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

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

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

revision 427 by ph10, Fri Aug 28 09:55:54 2009 UTC revision 512 by ph10, Tue Mar 30 11:11:52 2010 UTC
# Line 118  external symbols to prevent clashes. */ Line 118  external symbols to prevent clashes. */
118    
119  /* We also need the pcre_printint() function for printing out compiled  /* We also need the pcre_printint() function for printing out compiled
120  patterns. This function is in a separate file so that it can be included in  patterns. This function is in a separate file so that it can be included in
121  pcre_compile.c when that module is compiled with debugging enabled.  pcre_compile.c when that module is compiled with debugging enabled. It needs to
122    know which case is being compiled. */
 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. */  
123    
124    #define COMPILING_PCRETEST
125  #include "pcre_printint.src"  #include "pcre_printint.src"
126    
127  #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))  /* The definition of the macro PRINTABLE, which determines whether to print an
128    output character as-is or as a hex value when showing compiled patterns, is
129    contained in the printint.src file. We uses it here also, in cases when the
130    locale has not been explicitly changed, so as to get consistent output from
131    systems that differ in their output from isprint() even in the "C" locale. */
132    
133    #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
134    
135  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
136  testing the POSIX interface, though this is not available via the standard  testing the POSIX interface, though this is not available via the standard
# Line 1039  while (!done) Line 1040  while (!done)
1040  #endif  #endif
1041    
1042    const char *error;    const char *error;
1043      unsigned char *markptr;
1044    unsigned char *p, *pp, *ppp;    unsigned char *p, *pp, *ppp;
1045    unsigned char *to_file = NULL;    unsigned char *to_file = NULL;
1046    const unsigned char *tables = NULL;    const unsigned char *tables = NULL;
1047    unsigned long int true_size, true_study_size = 0;    unsigned long int true_size, true_study_size = 0;
1048    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
1049      int do_mark = 0;
1050    int do_study = 0;    int do_study = 0;
1051    int do_debug = debug;    int do_debug = debug;
1052    int do_G = 0;    int do_G = 0;
# Line 1225  while (!done) Line 1228  while (!done)
1228        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1229        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1230        case 'J': options |= PCRE_DUPNAMES; break;        case 'J': options |= PCRE_DUPNAMES; break;
1231          case 'K': do_mark = 1; break;
1232        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1233        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1234    
# Line 1305  while (!done) Line 1309  while (!done)
1309      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1310      if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;      if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1311      if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;      if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1312        if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY;
1313    
1314      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1315    
# Line 1418  while (!done) Line 1423  while (!done)
1423          true_study_size = ((pcre_study_data *)(extra->study_data))->size;          true_study_size = ((pcre_study_data *)(extra->study_data))->size;
1424        }        }
1425    
1426        /* If /K was present, we set up for handling MARK data. */
1427    
1428        if (do_mark)
1429          {
1430          if (extra == NULL)
1431            {
1432            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1433            extra->flags = 0;
1434            }
1435          extra->mark = &markptr;
1436          extra->flags |= PCRE_EXTRA_MARK;
1437          }
1438    
1439      /* If the 'F' option was present, we flip the bytes of all the integer      /* If the 'F' option was present, we flip the bytes of all the integer
1440      fields in the regex data block and the study block. This is to make it      fields in the regex data block and the study block. This is to make it
1441      possible to test PCRE's handling of byte-flipped patterns, e.g. those      possible to test PCRE's handling of byte-flipped patterns, e.g. those
# Line 1450  while (!done) Line 1468  while (!done)
1468          {          {
1469          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
1470          rsd->size = byteflip(rsd->size, sizeof(rsd->size));          rsd->size = byteflip(rsd->size, sizeof(rsd->size));
1471          rsd->options = byteflip(rsd->options, sizeof(rsd->options));          rsd->flags = byteflip(rsd->flags, sizeof(rsd->flags));
1472            rsd->minlength = byteflip(rsd->minlength, sizeof(rsd->minlength));
1473          }          }
1474        }        }
1475    
# Line 1627  while (!done) Line 1646  while (!done)
1646          else          else
1647            {            {
1648            uschar *start_bits = NULL;            uschar *start_bits = NULL;
1649            new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);            int minlength;
1650    
1651              new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength);
1652              fprintf(outfile, "Subject length lower bound = %d\n", minlength);
1653    
1654              new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);
1655            if (start_bits == NULL)            if (start_bits == NULL)
1656              fprintf(outfile, "No starting byte set\n");              fprintf(outfile, "No set of starting bytes\n");
1657            else            else
1658              {              {
1659              int i;              int i;
# Line 1969  while (!done) Line 1992  while (!done)
1992          continue;          continue;
1993    
1994          case 'N':          case 'N':
1995          options |= PCRE_NOTEMPTY;          if ((options & PCRE_NOTEMPTY) != 0)
1996              options = (options & ~PCRE_NOTEMPTY) | PCRE_NOTEMPTY_ATSTART;
1997            else
1998              options |= PCRE_NOTEMPTY;
1999          continue;          continue;
2000    
2001          case 'O':          case 'O':
# Line 1992  while (!done) Line 2018  while (!done)
2018          continue;          continue;
2019    
2020          case 'P':          case 'P':
2021          options |= ((options & PCRE_PARTIAL_SOFT) == 0)?          options |= ((options & PCRE_PARTIAL_SOFT) == 0)?
2022            PCRE_PARTIAL_SOFT : PCRE_PARTIAL_HARD;            PCRE_PARTIAL_SOFT : PCRE_PARTIAL_HARD;
2023          continue;          continue;
2024    
# Line 2135  while (!done) Line 2161  while (!done)
2161    
2162      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
2163        {        {
2164          markptr = NULL;
2165    
2166        if (timeitm > 0)        if (timeitm > 0)
2167          {          {
2168          register int i;          register int i;
# Line 2146  while (!done) Line 2174  while (!done)
2174            {            {
2175            int workspace[1000];            int workspace[1000];
2176            for (i = 0; i < timeitm; i++)            for (i = 0; i < timeitm; i++)
2177              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,              count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2178                options | g_notempty, use_offsets, use_size_offsets, workspace,                options | g_notempty, use_offsets, use_size_offsets, workspace,
2179                sizeof(workspace)/sizeof(int));                sizeof(workspace)/sizeof(int));
2180            }            }
# Line 2209  while (!done) Line 2237  while (!done)
2237        else if (all_use_dfa || use_dfa)        else if (all_use_dfa || use_dfa)
2238          {          {
2239          int workspace[1000];          int workspace[1000];
2240          count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,          count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2241            options | g_notempty, use_offsets, use_size_offsets, workspace,            options | g_notempty, use_offsets, use_size_offsets, workspace,
2242            sizeof(workspace)/sizeof(int));            sizeof(workspace)/sizeof(int));
2243          if (count == 0)          if (count == 0)
# Line 2280  while (!done) Line 2308  while (!done)
2308              }              }
2309            }            }
2310    
2311            if (markptr != NULL) fprintf(outfile, "MK: %s\n", markptr);
2312    
2313          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2314            {            {
2315            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
# Line 2363  while (!done) Line 2393  while (!done)
2393    
2394        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2395          {          {
2396          fprintf(outfile, "Partial match");          if (markptr == NULL) fprintf(outfile, "Partial match");
2397              else fprintf(outfile, "Partial match, mark=%s", markptr);
2398          if (use_size_offsets > 1)          if (use_size_offsets > 1)
2399            {            {
2400            fprintf(outfile, ": ");            fprintf(outfile, ": ");
2401            pchars(bptr + use_offsets[0], use_offsets[1] - use_offsets[0],            pchars(bptr + use_offsets[0], use_offsets[1] - use_offsets[0],
2402              outfile);              outfile);
2403            }            }
2404          fprintf(outfile, "\n");          fprintf(outfile, "\n");
2405          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2406          }          }
# Line 2430  while (!done) Line 2461  while (!done)
2461            {            {
2462            if (count == PCRE_ERROR_NOMATCH)            if (count == PCRE_ERROR_NOMATCH)
2463              {              {
2464              if (gmatched == 0) fprintf(outfile, "No match\n");              if (gmatched == 0)
2465                  {
2466                  if (markptr == NULL) fprintf(outfile, "No match\n");
2467                    else fprintf(outfile, "No match, mark = %s\n", markptr);
2468                  }
2469              }              }
2470            else fprintf(outfile, "Error %d\n", count);            else fprintf(outfile, "Error %d\n", count);
2471            break;  /* Out of the /g loop */            break;  /* Out of the /g loop */
# Line 2442  while (!done) Line 2477  while (!done)
2477        if (!do_g && !do_G) break;        if (!do_g && !do_G) break;
2478    
2479        /* 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
2480        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
2481        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
2482        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
2483        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
2484        character. */        character. */
2485    
# Line 2453  while (!done) Line 2488  while (!done)
2488        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2489          {          {
2490          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
2491          g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED;          g_notempty = PCRE_NOTEMPTY_ATSTART | PCRE_ANCHORED;
2492          }          }
2493    
2494        /* For /g, update the start offset, leaving the rest alone */        /* For /g, update the start offset, leaving the rest alone */

Legend:
Removed from v.427  
changed lines
  Added in v.512

  ViewVC Help
Powered by ViewVC 1.1.5