/[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 545 by ph10, Wed Jun 16 10:51:15 2010 UTC revision 612 by ph10, Sat Jul 2 15:20:59 2011 UTC
# Line 79  input mode under Windows. */ Line 79  input mode under Windows. */
79  #define fileno _fileno  #define fileno _fileno
80  #endif  #endif
81    
82    /* A user sent this fix for Borland Builder 5 under Windows. */
83    
84    #ifdef __BORLANDC__
85    #define _setmode(handle, mode) setmode(handle, mode)
86    #endif
87    
88    /* Not Windows */
89    
90  #else  #else
91  #include <sys/time.h>          /* These two includes are needed */  #include <sys/time.h>          /* These two includes are needed */
92  #include <sys/resource.h>      /* for setrlimit(). */  #include <sys/resource.h>      /* for setrlimit(). */
# Line 189  static uschar *buffer = NULL; Line 197  static uschar *buffer = NULL;
197  static uschar *dbuffer = NULL;  static uschar *dbuffer = NULL;
198  static uschar *pbuffer = NULL;  static uschar *pbuffer = NULL;
199    
200    /* Textual explanations for runtime error codes */
201    
202    static const char *errtexts[] = {
203      NULL,  /* 0 is no error */
204      NULL,  /* NOMATCH is handled specially */
205      "NULL argument passed",
206      "bad option value",
207      "magic number missing",
208      "unknown opcode - pattern overwritten?",
209      "no more memory",
210      NULL,  /* never returned by pcre_exec() or pcre_dfa_exec() */
211      "match limit exceeded",
212      "callout error code",
213      NULL,  /* BADUTF8 is handled specially */
214      "bad UTF-8 offset",
215      NULL,  /* PARTIAL is handled specially */
216      "not used - internal error",
217      "internal error - pattern overwritten?",
218      "bad count value",
219      "item unsupported for DFA matching",
220      "backreference condition or recursion test not supported for DFA matching",
221      "match limit not supported for DFA matching",
222      "workspace size exceeded in DFA matching",
223      "too much recursion for DFA matching",
224      "recursion limit exceeded",
225      "not used - internal error",
226      "invalid combination of newline options",
227      "bad offset value",
228      NULL  /* SHORTUTF8 is handled specially */
229    };
230    
231    
232  /*************************************************  /*************************************************
233  *         Alternate character tables             *  *         Alternate character tables             *
# Line 516  static const unsigned char tables1[] = { Line 555  static const unsigned char tables1[] = {
555    
556    
557    
558    
559    #ifndef HAVE_STRERROR
560    /*************************************************
561    *     Provide strerror() for non-ANSI libraries  *
562    *************************************************/
563    
564    /* Some old-fashioned systems still around (e.g. SunOS4) don't have strerror()
565    in their libraries, but can provide the same facility by this simple
566    alternative function. */
567    
568    extern int   sys_nerr;
569    extern char *sys_errlist[];
570    
571    char *
572    strerror(int n)
573    {
574    if (n < 0 || n >= sys_nerr) return "unknown error number";
575    return sys_errlist[n];
576    }
577    #endif /* HAVE_STRERROR */
578    
579    
580    
581    
582  /*************************************************  /*************************************************
583  *        Read or extend an input line            *  *        Read or extend an input line            *
584  *************************************************/  *************************************************/
# Line 1123  printf("  -p       use POSIX interface\n Line 1186  printf("  -p       use POSIX interface\n
1186  #endif  #endif
1187  printf("  -q       quiet: do not output PCRE version number at start\n");  printf("  -q       quiet: do not output PCRE version number at start\n");
1188  printf("  -S <n>   set stack size to <n> megabytes\n");  printf("  -S <n>   set stack size to <n> megabytes\n");
1189  printf("  -s       output store (memory) used information\n"  printf("  -s       force each pattern to be studied\n"
1190         "  -t       time compilation and execution\n");         "  -t       time compilation and execution\n");
1191  printf("  -t <n>   time compilation and execution, repeating <n> times\n");  printf("  -t <n>   time compilation and execution, repeating <n> times\n");
1192  printf("  -tm      time execution (matching) only\n");  printf("  -tm      time execution (matching) only\n");
# Line 1151  int timeit = 0; Line 1214  int timeit = 0;
1214  int timeitm = 0;  int timeitm = 0;
1215  int showinfo = 0;  int showinfo = 0;
1216  int showstore = 0;  int showstore = 0;
1217    int force_study = 0;
1218  int quiet = 0;  int quiet = 0;
1219  int size_offsets = 45;  int size_offsets = 45;
1220  int size_offsets_max;  int size_offsets_max;
# Line 1199  while (argc > 1 && argv[op][0] == '-') Line 1263  while (argc > 1 && argv[op][0] == '-')
1263    {    {
1264    unsigned char *endptr;    unsigned char *endptr;
1265    
1266    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-m") == 0) showstore = 1;
1267      showstore = 1;    else if (strcmp(argv[op], "-s") == 0) force_study = 1;
1268    else if (strcmp(argv[op], "-q") == 0) quiet = 1;    else if (strcmp(argv[op], "-q") == 0) quiet = 1;
1269    else if (strcmp(argv[op], "-b") == 0) debug = 1;    else if (strcmp(argv[op], "-b") == 0) debug = 1;
1270    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
# Line 1372  while (!done) Line 1436  while (!done)
1436    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
1437    int do_mark = 0;    int do_mark = 0;
1438    int do_study = 0;    int do_study = 0;
1439      int no_force_study = 0;
1440    int do_debug = debug;    int do_debug = debug;
1441    int do_G = 0;    int do_G = 0;
1442    int do_g = 0;    int do_g = 0;
# Line 1438  while (!done) Line 1503  while (!done)
1503          }          }
1504        }        }
1505    
1506      fprintf(outfile, "Compiled regex%s loaded from %s\n",      fprintf(outfile, "Compiled pattern%s loaded from %s\n",
1507        do_flip? " (byte-inverted)" : "", p);        do_flip? " (byte-inverted)" : "", p);
1508    
1509      /* Need to know if UTF-8 for printing data strings */      /* Need to know if UTF-8 for printing data strings */
# Line 1446  while (!done) Line 1511  while (!done)
1511      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1512      use_utf8 = (get_options & PCRE_UTF8) != 0;      use_utf8 = (get_options & PCRE_UTF8) != 0;
1513    
1514      /* Now see if there is any following study data */      /* Now see if there is any following study data. */
1515    
1516      if (true_study_size != 0)      if (true_study_size != 0)
1517        {        {
# Line 1560  while (!done) Line 1625  while (!done)
1625        case 'P': do_posix = 1; break;        case 'P': do_posix = 1; break;
1626  #endif  #endif
1627    
1628        case 'S': do_study = 1; break;        case 'S':
1629          if (do_study == 0) do_study = 1; else
1630            {
1631            do_study = 0;
1632            no_force_study = 1;
1633            }
1634          break;
1635    
1636        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1637        case 'W': options |= PCRE_UCP; break;        case 'W': options |= PCRE_UCP; break;
1638        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
1639          case 'Y': options |= PCRE_NO_START_OPTIMISE; break;
1640        case 'Z': debug_lengths = 0; break;        case 'Z': debug_lengths = 0; break;
1641        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1642        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
# Line 1743  while (!done) Line 1816  while (!done)
1816      true_size = ((real_pcre *)re)->size;      true_size = ((real_pcre *)re)->size;
1817      regex_gotten_store = gotten_store;      regex_gotten_store = gotten_store;
1818    
1819      /* If /S was present, study the regexp to generate additional info to      /* If -s or /S was present, study the regex to generate additional info to
1820      help with the matching. */      help with the matching, unless the pattern has the SS option, which
1821        suppresses the effect of /S (used for a few test patterns where studying is
1822        never sensible). */
1823    
1824      if (do_study)      if (do_study || (force_study && !no_force_study))
1825        {        {
1826        if (timeit > 0)        if (timeit > 0)
1827          {          {
# Line 1900  while (!done) Line 1975  while (!done)
1975        if (do_flip) all_options = byteflip(all_options, sizeof(all_options));        if (do_flip) all_options = byteflip(all_options, sizeof(all_options));
1976    
1977        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1978          else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s\n",          else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s\n",
1979            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1980            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1981            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
# Line 1916  while (!done) Line 1991  while (!done)
1991            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1992            ((get_options & PCRE_UCP) != 0)? " ucp" : "",            ((get_options & PCRE_UCP) != 0)? " ucp" : "",
1993            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1994              ((get_options & PCRE_NO_START_OPTIMIZE) != 0)? " no_start_optimize" : "",
1995            ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");            ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1996    
1997        if (jchanged) fprintf(outfile, "Duplicate name status changes\n");        if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
# Line 1983  while (!done) Line 2059  while (!done)
2059        /* Don't output study size; at present it is in any case a fixed        /* Don't output study size; at present it is in any case a fixed
2060        value, but it varies, depending on the computer architecture, and        value, but it varies, depending on the computer architecture, and
2061        so messes up the test suite. (And with the /F option, it might be        so messes up the test suite. (And with the /F option, it might be
2062        flipped.) */        flipped.) If study was forced by an external -s, don't show this
2063          information unless -i or -d was also present. This means that, except
2064          when auto-callouts are involved, the output from runs with and without
2065          -s should be identical. */
2066    
2067        if (do_study)        if (do_study || (force_study && showinfo && !no_force_study))
2068          {          {
2069          if (extra == NULL)          if (extra == NULL)
2070            fprintf(outfile, "Study returned NULL\n");            fprintf(outfile, "Study returned NULL\n");
# Line 2063  while (!done) Line 2142  while (!done)
2142            }            }
2143          else          else
2144            {            {
2145            fprintf(outfile, "Compiled regex written to %s\n", to_file);            fprintf(outfile, "Compiled pattern written to %s\n", to_file);
2146    
2147              /* If there is study data, write it, but verify the writing only
2148              if the studying was requested by /S, not just by -s. */
2149    
2150            if (extra != NULL)            if (extra != NULL)
2151              {              {
2152              if (fwrite(extra->study_data, 1, true_study_size, f) <              if (fwrite(extra->study_data, 1, true_study_size, f) <
# Line 2073  while (!done) Line 2156  while (!done)
2156                  strerror(errno));                  strerror(errno));
2157                }                }
2158              else fprintf(outfile, "Study data written to %s\n", to_file);              else fprintf(outfile, "Study data written to %s\n", to_file);
   
2159              }              }
2160            }            }
2161          fclose(f);          fclose(f);
# Line 2108  while (!done) Line 2190  while (!done)
2190      int getlist = 0;      int getlist = 0;
2191      int gmatched = 0;      int gmatched = 0;
2192      int start_offset = 0;      int start_offset = 0;
2193        int start_offset_sign = 1;
2194      int g_notempty = 0;      int g_notempty = 0;
2195      int use_dfa = 0;      int use_dfa = 0;
2196    
# Line 2240  while (!done) Line 2323  while (!done)
2323          continue;          continue;
2324    
2325          case '>':          case '>':
2326            if (*p == '-')
2327              {
2328              start_offset_sign = -1;
2329              p++;
2330              }
2331          while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';          while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';
2332            start_offset *= start_offset_sign;
2333          continue;          continue;
2334    
2335          case 'A':  /* Option setting */          case 'A':  /* Option setting */
# Line 2313  while (!done) Line 2402  while (!done)
2402  #endif  #endif
2403            use_dfa = 1;            use_dfa = 1;
2404          continue;          continue;
2405    #endif
2406    
2407    #if !defined NODFA
2408          case 'F':          case 'F':
2409          options |= PCRE_DFA_SHORTEST;          options |= PCRE_DFA_SHORTEST;
2410          continue;          continue;
# Line 2765  while (!done) Line 2856  while (!done)
2856        to advance the start offset, and continue. We won't be at the end of the        to advance the start offset, and continue. We won't be at the end of the
2857        string - that was checked before setting g_notempty.        string - that was checked before setting g_notempty.
2858    
2859        Complication arises in the case when the newline option is "any" or        Complication arises in the case when the newline convention is "any",
2860        "anycrlf". If the previous match was at the end of a line terminated by        "crlf", or "anycrlf". If the previous match was at the end of a line
2861        CRLF, an advance of one character just passes the \r, whereas we should        terminated by CRLF, an advance of one character just passes the \r,
2862        prefer the longer newline sequence, as does the code in pcre_exec().        whereas we should prefer the longer newline sequence, as does the code in
2863        Fudge the offset value to achieve this.        pcre_exec(). Fudge the offset value to achieve this. We check for a
2864          newline setting in the pattern; if none was set, use pcre_config() to
2865          find the default.
2866    
2867        Otherwise, in the case of UTF-8 matching, the advance must be one        Otherwise, in the case of UTF-8 matching, the advance must be one
2868        character, not one byte. */        character, not one byte. */
# Line 2794  while (!done) Line 2887  while (!done)
2887                      (d == -1)? PCRE_NEWLINE_ANY : 0;                      (d == -1)? PCRE_NEWLINE_ANY : 0;
2888              }              }
2889            if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||            if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2890                   (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_CRLF ||
2891                 (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)                 (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2892                &&                &&
2893                start_offset < len - 1 &&                start_offset < len - 1 &&
# Line 2804  while (!done) Line 2898  while (!done)
2898              {              {
2899              while (start_offset + onechar < len)              while (start_offset + onechar < len)
2900                {                {
2901                int tb = bptr[start_offset+onechar];                if ((bptr[start_offset+onechar] & 0xc0) != 0x80) break;
2902                if (tb <= 127) break;                onechar++;
               tb &= 0xc0;  
               if (tb != 0 && tb != 0xc0) onechar++;  
2903                }                }
2904              }              }
2905            use_offsets[1] = start_offset + onechar;            use_offsets[1] = start_offset + onechar;
2906            }            }
2907          else          else
2908            {            {
2909            if (count == PCRE_ERROR_NOMATCH)            switch(count)
2910              {              {
2911                case PCRE_ERROR_NOMATCH:
2912              if (gmatched == 0)              if (gmatched == 0)
2913                {                {
2914                if (markptr == NULL) fprintf(outfile, "No match\n");                if (markptr == NULL) fprintf(outfile, "No match\n");
2915                  else fprintf(outfile, "No match, mark = %s\n", markptr);                  else fprintf(outfile, "No match, mark = %s\n", markptr);
2916                }                }
2917                break;
2918    
2919                case PCRE_ERROR_BADUTF8:
2920                case PCRE_ERROR_SHORTUTF8:
2921                fprintf(outfile, "Error %d (%s UTF-8 string)", count,
2922                  (count == PCRE_ERROR_BADUTF8)? "bad" : "short");
2923                if (use_size_offsets >= 2)
2924                  fprintf(outfile, " offset=%d reason=%d", use_offsets[0],
2925                    use_offsets[1]);
2926                fprintf(outfile, "\n");
2927                break;
2928    
2929                default:
2930                if (count < 0 && (-count) < sizeof(errtexts)/sizeof(const char *))
2931                  fprintf(outfile, "Error %d (%s)\n", count, errtexts[-count]);
2932                else
2933                  fprintf(outfile, "Error %d (Unexpected value)\n", count);
2934                break;
2935              }              }
2936            else fprintf(outfile, "Error %d\n", count);  
2937            break;  /* Out of the /g loop */            break;  /* Out of the /g loop */
2938            }            }
2939          }          }

Legend:
Removed from v.545  
changed lines
  Added in v.612

  ViewVC Help
Powered by ViewVC 1.1.5