/[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 411 by ph10, Fri Apr 10 15:40:21 2009 UTC revision 598 by ph10, Sat May 7 15:37:31 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 118  external symbols to prevent clashes. */ Line 126  external symbols to prevent clashes. */
126    
127  /* We also need the pcre_printint() function for printing out compiled  /* We also need the pcre_printint() function for printing out compiled
128  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
129  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
130    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. */  
131    
132    #define COMPILING_PCRETEST
133  #include "pcre_printint.src"  #include "pcre_printint.src"
134    
135  #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))  /* The definition of the macro PRINTABLE, which determines whether to print an
136    output character as-is or as a hex value when showing compiled patterns, is
137    contained in the printint.src file. We uses it here also, in cases when the
138    locale has not been explicitly changed, so as to get consistent output from
139    systems that differ in their output from isprint() even in the "C" locale. */
140    
141    #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
142    
143  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
144  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 188  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             *
234    *************************************************/
235    
236    /* By default, the "tables" pointer when calling PCRE is set to NULL, thereby
237    using the default tables of the library. However, the T option can be used to
238    select alternate sets of tables, for different kinds of testing. Note also that
239    the L (locale) option also adjusts the tables. */
240    
241    /* This is the set of tables distributed as default with PCRE. It recognizes
242    only ASCII characters. */
243    
244    static const unsigned char tables0[] = {
245    
246    /* This table is a lower casing table. */
247    
248        0,  1,  2,  3,  4,  5,  6,  7,
249        8,  9, 10, 11, 12, 13, 14, 15,
250       16, 17, 18, 19, 20, 21, 22, 23,
251       24, 25, 26, 27, 28, 29, 30, 31,
252       32, 33, 34, 35, 36, 37, 38, 39,
253       40, 41, 42, 43, 44, 45, 46, 47,
254       48, 49, 50, 51, 52, 53, 54, 55,
255       56, 57, 58, 59, 60, 61, 62, 63,
256       64, 97, 98, 99,100,101,102,103,
257      104,105,106,107,108,109,110,111,
258      112,113,114,115,116,117,118,119,
259      120,121,122, 91, 92, 93, 94, 95,
260       96, 97, 98, 99,100,101,102,103,
261      104,105,106,107,108,109,110,111,
262      112,113,114,115,116,117,118,119,
263      120,121,122,123,124,125,126,127,
264      128,129,130,131,132,133,134,135,
265      136,137,138,139,140,141,142,143,
266      144,145,146,147,148,149,150,151,
267      152,153,154,155,156,157,158,159,
268      160,161,162,163,164,165,166,167,
269      168,169,170,171,172,173,174,175,
270      176,177,178,179,180,181,182,183,
271      184,185,186,187,188,189,190,191,
272      192,193,194,195,196,197,198,199,
273      200,201,202,203,204,205,206,207,
274      208,209,210,211,212,213,214,215,
275      216,217,218,219,220,221,222,223,
276      224,225,226,227,228,229,230,231,
277      232,233,234,235,236,237,238,239,
278      240,241,242,243,244,245,246,247,
279      248,249,250,251,252,253,254,255,
280    
281    /* This table is a case flipping table. */
282    
283        0,  1,  2,  3,  4,  5,  6,  7,
284        8,  9, 10, 11, 12, 13, 14, 15,
285       16, 17, 18, 19, 20, 21, 22, 23,
286       24, 25, 26, 27, 28, 29, 30, 31,
287       32, 33, 34, 35, 36, 37, 38, 39,
288       40, 41, 42, 43, 44, 45, 46, 47,
289       48, 49, 50, 51, 52, 53, 54, 55,
290       56, 57, 58, 59, 60, 61, 62, 63,
291       64, 97, 98, 99,100,101,102,103,
292      104,105,106,107,108,109,110,111,
293      112,113,114,115,116,117,118,119,
294      120,121,122, 91, 92, 93, 94, 95,
295       96, 65, 66, 67, 68, 69, 70, 71,
296       72, 73, 74, 75, 76, 77, 78, 79,
297       80, 81, 82, 83, 84, 85, 86, 87,
298       88, 89, 90,123,124,125,126,127,
299      128,129,130,131,132,133,134,135,
300      136,137,138,139,140,141,142,143,
301      144,145,146,147,148,149,150,151,
302      152,153,154,155,156,157,158,159,
303      160,161,162,163,164,165,166,167,
304      168,169,170,171,172,173,174,175,
305      176,177,178,179,180,181,182,183,
306      184,185,186,187,188,189,190,191,
307      192,193,194,195,196,197,198,199,
308      200,201,202,203,204,205,206,207,
309      208,209,210,211,212,213,214,215,
310      216,217,218,219,220,221,222,223,
311      224,225,226,227,228,229,230,231,
312      232,233,234,235,236,237,238,239,
313      240,241,242,243,244,245,246,247,
314      248,249,250,251,252,253,254,255,
315    
316    /* This table contains bit maps for various character classes. Each map is 32
317    bytes long and the bits run from the least significant end of each byte. The
318    classes that have their own maps are: space, xdigit, digit, upper, lower, word,
319    graph, print, punct, and cntrl. Other classes are built from combinations. */
320    
321      0x00,0x3e,0x00,0x00,0x01,0x00,0x00,0x00,
322      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
323      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
324      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
325    
326      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
327      0x7e,0x00,0x00,0x00,0x7e,0x00,0x00,0x00,
328      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
329      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
330    
331      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
332      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
333      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
334      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
335    
336      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
337      0xfe,0xff,0xff,0x07,0x00,0x00,0x00,0x00,
338      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
339      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
340    
341      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
342      0x00,0x00,0x00,0x00,0xfe,0xff,0xff,0x07,
343      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
344      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
345    
346      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
347      0xfe,0xff,0xff,0x87,0xfe,0xff,0xff,0x07,
348      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
349      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
350    
351      0x00,0x00,0x00,0x00,0xfe,0xff,0xff,0xff,
352      0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x7f,
353      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
354      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
355    
356      0x00,0x00,0x00,0x00,0xff,0xff,0xff,0xff,
357      0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x7f,
358      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
359      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
360    
361      0x00,0x00,0x00,0x00,0xfe,0xff,0x00,0xfc,
362      0x01,0x00,0x00,0xf8,0x01,0x00,0x00,0x78,
363      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
364      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
365    
366      0xff,0xff,0xff,0xff,0x00,0x00,0x00,0x00,
367      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x80,
368      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
369      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
370    
371    /* This table identifies various classes of character by individual bits:
372      0x01   white space character
373      0x02   letter
374      0x04   decimal digit
375      0x08   hexadecimal digit
376      0x10   alphanumeric or '_'
377      0x80   regular expression metacharacter or binary zero
378    */
379    
380      0x80,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   0-  7 */
381      0x00,0x01,0x01,0x00,0x01,0x01,0x00,0x00, /*   8- 15 */
382      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  16- 23 */
383      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  24- 31 */
384      0x01,0x00,0x00,0x00,0x80,0x00,0x00,0x00, /*    - '  */
385      0x80,0x80,0x80,0x80,0x00,0x00,0x80,0x00, /*  ( - /  */
386      0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c, /*  0 - 7  */
387      0x1c,0x1c,0x00,0x00,0x00,0x00,0x00,0x80, /*  8 - ?  */
388      0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  @ - G  */
389      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  H - O  */
390      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  P - W  */
391      0x12,0x12,0x12,0x80,0x80,0x00,0x80,0x10, /*  X - _  */
392      0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  ` - g  */
393      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  h - o  */
394      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  p - w  */
395      0x12,0x12,0x12,0x80,0x80,0x00,0x00,0x00, /*  x -127 */
396      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 128-135 */
397      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 136-143 */
398      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144-151 */
399      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 152-159 */
400      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160-167 */
401      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 168-175 */
402      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 176-183 */
403      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
404      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 192-199 */
405      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 200-207 */
406      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 208-215 */
407      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 216-223 */
408      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 224-231 */
409      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 232-239 */
410      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 240-247 */
411      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00};/* 248-255 */
412    
413    /* This is a set of tables that came orginally from a Windows user. It seems to
414    be at least an approximation of ISO 8859. In particular, there are characters
415    greater than 128 that are marked as spaces, letters, etc. */
416    
417    static const unsigned char tables1[] = {
418    0,1,2,3,4,5,6,7,
419    8,9,10,11,12,13,14,15,
420    16,17,18,19,20,21,22,23,
421    24,25,26,27,28,29,30,31,
422    32,33,34,35,36,37,38,39,
423    40,41,42,43,44,45,46,47,
424    48,49,50,51,52,53,54,55,
425    56,57,58,59,60,61,62,63,
426    64,97,98,99,100,101,102,103,
427    104,105,106,107,108,109,110,111,
428    112,113,114,115,116,117,118,119,
429    120,121,122,91,92,93,94,95,
430    96,97,98,99,100,101,102,103,
431    104,105,106,107,108,109,110,111,
432    112,113,114,115,116,117,118,119,
433    120,121,122,123,124,125,126,127,
434    128,129,130,131,132,133,134,135,
435    136,137,138,139,140,141,142,143,
436    144,145,146,147,148,149,150,151,
437    152,153,154,155,156,157,158,159,
438    160,161,162,163,164,165,166,167,
439    168,169,170,171,172,173,174,175,
440    176,177,178,179,180,181,182,183,
441    184,185,186,187,188,189,190,191,
442    224,225,226,227,228,229,230,231,
443    232,233,234,235,236,237,238,239,
444    240,241,242,243,244,245,246,215,
445    248,249,250,251,252,253,254,223,
446    224,225,226,227,228,229,230,231,
447    232,233,234,235,236,237,238,239,
448    240,241,242,243,244,245,246,247,
449    248,249,250,251,252,253,254,255,
450    0,1,2,3,4,5,6,7,
451    8,9,10,11,12,13,14,15,
452    16,17,18,19,20,21,22,23,
453    24,25,26,27,28,29,30,31,
454    32,33,34,35,36,37,38,39,
455    40,41,42,43,44,45,46,47,
456    48,49,50,51,52,53,54,55,
457    56,57,58,59,60,61,62,63,
458    64,97,98,99,100,101,102,103,
459    104,105,106,107,108,109,110,111,
460    112,113,114,115,116,117,118,119,
461    120,121,122,91,92,93,94,95,
462    96,65,66,67,68,69,70,71,
463    72,73,74,75,76,77,78,79,
464    80,81,82,83,84,85,86,87,
465    88,89,90,123,124,125,126,127,
466    128,129,130,131,132,133,134,135,
467    136,137,138,139,140,141,142,143,
468    144,145,146,147,148,149,150,151,
469    152,153,154,155,156,157,158,159,
470    160,161,162,163,164,165,166,167,
471    168,169,170,171,172,173,174,175,
472    176,177,178,179,180,181,182,183,
473    184,185,186,187,188,189,190,191,
474    224,225,226,227,228,229,230,231,
475    232,233,234,235,236,237,238,239,
476    240,241,242,243,244,245,246,215,
477    248,249,250,251,252,253,254,223,
478    192,193,194,195,196,197,198,199,
479    200,201,202,203,204,205,206,207,
480    208,209,210,211,212,213,214,247,
481    216,217,218,219,220,221,222,255,
482    0,62,0,0,1,0,0,0,
483    0,0,0,0,0,0,0,0,
484    32,0,0,0,1,0,0,0,
485    0,0,0,0,0,0,0,0,
486    0,0,0,0,0,0,255,3,
487    126,0,0,0,126,0,0,0,
488    0,0,0,0,0,0,0,0,
489    0,0,0,0,0,0,0,0,
490    0,0,0,0,0,0,255,3,
491    0,0,0,0,0,0,0,0,
492    0,0,0,0,0,0,12,2,
493    0,0,0,0,0,0,0,0,
494    0,0,0,0,0,0,0,0,
495    254,255,255,7,0,0,0,0,
496    0,0,0,0,0,0,0,0,
497    255,255,127,127,0,0,0,0,
498    0,0,0,0,0,0,0,0,
499    0,0,0,0,254,255,255,7,
500    0,0,0,0,0,4,32,4,
501    0,0,0,128,255,255,127,255,
502    0,0,0,0,0,0,255,3,
503    254,255,255,135,254,255,255,7,
504    0,0,0,0,0,4,44,6,
505    255,255,127,255,255,255,127,255,
506    0,0,0,0,254,255,255,255,
507    255,255,255,255,255,255,255,127,
508    0,0,0,0,254,255,255,255,
509    255,255,255,255,255,255,255,255,
510    0,2,0,0,255,255,255,255,
511    255,255,255,255,255,255,255,127,
512    0,0,0,0,255,255,255,255,
513    255,255,255,255,255,255,255,255,
514    0,0,0,0,254,255,0,252,
515    1,0,0,248,1,0,0,120,
516    0,0,0,0,254,255,255,255,
517    0,0,128,0,0,0,128,0,
518    255,255,255,255,0,0,0,0,
519    0,0,0,0,0,0,0,128,
520    255,255,255,255,0,0,0,0,
521    0,0,0,0,0,0,0,0,
522    128,0,0,0,0,0,0,0,
523    0,1,1,0,1,1,0,0,
524    0,0,0,0,0,0,0,0,
525    0,0,0,0,0,0,0,0,
526    1,0,0,0,128,0,0,0,
527    128,128,128,128,0,0,128,0,
528    28,28,28,28,28,28,28,28,
529    28,28,0,0,0,0,0,128,
530    0,26,26,26,26,26,26,18,
531    18,18,18,18,18,18,18,18,
532    18,18,18,18,18,18,18,18,
533    18,18,18,128,128,0,128,16,
534    0,26,26,26,26,26,26,18,
535    18,18,18,18,18,18,18,18,
536    18,18,18,18,18,18,18,18,
537    18,18,18,128,128,0,0,0,
538    0,0,0,0,0,1,0,0,
539    0,0,0,0,0,0,0,0,
540    0,0,0,0,0,0,0,0,
541    0,0,0,0,0,0,0,0,
542    1,0,0,0,0,0,0,0,
543    0,0,18,0,0,0,0,0,
544    0,0,20,20,0,18,0,0,
545    0,20,18,0,0,0,0,0,
546    18,18,18,18,18,18,18,18,
547    18,18,18,18,18,18,18,18,
548    18,18,18,18,18,18,18,0,
549    18,18,18,18,18,18,18,18,
550    18,18,18,18,18,18,18,18,
551    18,18,18,18,18,18,18,18,
552    18,18,18,18,18,18,18,0,
553    18,18,18,18,18,18,18,18
554    };
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  /*************************************************  /*************************************************
# Line 220  uschar *here = start; Line 609  uschar *here = start;
609    
610  for (;;)  for (;;)
611    {    {
612    int rlen = buffer_size - (here - buffer);    int rlen = (int)(buffer_size - (here - buffer));
613    
614    if (rlen > 1000)    if (rlen > 1000)
615      {      {
# Line 250  for (;;) Line 639  for (;;)
639      /* Read the next line by normal means, prompting if the file is stdin. */      /* Read the next line by normal means, prompting if the file is stdin. */
640    
641        {        {
642        if (f == stdin) printf(prompt);        if (f == stdin) printf("%s", prompt);
643        if (fgets((char *)here, rlen,  f) == NULL)        if (fgets((char *)here, rlen,  f) == NULL)
644          return (here == start)? NULL : start;          return (here == start)? NULL : start;
645        }        }
# Line 740  return 0; Line 1129  return 0;
1129  *         Check newline indicator                *  *         Check newline indicator                *
1130  *************************************************/  *************************************************/
1131    
1132  /* This is used both at compile and run-time to check for <xxx> escapes, where  /* This is used both at compile and run-time to check for <xxx> escapes. Print
1133  xxx is LF, CR, CRLF, ANYCRLF, or ANY. Print a message and return 0 if there is  a message and return 0 if there is no match.
 no match.  
1134    
1135  Arguments:  Arguments:
1136    p           points after the leading '<'    p           points after the leading '<'
# Line 1039  while (!done) Line 1427  while (!done)
1427  #endif  #endif
1428    
1429    const char *error;    const char *error;
1430      unsigned char *markptr;
1431    unsigned char *p, *pp, *ppp;    unsigned char *p, *pp, *ppp;
1432    unsigned char *to_file = NULL;    unsigned char *to_file = NULL;
1433    const unsigned char *tables = NULL;    const unsigned char *tables = NULL;
1434    unsigned long int true_size, true_study_size = 0;    unsigned long int true_size, true_study_size = 0;
1435    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
1436      int do_mark = 0;
1437    int do_study = 0;    int do_study = 0;
1438    int do_debug = debug;    int do_debug = debug;
1439    int do_G = 0;    int do_G = 0;
# Line 1161  while (!done) Line 1551  while (!done)
1551      }      }
1552    
1553    pp = p;    pp = p;
1554    poffset = p - buffer;    poffset = (int)(p - buffer);
1555    
1556    for(;;)    for(;;)
1557      {      {
# Line 1225  while (!done) Line 1615  while (!done)
1615        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1616        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1617        case 'J': options |= PCRE_DUPNAMES; break;        case 'J': options |= PCRE_DUPNAMES; break;
1618          case 'K': do_mark = 1; break;
1619        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1620        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1621    
# Line 1234  while (!done) Line 1625  while (!done)
1625    
1626        case 'S': do_study = 1; break;        case 'S': do_study = 1; break;
1627        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1628          case 'W': options |= PCRE_UCP; break;
1629        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
1630          case 'Y': options |= PCRE_NO_START_OPTIMISE; break;
1631        case 'Z': debug_lengths = 0; break;        case 'Z': debug_lengths = 0; break;
1632        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1633        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
1634    
1635          case 'T':
1636          switch (*pp++)
1637            {
1638            case '0': tables = tables0; break;
1639            case '1': tables = tables1; break;
1640    
1641            case '\r':
1642            case '\n':
1643            case ' ':
1644            case 0:
1645            fprintf(outfile, "** Missing table number after /T\n");
1646            goto SKIP_DATA;
1647    
1648            default:
1649            fprintf(outfile, "** Bad table number \"%c\" after /T\n", pp[-1]);
1650            goto SKIP_DATA;
1651            }
1652          break;
1653    
1654        case 'L':        case 'L':
1655        ppp = pp;        ppp = pp;
1656        /* The '\r' test here is so that it works on Windows. */        /* The '\r' test here is so that it works on Windows. */
# Line 1264  while (!done) Line 1676  while (!done)
1676    
1677        case '<':        case '<':
1678          {          {
1679          if (strncmp((char *)pp, "JS>", 3) == 0)          if (strncmpic(pp, (uschar *)"JS>", 3) == 0)
1680            {            {
1681            options |= PCRE_JAVASCRIPT_COMPAT;            options |= PCRE_JAVASCRIPT_COMPAT;
1682            pp += 3;            pp += 3;
# Line 1305  while (!done) Line 1717  while (!done)
1717      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1718      if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;      if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1719      if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;      if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1720        if ((options & PCRE_UCP) != 0) cflags |= REG_UCP;
1721        if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY;
1722    
1723      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1724    
# Line 1325  while (!done) Line 1739  while (!done)
1739  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1740    
1741      {      {
1742        unsigned long int get_options;
1743    
1744      if (timeit > 0)      if (timeit > 0)
1745        {        {
1746        register int i;        register int i;
# Line 1368  while (!done) Line 1784  while (!done)
1784        goto CONTINUE;        goto CONTINUE;
1785        }        }
1786    
1787      /* Compilation succeeded; print data if required. There are now two      /* Compilation succeeded. It is now possible to set the UTF-8 option from
1788      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
1789      returns only limited data. Check that it agrees with the newer one. */      lines. */
1790    
1791        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1792        if ((get_options & PCRE_UTF8) != 0) use_utf8 = 1;
1793    
1794        /* Print information if required. There are now two info-returning
1795        functions. The old one has a limited interface and returns only limited
1796        data. Check that it agrees with the newer one. */
1797    
1798      if (log_store)      if (log_store)
1799        fprintf(outfile, "Memory allocation (code space): %d\n",        fprintf(outfile, "Memory allocation (code space): %d\n",
# Line 1409  while (!done) Line 1832  while (!done)
1832          true_study_size = ((pcre_study_data *)(extra->study_data))->size;          true_study_size = ((pcre_study_data *)(extra->study_data))->size;
1833        }        }
1834    
1835        /* If /K was present, we set up for handling MARK data. */
1836    
1837        if (do_mark)
1838          {
1839          if (extra == NULL)
1840            {
1841            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1842            extra->flags = 0;
1843            }
1844          extra->mark = &markptr;
1845          extra->flags |= PCRE_EXTRA_MARK;
1846          }
1847    
1848      /* 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
1849      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
1850      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 1441  while (!done) Line 1877  while (!done)
1877          {          {
1878          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
1879          rsd->size = byteflip(rsd->size, sizeof(rsd->size));          rsd->size = byteflip(rsd->size, sizeof(rsd->size));
1880          rsd->options = byteflip(rsd->options, sizeof(rsd->options));          rsd->flags = byteflip(rsd->flags, sizeof(rsd->flags));
1881            rsd->minlength = byteflip(rsd->minlength, sizeof(rsd->minlength));
1882          }          }
1883        }        }
1884    
# Line 1455  while (!done) Line 1892  while (!done)
1892        pcre_printint(re, outfile, debug_lengths);        pcre_printint(re, outfile, debug_lengths);
1893        }        }
1894    
1895        /* We already have the options in get_options (see above) */
1896    
1897      if (do_showinfo)      if (do_showinfo)
1898        {        {
1899        unsigned long int get_options, all_options;        unsigned long int all_options;
1900  #if !defined NOINFOCHECK  #if !defined NOINFOCHECK
1901        int old_first_char, old_options, old_count;        int old_first_char, old_options, old_count;
1902  #endif  #endif
# Line 1466  while (!done) Line 1905  while (!done)
1905        int nameentrysize, namecount;        int nameentrysize, namecount;
1906        const uschar *nametable;        const uschar *nametable;
1907    
       new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);  
1908        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1909        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
1910        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);
# Line 1526  while (!done) Line 1964  while (!done)
1964        if (do_flip) all_options = byteflip(all_options, sizeof(all_options));        if (do_flip) all_options = byteflip(all_options, sizeof(all_options));
1965    
1966        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1967          else fprintf(outfile, "Options:%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",
1968            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1969            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1970            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
# Line 1540  while (!done) Line 1978  while (!done)
1978            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1979            ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",            ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1980            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1981              ((get_options & PCRE_UCP) != 0)? " ucp" : "",
1982            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1983              ((get_options & PCRE_NO_START_OPTIMIZE) != 0)? " no_start_optimize" : "",
1984            ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");            ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1985    
1986        if (jchanged) fprintf(outfile, "Duplicate name status changes\n");        if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
# Line 1617  while (!done) Line 2057  while (!done)
2057          else          else
2058            {            {
2059            uschar *start_bits = NULL;            uschar *start_bits = NULL;
2060            new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);            int minlength;
2061    
2062              new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength);
2063              fprintf(outfile, "Subject length lower bound = %d\n", minlength);
2064    
2065              new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);
2066            if (start_bits == NULL)            if (start_bits == NULL)
2067              fprintf(outfile, "No starting byte set\n");              fprintf(outfile, "No set of starting bytes\n");
2068            else            else
2069              {              {
2070              int i;              int i;
# Line 1702  while (!done) Line 2146  while (!done)
2146    
2147        new_free(re);        new_free(re);
2148        if (extra != NULL) new_free(extra);        if (extra != NULL) new_free(extra);
2149        if (tables != NULL) new_free((void *)tables);        if (locale_set)
2150            {
2151            new_free((void *)tables);
2152            setlocale(LC_CTYPE, "C");
2153            locale_set = 0;
2154            }
2155        continue;  /* With next regex */        continue;  /* With next regex */
2156        }        }
2157      }        /* End of non-POSIX compile */      }        /* End of non-POSIX compile */
# Line 1724  while (!done) Line 2173  while (!done)
2173      int getlist = 0;      int getlist = 0;
2174      int gmatched = 0;      int gmatched = 0;
2175      int start_offset = 0;      int start_offset = 0;
2176        int start_offset_sign = 1;
2177      int g_notempty = 0;      int g_notempty = 0;
2178      int use_dfa = 0;      int use_dfa = 0;
2179    
# Line 1751  while (!done) Line 2201  while (!done)
2201        {        {
2202        if (extend_inputline(infile, buffer + len, "data> ") == NULL)        if (extend_inputline(infile, buffer + len, "data> ") == NULL)
2203          {          {
2204          if (len > 0) break;          if (len > 0)    /* Reached EOF without hitting a newline */
2205              {
2206              fprintf(outfile, "\n");
2207              break;
2208              }
2209          done = 1;          done = 1;
2210          goto CONTINUE;          goto CONTINUE;
2211          }          }
# Line 1852  while (!done) Line 2306  while (!done)
2306          continue;          continue;
2307    
2308          case '>':          case '>':
2309            if (*p == '-')
2310              {
2311              start_offset_sign = -1;
2312              p++;
2313              }
2314          while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';          while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';
2315            start_offset *= start_offset_sign;
2316          continue;          continue;
2317    
2318          case 'A':  /* Option setting */          case 'A':  /* Option setting */
# Line 1925  while (!done) Line 2385  while (!done)
2385  #endif  #endif
2386            use_dfa = 1;            use_dfa = 1;
2387          continue;          continue;
2388    #endif
2389    
2390    #if !defined NODFA
2391          case 'F':          case 'F':
2392          options |= PCRE_DFA_SHORTEST;          options |= PCRE_DFA_SHORTEST;
2393          continue;          continue;
# Line 1959  while (!done) Line 2421  while (!done)
2421          continue;          continue;
2422    
2423          case 'N':          case 'N':
2424          options |= PCRE_NOTEMPTY;          if ((options & PCRE_NOTEMPTY) != 0)
2425              options = (options & ~PCRE_NOTEMPTY) | PCRE_NOTEMPTY_ATSTART;
2426            else
2427              options |= PCRE_NOTEMPTY;
2428          continue;          continue;
2429    
2430          case 'O':          case 'O':
# Line 1982  while (!done) Line 2447  while (!done)
2447          continue;          continue;
2448    
2449          case 'P':          case 'P':
2450          options |= PCRE_PARTIAL;          options |= ((options & PCRE_PARTIAL_SOFT) == 0)?
2451              PCRE_PARTIAL_SOFT : PCRE_PARTIAL_HARD;
2452          continue;          continue;
2453    
2454          case 'Q':          case 'Q':
# Line 2041  while (!done) Line 2507  while (!done)
2507        *q++ = c;        *q++ = c;
2508        }        }
2509      *q = 0;      *q = 0;
2510      len = q - dbuffer;      len = (int)(q - dbuffer);
2511    
2512      /* Move the data to the end of the buffer so that a read over the end of      /* Move the data to the end of the buffer so that a read over the end of
2513      the buffer will be seen by valgrind, even if it doesn't cause a crash. If      the buffer will be seen by valgrind, even if it doesn't cause a crash. If
# Line 2124  while (!done) Line 2590  while (!done)
2590    
2591      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
2592        {        {
2593          markptr = NULL;
2594    
2595        if (timeitm > 0)        if (timeitm > 0)
2596          {          {
2597          register int i;          register int i;
# Line 2135  while (!done) Line 2603  while (!done)
2603            {            {
2604            int workspace[1000];            int workspace[1000];
2605            for (i = 0; i < timeitm; i++)            for (i = 0; i < timeitm; i++)
2606              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,              count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2607                options | g_notempty, use_offsets, use_size_offsets, workspace,                options | g_notempty, use_offsets, use_size_offsets, workspace,
2608                sizeof(workspace)/sizeof(int));                sizeof(workspace)/sizeof(int));
2609            }            }
# Line 2198  while (!done) Line 2666  while (!done)
2666        else if (all_use_dfa || use_dfa)        else if (all_use_dfa || use_dfa)
2667          {          {
2668          int workspace[1000];          int workspace[1000];
2669          count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,          count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2670            options | g_notempty, use_offsets, use_size_offsets, workspace,            options | g_notempty, use_offsets, use_size_offsets, workspace,
2671            sizeof(workspace)/sizeof(int));            sizeof(workspace)/sizeof(int));
2672          if (count == 0)          if (count == 0)
# Line 2269  while (!done) Line 2737  while (!done)
2737              }              }
2738            }            }
2739    
2740            if (markptr != NULL) fprintf(outfile, "MK: %s\n", markptr);
2741    
2742          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2743            {            {
2744            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
# Line 2352  while (!done) Line 2822  while (!done)
2822    
2823        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2824          {          {
2825          fprintf(outfile, "Partial match");          if (markptr == NULL) fprintf(outfile, "Partial match");
2826  #if !defined NODFA            else fprintf(outfile, "Partial match, mark=%s", markptr);
2827          if ((all_use_dfa || use_dfa) && use_size_offsets > 2)          if (use_size_offsets > 1)
2828            fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],            {
2829              bptr + use_offsets[0]);            fprintf(outfile, ": ");
2830  #endif            pchars(bptr + use_offsets[0], use_offsets[1] - use_offsets[0],
2831                outfile);
2832              }
2833          fprintf(outfile, "\n");          fprintf(outfile, "\n");
2834          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2835          }          }
# Line 2367  while (!done) Line 2839  while (!done)
2839        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
2840        string - that was checked before setting g_notempty.        string - that was checked before setting g_notempty.
2841    
2842        Complication arises in the case when the newline option is "any" or        Complication arises in the case when the newline convention is "any",
2843        "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
2844        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,
2845        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
2846        Fudge the offset value to achieve this.        pcre_exec(). Fudge the offset value to achieve this. We check for a
2847          newline setting in the pattern; if none was set, use pcre_config() to
2848          find the default.
2849    
2850        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
2851        character, not one byte. */        character, not one byte. */
# Line 2396  while (!done) Line 2870  while (!done)
2870                      (d == -1)? PCRE_NEWLINE_ANY : 0;                      (d == -1)? PCRE_NEWLINE_ANY : 0;
2871              }              }
2872            if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||            if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2873                   (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_CRLF ||
2874                 (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)                 (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2875                &&                &&
2876                start_offset < len - 1 &&                start_offset < len - 1 &&
# Line 2406  while (!done) Line 2881  while (!done)
2881              {              {
2882              while (start_offset + onechar < len)              while (start_offset + onechar < len)
2883                {                {
2884                int tb = bptr[start_offset+onechar];                if ((bptr[start_offset+onechar] & 0xc0) != 0x80) break;
2885                if (tb <= 127) break;                onechar++;
               tb &= 0xc0;  
               if (tb != 0 && tb != 0xc0) onechar++;  
2886                }                }
2887              }              }
2888            use_offsets[1] = start_offset + onechar;            use_offsets[1] = start_offset + onechar;
2889            }            }
2890          else          else
2891            {            {
2892            if (count == PCRE_ERROR_NOMATCH)            switch(count)
2893              {              {
2894              if (gmatched == 0) fprintf(outfile, "No match\n");              case PCRE_ERROR_NOMATCH:
2895                if (gmatched == 0)
2896                  {
2897                  if (markptr == NULL) fprintf(outfile, "No match\n");
2898                    else fprintf(outfile, "No match, mark = %s\n", markptr);
2899                  }
2900                break;
2901    
2902                case PCRE_ERROR_BADUTF8:
2903                case PCRE_ERROR_SHORTUTF8:
2904                fprintf(outfile, "Error %d (%s UTF-8 string)", count,
2905                  (count == PCRE_ERROR_BADUTF8)? "bad" : "short");
2906                if (use_size_offsets >= 2)
2907                  fprintf(outfile, " offset=%d reason=%d", use_offsets[0],
2908                    use_offsets[1]);
2909                fprintf(outfile, "\n");
2910                break;
2911    
2912                default:
2913                fprintf(outfile, "Error %d (%s)\n", count, errtexts[-count]);
2914                break;
2915              }              }
2916            else fprintf(outfile, "Error %d\n", count);  
2917            break;  /* Out of the /g loop */            break;  /* Out of the /g loop */
2918            }            }
2919          }          }
# Line 2430  while (!done) Line 2923  while (!done)
2923        if (!do_g && !do_G) break;        if (!do_g && !do_G) break;
2924    
2925        /* 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
2926        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
2927        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
2928        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
2929        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
2930        character. */        character. */
2931    
# Line 2441  while (!done) Line 2934  while (!done)
2934        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2935          {          {
2936          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
2937          g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED;          g_notempty = PCRE_NOTEMPTY_ATSTART | PCRE_ANCHORED;
2938          }          }
2939    
2940        /* For /g, update the start offset, leaving the rest alone */        /* For /g, update the start offset, leaving the rest alone */
# Line 2468  while (!done) Line 2961  while (!done)
2961    
2962    if (re != NULL) new_free(re);    if (re != NULL) new_free(re);
2963    if (extra != NULL) new_free(extra);    if (extra != NULL) new_free(extra);
2964    if (tables != NULL)    if (locale_set)
2965      {      {
2966      new_free((void *)tables);      new_free((void *)tables);
2967      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");

Legend:
Removed from v.411  
changed lines
  Added in v.598

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12