesn't support the
-   timing, showing, or debugging options. */
+   timing, showing, or debugging options, nor the ability to pass over
+   local character tables. */
+ #if !defined NOPOSIX
    if (posix || do_posix)
      {
      int rc;
      int cflags = 0;
      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
+     if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
+     if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
+     if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
      rc = regcomp(&preg, (char *)p, cflags);
      /* Compilation failed; go back for another re, skipping to blank line
-Line 426 
 for (;;)
+Line 1203 
 for (;;)
      if (rc != 0)
        {
-       (void)regerror(rc, &preg, (char *)buffer, sizeof(buffer));
+       (void)regerror(rc, &preg, (char *)buffer, buffer_size);
        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
        goto SKIP_DATA;
        }
-Line 435 
 for (;;)
+Line 1212 
 for (;;)
    /* Handle compiling via the native interface */
    else
+ #endif  /* !defined NOPOSIX */
      {
-     if (timeit)
+     if (timeit > 0)
        {
        register int i;
        clock_t time_taken;
        clock_t start_time = clock();
-       for (i = 0; i < 4000; i++)
+       for (i = 0; i < timeit; i++)
          {
-         re = pcre_compile((char *)p, options, &error, &erroroffset);
+         re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
          if (re != NULL) free(re);
          }
        time_taken = clock() - start_time;
-       fprintf(outfile, "Compile time %.2f milliseconds\n",
+       fprintf(outfile, "Compile time %.4f milliseconds\n",
-         ((double)time_taken)/(4 * CLOCKS_PER_SEC));
+         (((double)time_taken * 1000.0) / (double)timeit) /
+           (double)CLOCKS_PER_SEC);
        }
-     re = pcre_compile((char *)p, options, &error, &erroroffset);
+     re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
      /* Compilation failed; go back for another re, skipping to blank line
      if non-interactive. */
-Line 464 
 for (;;)
+Line 1244 
 for (;;)
          {
          for (;;)
            {
-           if (fgets((char *)buffer, sizeof(buffer), infile) == NULL)
+           if (extend_inputline(infile, buffer) == NULL)
-             goto END_OFF;
+             {
+             done = 1;
+             goto CONTINUE;
+             }
            len = (int)strlen((char *)buffer);
            while (len > 0 && isspace(buffer[len-1])) len--;
            if (len == 0) break;
            }
          fprintf(outfile, "\n");
          }
-       continue;
+       goto CONTINUE;
        }
-     /* Compilation succeeded; print data if required */
+     /* Compilation succeeded; print data if required. There are now two
+     info-returning functions. The old one has a limited interface and
+     returns only limited data. Check that it agrees with the newer one. */
+     if (log_store)
+       fprintf(outfile, "Memory allocation (code space): %d\n",
+         (int)(gotten_store -
+               sizeof(real_pcre) -
+               ((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size));
+     /* Extract the size for possible writing before possibly flipping it,
+     and remember the store that was got. */
+     true_size = ((real_pcre *)re)->size;
+     regex_gotten_store = gotten_store;
+     /* If /S was present, study the regexp to generate additional info to
+     help with the matching. */
+     if (do_study)
+       {
+       if (timeit > 0)
+         {
+         register int i;
+         clock_t time_taken;
+         clock_t start_time = clock();
+         for (i = 0; i < timeit; i++)
+           extra = pcre_study(re, study_options, &error);
+         time_taken = clock() - start_time;
+         if (extra != NULL) free(extra);
+         fprintf(outfile, "  Study time %.4f milliseconds\n",
+           (((double)time_taken * 1000.0) / (double)timeit) /
+             (double)CLOCKS_PER_SEC);
+         }
+       extra = pcre_study(re, study_options, &error);
+       if (error != NULL)
+         fprintf(outfile, "Failed to study: %s\n", error);
+       else if (extra != NULL)
+         true_study_size = ((pcre_study_data *)(extra->study_data))->size;
+       }
+     /* If the 'F' option was present, we flip the bytes of all the integer
+     fields in the regex data block and the study block. This is to make it
+     possible to test PCRE's handling of byte-flipped patterns, e.g. those
+     compiled on a different architecture. */
+     if (do_flip)
+       {
+       real_pcre *rre = (real_pcre *)re;
+       rre->magic_number = byteflip(rre->magic_number, sizeof(rre->magic_number));
+       rre->size = byteflip(rre->size, sizeof(rre->size));
+       rre->options = byteflip(rre->options, sizeof(rre->options));
+       rre->top_bracket = byteflip(rre->top_bracket, sizeof(rre->top_bracket));
+       rre->top_backref = byteflip(rre->top_backref, sizeof(rre->top_backref));
+       rre->first_byte = byteflip(rre->first_byte, sizeof(rre->first_byte));
+       rre->req_byte = byteflip(rre->req_byte, sizeof(rre->req_byte));
+       rre->name_table_offset = byteflip(rre->name_table_offset,
+         sizeof(rre->name_table_offset));
+       rre->name_entry_size = byteflip(rre->name_entry_size,
+         sizeof(rre->name_entry_size));
+       rre->name_count = byteflip(rre->name_count, sizeof(rre->name_count));
+       if (extra != NULL)
+         {
+         pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
+         rsd->size = byteflip(rsd->size, sizeof(rsd->size));
+         rsd->options = byteflip(rsd->options, sizeof(rsd->options));
+         }
+       }
+     /* Extract information from the compiled data if required */
+     SHOW_INFO:
+     if (do_debug)
+       {
+       fprintf(outfile, "------------------------------------------------------------------\n");
+       pcre_printint(re, outfile);
+       }
+     if (do_showinfo)
+       {
+       unsigned long int get_options, all_options;
+ #if !defined NOINFOCHECK
+       int old_first_char, old_options, old_count;
+ #endif
+       int count, backrefmax, first_char, need_char;
+       int nameentrysize, namecount;
+       const uschar *nametable;
+       new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
+       new_info(re, NULL, PCRE_INFO_SIZE, &size);
+       new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
+       new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);
+       new_info(re, NULL, PCRE_INFO_FIRSTBYTE, &first_char);
+       new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char);
+       new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);
+       new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);
+       new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);
+ #if !defined NOINFOCHECK
+       old_count = pcre_info(re, &old_options, &old_first_char);
+       if (count < 0) fprintf(outfile,
+         "Error %d from pcre_info()\n", count);
+       else
+         {
+         if (old_count != count) fprintf(outfile,
+           "Count disagreement: pcre_fullinfo=%d pcre_info=%d\n", count,
+             old_count);
+         if (old_first_char != first_char) fprintf(outfile,
+           "First char disagreement: pcre_fullinfo=%d pcre_info=%d\n",
+             first_char, old_first_char);
+         if (old_options != (int)get_options) fprintf(outfile,
+           "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",
+             get_options, old_options);
+         }
+ #endif
+       if (size != regex_gotten_store) fprintf(outfile,
+         "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",
+         (int)size, (int)regex_gotten_store);
+       fprintf(outfile, "Capturing subpattern count = %d\n", count);
+       if (backrefmax > 0)
+         fprintf(outfile, "Max back reference = %d\n", backrefmax);
+       if (namecount > 0)
+         {
+         fprintf(outfile, "Named capturing subpatterns:\n");
+         while (namecount-- > 0)
+           {
+           fprintf(outfile, "  %s %*s%3d\n", nametable + 2,
+             nameentrysize - 3 - (int)strlen((char *)nametable + 2), "",
+             GET2(nametable, 0));
+           nametable += nameentrysize;
+           }
+         }
+       /* The NOPARTIAL bit is a private bit in the options, so we have
+       to fish it out via out back door */
+       all_options = ((real_pcre *)re)->options;
+       if (do_flip)
+         {
+         all_options = byteflip(all_options, sizeof(all_options));
+          }
+       if ((all_options & PCRE_NOPARTIAL) != 0)
+         fprintf(outfile, "Partial matching not supported\n");
+       if (get_options == 0) fprintf(outfile, "No options\n");
+         else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s%s%s\n",
+           ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
+           ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
+           ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
+           ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",
+           ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",
+           ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",
+           ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
+           ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
+           ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
+           ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
+           ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
+           ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
+           ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
+       switch (get_options & PCRE_NEWLINE_BITS)
+         {
+         case PCRE_NEWLINE_CR:
+         fprintf(outfile, "Forced newline sequence: CR\n");
+         break;
+         case PCRE_NEWLINE_LF:
+         fprintf(outfile, "Forced newline sequence: LF\n");
+         break;
+         case PCRE_NEWLINE_CRLF:
+         fprintf(outfile, "Forced newline sequence: CRLF\n");
+         break;
+         case PCRE_NEWLINE_ANY:
+         fprintf(outfile, "Forced newline sequence: ANY\n");
+         break;
+         default:
+         break;
+         }
+       if (first_char == -1)
+         {
+         fprintf(outfile, "First char at start or follows newline\n");
+         }
+       else if (first_char < 0)
+         {
+         fprintf(outfile, "No first char\n");
+         }
+       else
+         {
+         int ch = first_char & 255;
+         const char *caseless = ((first_char & REQ_CASELESS) == 0)?
+           "" : " (caseless)";
+         if (PRINTHEX(ch))
+           fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);
+         else
+           fprintf(outfile, "First char = %d%s\n", ch, caseless);
+         }
+       if (need_char < 0)
+         {
+         fprintf(outfile, "No need char\n");
+         }
+       else
+         {
+         int ch = need_char & 255;
+         const char *caseless = ((need_char & REQ_CASELESS) == 0)?
+           "" : " (caseless)";
+         if (PRINTHEX(ch))
+           fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);
+         else
+           fprintf(outfile, "Need char = %d%s\n", ch, caseless);
+         }
-     if (showinfo || do_debug)
+       /* Don't output study size; at present it is in any case a fixed
-       {
+       value, but it varies, depending on the computer architecture, and
-       int first_char, count;
+       so messes up the test suite. (And with the /F option, it might be
+       flipped.) */
-       if (debug || do_debug) print_internals(re);
-       count = pcre_info(re, &options, &first_char);
+       if (do_study)
-       if (count < 0) fprintf(outfile,
-         "Error %d while reading info\n", count);
-       else
          {
-         fprintf(outfile, "Identifying subpattern count = %d\n", count);
+         if (extra == NULL)
-         if (options == 0) fprintf(outfile, "No options\n");
+           fprintf(outfile, "Study returned NULL\n");
-           else fprintf(outfile, "Options:%s%s%s%s%s%s%s\n",
-             ((options & PCRE_ANCHORED) != 0)? " anchored" : "",
-             ((options & PCRE_CASELESS) != 0)? " caseless" : "",
-             ((options & PCRE_EXTENDED) != 0)? " extended" : "",
-             ((options & PCRE_MULTILINE) != 0)? " multiline" : "",
-             ((options & PCRE_DOTALL) != 0)? " dotall" : "",
-             ((options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
-             ((options & PCRE_EXTRA) != 0)? " extra" : "");
-         if (first_char == -1)
-           {
-           fprintf(outfile, "First char at start or follows \\n\n");
-           }
-         else if (first_char < 0)
-           {
-           fprintf(outfile, "No first char\n");
-           }
          else
            {
-           if (isprint(first_char))
+           uschar *start_bits = NULL;
-             fprintf(outfile, "First char = \'%c\'\n", first_char);
+           new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);
+           if (start_bits == NULL)
+             fprintf(outfile, "No starting byte set\n");
            else
-             fprintf(outfile, "First char = %d\n", first_char);
+             {
+             int i;
+             int c = 24;
+             fprintf(outfile, "Starting byte set: ");
+             for (i = 0; i < 256; i++)
+               {
+               if ((start_bits[i/8] & (1<<(i&7))) != 0)
+                 {
+                 if (c > 75)
+                   {
+                   fprintf(outfile, "\n  ");
+                   c = 2;
+                   }
+                 if (PRINTHEX(i) && i != ' ')
+                   {
+                   fprintf(outfile, "%c ", i);
+                   c += 2;
+                   }
+                 else
+                   {
+                   fprintf(outfile, "\\x%02x ", i);
+                   c += 5;
+                   }
+                 }
+               }
+             fprintf(outfile, "\n");
+             }
            }
          }
        }
-     /* If /S was present, study the regexp to generate additional info to
+     /* If the '>' option was present, we write out the regex to a file, and
-     help with the matching. */
+     that is all. The first 8 bytes of the file are the regex length and then
+     the study length, in big-endian order. */
-     if (do_study)
+     if (to_file != NULL)
        {
-       if (timeit)
+       FILE *f = fopen((char *)to_file, "wb");
+       if (f == NULL)
          {
-         register int i;
+         fprintf(outfile, "Unable to open %s: %s\n", to_file, strerror(errno));
-         clock_t time_taken;
-         clock_t start_time = clock();
-         for (i = 0; i < 4000; i++)
-           extra = pcre_study(re, study_options, &error);
-         time_taken = clock() - start_time;
-         if (extra != NULL) free(extra);
-         fprintf(outfile, "  Study time %.2f milliseconds\n",
-           ((double)time_taken)/(4 * CLOCKS_PER_SEC));
          }
+       else
-       extra = pcre_study(re, study_options, &error);
-       if (error != NULL)
-         fprintf(outfile, "Failed to study: %s\n", error);
-       else if (extra == NULL)
-         fprintf(outfile, "Study returned NULL\n");
-       /* This looks at internal information. A bit kludgy to do it this
-       way, but it is useful for testing. */
-       else if (showinfo || do_debug)
          {
-         real_pcre_extra *xx = (real_pcre_extra *)extra;
+         uschar sbuf[8];
-         if ((xx->options & PCRE_STUDY_MAPPED) == 0)
+         sbuf[0] = (true_size >> 24)  & 255;
-           fprintf(outfile, "No starting character set\n");
+         sbuf[1] = (true_size >> 16)  & 255;
+         sbuf[2] = (true_size >>  8)  & 255;
+         sbuf[3] = (true_size)  & 255;
+         sbuf[4] = (true_study_size >> 24)  & 255;
+         sbuf[5] = (true_study_size >> 16)  & 255;
+         sbuf[6] = (true_study_size >>  8)  & 255;
+         sbuf[7] = (true_study_size)  & 255;
+         if (fwrite(sbuf, 1, 8, f) < 8 ||
+             fwrite(re, 1, true_size, f) < true_size)
+           {
+           fprintf(outfile, "Write error on %s: %s\n", to_file, strerror(errno));
+           }
          else
            {
-           int i;
+           fprintf(outfile, "Compiled regex written to %s\n", to_file);
-           int c = 24;
+           if (extra != NULL)
-           fprintf(outfile, "Starting character set: ");
-           for (i = 0; i < 256; i++)
              {
-             if ((xx->start_bits[i/8] & (1<<(i%8))) != 0)
+             if (fwrite(extra->study_data, 1, true_study_size, f) <
+                 true_study_size)
                {
-               if (c > 75)
+               fprintf(outfile, "Write error on %s: %s\n", to_file,
-                 {
+                 strerror(errno));
-                 fprintf(outfile, "\n  ");
-                 c = 2;
-                 }
-               if (isprint(i) && i != ' ')
-                 {
-                 fprintf(outfile, "%c ", i);
-                 c += 2;
-                 }
-               else
-                 {
-                 fprintf(outfile, "\\x%02x ", i);
-                 c += 5;
-                 }
                }
+             else fprintf(outfile, "Study data written to %s\n", to_file);
              }
-           fprintf(outfile, "\n");
            }
+         fclose(f);
          }
+       new_free(re);
+       if (extra != NULL) new_free(extra);
+       if (tables != NULL) new_free((void *)tables);
+       continue;  /* With next regex */
        }
-     }
+     }        /* End of non-POSIX compile */
    /* Read data lines and test them */
    for (;;)
      {
-     unsigned char *q;
+     uschar *q;
+     uschar *bptr = dbuffer;
+     int *use_offsets = offsets;
+     int use_size_offsets = size_offsets;
+     int callout_data = 0;
+     int callout_data_set = 0;
      int count, c;
-     int offsets[30];
+     int copystrings = 0;
-     int size_offsets = sizeof(offsets)/sizeof(int);
+     int find_match_limit = 0;
+     int getstrings = 0;
+     int getlist = 0;
+     int gmatched = 0;
+     int start_offset = 0;
+     int g_notempty = 0;
+     int use_dfa = 0;
      options = 0;
-     if (infile == stdin) printf("  data> ");
+     *copynames = 0;
-     if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) goto END_OFF;
+     *getnames = 0;
-     if (infile != stdin) fprintf(outfile, (char *)buffer);
+     copynamesptr = copynames;
+     getnamesptr = getnames;
+     pcre_callout = callout;
+     first_callout = 1;
+     callout_extra = 0;
+     callout_count = 0;
+     callout_fail_count = 999999;
+     callout_fail_id = -1;
+     show_malloc = 0;
+     if (extra != NULL) extra->flags &=
+       ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
+     len = 0;
+     for (;;)
+       {
+       if (infile == stdin) printf("data> ");
+       if (extend_inputline(infile, buffer + len) == NULL)
+         {
+         if (len > 0) break;
+         done = 1;
+         goto CONTINUE;
+         }
+       if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
+       len = (int)strlen((char *)buffer);
+       if (buffer[len-1] == '\n') break;
+       }
-     len = (int)strlen((char *)buffer);
      while (len > 0 && isspace(buffer[len-1])) len--;
      buffer[len] = 0;
      if (len == 0) break;
-Line 608 
 for (;;)
+Line 1650 
 for (;;)
        {
        int i = 0;
        int n = 0;
        if (c == '\\') switch ((c = *p++))
          {
          case 'a': c =    7; break;
-Line 624 
 for (;;)
+Line 1667 
 for (;;)
          c -= '0';
          while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')
            c = c * 8 + *p++ - '0';
+ #if !defined NOUTF8
+         if (use_utf8 && c > 255)
+           {
+           unsigned char buff8[8];
+           int ii, utn;
+           utn = ord2utf8(c, buff8);
+           for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
+           c = buff8[ii];   /* Last byte */
+           }
+ #endif
          break;
          case 'x':
+         /* Handle \x{..} specially - new Perl thing for utf8 */
+ #if !defined NOUTF8
+         if (*p == '{')
+           {
+           unsigned char *pt = p;
+           c = 0;
+           while (isxdigit(*(++pt)))
+             c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W');
+           if (*pt == '}')
+             {
+             unsigned char buff8[8];
+             int ii, utn;
+             utn = ord2utf8(c, buff8);
+             for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
+             c = buff8[ii];   /* Last byte */
+             p = pt + 1;
+             break;
+             }
+           /* Not correct form; fall through */
+           }
+ #endif
+         /* Ordinary \x */
          c = 0;
          while (i++ < 2 && isxdigit(*p))
            {
-Line 635 
 for (;;)
+Line 1715 
 for (;;)
            }
          break;
-         case 0:   /* Allows for an empty line */
+         case 0:   /* \ followed by EOF allows for an empty line */
          p--;
          continue;
+         case '>':
+         while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';
+         continue;
          case 'A':  /* Option setting */
          options |= PCRE_ANCHORED;
          continue;
-Line 647 
 for (;;)
+Line 1731 
 for (;;)
          options |= PCRE_NOTBOL;
          continue;
-         case 'E':
+         case 'C':
-         options |= PCRE_DOLLAR_ENDONLY;
+         if (isdigit(*p))    /* Set copy string */
+           {
+           while(isdigit(*p)) n = n * 10 + *p++ - '0';
+           copystrings |= 1 << n;
+           }
+         else if (isalnum(*p))
+           {
+           uschar *npp = copynamesptr;
+           while (isalnum(*p)) *npp++ = *p++;
+           *npp++ = 0;
+           *npp = 0;
+           n = pcre_get_stringnumber(re, (char *)copynamesptr);
+           if (n < 0)
+             fprintf(outfile, "no parentheses with name \"%s\"\n", copynamesptr);
+           copynamesptr = npp;
+           }
+         else if (*p == '+')
+           {
+           callout_extra = 1;
+           p++;
+           }
+         else if (*p == '-')
+           {
+           pcre_callout = NULL;
+           p++;
+           }
+         else if (*p == '!')
+           {
+           callout_fail_id = 0;
+           p++;
+           while(isdigit(*p))
+             callout_fail_id = callout_fail_id * 10 + *p++ - '0';
+           callout_fail_count = 0;
+           if (*p == '!')
+             {
+             p++;
+             while(isdigit(*p))
+               callout_fail_count = callout_fail_count * 10 + *p++ - '0';
+             }
+           }
+         else if (*p == '*')
+           {
+           int sign = 1;
+           callout_data = 0;
+           if (*(++p) == '-') { sign = -1; p++; }
+           while(isdigit(*p))
+             callout_data = callout_data * 10 + *p++ - '0';
+           callout_data *= sign;
+           callout_data_set = 1;
+           }
+         continue;
+ #if !defined NODFA
+         case 'D':
+ #if !defined NOPOSIX
+         if (posix || do_posix)
+           printf("** Can't use dfa matching in POSIX mode: \\D ignored\n");
+         else
+ #endif
+           use_dfa = 1;
+         continue;
+         case 'F':
+         options |= PCRE_DFA_SHORTEST;
+         continue;
+ #endif
+         case 'G':
+         if (isdigit(*p))
+           {
+           while(isdigit(*p)) n = n * 10 + *p++ - '0';
+           getstrings |= 1 << n;
+           }
+         else if (isalnum(*p))
+           {
+           uschar *npp = getnamesptr;
+           while (isalnum(*p)) *npp++ = *p++;
+           *npp++ = 0;
+           *npp = 0;
+           n = pcre_get_stringnumber(re, (char *)getnamesptr);
+           if (n < 0)
+             fprintf(outfile, "no parentheses with name \"%s\"\n", getnamesptr);
+           getnamesptr = npp;
+           }
          continue;
-         case 'I':
+         case 'L':
-         options |= PCRE_CASELESS;
+         getlist = 1;
          continue;
          case 'M':
-         options |= PCRE_MULTILINE;
+         find_match_limit = 1;
          continue;
-         case 'S':
+         case 'N':
-         options |= PCRE_DOTALL;
+         options |= PCRE_NOTEMPTY;
          continue;
          case 'O':
          while(isdigit(*p)) n = n * 10 + *p++ - '0';
-         if (n <= (int)(sizeof(offsets)/sizeof(int))) size_offsets = n;
+         if (n > size_offsets_max)
+           {
+           size_offsets_max = n;
+           free(offsets);
+           use_offsets = offsets = (int *)malloc(size_offsets_max * sizeof(int));
+           if (offsets == NULL)
+             {
+             printf("** Failed to get %d bytes of memory for offsets vector\n",
+               size_offsets_max * sizeof(int));
+             yield = 1;
+             goto EXIT;
+             }
+           }
+         use_size_offsets = n;
+         if (n == 0) use_offsets = NULL;   /* Ensures it can't write to it */
+         continue;
+         case 'P':
+         options |= PCRE_PARTIAL;
+         continue;
+         case 'Q':
+         while(isdigit(*p)) n = n * 10 + *p++ - '0';
+         if (extra == NULL)
+           {
+           extra = (pcre_extra *)malloc(sizeof(pcre_extra));
+           extra->flags = 0;
+           }
+         extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
+         extra->match_limit_recursion = n;
+         continue;
+         case 'q':
+         while(isdigit(*p)) n = n * 10 + *p++ - '0';
+         if (extra == NULL)
+           {
+           extra = (pcre_extra *)malloc(sizeof(pcre_extra));
+           extra->flags = 0;
+           }
+         extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
+         extra->match_limit = n;
+         continue;
+ #if !defined NODFA
+         case 'R':
+         options |= PCRE_DFA_RESTART;
+         continue;
+ #endif
+         case 'S':
+         show_malloc = 1;
          continue;
          case 'Z':
          options |= PCRE_NOTEOL;
          continue;
+         case '?':
+         options |= PCRE_NO_UTF8_CHECK;
+         continue;
+         case '<':
+           {
+           int x = check_newline(p, outfile);
+           if (x == 0) goto NEXT_DATA;
+           options |= x;
+           while (*p++ != '>');
+           }
+         continue;
          }
        *q++ = c;
        }
      *q = 0;
      len = q - dbuffer;
+     if ((all_use_dfa || use_dfa) && find_match_limit)
+       {
+       printf("**Match limit not relevant for DFA matching: ignored\n");
+       find_match_limit = 0;
+       }
      /* Handle matching via the POSIX interface, which does not
-     support timing. */
+     support timing or playing with the match limit or callout data. */
+ #if !defined NOPOSIX
      if (posix || do_posix)
        {
        int rc;
        int eflags = 0;
-       regmatch_t pmatch[30];
+       regmatch_t *pmatch = NULL;
+       if (use_size_offsets > 0)
+         pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);
        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;
        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;
-       rc = regexec(&preg, (char *)dbuffer, sizeof(pmatch)/sizeof(regmatch_t),
+       rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);
-         pmatch, eflags);
        if (rc != 0)
          {
-         (void)regerror(rc, &preg, (char *)buffer, sizeof(buffer));
+         (void)regerror(rc, &preg, (char *)buffer, buffer_size);
          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
          }
+       else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
+               != 0)
+         {
+         fprintf(outfile, "Matched with REG_NOSUB\n");
+         }
        else
          {
          size_t i;
-         for (i = 0; i < sizeof(pmatch)/sizeof(regmatch_t); i++)
+         for (i = 0; i < (size_t)use_size_offsets; i++)
            {
            if (pmatch[i].rm_so >= 0)
              {
-             fprintf(outfile, "%2d: ", i);
+             fprintf(outfile, "%2d: ", (int)i);
-             pchars(dbuffer + pmatch[i].rm_so,
+             (void)pchars(dbuffer + pmatch[i].rm_so,
-               pmatch[i].rm_eo - pmatch[i].rm_so);
+               pmatch[i].rm_eo - pmatch[i].rm_so, outfile);
              fprintf(outfile, "\n");
+             if (i == 0 && do_showrest)
+               {
+               fprintf(outfile, " 0+ ");
+               (void)pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo,
+                 outfile);
+               fprintf(outfile, "\n");
+               }
              }
            }
          }
+       free(pmatch);
        }
-     /* Handle matching via the native interface */
+     /* Handle matching via the native interface - repeats for /g and /G */
      else
+ #endif  /* !defined NOPOSIX */
+     for (;; gmatched++)    /* Loop for /g or /G */
        {
-       if (timeit)
+       if (timeitm > 0)
          {
          register int i;
          clock_t time_taken;
          clock_t start_time = clock();
-         for (i = 0; i < 4000; i++)
-           count = pcre_exec(re, extra, (char *)dbuffer, len, options, offsets,
+ #if !defined NODFA
-             size_offsets);
+         if (all_use_dfa || use_dfa)
+           {
+           int workspace[1000];
+           for (i = 0; i < timeitm; i++)
+             count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
+               options | g_notempty, use_offsets, use_size_offsets, workspace,
+               sizeof(workspace)/sizeof(int));
+           }
+         else
+ #endif
+         for (i = 0; i < timeitm; i++)
+           count = pcre_exec(re, extra, (char *)bptr, len,
+             start_offset, options | g_notempty, use_offsets, use_size_offsets);
          time_taken = clock() - start_time;
-         fprintf(outfile, "Execute time %.2f milliseconds\n",
+         fprintf(outfile, "Execute time %.4f milliseconds\n",
-           ((double)time_taken)/(4 * CLOCKS_PER_SEC));
+           (((double)time_taken * 1000.0) / (double)timeitm) /
+             (double)CLOCKS_PER_SEC);
+         }
+       /* If find_match_limit is set, we want to do repeated matches with
+       varying limits in order to find the minimum value for the match limit and
+       for the recursion limit. */
+       if (find_match_limit)
+         {
+         if (extra == NULL)
+           {
+           extra = (pcre_extra *)malloc(sizeof(pcre_extra));
+           extra->flags = 0;
+           }
+         (void)check_match_limit(re, extra, bptr, len, start_offset,
+           options|g_notempty, use_offsets, use_size_offsets,
+           PCRE_EXTRA_MATCH_LIMIT, &(extra->match_limit),
+           PCRE_ERROR_MATCHLIMIT, "match()");
+         count = check_match_limit(re, extra, bptr, len, start_offset,
+           options|g_notempty, use_offsets, use_size_offsets,
+           PCRE_EXTRA_MATCH_LIMIT_RECURSION, &(extra->match_limit_recursion),
+           PCRE_ERROR_RECURSIONLIMIT, "match() recursion");
+         }
+       /* If callout_data is set, use the interface with additional data */
+       else if (callout_data_set)
+         {
+         if (extra == NULL)
+           {
+           extra = (pcre_extra *)malloc(sizeof(pcre_extra));
+           extra->flags = 0;
+           }
+         extra->flags |= PCRE_EXTRA_CALLOUT_DATA;
+         extra->callout_data = &callout_data;
+         count = pcre_exec(re, extra, (char *)bptr, len, start_offset,
+           options | g_notempty, use_offsets, use_size_offsets);
+         extra->flags &= ~PCRE_EXTRA_CALLOUT_DATA;
          }
-       count = pcre_exec(re, extra, (char *)dbuffer, len, options, offsets,
+       /* The normal case is just to do the match once, with the default
-         size_offsets);
+       value of match_limit. */
+ #if !defined NODFA
+       else if (all_use_dfa || use_dfa)
+         {
+         int workspace[1000];
+         count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
+           options | g_notempty, use_offsets, use_size_offsets, workspace,
+           sizeof(workspace)/sizeof(int));
+         if (count == 0)
+           {
+           fprintf(outfile, "Matched, but too many subsidiary matches\n");
+           count = use_size_offsets/2;
+           }
+         }
+ #endif
-       if (count == 0)
+       else
          {
-         fprintf(outfile, "Matched, but too many substrings\n");
+         count = pcre_exec(re, extra, (char *)bptr, len,
-         count = size_offsets/2;
+           start_offset, options | g_notempty, use_offsets, use_size_offsets);
+         if (count == 0)
+           {
+           fprintf(outfile, "Matched, but too many substrings\n");
+           count = use_size_offsets/3;
+           }
          }
+       /* Matched */
        if (count >= 0)
          {
-         int i;
+         int i, maxcount;
-         count *= 2;
-         for (i = 0; i < count; i += 2)
+ #if !defined NODFA
+         if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else
+ #endif
+           maxcount = use_size_offsets/3;
+         /* This is a check against a lunatic return value. */
+         if (count > maxcount)
+           {
+           fprintf(outfile,
+             "** PCRE error: returned count %d is too big for offset size %d\n",
+             count, use_size_offsets);
+           count = use_size_offsets/3;
+           if (do_g || do_G)
+             {
+             fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G');
+             do_g = do_G = FALSE;        /* Break g/G loop */
+             }
+           }
+         for (i = 0; i < count * 2; i += 2)
            {
-           if (offsets[i] < 0)
+           if (use_offsets[i] < 0)
              fprintf(outfile, "%2d: <unset>\n", i/2);
            else
              {
              fprintf(outfile, "%2d: ", i/2);
-             pchars(dbuffer + offsets[i], offsets[i+1] - offsets[i]);
+             (void)pchars(bptr + use_offsets[i],
+               use_offsets[i+1] - use_offsets[i], outfile);
              fprintf(outfile, "\n");
+             if (i == 0)
+               {
+               if (do_showrest)
+                 {
+                 fprintf(outfile, " 0+ ");
+                 (void)pchars(bptr + use_offsets[i+1], len - use_offsets[i+1],
+                   outfile);
+                 fprintf(outfile, "\n");
+                 }
+               }
+             }
+           }
+         for (i = 0; i < 32; i++)
+           {
+           if ((copystrings & (1 << i)) != 0)
+             {
+             char copybuffer[256];
+             int rc = pcre_copy_substring((char *)bptr, use_offsets, count,
+               i, copybuffer, sizeof(copybuffer));
+             if (rc < 0)
+               fprintf(outfile, "copy substring %d failed %d\n", i, rc);
+             else
+               fprintf(outfile, "%2dC %s (%d)\n", i, copybuffer, rc);
+             }
+           }
+         for (copynamesptr = copynames;
+              *copynamesptr != 0;
+              copynamesptr += (int)strlen((char*)copynamesptr) + 1)
+           {
+           char copybuffer[256];
+           int rc = pcre_copy_named_substring(re, (char *)bptr, use_offsets,
+             count, (char *)copynamesptr, copybuffer, sizeof(copybuffer));
+           if (rc < 0)
+             fprintf(outfile, "copy substring %s failed %d\n", copynamesptr, rc);
+           else
+             fprintf(outfile, "  C %s (%d) %s\n", copybuffer, rc, copynamesptr);
+           }
+         for (i = 0; i < 32; i++)
+           {
+           if ((getstrings & (1 << i)) != 0)
+             {
+             const char *substring;
+             int rc = pcre_get_substring((char *)bptr, use_offsets, count,
+               i, &substring);
+             if (rc < 0)
+               fprintf(outfile, "get substring %d failed %d\n", i, rc);
+             else
+               {
+               fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
+               pcre_free_substring(substring);
+               }
+             }
+           }
+         for (getnamesptr = getnames;
+              *getnamesptr != 0;
+              getnamesptr += (int)strlen((char*)getnamesptr) + 1)
+           {
+           const char *substring;
+           int rc = pcre_get_named_substring(re, (char *)bptr, use_offsets,
+             count, (char *)getnamesptr, &substring);
+           if (rc < 0)
+             fprintf(outfile, "copy substring %s failed %d\n", getnamesptr, rc);
+           else
+             {
+             fprintf(outfile, "  G %s (%d) %s\n", substring, rc, getnamesptr);
+             pcre_free_substring(substring);
+             }
+           }
+         if (getlist)
+           {
+           const char **stringlist;
+           int rc = pcre_get_substring_list((char *)bptr, use_offsets, count,
+             &stringlist);
+           if (rc < 0)
+             fprintf(outfile, "get substring list failed %d\n", rc);
+           else
+             {
+             for (i = 0; i < count; i++)
+               fprintf(outfile, "%2dL %s\n", i, stringlist[i]);
+             if (stringlist[i] != NULL)
+               fprintf(outfile, "string list not terminated by NULL\n");
+             /* free((void *)stringlist); */
+             pcre_free_substring_list(stringlist);
              }
            }
          }
+       /* There was a partial match */
+       else if (count == PCRE_ERROR_PARTIAL)
+         {
+         fprintf(outfile, "Partial match");
+ #if !defined NODFA
+         if ((all_use_dfa || use_dfa) && use_size_offsets > 2)
+           fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],
+             bptr + use_offsets[0]);
+ #endif
+         fprintf(outfile, "\n");
+         break;  /* Out of the /g loop */
+         }
+       /* Failed to match. If this is a /g or /G loop and we previously set
+       g_notempty after a null match, this is not necessarily the end.
+       We want to advance the start offset, and continue. In the case of UTF-8
+       matching, the advance must be one character, not one byte. Fudge the
+       offset values to achieve this. We won't be at the end of the string -
+       that was checked before setting g_notempty. */
        else
          {
-         if (count == -1) fprintf(outfile, "No match\n");
+         if (g_notempty != 0)
+           {
+           int onechar = 1;
+           use_offsets[0] = start_offset;
+           if (use_utf8)
+             {
+             while (start_offset + onechar < len)
+               {
+               int tb = bptr[start_offset+onechar];
+               if (tb <= 127) break;
+               tb &= 0xc0;
+               if (tb != 0 && tb != 0xc0) onechar++;
+               }
+             }
+           use_offsets[1] = start_offset + onechar;
+           }
+         else
+           {
+           if (count == PCRE_ERROR_NOMATCH)
+             {
+             if (gmatched == 0) fprintf(outfile, "No match\n");
+             }
            else fprintf(outfile, "Error %d\n", count);
+           break;  /* Out of the /g loop */
+           }
          }
-       }
-     }
+       /* If not /g or /G we are done */
+       if (!do_g && !do_G) break;
+       /* If we have matched an empty string, first check to see if we are at
+       the end of the subject. If so, the /g loop is over. Otherwise, mimic
+       what Perl's /g options does. This turns out to be rather cunning. First
+       we set PCRE_NOTEMPTY and PCRE_ANCHORED and try the match again at the
+       same point. If this fails (picked up above) we advance to the next
+       character. */
+       g_notempty = 0;
+       if (use_offsets[0] == use_offsets[1])
+         {
+         if (use_offsets[0] == len) break;
+         g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED;
+         }
+       /* For /g, update the start offset, leaving the rest alone */
+       if (do_g) start_offset = use_offsets[1];
+       /* For /G, update the pointer and length */
+       else
+         {
+         bptr += use_offsets[1];
+         len -= use_offsets[1];
+         }
+       }  /* End of loop for /g and /G */
+     NEXT_DATA: continue;
+     }    /* End of loop for data lines */
+   CONTINUE:
+ #if !defined NOPOSIX
    if (posix || do_posix) regfree(&preg);
-   if (re != NULL) free(re);
+ #endif
-   if (extra != NULL) free(extra);
+   if (re != NULL) new_free(re);
+   if (extra != NULL) new_free(extra);
+   if (tables != NULL)
+     {
+     new_free((void *)tables);
+     setlocale(LC_CTYPE, "C");
+     locale_set = 0;
+     }
    }
- END_OFF:
+ if (infile == stdin) fprintf(outfile, "\n");
- fprintf(outfile, "\n");
- return 0;
+ EXIT:
+ if (infile != NULL && infile != stdin) fclose(infile);
+ if (outfile != NULL && outfile != stdout) fclose(outfile);
+ free(buffer);
+ free(dbuffer);
+ free(pbuffer);
+ free(offsets);
+ return yield;
  }
- /* End */
+ /* End of pcretest.c */

 Legend:



Removed from v.9
 


changed lines


 
Added in v.107
 Legend:



Removed from v.9
 


changed lines


 
Added in v.107
-Removed from v.9
+Added in v.107

webmaster@exim.org	ViewVC Help
Powered by ViewVC 1.1.12