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

Diff of /code/trunk/pcretest.c

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

revision 355 by ph10, Mon Jul 7 17:45:23 2008 UTC revision 455 by ph10, Sat Sep 26 19:12:32 2009 UTC
# Line 71  input mode under Windows. */ Line 71  input mode under Windows. */
71  #define INPUT_MODE   "r"  #define INPUT_MODE   "r"
72  #define OUTPUT_MODE  "wb"  #define OUTPUT_MODE  "wb"
73    
74  #define isatty _isatty         /* This is what Windows calls them, I'm told */  #ifndef isatty
75    #define isatty _isatty         /* This is what Windows calls them, I'm told, */
76    #endif                         /* though in some environments they seem to   */
77                                   /* be already defined, hence the #ifndefs.    */
78    #ifndef fileno
79  #define fileno _fileno  #define fileno _fileno
80    #endif
81    
82  #else  #else
83  #include <sys/time.h>          /* These two includes are needed */  #include <sys/time.h>          /* These two includes are needed */
# Line 785  printf(" -dfa force DFA matching fo Line 790  printf(" -dfa force DFA matching fo
790  #endif  #endif
791  printf("  -help    show usage information\n");  printf("  -help    show usage information\n");
792  printf("  -i       show information about compiled patterns\n"  printf("  -i       show information about compiled patterns\n"
793           "  -M       find MATCH_LIMIT minimum for each subject\n"
794         "  -m       output memory used information\n"         "  -m       output memory used information\n"
795         "  -o <n>   set size of offsets vector to <n>\n");         "  -o <n>   set size of offsets vector to <n>\n");
796  #if !defined NOPOSIX  #if !defined NOPOSIX
# Line 814  int main(int argc, char **argv) Line 820  int main(int argc, char **argv)
820  FILE *infile = stdin;  FILE *infile = stdin;
821  int options = 0;  int options = 0;
822  int study_options = 0;  int study_options = 0;
823    int default_find_match_limit = FALSE;
824  int op = 1;  int op = 1;
825  int timeit = 0;  int timeit = 0;
826  int timeitm = 0;  int timeitm = 0;
# Line 873  while (argc > 1 && argv[op][0] == '-') Line 880  while (argc > 1 && argv[op][0] == '-')
880    else if (strcmp(argv[op], "-b") == 0) debug = 1;    else if (strcmp(argv[op], "-b") == 0) debug = 1;
881    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
882    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
883      else if (strcmp(argv[op], "-M") == 0) default_find_match_limit = TRUE;
884  #if !defined NODFA  #if !defined NODFA
885    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
886  #endif  #endif
# Line 925  while (argc > 1 && argv[op][0] == '-') Line 933  while (argc > 1 && argv[op][0] == '-')
933    else if (strcmp(argv[op], "-C") == 0)    else if (strcmp(argv[op], "-C") == 0)
934      {      {
935      int rc;      int rc;
936        unsigned long int lrc;
937      printf("PCRE version %s\n", pcre_version());      printf("PCRE version %s\n", pcre_version());
938      printf("Compiled with\n");      printf("Compiled with\n");
939      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
# Line 932  while (argc > 1 && argv[op][0] == '-') Line 941  while (argc > 1 && argv[op][0] == '-')
941      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
942      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
943      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
944      printf("  Newline sequence is %s\n", (rc == '\r')? "CR" :      /* Note that these values are always the ASCII values, even
945        (rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" :      in EBCDIC environments. CR is 13 and NL is 10. */
946        printf("  Newline sequence is %s\n", (rc == 13)? "CR" :
947          (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
948        (rc == -2)? "ANYCRLF" :        (rc == -2)? "ANYCRLF" :
949        (rc == -1)? "ANY" : "???");        (rc == -1)? "ANY" : "???");
950      (void)pcre_config(PCRE_CONFIG_BSR, &rc);      (void)pcre_config(PCRE_CONFIG_BSR, &rc);
# Line 943  while (argc > 1 && argv[op][0] == '-') Line 954  while (argc > 1 && argv[op][0] == '-')
954      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
955      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
956      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
957      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &lrc);
958      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %ld\n", lrc);
959      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &lrc);
960      printf("  Default recursion depth limit = %d\n", rc);      printf("  Default recursion depth limit = %ld\n", lrc);
961      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
962      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
963      goto EXIT;      goto EXIT;
# Line 1294  while (!done) Line 1305  while (!done)
1305      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1306      if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;      if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1307      if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;      if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1308        if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY;
1309    
1310      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1311    
# Line 1314  while (!done) Line 1326  while (!done)
1326  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1327    
1328      {      {
1329        unsigned long int get_options;
1330    
1331      if (timeit > 0)      if (timeit > 0)
1332        {        {
1333        register int i;        register int i;
# Line 1357  while (!done) Line 1371  while (!done)
1371        goto CONTINUE;        goto CONTINUE;
1372        }        }
1373    
1374      /* Compilation succeeded; print data if required. There are now two      /* Compilation succeeded. It is now possible to set the UTF-8 option from
1375      info-returning functions. The old one has a limited interface and      within the regex; check for this so that we know how to process the data
1376      returns only limited data. Check that it agrees with the newer one. */      lines. */
1377    
1378        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1379        if ((get_options & PCRE_UTF8) != 0) use_utf8 = 1;
1380    
1381        /* Print information if required. There are now two info-returning
1382        functions. The old one has a limited interface and returns only limited
1383        data. Check that it agrees with the newer one. */
1384    
1385      if (log_store)      if (log_store)
1386        fprintf(outfile, "Memory allocation (code space): %d\n",        fprintf(outfile, "Memory allocation (code space): %d\n",
# Line 1430  while (!done) Line 1451  while (!done)
1451          {          {
1452          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
1453          rsd->size = byteflip(rsd->size, sizeof(rsd->size));          rsd->size = byteflip(rsd->size, sizeof(rsd->size));
1454          rsd->options = byteflip(rsd->options, sizeof(rsd->options));          rsd->flags = byteflip(rsd->flags, sizeof(rsd->flags));
1455            rsd->minlength = byteflip(rsd->minlength, sizeof(rsd->minlength));
1456          }          }
1457        }        }
1458    
# Line 1444  while (!done) Line 1466  while (!done)
1466        pcre_printint(re, outfile, debug_lengths);        pcre_printint(re, outfile, debug_lengths);
1467        }        }
1468    
1469        /* We already have the options in get_options (see above) */
1470    
1471      if (do_showinfo)      if (do_showinfo)
1472        {        {
1473        unsigned long int get_options, all_options;        unsigned long int all_options;
1474  #if !defined NOINFOCHECK  #if !defined NOINFOCHECK
1475        int old_first_char, old_options, old_count;        int old_first_char, old_options, old_count;
1476  #endif  #endif
# Line 1455  while (!done) Line 1479  while (!done)
1479        int nameentrysize, namecount;        int nameentrysize, namecount;
1480        const uschar *nametable;        const uschar *nametable;
1481    
       new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);  
1482        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1483        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
1484        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);
# Line 1606  while (!done) Line 1629  while (!done)
1629          else          else
1630            {            {
1631            uschar *start_bits = NULL;            uschar *start_bits = NULL;
1632              int minlength;
1633    
1634              new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength);
1635              fprintf(outfile, "Subject length lower bound = %d\n", minlength);
1636    
1637            new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);            new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);
   
1638            if (start_bits == NULL)            if (start_bits == NULL)
1639              fprintf(outfile, "No starting byte set\n");              fprintf(outfile, "No set of starting bytes\n");
1640            else            else
1641              {              {
1642              int i;              int i;
# Line 1708  while (!done) Line 1735  while (!done)
1735      int callout_data_set = 0;      int callout_data_set = 0;
1736      int count, c;      int count, c;
1737      int copystrings = 0;      int copystrings = 0;
1738      int find_match_limit = 0;      int find_match_limit = default_find_match_limit;
1739      int getstrings = 0;      int getstrings = 0;
1740      int getlist = 0;      int getlist = 0;
1741      int gmatched = 0;      int gmatched = 0;
# Line 1807  while (!done) Line 1834  while (!done)
1834              unsigned char buff8[8];              unsigned char buff8[8];
1835              int ii, utn;              int ii, utn;
1836              if (use_utf8)              if (use_utf8)
1837                {                {
1838                utn = ord2utf8(c, buff8);                utn = ord2utf8(c, buff8);
1839                for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];                for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1840                c = buff8[ii];   /* Last byte */                c = buff8[ii];   /* Last byte */
1841                }                }
1842              else              else
1843               {               {
1844               if (c > 255)               if (c > 255)
1845                 fprintf(outfile, "** Character \\x{%x} is greater than 255 and "                 fprintf(outfile, "** Character \\x{%x} is greater than 255 and "
1846                   "UTF-8 mode is not enabled.\n"                   "UTF-8 mode is not enabled.\n"
1847                   "** Truncation will probably give the wrong result.\n", c);                   "** Truncation will probably give the wrong result.\n", c);
1848               }               }
1849              p = pt + 1;              p = pt + 1;
1850              break;              break;
1851              }              }
# Line 1948  while (!done) Line 1975  while (!done)
1975          continue;          continue;
1976    
1977          case 'N':          case 'N':
1978          options |= PCRE_NOTEMPTY;          if ((options & PCRE_NOTEMPTY) != 0)
1979              options = (options & ~PCRE_NOTEMPTY) | PCRE_NOTEMPTY_ATSTART;
1980            else
1981              options |= PCRE_NOTEMPTY;
1982          continue;          continue;
1983    
1984          case 'O':          case 'O':
# Line 1971  while (!done) Line 2001  while (!done)
2001          continue;          continue;
2002    
2003          case 'P':          case 'P':
2004          options |= PCRE_PARTIAL;          options |= ((options & PCRE_PARTIAL_SOFT) == 0)?
2005              PCRE_PARTIAL_SOFT : PCRE_PARTIAL_HARD;
2006          continue;          continue;
2007    
2008          case 'Q':          case 'Q':
# Line 2006  while (!done) Line 2037  while (!done)
2037          show_malloc = 1;          show_malloc = 1;
2038          continue;          continue;
2039    
2040            case 'Y':
2041            options |= PCRE_NO_START_OPTIMIZE;
2042            continue;
2043    
2044          case 'Z':          case 'Z':
2045          options |= PCRE_NOTEOL;          options |= PCRE_NOTEOL;
2046          continue;          continue;
# Line 2028  while (!done) Line 2063  while (!done)
2063      *q = 0;      *q = 0;
2064      len = q - dbuffer;      len = q - dbuffer;
2065    
2066        /* Move the data to the end of the buffer so that a read over the end of
2067        the buffer will be seen by valgrind, even if it doesn't cause a crash. If
2068        we are using the POSIX interface, we must include the terminating zero. */
2069    
2070    #if !defined NOPOSIX
2071        if (posix || do_posix)
2072          {
2073          memmove(bptr + buffer_size - len - 1, bptr, len + 1);
2074          bptr += buffer_size - len - 1;
2075          }
2076        else
2077    #endif
2078          {
2079          memmove(bptr + buffer_size - len, bptr, len);
2080          bptr += buffer_size - len;
2081          }
2082    
2083      if ((all_use_dfa || use_dfa) && find_match_limit)      if ((all_use_dfa || use_dfa) && find_match_limit)
2084        {        {
2085        printf("**Match limit not relevant for DFA matching: ignored\n");        printf("**Match limit not relevant for DFA matching: ignored\n");
# Line 2047  while (!done) Line 2099  while (!done)
2099          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);
2100        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;
2101        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;
2102          if ((options & PCRE_NOTEMPTY) != 0) eflags |= REG_NOTEMPTY;
2103    
2104        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);
2105    
# Line 2102  while (!done) Line 2155  while (!done)
2155            {            {
2156            int workspace[1000];            int workspace[1000];
2157            for (i = 0; i < timeitm; i++)            for (i = 0; i < timeitm; i++)
2158              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,              count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2159                options | g_notempty, use_offsets, use_size_offsets, workspace,                options | g_notempty, use_offsets, use_size_offsets, workspace,
2160                sizeof(workspace)/sizeof(int));                sizeof(workspace)/sizeof(int));
2161            }            }
# Line 2165  while (!done) Line 2218  while (!done)
2218        else if (all_use_dfa || use_dfa)        else if (all_use_dfa || use_dfa)
2219          {          {
2220          int workspace[1000];          int workspace[1000];
2221          count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,          count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2222            options | g_notempty, use_offsets, use_size_offsets, workspace,            options | g_notempty, use_offsets, use_size_offsets, workspace,
2223            sizeof(workspace)/sizeof(int));            sizeof(workspace)/sizeof(int));
2224          if (count == 0)          if (count == 0)
# Line 2320  while (!done) Line 2373  while (!done)
2373        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2374          {          {
2375          fprintf(outfile, "Partial match");          fprintf(outfile, "Partial match");
2376  #if !defined NODFA          if (use_size_offsets > 1)
2377          if ((all_use_dfa || use_dfa) && use_size_offsets > 2)            {
2378            fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],            fprintf(outfile, ": ");
2379              bptr + use_offsets[0]);            pchars(bptr + use_offsets[0], use_offsets[1] - use_offsets[0],
2380  #endif              outfile);
2381              }
2382          fprintf(outfile, "\n");          fprintf(outfile, "\n");
2383          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2384          }          }
# Line 2354  while (!done) Line 2408  while (!done)
2408              {              {
2409              int d;              int d;
2410              (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);              (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);
2411              obits = (d == '\r')? PCRE_NEWLINE_CR :              /* Note that these values are always the ASCII ones, even in
2412                      (d == '\n')? PCRE_NEWLINE_LF :              EBCDIC environments. CR = 13, NL = 10. */
2413                      (d == ('\r'<<8 | '\n'))? PCRE_NEWLINE_CRLF :              obits = (d == 13)? PCRE_NEWLINE_CR :
2414                        (d == 10)? PCRE_NEWLINE_LF :
2415                        (d == (13<<8 | 10))? PCRE_NEWLINE_CRLF :
2416                      (d == -2)? PCRE_NEWLINE_ANYCRLF :                      (d == -2)? PCRE_NEWLINE_ANYCRLF :
2417                      (d == -1)? PCRE_NEWLINE_ANY : 0;                      (d == -1)? PCRE_NEWLINE_ANY : 0;
2418              }              }
# Line 2395  while (!done) Line 2451  while (!done)
2451        if (!do_g && !do_G) break;        if (!do_g && !do_G) break;
2452    
2453        /* If we have matched an empty string, first check to see if we are at        /* If we have matched an empty string, first check to see if we are at
2454        the end of the subject. If so, the /g loop is over. Otherwise, mimic        the end of the subject. If so, the /g loop is over. Otherwise, mimic what
2455        what Perl's /g options does. This turns out to be rather cunning. First        Perl's /g options does. This turns out to be rather cunning. First we set
2456        we set PCRE_NOTEMPTY and PCRE_ANCHORED and try the match again at the        PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED and try the match again at the
2457        same point. If this fails (picked up above) we advance to the next        same point. If this fails (picked up above) we advance to the next
2458        character. */        character. */
2459    
# Line 2406  while (!done) Line 2462  while (!done)
2462        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2463          {          {
2464          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
2465          g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED;          g_notempty = PCRE_NOTEMPTY_ATSTART | PCRE_ANCHORED;
2466          }          }
2467    
2468        /* For /g, update the start offset, leaving the rest alone */        /* For /g, update the start offset, leaving the rest alone */
Legend:
Removed from v.355  
changed lines
  Added in v.455
webmaster@exim.org
 ViewVC Help
Powered by ViewVC 1.1.12