/[pcre]/code/tags/pcre-3.7/pcretest.c
ViewVC logotype

Diff of /code/tags/pcre-3.7/pcretest.c

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

revision 3 by nigel, Sat Feb 24 21:38:01 2007 UTC revision 35 by nigel, Sat Feb 24 21:39:05 2007 UTC
# Line 7  Line 7 
7  #include <string.h>  #include <string.h>
8  #include <stdlib.h>  #include <stdlib.h>
9  #include <time.h>  #include <time.h>
10    #include <locale.h>
11    
12  /* Use the internal info for displaying the results of pcre_study(). */  /* Use the internal info for displaying the results of pcre_study(). */
13    
# Line 21  Line 22 
22  #endif  #endif
23  #endif  #endif
24    
25    #define LOOPREPEAT 20000
26    
27    
28  static FILE *outfile;  static FILE *outfile;
29  static int log_store = 0;  static int log_store = 0;
# Line 30  static int log_store = 0; Line 33  static int log_store = 0;
33  /* Debugging function to print the internal form of the regex. This is the same  /* Debugging function to print the internal form of the regex. This is the same
34  code as contained in pcre.c under the DEBUG macro. */  code as contained in pcre.c under the DEBUG macro. */
35    
36  static char *OP_names[] = { "End", "\\A", "\\B", "\\b", "\\D", "\\d",  static const char *OP_names[] = {
37    "\\S", "\\s", "\\W", "\\w", "Cut", "\\Z", "^", "$", "Any", "chars",    "End", "\\A", "\\B", "\\b", "\\D", "\\d",
38    "not",    "\\S", "\\s", "\\W", "\\w", "\\Z", "\\z",
39      "Opt", "^", "$", "Any", "chars", "not",
40    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",
41    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",
42    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",
43    "*", "*?", "+", "+?", "?", "??", "{", "{",    "*", "*?", "+", "+?", "?", "??", "{", "{",
44    "class", "Ref",    "class", "Ref",
45    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", "Once",    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",
46      "AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cref",
47    "Brazero", "Braminzero", "Bra"    "Brazero", "Braminzero", "Bra"
48  };  };
49    
50    
51  static void print_internals(pcre *re)  static void print_internals(pcre *re, FILE *outfile)
52  {  {
53  unsigned char *code = ((real_pcre *)re)->code;  unsigned char *code = ((real_pcre *)re)->code;
54    
55  printf("------------------------------------------------------------------\n");  fprintf(outfile, "------------------------------------------------------------------\n");
56    
57  for(;;)  for(;;)
58    {    {
59    int c;    int c;
60    int charlength;    int charlength;
61    
62    printf("%3d ", code - ((real_pcre *)re)->code);    fprintf(outfile, "%3d ", (int)(code - ((real_pcre *)re)->code));
63    
64    if (*code >= OP_BRA)    if (*code >= OP_BRA)
65      {      {
66      printf("%3d Bra %d", (code[1] << 8) + code[2], *code - OP_BRA);      fprintf(outfile, "%3d Bra %d", (code[1] << 8) + code[2], *code - OP_BRA);
67      code += 2;      code += 2;
68      }      }
69    
70    else switch(*code)    else switch(*code)
71      {      {
72      case OP_END:      case OP_END:
73      printf("    %s\n", OP_names[*code]);      fprintf(outfile, "    %s\n", OP_names[*code]);
74      printf("------------------------------------------------------------------\n");      fprintf(outfile, "------------------------------------------------------------------\n");
75      return;      return;
76    
77        case OP_OPT:
78        fprintf(outfile, " %.2x %s", code[1], OP_names[*code]);
79        code++;
80        break;
81    
82        case OP_COND:
83        fprintf(outfile, "%3d Cond", (code[1] << 8) + code[2]);
84        code += 2;
85        break;
86    
87        case OP_CREF:
88        fprintf(outfile, " %.2d %s", code[1], OP_names[*code]);
89        code++;
90        break;
91    
92      case OP_CHARS:      case OP_CHARS:
93      charlength = *(++code);      charlength = *(++code);
94      printf("%3d ", charlength);      fprintf(outfile, "%3d ", charlength);
95      while (charlength-- > 0)      while (charlength-- > 0)
96        if (isprint(c = *(++code))) printf("%c", c); else printf("\\x%02x", c);        if (isprint(c = *(++code))) fprintf(outfile, "%c", c);
97            else fprintf(outfile, "\\x%02x", c);
98      break;      break;
99    
100      case OP_KETRMAX:      case OP_KETRMAX:
# Line 82  for(;;) Line 103  for(;;)
103      case OP_KET:      case OP_KET:
104      case OP_ASSERT:      case OP_ASSERT:
105      case OP_ASSERT_NOT:      case OP_ASSERT_NOT:
106        case OP_ASSERTBACK:
107        case OP_ASSERTBACK_NOT:
108      case OP_ONCE:      case OP_ONCE:
109      printf("%3d %s", (code[1] << 8) + code[2], OP_names[*code]);      fprintf(outfile, "%3d %s", (code[1] << 8) + code[2], OP_names[*code]);
110        code += 2;
111        break;
112    
113        case OP_REVERSE:
114        fprintf(outfile, "%3d %s", (code[1] << 8) + code[2], OP_names[*code]);
115      code += 2;      code += 2;
116      break;      break;
117    
# Line 100  for(;;) Line 128  for(;;)
128      case OP_TYPEQUERY:      case OP_TYPEQUERY:
129      case OP_TYPEMINQUERY:      case OP_TYPEMINQUERY:
130      if (*code >= OP_TYPESTAR)      if (*code >= OP_TYPESTAR)
131        printf("    %s", OP_names[code[1]]);        fprintf(outfile, "    %s", OP_names[code[1]]);
132      else if (isprint(c = code[1])) printf("    %c", c);      else if (isprint(c = code[1])) fprintf(outfile, "    %c", c);
133        else printf("    \\x%02x", c);        else fprintf(outfile, "    \\x%02x", c);
134      printf("%s", OP_names[*code++]);      fprintf(outfile, "%s", OP_names[*code++]);
135      break;      break;
136    
137      case OP_EXACT:      case OP_EXACT:
138      case OP_UPTO:      case OP_UPTO:
139      case OP_MINUPTO:      case OP_MINUPTO:
140      if (isprint(c = code[3])) printf("    %c{", c);      if (isprint(c = code[3])) fprintf(outfile, "    %c{", c);
141        else printf("    \\x%02x{", c);        else fprintf(outfile, "    \\x%02x{", c);
142      if (*code != OP_EXACT) printf(",");      if (*code != OP_EXACT) fprintf(outfile, ",");
143      printf("%d}", (code[1] << 8) + code[2]);      fprintf(outfile, "%d}", (code[1] << 8) + code[2]);
144      if (*code == OP_MINUPTO) printf("?");      if (*code == OP_MINUPTO) fprintf(outfile, "?");
145      code += 3;      code += 3;
146      break;      break;
147    
148      case OP_TYPEEXACT:      case OP_TYPEEXACT:
149      case OP_TYPEUPTO:      case OP_TYPEUPTO:
150      case OP_TYPEMINUPTO:      case OP_TYPEMINUPTO:
151      printf("    %s{", OP_names[code[3]]);      fprintf(outfile, "    %s{", OP_names[code[3]]);
152      if (*code != OP_TYPEEXACT) printf(",");      if (*code != OP_TYPEEXACT) fprintf(outfile, "0,");
153      printf("%d}", (code[1] << 8) + code[2]);      fprintf(outfile, "%d}", (code[1] << 8) + code[2]);
154      if (*code == OP_TYPEMINUPTO) printf("?");      if (*code == OP_TYPEMINUPTO) fprintf(outfile, "?");
155      code += 3;      code += 3;
156      break;      break;
157    
158      case OP_NOT:      case OP_NOT:
159      if (isprint(c = *(++code))) printf("    [^%c]", c);      if (isprint(c = *(++code))) fprintf(outfile, "    [^%c]", c);
160        else printf("    [^\\x%02x]", c);        else fprintf(outfile, "    [^\\x%02x]", c);
161      break;      break;
162    
163      case OP_NOTSTAR:      case OP_NOTSTAR:
# Line 138  for(;;) Line 166  for(;;)
166      case OP_NOTMINPLUS:      case OP_NOTMINPLUS:
167      case OP_NOTQUERY:      case OP_NOTQUERY:
168      case OP_NOTMINQUERY:      case OP_NOTMINQUERY:
169      if (isprint(c = code[1])) printf("    [^%c]", c);      if (isprint(c = code[1])) fprintf(outfile, "    [^%c]", c);
170        else printf("    [^\\x%02x]", c);        else fprintf(outfile, "    [^\\x%02x]", c);
171      printf("%s", OP_names[*code++]);      fprintf(outfile, "%s", OP_names[*code++]);
172      break;      break;
173    
174      case OP_NOTEXACT:      case OP_NOTEXACT:
175      case OP_NOTUPTO:      case OP_NOTUPTO:
176      case OP_NOTMINUPTO:      case OP_NOTMINUPTO:
177      if (isprint(c = code[3])) printf("    [^%c]{", c);      if (isprint(c = code[3])) fprintf(outfile, "    [^%c]{", c);
178        else printf("    [^\\x%02x]{", c);        else fprintf(outfile, "    [^\\x%02x]{", c);
179      if (*code != OP_NOTEXACT) printf(",");      if (*code != OP_NOTEXACT) fprintf(outfile, ",");
180      printf("%d}", (code[1] << 8) + code[2]);      fprintf(outfile, "%d}", (code[1] << 8) + code[2]);
181      if (*code == OP_NOTMINUPTO) printf("?");      if (*code == OP_NOTMINUPTO) fprintf(outfile, "?");
182      code += 3;      code += 3;
183      break;      break;
184    
185      case OP_REF:      case OP_REF:
186      printf("    \\%d", *(++code));      fprintf(outfile, "    \\%d", *(++code));
187      break;      code++;
188        goto CLASS_REF_REPEAT;
189    
190      case OP_CLASS:      case OP_CLASS:
191        {        {
192        int i, min, max;        int i, min, max;
   
193        code++;        code++;
194        printf("    [");        fprintf(outfile, "    [");
195    
196        for (i = 0; i < 256; i++)        for (i = 0; i < 256; i++)
197          {          {
# Line 172  for(;;) Line 200  for(;;)
200            int j;            int j;
201            for (j = i+1; j < 256; j++)            for (j = i+1; j < 256; j++)
202              if ((code[j/8] & (1 << (j&7))) == 0) break;              if ((code[j/8] & (1 << (j&7))) == 0) break;
203            if (i == '-' || i == ']') printf("\\");            if (i == '-' || i == ']') fprintf(outfile, "\\");
204            if (isprint(i)) printf("%c", i); else printf("\\x%02x", i);            if (isprint(i)) fprintf(outfile, "%c", i); else fprintf(outfile, "\\x%02x", i);
205            if (--j > i)            if (--j > i)
206              {              {
207              printf("-");              fprintf(outfile, "-");
208              if (j == '-' || j == ']') printf("\\");              if (j == '-' || j == ']') fprintf(outfile, "\\");
209              if (isprint(j)) printf("%c", j); else printf("\\x%02x", j);              if (isprint(j)) fprintf(outfile, "%c", j); else fprintf(outfile, "\\x%02x", j);
210              }              }
211            i = j;            i = j;
212            }            }
213          }          }
214        printf("]");        fprintf(outfile, "]");
215        code += 32;        code += 32;
216    
217          CLASS_REF_REPEAT:
218    
219        switch(*code)        switch(*code)
220          {          {
221          case OP_CRSTAR:          case OP_CRSTAR:
# Line 194  for(;;) Line 224  for(;;)
224          case OP_CRMINPLUS:          case OP_CRMINPLUS:
225          case OP_CRQUERY:          case OP_CRQUERY:
226          case OP_CRMINQUERY:          case OP_CRMINQUERY:
227          printf("%s", OP_names[*code]);          fprintf(outfile, "%s", OP_names[*code]);
228          break;          break;
229    
230          case OP_CRRANGE:          case OP_CRRANGE:
231          case OP_CRMINRANGE:          case OP_CRMINRANGE:
232          min = (code[1] << 8) + code[2];          min = (code[1] << 8) + code[2];
233          max = (code[3] << 8) + code[4];          max = (code[3] << 8) + code[4];
234          if (max == 0) printf("{%d,}", min);          if (max == 0) fprintf(outfile, "{%d,}", min);
235          else printf("{%d,%d}", min, max);          else fprintf(outfile, "{%d,%d}", min, max);
236          if (*code == OP_CRMINRANGE) printf("?");          if (*code == OP_CRMINRANGE) fprintf(outfile, "?");
237          code += 4;          code += 4;
238          break;          break;
239    
# Line 216  for(;;) Line 246  for(;;)
246      /* Anything else is just a one-node item */      /* Anything else is just a one-node item */
247    
248      default:      default:
249      printf("    %s", OP_names[*code]);      fprintf(outfile, "    %s", OP_names[*code]);
250      break;      break;
251      }      }
252    
253    code++;    code++;
254    printf("\n");    fprintf(outfile, "\n");
255    }    }
256  }  }
257    
# Line 244  compiled re. */ Line 274  compiled re. */
274    
275  static void *new_malloc(size_t size)  static void *new_malloc(size_t size)
276  {  {
277  if (log_store) fprintf(outfile, "Store size request: %d\n", (int)size);  if (log_store)
278      fprintf(outfile, "Memory allocation (code space): %d\n",
279        (int)((int)size - offsetof(real_pcre, code[0])));
280  return malloc(size);  return malloc(size);
281  }  }
282    
# Line 262  int study_options = 0; Line 294  int study_options = 0;
294  int op = 1;  int op = 1;
295  int timeit = 0;  int timeit = 0;
296  int showinfo = 0;  int showinfo = 0;
297    int showstore = 0;
298  int posix = 0;  int posix = 0;
299  int debug = 0;  int debug = 0;
300    int done = 0;
301  unsigned char buffer[30000];  unsigned char buffer[30000];
302  unsigned char dbuffer[1024];  unsigned char dbuffer[1024];
303    
# Line 275  outfile = stdout; Line 309  outfile = stdout;
309    
310  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
311    {    {
312    if (strcmp(argv[op], "-s") == 0) log_store = 1;    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
313        showstore = 1;
314    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-t") == 0) timeit = 1;
315    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
316    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
# Line 283  while (argc > 1 && argv[op][0] == '-') Line 318  while (argc > 1 && argv[op][0] == '-')
318    else    else
319      {      {
320      printf("*** Unknown option %s\n", argv[op]);      printf("*** Unknown option %s\n", argv[op]);
321        printf("Usage: pcretest [-d] [-i] [-p] [-s] [-t] [<input> [<output>]]\n");
322        printf("  -d   debug: show compiled code; implies -i\n"
323               "  -i   show information about compiled pattern\n"
324               "  -p   use POSIX interface\n"
325               "  -s   output store information\n"
326               "  -t   time compilation and execution\n");
327      return 1;      return 1;
328      }      }
329    op++;    op++;
# Line 315  if (argc > 2) Line 356  if (argc > 2)
356    
357  pcre_malloc = new_malloc;  pcre_malloc = new_malloc;
358    
359  /* Heading line, then prompt for first re if stdin */  /* Heading line, then prompt for first regex if stdin */
360    
 fprintf(outfile, "Testing Perl-Compatible Regular Expressions\n");  
361  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  fprintf(outfile, "PCRE version %s\n\n", pcre_version());
362    
363  /* Main loop */  /* Main loop */
364    
365  for (;;)  while (!done)
366    {    {
367    pcre *re = NULL;    pcre *re = NULL;
368    pcre_extra *extra = NULL;    pcre_extra *extra = NULL;
369    regex_t preg;    regex_t preg;
370    char *error;    const char *error;
371    unsigned char *p, *pp;    unsigned char *p, *pp, *ppp;
372      unsigned const char *tables = NULL;
373    int do_study = 0;    int do_study = 0;
374    int do_debug = 0;    int do_debug = debug;
375      int do_G = 0;
376      int do_g = 0;
377      int do_showinfo = showinfo;
378      int do_showrest = 0;
379    int do_posix = 0;    int do_posix = 0;
380    int erroroffset, len, delimiter;    int erroroffset, len, delimiter;
381    
382    if (infile == stdin) printf("  re> ");    if (infile == stdin) printf("  re> ");
383    if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) break;    if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) break;
384    if (infile != stdin) fprintf(outfile, (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
385    
386    p = buffer;    p = buffer;
387    while (isspace(*p)) p++;    while (isspace(*p)) p++;
# Line 347  for (;;) Line 392  for (;;)
392    
393    delimiter = *p++;    delimiter = *p++;
394    
395    if (isalnum(delimiter))    if (isalnum(delimiter) || delimiter == '\\')
396      {      {
397      fprintf(outfile, "** Delimiter must not be alphameric\n");      fprintf(outfile, "** Delimiter must not be alphameric or \\\n");
398      goto SKIP_DATA;      goto SKIP_DATA;
399      }      }
400    
# Line 357  for (;;) Line 402  for (;;)
402    
403    for(;;)    for(;;)
404      {      {
405      while (*pp != 0 && *pp != delimiter) pp++;      while (*pp != 0)
406          {
407          if (*pp == '\\' && pp[1] != 0) pp++;
408            else if (*pp == delimiter) break;
409          pp++;
410          }
411      if (*pp != 0) break;      if (*pp != 0) break;
412    
413      len = sizeof(buffer) - (pp - buffer);      len = sizeof(buffer) - (pp - buffer);
# Line 371  for (;;) Line 421  for (;;)
421      if (fgets((char *)pp, len, infile) == NULL)      if (fgets((char *)pp, len, infile) == NULL)
422        {        {
423        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
424        goto END_OFF;        done = 1;
425          goto CONTINUE;
426        }        }
427      if (infile != stdin) fprintf(outfile, (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
428      }      }
429    
430      /* If the first character after the delimiter is backslash, make
431      the pattern end with backslash. This is purely to provide a way
432      of testing for the error message when a pattern ends with backslash. */
433    
434      if (pp[1] == '\\') *pp++ = '\\';
435    
436    /* Terminate the pattern at the delimiter */    /* Terminate the pattern at the delimiter */
437    
438    *pp++ = 0;    *pp++ = 0;
# Line 384  for (;;) Line 441  for (;;)
441    
442    options = 0;    options = 0;
443    study_options = 0;    study_options = 0;
444      log_store = showstore;  /* default from command line */
445    
446    while (*pp != 0)    while (*pp != 0)
447      {      {
448      switch (*pp++)      switch (*pp++)
449        {        {
450          case 'g': do_g = 1; break;
451        case 'i': options |= PCRE_CASELESS; break;        case 'i': options |= PCRE_CASELESS; break;
452        case 'm': options |= PCRE_MULTILINE; break;        case 'm': options |= PCRE_MULTILINE; break;
453        case 's': options |= PCRE_DOTALL; break;        case 's': options |= PCRE_DOTALL; break;
454        case 'x': options |= PCRE_EXTENDED; break;        case 'x': options |= PCRE_EXTENDED; break;
455    
456          case '+': do_showrest = 1; break;
457        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
458        case 'D': do_debug = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
459        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
460          case 'G': do_G = 1; break;
461          case 'I': do_showinfo = 1; break;
462          case 'M': log_store = 1; break;
463        case 'P': do_posix = 1; break;        case 'P': do_posix = 1; break;
464        case 'S': do_study = 1; break;        case 'S': do_study = 1; break;
465        case 'I': study_options |= PCRE_CASELESS; break;        case 'U': options |= PCRE_UNGREEDY; break;
466        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
467    
468          case 'L':
469          ppp = pp;
470          while (*ppp != '\n' && *ppp != ' ') ppp++;
471          *ppp = 0;
472          if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
473            {
474            fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
475            goto SKIP_DATA;
476            }
477          tables = pcre_maketables();
478          pp = ppp;
479          break;
480    
481        case '\n': case ' ': break;        case '\n': case ' ': break;
482        default:        default:
483        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);
# Line 406  for (;;) Line 485  for (;;)
485        }        }
486      }      }
487    
488    /* Handle compiing via the POSIX interface, which doesn't support the    /* Handle compiling via the POSIX interface, which doesn't support the
489    timing, showing, or debugging options. */    timing, showing, or debugging options, nor the ability to pass over
490      local character tables. */
491    
492    if (posix || do_posix)    if (posix || do_posix)
493      {      {
# Line 437  for (;;) Line 517  for (;;)
517        register int i;        register int i;
518        clock_t time_taken;        clock_t time_taken;
519        clock_t start_time = clock();        clock_t start_time = clock();
520        for (i = 0; i < 4000; i++)        for (i = 0; i < LOOPREPEAT; i++)
521          {          {
522          re = pcre_compile((char *)p, options, &error, &erroroffset);          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
523          if (re != NULL) free(re);          if (re != NULL) free(re);
524          }          }
525        time_taken = clock() - start_time;        time_taken = clock() - start_time;
526        fprintf(outfile, "Compile time %.2f milliseconds\n",        fprintf(outfile, "Compile time %.3f milliseconds\n",
527          ((double)time_taken)/(4 * CLOCKS_PER_SEC));          ((double)time_taken * 1000.0) /
528            ((double)LOOPREPEAT * (double)CLOCKS_PER_SEC));
529        }        }
530    
531      re = pcre_compile((char *)p, options, &error, &erroroffset);      re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
532    
533      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
534      if non-interactive. */      if non-interactive. */
# Line 461  for (;;) Line 542  for (;;)
542          for (;;)          for (;;)
543            {            {
544            if (fgets((char *)buffer, sizeof(buffer), infile) == NULL)            if (fgets((char *)buffer, sizeof(buffer), infile) == NULL)
545              goto END_OFF;              {
546                done = 1;
547                goto CONTINUE;
548                }
549            len = (int)strlen((char *)buffer);            len = (int)strlen((char *)buffer);
550            while (len > 0 && isspace(buffer[len-1])) len--;            while (len > 0 && isspace(buffer[len-1])) len--;
551            if (len == 0) break;            if (len == 0) break;
552            }            }
553          fprintf(outfile, "\n");          fprintf(outfile, "\n");
554          }          }
555        continue;        goto CONTINUE;
556        }        }
557    
558      /* Compilation succeeded; print data if required */      /* Compilation succeeded; print data if required */
559    
560      if (showinfo || do_debug)      if (do_showinfo)
561        {        {
562        int first_char, count;        int first_char, count;
563    
564        if (debug || do_debug) print_internals(re);        if (do_debug) print_internals(re, outfile);
565    
566        count = pcre_info(re, &options, &first_char);        count = pcre_info(re, &options, &first_char);
567        if (count < 0) fprintf(outfile,        if (count < 0) fprintf(outfile,
# Line 486  for (;;) Line 570  for (;;)
570          {          {
571          fprintf(outfile, "Identifying subpattern count = %d\n", count);          fprintf(outfile, "Identifying subpattern count = %d\n", count);
572          if (options == 0) fprintf(outfile, "No options\n");          if (options == 0) fprintf(outfile, "No options\n");
573            else fprintf(outfile, "Options:%s%s%s%s%s%s%s\n",            else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s\n",
574              ((options & PCRE_ANCHORED) != 0)? " anchored" : "",              ((options & PCRE_ANCHORED) != 0)? " anchored" : "",
575              ((options & PCRE_CASELESS) != 0)? " caseless" : "",              ((options & PCRE_CASELESS) != 0)? " caseless" : "",
576              ((options & PCRE_EXTENDED) != 0)? " extended" : "",              ((options & PCRE_EXTENDED) != 0)? " extended" : "",
577              ((options & PCRE_MULTILINE) != 0)? " multiline" : "",              ((options & PCRE_MULTILINE) != 0)? " multiline" : "",
578              ((options & PCRE_DOTALL) != 0)? " dotall" : "",              ((options & PCRE_DOTALL) != 0)? " dotall" : "",
579              ((options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",              ((options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
580              ((options & PCRE_EXTRA) != 0)? " extra" : "");              ((options & PCRE_EXTRA) != 0)? " extra" : "",
581                ((options & PCRE_UNGREEDY) != 0)? " ungreedy" : "");
582          if (first_char == -1)          if (first_char == -1)
583            {            {
584            fprintf(outfile, "First char at start or follows \\n\n");            fprintf(outfile, "First char at start or follows \\n\n");
# Line 522  for (;;) Line 607  for (;;)
607          register int i;          register int i;
608          clock_t time_taken;          clock_t time_taken;
609          clock_t start_time = clock();          clock_t start_time = clock();
610          for (i = 0; i < 4000; i++)          for (i = 0; i < LOOPREPEAT; i++)
611            extra = pcre_study(re, study_options, &error);            extra = pcre_study(re, study_options, &error);
612          time_taken = clock() - start_time;          time_taken = clock() - start_time;
613          if (extra != NULL) free(extra);          if (extra != NULL) free(extra);
614          fprintf(outfile, "  Study time %.2f milliseconds\n",          fprintf(outfile, "  Study time %.3f milliseconds\n",
615            ((double)time_taken)/(4 * CLOCKS_PER_SEC));            ((double)time_taken * 1000.0)/
616              ((double)LOOPREPEAT * (double)CLOCKS_PER_SEC));
617          }          }
618    
619        extra = pcre_study(re, study_options, &error);        extra = pcre_study(re, study_options, &error);
# Line 539  for (;;) Line 625  for (;;)
625        /* This looks at internal information. A bit kludgy to do it this        /* This looks at internal information. A bit kludgy to do it this
626        way, but it is useful for testing. */        way, but it is useful for testing. */
627    
628        else if (showinfo || do_debug)        else if (do_showinfo)
629          {          {
630          real_pcre_extra *xx = (real_pcre_extra *)extra;          real_pcre_extra *xx = (real_pcre_extra *)extra;
631          if ((xx->options & PCRE_STUDY_MAPPED) == 0)          if ((xx->options & PCRE_STUDY_MAPPED) == 0)
# Line 580  for (;;) Line 666  for (;;)
666    
667    for (;;)    for (;;)
668      {      {
669      unsigned char *pp;      unsigned char *q;
670        unsigned char *bptr = dbuffer;
671      int count, c;      int count, c;
672      int offsets[30];      int copystrings = 0;
673        int getstrings = 0;
674        int getlist = 0;
675        int start_offset = 0;
676        int offsets[45];
677      int size_offsets = sizeof(offsets)/sizeof(int);      int size_offsets = sizeof(offsets)/sizeof(int);
678    
679      options = 0;      options = 0;
680    
681      if (infile == stdin) printf("  data> ");      if (infile == stdin) printf("data> ");
682      if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) goto END_OFF;      if (fgets((char *)buffer, sizeof(buffer), infile) == NULL)
683      if (infile != stdin) fprintf(outfile, (char *)buffer);        {
684          done = 1;
685          goto CONTINUE;
686          }
687        if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
688    
689      len = (int)strlen((char *)buffer);      len = (int)strlen((char *)buffer);
690      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
# Line 599  for (;;) Line 694  for (;;)
694      p = buffer;      p = buffer;
695      while (isspace(*p)) p++;      while (isspace(*p)) p++;
696    
697      pp = dbuffer;      q = dbuffer;
698      while ((c = *p++) != 0)      while ((c = *p++) != 0)
699        {        {
700        int i = 0;        int i = 0;
# Line 643  for (;;) Line 738  for (;;)
738          options |= PCRE_NOTBOL;          options |= PCRE_NOTBOL;
739          continue;          continue;
740    
741          case 'E':          case 'C':
742          options |= PCRE_DOLLAR_ENDONLY;          while(isdigit(*p)) n = n * 10 + *p++ - '0';
743          continue;          copystrings |= 1 << n;
   
         case 'I':  
         options |= PCRE_CASELESS;  
744          continue;          continue;
745    
746          case 'M':          case 'G':
747          options |= PCRE_MULTILINE;          while(isdigit(*p)) n = n * 10 + *p++ - '0';
748            getstrings |= 1 << n;
749          continue;          continue;
750    
751          case 'S':          case 'L':
752          options |= PCRE_DOTALL;          getlist = 1;
753          continue;          continue;
754    
755          case 'O':          case 'O':
756          while(isdigit(*p)) n = n * 10 + *p++ - '0';          while(isdigit(*p)) n = n * 10 + *p++ - '0';
757          if (n <= sizeof(offsets)/sizeof(int)) size_offsets = n;          if (n <= (int)(sizeof(offsets)/sizeof(int))) size_offsets = n;
758          continue;          continue;
759    
760          case 'Z':          case 'Z':
761          options |= PCRE_NOTEOL;          options |= PCRE_NOTEOL;
762          continue;          continue;
763          }          }
764        *pp++ = c;        *q++ = c;
765        }        }
766      *pp = 0;      *q = 0;
767      len = pp - dbuffer;      len = q - dbuffer;
768    
769      /* Handle matching via the POSIX interface, which does not      /* Handle matching via the POSIX interface, which does not
770      support timing. */      support timing. */
# Line 684  for (;;) Line 777  for (;;)
777        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;
778        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;
779    
780        rc = regexec(&preg, (char *)dbuffer, sizeof(pmatch)/sizeof(regmatch_t),        rc = regexec(&preg, (unsigned char *)bptr,
781          pmatch, eflags);          sizeof(pmatch)/sizeof(regmatch_t), pmatch, eflags);
782    
783        if (rc != 0)        if (rc != 0)
784          {          {
# Line 694  for (;;) Line 787  for (;;)
787          }          }
788        else        else
789          {          {
790          int i;          size_t i;
791          for (i = 0; i < sizeof(pmatch)/sizeof(regmatch_t); i++)          for (i = 0; i < sizeof(pmatch)/sizeof(regmatch_t); i++)
792            {            {
793            if (pmatch[i].rm_so >= 0)            if (pmatch[i].rm_so >= 0)
794              {              {
795              fprintf(outfile, "%2d: ", i);              fprintf(outfile, "%2d: ", (int)i);
796              pchars(dbuffer + pmatch[i].rm_so,              pchars(dbuffer + pmatch[i].rm_so,
797                pmatch[i].rm_eo - pmatch[i].rm_so);                pmatch[i].rm_eo - pmatch[i].rm_so);
798              fprintf(outfile, "\n");              fprintf(outfile, "\n");
799                if (i == 0 && do_showrest)
800                  {
801                  fprintf(outfile, " 0+ ");
802                  pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo);
803                  fprintf(outfile, "\n");
804                  }
805              }              }
806            }            }
807          }          }
808        }        }
809    
810      /* Handle matching via the native interface */      /* Handle matching via the native interface - repeats for /g and /G */
811    
812      else      else for (;;)
813        {        {
814        if (timeit)        if (timeit)
815          {          {
816          register int i;          register int i;
817          clock_t time_taken;          clock_t time_taken;
818          clock_t start_time = clock();          clock_t start_time = clock();
819          for (i = 0; i < 4000; i++)          for (i = 0; i < LOOPREPEAT; i++)
820            count = pcre_exec(re, extra, (char *)dbuffer, len, options, offsets,            count = pcre_exec(re, extra, (char *)bptr, len,
821              size_offsets);              (do_g? start_offset : 0), options, offsets, size_offsets);
822          time_taken = clock() - start_time;          time_taken = clock() - start_time;
823          fprintf(outfile, "Execute time %.2f milliseconds\n",          fprintf(outfile, "Execute time %.3f milliseconds\n",
824            ((double)time_taken)/(4 * CLOCKS_PER_SEC));            ((double)time_taken * 1000.0)/
825              ((double)LOOPREPEAT * (double)CLOCKS_PER_SEC));
826          }          }
827    
828        count = pcre_exec(re, extra, (char *)dbuffer, len, options, offsets,        count = pcre_exec(re, extra, (char *)bptr, len,
829          size_offsets);          (do_g? start_offset : 0), options, offsets, size_offsets);
830    
831        if (count == 0)        if (count == 0)
832          {          {
833          fprintf(outfile, "Matched, but too many substrings\n");          fprintf(outfile, "Matched, but too many substrings\n");
834          count = size_offsets/2;          count = size_offsets/3;
835          }          }
836    
837        if (count >= 0)        if (count >= 0)
838          {          {
839          int i;          int i;
840          count *= 2;          for (i = 0; i < count * 2; i += 2)
         for (i = 0; i < count; i += 2)  
841            {            {
842            if (offsets[i] < 0)            if (offsets[i] < 0)
843              fprintf(outfile, "%2d: <unset>\n", i/2);              fprintf(outfile, "%2d: <unset>\n", i/2);
844            else            else
845              {              {
846              fprintf(outfile, "%2d: ", i/2);              fprintf(outfile, "%2d: ", i/2);
847              pchars(dbuffer + offsets[i], offsets[i+1] - offsets[i]);              pchars(bptr + offsets[i], offsets[i+1] - offsets[i]);
848              fprintf(outfile, "\n");              fprintf(outfile, "\n");
849                if (i == 0)
850                  {
851                  start_offset = offsets[1];
852                  if (do_showrest)
853                    {
854                    fprintf(outfile, " 0+ ");
855                    pchars(bptr + offsets[i+1], len - offsets[i+1]);
856                    fprintf(outfile, "\n");
857                    }
858                  }
859                }
860              }
861    
862            for (i = 0; i < 32; i++)
863              {
864              if ((copystrings & (1 << i)) != 0)
865                {
866                char buffer[16];
867                int rc = pcre_copy_substring((char *)bptr, offsets, count,
868                  i, buffer, sizeof(buffer));
869                if (rc < 0)
870                  fprintf(outfile, "copy substring %d failed %d\n", i, rc);
871                else
872                  fprintf(outfile, "%2dC %s (%d)\n", i, buffer, rc);
873              }              }
874            }            }
875    
876            for (i = 0; i < 32; i++)
877              {
878              if ((getstrings & (1 << i)) != 0)
879                {
880                const char *substring;
881                int rc = pcre_get_substring((char *)bptr, offsets, count,
882                  i, &substring);
883                if (rc < 0)
884                  fprintf(outfile, "get substring %d failed %d\n", i, rc);
885                else
886                  {
887                  fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
888                  free((void *)substring);
889                  }
890                }
891              }
892    
893            if (getlist)
894              {
895              const char **stringlist;
896              int rc = pcre_get_substring_list((char *)bptr, offsets, count,
897                &stringlist);
898              if (rc < 0)
899                fprintf(outfile, "get substring list failed %d\n", rc);
900              else
901                {
902                for (i = 0; i < count; i++)
903                  fprintf(outfile, "%2dL %s\n", i, stringlist[i]);
904                if (stringlist[i] != NULL)
905                  fprintf(outfile, "string list not terminated by NULL\n");
906                free((void *)stringlist);
907                }
908              }
909    
910          }          }
911        else        else
912          {          {
913          if (count == -1) fprintf(outfile, "No match\n");          if (start_offset == 0)
914            else fprintf(outfile, "Error %d\n", count);            {
915              if (count == -1) fprintf(outfile, "No match\n");
916                else fprintf(outfile, "Error %d\n", count);
917              }
918            start_offset = -1;
919            }
920    
921          if ((!do_g && !do_G) || start_offset <= 0) break;
922          if (do_G)
923            {
924            bptr += start_offset;
925            len -= start_offset;
926          }          }
927        }        }
928      }      }
929    
930      CONTINUE:
931    if (posix || do_posix) regfree(&preg);    if (posix || do_posix) regfree(&preg);
932    if (re != NULL) free(re);    if (re != NULL) free(re);
933    if (extra != NULL) free(extra);    if (extra != NULL) free(extra);
934      if (tables != NULL)
935        {
936        free((void *)tables);
937        setlocale(LC_CTYPE, "C");
938        }
939    }    }
940    
 END_OFF:  
941  fprintf(outfile, "\n");  fprintf(outfile, "\n");
942  return 0;  return 0;
943  }  }
Legend:
Removed from v.3  
changed lines
  Added in v.35
  ViewVC Help
Powered by ViewVC 1.1.5