/[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 259 by ph10, Wed Sep 19 09:52:44 2007 UTC revision 391 by ph10, Tue Mar 17 21:16:01 2009 UTC
# Line 48  POSSIBILITY OF SUCH DAMAGE. Line 48  POSSIBILITY OF SUCH DAMAGE.
48  #include <locale.h>  #include <locale.h>
49  #include <errno.h>  #include <errno.h>
50    
51    #ifdef SUPPORT_LIBREADLINE
52    #ifdef HAVE_UNISTD_H
53    #include <unistd.h>
54    #endif
55    #include <readline/readline.h>
56    #include <readline/history.h>
57    #endif
58    
59    
60  /* A number of things vary for Windows builds. Originally, pcretest opened its  /* A number of things vary for Windows builds. Originally, pcretest opened its
61  input and output without "b"; then I was told that "b" was needed in some  input and output without "b"; then I was told that "b" was needed in some
# Line 63  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 */
75    #define fileno _fileno
76    
77  #else  #else
78  #include <sys/time.h>          /* These two includes are needed */  #include <sys/time.h>          /* These two includes are needed */
79  #include <sys/resource.h>      /* for setrlimit(). */  #include <sys/resource.h>      /* for setrlimit(). */
# Line 83  appropriately for an application, not fo Line 94  appropriately for an application, not fo
94  #include "pcre.h"  #include "pcre.h"
95  #include "pcre_internal.h"  #include "pcre_internal.h"
96    
97  /* We need access to the data tables that PCRE uses. So as not to have to keep  /* We need access to some of the data tables that PCRE uses. So as not to have
98  two copies, we include the source file here, changing the names of the external  to keep two copies, we include the source file here, changing the names of the
99  symbols to prevent clashes. */  external symbols to prevent clashes. */
100    
101    #define _pcre_ucp_gentype      ucp_gentype
102  #define _pcre_utf8_table1      utf8_table1  #define _pcre_utf8_table1      utf8_table1
103  #define _pcre_utf8_table1_size utf8_table1_size  #define _pcre_utf8_table1_size utf8_table1_size
104  #define _pcre_utf8_table2      utf8_table2  #define _pcre_utf8_table2      utf8_table2
# Line 189  optimal way of handling this, but hey, t Line 201  optimal way of handling this, but hey, t
201  Arguments:  Arguments:
202    f            the file to read    f            the file to read
203    start        where in buffer to start (this *must* be within buffer)    start        where in buffer to start (this *must* be within buffer)
204      prompt       for stdin or readline()
205    
206  Returns:       pointer to the start of new data  Returns:       pointer to the start of new data
207                 could be a copy of start, or could be moved                 could be a copy of start, or could be moved
# Line 196  Returns:       pointer to the start of n Line 209  Returns:       pointer to the start of n
209  */  */
210    
211  static uschar *  static uschar *
212  extend_inputline(FILE *f, uschar *start)  extend_inputline(FILE *f, uschar *start, const char *prompt)
213  {  {
214  uschar *here = start;  uschar *here = start;
215    
# Line 207  for (;;) Line 220  for (;;)
220    if (rlen > 1000)    if (rlen > 1000)
221      {      {
222      int dlen;      int dlen;
223      if (fgets((char *)here, rlen,  f) == NULL)  
224        return (here == start)? NULL : start;      /* If libreadline support is required, use readline() to read a line if the
225        input is a terminal. Note that readline() removes the trailing newline, so
226        we must put it back again, to be compatible with fgets(). */
227    
228    #ifdef SUPPORT_LIBREADLINE
229        if (isatty(fileno(f)))
230          {
231          size_t len;
232          char *s = readline(prompt);
233          if (s == NULL) return (here == start)? NULL : start;
234          len = strlen(s);
235          if (len > 0) add_history(s);
236          if (len > rlen - 1) len = rlen - 1;
237          memcpy(here, s, len);
238          here[len] = '\n';
239          here[len+1] = 0;
240          free(s);
241          }
242        else
243    #endif
244    
245        /* Read the next line by normal means, prompting if the file is stdin. */
246    
247          {
248          if (f == stdin) printf(prompt);
249          if (fgets((char *)here, rlen,  f) == NULL)
250            return (here == start)? NULL : start;
251          }
252    
253      dlen = (int)strlen((char *)here);      dlen = (int)strlen((char *)here);
254      if (dlen > 0 && here[dlen - 1] == '\n') return start;      if (dlen > 0 && here[dlen - 1] == '\n') return start;
255      here += dlen;      here += dlen;
# Line 728  return 0; Line 769  return 0;
769  static void  static void
770  usage(void)  usage(void)
771  {  {
772  printf("Usage:     pcretest [options] [<input> [<output>]]\n");  printf("Usage:     pcretest [options] [<input file> [<output file>]]\n\n");
773    printf("Input and output default to stdin and stdout.\n");
774    #ifdef SUPPORT_LIBREADLINE
775    printf("If input is a terminal, readline() is used to read from it.\n");
776    #else
777    printf("This version of pcretest is not linked with readline().\n");
778    #endif
779    printf("\nOptions:\n");
780  printf("  -b       show compiled code (bytecode)\n");  printf("  -b       show compiled code (bytecode)\n");
781  printf("  -C       show PCRE compile-time options and exit\n");  printf("  -C       show PCRE compile-time options and exit\n");
782  printf("  -d       debug: show compiled code and information (-b and -i)\n");  printf("  -d       debug: show compiled code and information (-b and -i)\n");
# Line 737  printf("  -dfa     force DFA matching fo Line 785  printf("  -dfa     force DFA matching fo
785  #endif  #endif
786  printf("  -help    show usage information\n");  printf("  -help    show usage information\n");
787  printf("  -i       show information about compiled patterns\n"  printf("  -i       show information about compiled patterns\n"
788           "  -M       find MATCH_LIMIT minimum for each subject\n"
789         "  -m       output memory used information\n"         "  -m       output memory used information\n"
790         "  -o <n>   set size of offsets vector to <n>\n");         "  -o <n>   set size of offsets vector to <n>\n");
791  #if !defined NOPOSIX  #if !defined NOPOSIX
# Line 766  int main(int argc, char **argv) Line 815  int main(int argc, char **argv)
815  FILE *infile = stdin;  FILE *infile = stdin;
816  int options = 0;  int options = 0;
817  int study_options = 0;  int study_options = 0;
818    int default_find_match_limit = FALSE;
819  int op = 1;  int op = 1;
820  int timeit = 0;  int timeit = 0;
821  int timeitm = 0;  int timeitm = 0;
# Line 825  while (argc > 1 && argv[op][0] == '-') Line 875  while (argc > 1 && argv[op][0] == '-')
875    else if (strcmp(argv[op], "-b") == 0) debug = 1;    else if (strcmp(argv[op], "-b") == 0) debug = 1;
876    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
877    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
878      else if (strcmp(argv[op], "-M") == 0) default_find_match_limit = TRUE;
879  #if !defined NODFA  #if !defined NODFA
880    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
881  #endif  #endif
# Line 877  while (argc > 1 && argv[op][0] == '-') Line 928  while (argc > 1 && argv[op][0] == '-')
928    else if (strcmp(argv[op], "-C") == 0)    else if (strcmp(argv[op], "-C") == 0)
929      {      {
930      int rc;      int rc;
931        unsigned long int lrc;
932      printf("PCRE version %s\n", pcre_version());      printf("PCRE version %s\n", pcre_version());
933      printf("Compiled with\n");      printf("Compiled with\n");
934      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
# Line 884  while (argc > 1 && argv[op][0] == '-') Line 936  while (argc > 1 && argv[op][0] == '-')
936      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
937      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
938      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
939      printf("  Newline sequence is %s\n", (rc == '\r')? "CR" :      /* Note that these values are always the ASCII values, even
940        (rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" :      in EBCDIC environments. CR is 13 and NL is 10. */
941        printf("  Newline sequence is %s\n", (rc == 13)? "CR" :
942          (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
943        (rc == -2)? "ANYCRLF" :        (rc == -2)? "ANYCRLF" :
944        (rc == -1)? "ANY" : "???");        (rc == -1)? "ANY" : "???");
945      (void)pcre_config(PCRE_CONFIG_BSR, &rc);      (void)pcre_config(PCRE_CONFIG_BSR, &rc);
# Line 895  while (argc > 1 && argv[op][0] == '-') Line 949  while (argc > 1 && argv[op][0] == '-')
949      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
950      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
951      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
952      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &lrc);
953      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %ld\n", lrc);
954      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &lrc);
955      printf("  Default recursion depth limit = %d\n", rc);      printf("  Default recursion depth limit = %ld\n", lrc);
956      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
957      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
958      goto EXIT;      goto EXIT;
# Line 997  while (!done) Line 1051  while (!done)
1051    use_utf8 = 0;    use_utf8 = 0;
1052    debug_lengths = 1;    debug_lengths = 1;
1053    
1054    if (infile == stdin) printf("  re> ");    if (extend_inputline(infile, buffer, "  re> ") == NULL) break;
   if (extend_inputline(infile, buffer) == NULL) break;  
1055    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1056    fflush(outfile);    fflush(outfile);
1057    
# Line 1098  while (!done) Line 1151  while (!done)
1151    
1152    if (isalnum(delimiter) || delimiter == '\\')    if (isalnum(delimiter) || delimiter == '\\')
1153      {      {
1154      fprintf(outfile, "** Delimiter must not be alphameric or \\\n");      fprintf(outfile, "** Delimiter must not be alphanumeric or \\\n");
1155      goto SKIP_DATA;      goto SKIP_DATA;
1156      }      }
1157    
# Line 1114  while (!done) Line 1167  while (!done)
1167        pp++;        pp++;
1168        }        }
1169      if (*pp != 0) break;      if (*pp != 0) break;
1170      if (infile == stdin) printf("    > ");      if ((pp = extend_inputline(infile, pp, "    > ")) == NULL)
     if ((pp = extend_inputline(infile, pp)) == NULL)  
1171        {        {
1172        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
1173        done = 1;        done = 1;
# Line 1207  while (!done) Line 1259  while (!done)
1259    
1260        case '<':        case '<':
1261          {          {
1262          int x = check_newline(pp, outfile);          if (strncmp((char *)pp, "JS>", 3) == 0)
1263          if (x == 0) goto SKIP_DATA;            {
1264          options |= x;            options |= PCRE_JAVASCRIPT_COMPAT;
1265          while (*pp++ != '>');            pp += 3;
1266              }
1267            else
1268              {
1269              int x = check_newline(pp, outfile);
1270              if (x == 0) goto SKIP_DATA;
1271              options |= x;
1272              while (*pp++ != '>');
1273              }
1274          }          }
1275        break;        break;
1276    
# Line 1289  while (!done) Line 1349  while (!done)
1349          {          {
1350          for (;;)          for (;;)
1351            {            {
1352            if (extend_inputline(infile, buffer) == NULL)            if (extend_inputline(infile, buffer, NULL) == NULL)
1353              {              {
1354              done = 1;              done = 1;
1355              goto CONTINUE;              goto CONTINUE;
# Line 1654  while (!done) Line 1714  while (!done)
1714      int callout_data_set = 0;      int callout_data_set = 0;
1715      int count, c;      int count, c;
1716      int copystrings = 0;      int copystrings = 0;
1717      int find_match_limit = 0;      int find_match_limit = default_find_match_limit;
1718      int getstrings = 0;      int getstrings = 0;
1719      int getlist = 0;      int getlist = 0;
1720      int gmatched = 0;      int gmatched = 0;
# Line 1684  while (!done) Line 1744  while (!done)
1744      len = 0;      len = 0;
1745      for (;;)      for (;;)
1746        {        {
1747        if (infile == stdin) printf("data> ");        if (extend_inputline(infile, buffer + len, "data> ") == NULL)
       if (extend_inputline(infile, buffer + len) == NULL)  
1748          {          {
1749          if (len > 0) break;          if (len > 0) break;
1750          done = 1;          done = 1;
# Line 1753  while (!done) Line 1812  while (!done)
1812              {              {
1813              unsigned char buff8[8];              unsigned char buff8[8];
1814              int ii, utn;              int ii, utn;
1815              utn = ord2utf8(c, buff8);              if (use_utf8)
1816              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];                {
1817              c = buff8[ii];   /* Last byte */                utn = ord2utf8(c, buff8);
1818                  for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1819                  c = buff8[ii];   /* Last byte */
1820                  }
1821                else
1822                 {
1823                 if (c > 255)
1824                   fprintf(outfile, "** Character \\x{%x} is greater than 255 and "
1825                     "UTF-8 mode is not enabled.\n"
1826                     "** Truncation will probably give the wrong result.\n", c);
1827                 }
1828              p = pt + 1;              p = pt + 1;
1829              break;              break;
1830              }              }
# Line 1942  while (!done) Line 2011  while (!done)
2011          case 'S':          case 'S':
2012          show_malloc = 1;          show_malloc = 1;
2013          continue;          continue;
2014    
2015            case 'Y':
2016            options |= PCRE_NO_START_OPTIMIZE;
2017            continue;
2018    
2019          case 'Z':          case 'Z':
2020          options |= PCRE_NOTEOL;          options |= PCRE_NOTEOL;
# Line 1965  while (!done) Line 2038  while (!done)
2038      *q = 0;      *q = 0;
2039      len = q - dbuffer;      len = q - dbuffer;
2040    
2041        /* Move the data to the end of the buffer so that a read over the end of
2042        the buffer will be seen by valgrind, even if it doesn't cause a crash. If
2043        we are using the POSIX interface, we must include the terminating zero. */
2044    
2045    #if !defined NOPOSIX
2046        if (posix || do_posix)
2047          {
2048          memmove(bptr + buffer_size - len - 1, bptr, len + 1);
2049          bptr += buffer_size - len - 1;
2050          }
2051        else
2052    #endif
2053          {
2054          memmove(bptr + buffer_size - len, bptr, len);
2055          bptr += buffer_size - len;
2056          }
2057    
2058      if ((all_use_dfa || use_dfa) && find_match_limit)      if ((all_use_dfa || use_dfa) && find_match_limit)
2059        {        {
2060        printf("**Match limit not relevant for DFA matching: ignored\n");        printf("**Match limit not relevant for DFA matching: ignored\n");
# Line 1984  while (!done) Line 2074  while (!done)
2074          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);
2075        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;
2076        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;
2077          if ((options & PCRE_NOTEMPTY) != 0) eflags |= REG_NOTEMPTY;
2078    
2079        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);
2080    
# Line 2291  while (!done) Line 2382  while (!done)
2382              {              {
2383              int d;              int d;
2384              (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);              (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);
2385              obits = (d == '\r')? PCRE_NEWLINE_CR :              /* Note that these values are always the ASCII ones, even in
2386                      (d == '\n')? PCRE_NEWLINE_LF :              EBCDIC environments. CR = 13, NL = 10. */
2387                      (d == ('\r'<<8 | '\n'))? PCRE_NEWLINE_CRLF :              obits = (d == 13)? PCRE_NEWLINE_CR :
2388                        (d == 10)? PCRE_NEWLINE_LF :
2389                        (d == (13<<8 | 10))? PCRE_NEWLINE_CRLF :
2390                      (d == -2)? PCRE_NEWLINE_ANYCRLF :                      (d == -2)? PCRE_NEWLINE_ANYCRLF :
2391                      (d == -1)? PCRE_NEWLINE_ANY : 0;                      (d == -1)? PCRE_NEWLINE_ANY : 0;
2392              }              }

Legend:
Removed from v.259  
changed lines
  Added in v.391

  ViewVC Help
Powered by ViewVC 1.1.5