/[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 416 by ph10, Sat Apr 11 14:34:02 2009 UTC revision 566 by ph10, Wed Nov 3 18:32:55 2010 UTC
# Line 118  external symbols to prevent clashes. */ Line 118  external symbols to prevent clashes. */
118    
119  /* We also need the pcre_printint() function for printing out compiled  /* We also need the pcre_printint() function for printing out compiled
120  patterns. This function is in a separate file so that it can be included in  patterns. This function is in a separate file so that it can be included in
121  pcre_compile.c when that module is compiled with debugging enabled.  pcre_compile.c when that module is compiled with debugging enabled. It needs to
122    know which case is being compiled. */
 The definition of the macro PRINTABLE, which determines whether to print an  
 output character as-is or as a hex value when showing compiled patterns, is  
 contained in this file. We uses it here also, in cases when the locale has not  
 been explicitly changed, so as to get consistent output from systems that  
 differ in their output from isprint() even in the "C" locale. */  
123    
124    #define COMPILING_PCRETEST
125  #include "pcre_printint.src"  #include "pcre_printint.src"
126    
127  #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))  /* The definition of the macro PRINTABLE, which determines whether to print an
128    output character as-is or as a hex value when showing compiled patterns, is
129    contained in the printint.src file. We uses it here also, in cases when the
130    locale has not been explicitly changed, so as to get consistent output from
131    systems that differ in their output from isprint() even in the "C" locale. */
132    
133    #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
134    
135  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
136  testing the POSIX interface, though this is not available via the standard  testing the POSIX interface, though this is not available via the standard
# Line 189  static uschar *dbuffer = NULL; Line 190  static uschar *dbuffer = NULL;
190  static uschar *pbuffer = NULL;  static uschar *pbuffer = NULL;
191    
192    
193    /*************************************************
194    *         Alternate character tables             *
195    *************************************************/
196    
197    /* By default, the "tables" pointer when calling PCRE is set to NULL, thereby
198    using the default tables of the library. However, the T option can be used to
199    select alternate sets of tables, for different kinds of testing. Note also that
200    the L (locale) option also adjusts the tables. */
201    
202    /* This is the set of tables distributed as default with PCRE. It recognizes
203    only ASCII characters. */
204    
205    static const unsigned char tables0[] = {
206    
207    /* This table is a lower casing table. */
208    
209        0,  1,  2,  3,  4,  5,  6,  7,
210        8,  9, 10, 11, 12, 13, 14, 15,
211       16, 17, 18, 19, 20, 21, 22, 23,
212       24, 25, 26, 27, 28, 29, 30, 31,
213       32, 33, 34, 35, 36, 37, 38, 39,
214       40, 41, 42, 43, 44, 45, 46, 47,
215       48, 49, 50, 51, 52, 53, 54, 55,
216       56, 57, 58, 59, 60, 61, 62, 63,
217       64, 97, 98, 99,100,101,102,103,
218      104,105,106,107,108,109,110,111,
219      112,113,114,115,116,117,118,119,
220      120,121,122, 91, 92, 93, 94, 95,
221       96, 97, 98, 99,100,101,102,103,
222      104,105,106,107,108,109,110,111,
223      112,113,114,115,116,117,118,119,
224      120,121,122,123,124,125,126,127,
225      128,129,130,131,132,133,134,135,
226      136,137,138,139,140,141,142,143,
227      144,145,146,147,148,149,150,151,
228      152,153,154,155,156,157,158,159,
229      160,161,162,163,164,165,166,167,
230      168,169,170,171,172,173,174,175,
231      176,177,178,179,180,181,182,183,
232      184,185,186,187,188,189,190,191,
233      192,193,194,195,196,197,198,199,
234      200,201,202,203,204,205,206,207,
235      208,209,210,211,212,213,214,215,
236      216,217,218,219,220,221,222,223,
237      224,225,226,227,228,229,230,231,
238      232,233,234,235,236,237,238,239,
239      240,241,242,243,244,245,246,247,
240      248,249,250,251,252,253,254,255,
241    
242    /* This table is a case flipping table. */
243    
244        0,  1,  2,  3,  4,  5,  6,  7,
245        8,  9, 10, 11, 12, 13, 14, 15,
246       16, 17, 18, 19, 20, 21, 22, 23,
247       24, 25, 26, 27, 28, 29, 30, 31,
248       32, 33, 34, 35, 36, 37, 38, 39,
249       40, 41, 42, 43, 44, 45, 46, 47,
250       48, 49, 50, 51, 52, 53, 54, 55,
251       56, 57, 58, 59, 60, 61, 62, 63,
252       64, 97, 98, 99,100,101,102,103,
253      104,105,106,107,108,109,110,111,
254      112,113,114,115,116,117,118,119,
255      120,121,122, 91, 92, 93, 94, 95,
256       96, 65, 66, 67, 68, 69, 70, 71,
257       72, 73, 74, 75, 76, 77, 78, 79,
258       80, 81, 82, 83, 84, 85, 86, 87,
259       88, 89, 90,123,124,125,126,127,
260      128,129,130,131,132,133,134,135,
261      136,137,138,139,140,141,142,143,
262      144,145,146,147,148,149,150,151,
263      152,153,154,155,156,157,158,159,
264      160,161,162,163,164,165,166,167,
265      168,169,170,171,172,173,174,175,
266      176,177,178,179,180,181,182,183,
267      184,185,186,187,188,189,190,191,
268      192,193,194,195,196,197,198,199,
269      200,201,202,203,204,205,206,207,
270      208,209,210,211,212,213,214,215,
271      216,217,218,219,220,221,222,223,
272      224,225,226,227,228,229,230,231,
273      232,233,234,235,236,237,238,239,
274      240,241,242,243,244,245,246,247,
275      248,249,250,251,252,253,254,255,
276    
277    /* This table contains bit maps for various character classes. Each map is 32
278    bytes long and the bits run from the least significant end of each byte. The
279    classes that have their own maps are: space, xdigit, digit, upper, lower, word,
280    graph, print, punct, and cntrl. Other classes are built from combinations. */
281    
282      0x00,0x3e,0x00,0x00,0x01,0x00,0x00,0x00,
283      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
284      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
285      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
286    
287      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
288      0x7e,0x00,0x00,0x00,0x7e,0x00,0x00,0x00,
289      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
290      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
291    
292      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
293      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
294      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
295      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
296    
297      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
298      0xfe,0xff,0xff,0x07,0x00,0x00,0x00,0x00,
299      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
300      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
301    
302      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
303      0x00,0x00,0x00,0x00,0xfe,0xff,0xff,0x07,
304      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
305      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
306    
307      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
308      0xfe,0xff,0xff,0x87,0xfe,0xff,0xff,0x07,
309      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
310      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
311    
312      0x00,0x00,0x00,0x00,0xfe,0xff,0xff,0xff,
313      0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x7f,
314      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
315      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
316    
317      0x00,0x00,0x00,0x00,0xff,0xff,0xff,0xff,
318      0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x7f,
319      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
320      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
321    
322      0x00,0x00,0x00,0x00,0xfe,0xff,0x00,0xfc,
323      0x01,0x00,0x00,0xf8,0x01,0x00,0x00,0x78,
324      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
325      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
326    
327      0xff,0xff,0xff,0xff,0x00,0x00,0x00,0x00,
328      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x80,
329      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
330      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
331    
332    /* This table identifies various classes of character by individual bits:
333      0x01   white space character
334      0x02   letter
335      0x04   decimal digit
336      0x08   hexadecimal digit
337      0x10   alphanumeric or '_'
338      0x80   regular expression metacharacter or binary zero
339    */
340    
341      0x80,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   0-  7 */
342      0x00,0x01,0x01,0x00,0x01,0x01,0x00,0x00, /*   8- 15 */
343      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  16- 23 */
344      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  24- 31 */
345      0x01,0x00,0x00,0x00,0x80,0x00,0x00,0x00, /*    - '  */
346      0x80,0x80,0x80,0x80,0x00,0x00,0x80,0x00, /*  ( - /  */
347      0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c, /*  0 - 7  */
348      0x1c,0x1c,0x00,0x00,0x00,0x00,0x00,0x80, /*  8 - ?  */
349      0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  @ - G  */
350      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  H - O  */
351      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  P - W  */
352      0x12,0x12,0x12,0x80,0x80,0x00,0x80,0x10, /*  X - _  */
353      0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  ` - g  */
354      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  h - o  */
355      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  p - w  */
356      0x12,0x12,0x12,0x80,0x80,0x00,0x00,0x00, /*  x -127 */
357      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 128-135 */
358      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 136-143 */
359      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144-151 */
360      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 152-159 */
361      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160-167 */
362      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 168-175 */
363      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 176-183 */
364      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
365      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 192-199 */
366      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 200-207 */
367      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 208-215 */
368      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 216-223 */
369      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 224-231 */
370      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 232-239 */
371      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 240-247 */
372      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00};/* 248-255 */
373    
374    /* This is a set of tables that came orginally from a Windows user. It seems to
375    be at least an approximation of ISO 8859. In particular, there are characters
376    greater than 128 that are marked as spaces, letters, etc. */
377    
378    static const unsigned char tables1[] = {
379    0,1,2,3,4,5,6,7,
380    8,9,10,11,12,13,14,15,
381    16,17,18,19,20,21,22,23,
382    24,25,26,27,28,29,30,31,
383    32,33,34,35,36,37,38,39,
384    40,41,42,43,44,45,46,47,
385    48,49,50,51,52,53,54,55,
386    56,57,58,59,60,61,62,63,
387    64,97,98,99,100,101,102,103,
388    104,105,106,107,108,109,110,111,
389    112,113,114,115,116,117,118,119,
390    120,121,122,91,92,93,94,95,
391    96,97,98,99,100,101,102,103,
392    104,105,106,107,108,109,110,111,
393    112,113,114,115,116,117,118,119,
394    120,121,122,123,124,125,126,127,
395    128,129,130,131,132,133,134,135,
396    136,137,138,139,140,141,142,143,
397    144,145,146,147,148,149,150,151,
398    152,153,154,155,156,157,158,159,
399    160,161,162,163,164,165,166,167,
400    168,169,170,171,172,173,174,175,
401    176,177,178,179,180,181,182,183,
402    184,185,186,187,188,189,190,191,
403    224,225,226,227,228,229,230,231,
404    232,233,234,235,236,237,238,239,
405    240,241,242,243,244,245,246,215,
406    248,249,250,251,252,253,254,223,
407    224,225,226,227,228,229,230,231,
408    232,233,234,235,236,237,238,239,
409    240,241,242,243,244,245,246,247,
410    248,249,250,251,252,253,254,255,
411    0,1,2,3,4,5,6,7,
412    8,9,10,11,12,13,14,15,
413    16,17,18,19,20,21,22,23,
414    24,25,26,27,28,29,30,31,
415    32,33,34,35,36,37,38,39,
416    40,41,42,43,44,45,46,47,
417    48,49,50,51,52,53,54,55,
418    56,57,58,59,60,61,62,63,
419    64,97,98,99,100,101,102,103,
420    104,105,106,107,108,109,110,111,
421    112,113,114,115,116,117,118,119,
422    120,121,122,91,92,93,94,95,
423    96,65,66,67,68,69,70,71,
424    72,73,74,75,76,77,78,79,
425    80,81,82,83,84,85,86,87,
426    88,89,90,123,124,125,126,127,
427    128,129,130,131,132,133,134,135,
428    136,137,138,139,140,141,142,143,
429    144,145,146,147,148,149,150,151,
430    152,153,154,155,156,157,158,159,
431    160,161,162,163,164,165,166,167,
432    168,169,170,171,172,173,174,175,
433    176,177,178,179,180,181,182,183,
434    184,185,186,187,188,189,190,191,
435    224,225,226,227,228,229,230,231,
436    232,233,234,235,236,237,238,239,
437    240,241,242,243,244,245,246,215,
438    248,249,250,251,252,253,254,223,
439    192,193,194,195,196,197,198,199,
440    200,201,202,203,204,205,206,207,
441    208,209,210,211,212,213,214,247,
442    216,217,218,219,220,221,222,255,
443    0,62,0,0,1,0,0,0,
444    0,0,0,0,0,0,0,0,
445    32,0,0,0,1,0,0,0,
446    0,0,0,0,0,0,0,0,
447    0,0,0,0,0,0,255,3,
448    126,0,0,0,126,0,0,0,
449    0,0,0,0,0,0,0,0,
450    0,0,0,0,0,0,0,0,
451    0,0,0,0,0,0,255,3,
452    0,0,0,0,0,0,0,0,
453    0,0,0,0,0,0,12,2,
454    0,0,0,0,0,0,0,0,
455    0,0,0,0,0,0,0,0,
456    254,255,255,7,0,0,0,0,
457    0,0,0,0,0,0,0,0,
458    255,255,127,127,0,0,0,0,
459    0,0,0,0,0,0,0,0,
460    0,0,0,0,254,255,255,7,
461    0,0,0,0,0,4,32,4,
462    0,0,0,128,255,255,127,255,
463    0,0,0,0,0,0,255,3,
464    254,255,255,135,254,255,255,7,
465    0,0,0,0,0,4,44,6,
466    255,255,127,255,255,255,127,255,
467    0,0,0,0,254,255,255,255,
468    255,255,255,255,255,255,255,127,
469    0,0,0,0,254,255,255,255,
470    255,255,255,255,255,255,255,255,
471    0,2,0,0,255,255,255,255,
472    255,255,255,255,255,255,255,127,
473    0,0,0,0,255,255,255,255,
474    255,255,255,255,255,255,255,255,
475    0,0,0,0,254,255,0,252,
476    1,0,0,248,1,0,0,120,
477    0,0,0,0,254,255,255,255,
478    0,0,128,0,0,0,128,0,
479    255,255,255,255,0,0,0,0,
480    0,0,0,0,0,0,0,128,
481    255,255,255,255,0,0,0,0,
482    0,0,0,0,0,0,0,0,
483    128,0,0,0,0,0,0,0,
484    0,1,1,0,1,1,0,0,
485    0,0,0,0,0,0,0,0,
486    0,0,0,0,0,0,0,0,
487    1,0,0,0,128,0,0,0,
488    128,128,128,128,0,0,128,0,
489    28,28,28,28,28,28,28,28,
490    28,28,0,0,0,0,0,128,
491    0,26,26,26,26,26,26,18,
492    18,18,18,18,18,18,18,18,
493    18,18,18,18,18,18,18,18,
494    18,18,18,128,128,0,128,16,
495    0,26,26,26,26,26,26,18,
496    18,18,18,18,18,18,18,18,
497    18,18,18,18,18,18,18,18,
498    18,18,18,128,128,0,0,0,
499    0,0,0,0,0,1,0,0,
500    0,0,0,0,0,0,0,0,
501    0,0,0,0,0,0,0,0,
502    0,0,0,0,0,0,0,0,
503    1,0,0,0,0,0,0,0,
504    0,0,18,0,0,0,0,0,
505    0,0,20,20,0,18,0,0,
506    0,20,18,0,0,0,0,0,
507    18,18,18,18,18,18,18,18,
508    18,18,18,18,18,18,18,18,
509    18,18,18,18,18,18,18,0,
510    18,18,18,18,18,18,18,18,
511    18,18,18,18,18,18,18,18,
512    18,18,18,18,18,18,18,18,
513    18,18,18,18,18,18,18,0,
514    18,18,18,18,18,18,18,18
515    };
516    
517    
518    
519    
520    #ifndef HAVE_STRERROR
521    /*************************************************
522    *     Provide strerror() for non-ANSI libraries  *
523    *************************************************/
524    
525    /* Some old-fashioned systems still around (e.g. SunOS4) don't have strerror()
526    in their libraries, but can provide the same facility by this simple
527    alternative function. */
528    
529    extern int   sys_nerr;
530    extern char *sys_errlist[];
531    
532    char *
533    strerror(int n)
534    {
535    if (n < 0 || n >= sys_nerr) return "unknown error number";
536    return sys_errlist[n];
537    }
538    #endif /* HAVE_STRERROR */
539    
540    
541    
542    
543  /*************************************************  /*************************************************
544  *        Read or extend an input line            *  *        Read or extend an input line            *
# Line 220  uschar *here = start; Line 570  uschar *here = start;
570    
571  for (;;)  for (;;)
572    {    {
573    int rlen = buffer_size - (here - buffer);    int rlen = (int)(buffer_size - (here - buffer));
574    
575    if (rlen > 1000)    if (rlen > 1000)
576      {      {
# Line 250  for (;;) Line 600  for (;;)
600      /* 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. */
601    
602        {        {
603        if (f == stdin) printf(prompt);        if (f == stdin) printf("%s", prompt);
604        if (fgets((char *)here, rlen,  f) == NULL)        if (fgets((char *)here, rlen,  f) == NULL)
605          return (here == start)? NULL : start;          return (here == start)? NULL : start;
606        }        }
# Line 740  return 0; Line 1090  return 0;
1090  *         Check newline indicator                *  *         Check newline indicator                *
1091  *************************************************/  *************************************************/
1092    
1093  /* 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
1094  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.  
1095    
1096  Arguments:  Arguments:
1097    p           points after the leading '<'    p           points after the leading '<'
# Line 1039  while (!done) Line 1388  while (!done)
1388  #endif  #endif
1389    
1390    const char *error;    const char *error;
1391      unsigned char *markptr;
1392    unsigned char *p, *pp, *ppp;    unsigned char *p, *pp, *ppp;
1393    unsigned char *to_file = NULL;    unsigned char *to_file = NULL;
1394    const unsigned char *tables = NULL;    const unsigned char *tables = NULL;
1395    unsigned long int true_size, true_study_size = 0;    unsigned long int true_size, true_study_size = 0;
1396    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
1397      int do_mark = 0;
1398    int do_study = 0;    int do_study = 0;
1399    int do_debug = debug;    int do_debug = debug;
1400    int do_G = 0;    int do_G = 0;
# Line 1161  while (!done) Line 1512  while (!done)
1512      }      }
1513    
1514    pp = p;    pp = p;
1515    poffset = p - buffer;    poffset = (int)(p - buffer);
1516    
1517    for(;;)    for(;;)
1518      {      {
# Line 1225  while (!done) Line 1576  while (!done)
1576        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1577        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1578        case 'J': options |= PCRE_DUPNAMES; break;        case 'J': options |= PCRE_DUPNAMES; break;
1579          case 'K': do_mark = 1; break;
1580        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1581        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1582    
# Line 1234  while (!done) Line 1586  while (!done)
1586    
1587        case 'S': do_study = 1; break;        case 'S': do_study = 1; break;
1588        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1589          case 'W': options |= PCRE_UCP; break;
1590        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
1591        case 'Z': debug_lengths = 0; break;        case 'Z': debug_lengths = 0; break;
1592        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1593        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
1594    
1595          case 'T':
1596          switch (*pp++)
1597            {
1598            case '0': tables = tables0; break;
1599            case '1': tables = tables1; break;
1600    
1601            case '\r':
1602            case '\n':
1603            case ' ':
1604            case 0:
1605            fprintf(outfile, "** Missing table number after /T\n");
1606            goto SKIP_DATA;
1607    
1608            default:
1609            fprintf(outfile, "** Bad table number \"%c\" after /T\n", pp[-1]);
1610            goto SKIP_DATA;
1611            }
1612          break;
1613    
1614        case 'L':        case 'L':
1615        ppp = pp;        ppp = pp;
1616        /* 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 1636  while (!done)
1636    
1637        case '<':        case '<':
1638          {          {
1639          if (strncmp((char *)pp, "JS>", 3) == 0)          if (strncmpic(pp, (uschar *)"JS>", 3) == 0)
1640            {            {
1641            options |= PCRE_JAVASCRIPT_COMPAT;            options |= PCRE_JAVASCRIPT_COMPAT;
1642            pp += 3;            pp += 3;
# Line 1305  while (!done) Line 1677  while (!done)
1677      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1678      if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;      if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1679      if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;      if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1680        if ((options & PCRE_UCP) != 0) cflags |= REG_UCP;
1681        if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY;
1682    
1683      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1684    
# Line 1418  while (!done) Line 1792  while (!done)
1792          true_study_size = ((pcre_study_data *)(extra->study_data))->size;          true_study_size = ((pcre_study_data *)(extra->study_data))->size;
1793        }        }
1794    
1795        /* If /K was present, we set up for handling MARK data. */
1796    
1797        if (do_mark)
1798          {
1799          if (extra == NULL)
1800            {
1801            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1802            extra->flags = 0;
1803            }
1804          extra->mark = &markptr;
1805          extra->flags |= PCRE_EXTRA_MARK;
1806          }
1807    
1808      /* 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
1809      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
1810      possible to test PCRE's handling of byte-flipped patterns, e.g. those      possible to test PCRE's handling of byte-flipped patterns, e.g. those
# Line 1450  while (!done) Line 1837  while (!done)
1837          {          {
1838          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
1839          rsd->size = byteflip(rsd->size, sizeof(rsd->size));          rsd->size = byteflip(rsd->size, sizeof(rsd->size));
1840          rsd->options = byteflip(rsd->options, sizeof(rsd->options));          rsd->flags = byteflip(rsd->flags, sizeof(rsd->flags));
1841            rsd->minlength = byteflip(rsd->minlength, sizeof(rsd->minlength));
1842          }          }
1843        }        }
1844    
# Line 1536  while (!done) Line 1924  while (!done)
1924        if (do_flip) all_options = byteflip(all_options, sizeof(all_options));        if (do_flip) all_options = byteflip(all_options, sizeof(all_options));
1925    
1926        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
1927          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\n",
1928            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1929            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1930            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
# Line 1550  while (!done) Line 1938  while (!done)
1938            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1939            ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",            ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1940            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1941              ((get_options & PCRE_UCP) != 0)? " ucp" : "",
1942            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1943            ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");            ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1944    
# Line 1627  while (!done) Line 2016  while (!done)
2016          else          else
2017            {            {
2018            uschar *start_bits = NULL;            uschar *start_bits = NULL;
2019            new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);            int minlength;
2020    
2021              new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength);
2022              fprintf(outfile, "Subject length lower bound = %d\n", minlength);
2023    
2024              new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);
2025            if (start_bits == NULL)            if (start_bits == NULL)
2026              fprintf(outfile, "No starting byte set\n");              fprintf(outfile, "No set of starting bytes\n");
2027            else            else
2028              {              {
2029              int i;              int i;
# Line 1712  while (!done) Line 2105  while (!done)
2105    
2106        new_free(re);        new_free(re);
2107        if (extra != NULL) new_free(extra);        if (extra != NULL) new_free(extra);
2108        if (tables != NULL) new_free((void *)tables);        if (locale_set)
2109            {
2110            new_free((void *)tables);
2111            setlocale(LC_CTYPE, "C");
2112            locale_set = 0;
2113            }
2114        continue;  /* With next regex */        continue;  /* With next regex */
2115        }        }
2116      }        /* End of non-POSIX compile */      }        /* End of non-POSIX compile */
# Line 1761  while (!done) Line 2159  while (!done)
2159        {        {
2160        if (extend_inputline(infile, buffer + len, "data> ") == NULL)        if (extend_inputline(infile, buffer + len, "data> ") == NULL)
2161          {          {
2162          if (len > 0) break;          if (len > 0)    /* Reached EOF without hitting a newline */
2163              {
2164              fprintf(outfile, "\n");
2165              break;
2166              }
2167          done = 1;          done = 1;
2168          goto CONTINUE;          goto CONTINUE;
2169          }          }
# Line 1935  while (!done) Line 2337  while (!done)
2337  #endif  #endif
2338            use_dfa = 1;            use_dfa = 1;
2339          continue;          continue;
2340    #endif
2341    
2342    #if !defined NODFA
2343          case 'F':          case 'F':
2344          options |= PCRE_DFA_SHORTEST;          options |= PCRE_DFA_SHORTEST;
2345          continue;          continue;
# Line 1969  while (!done) Line 2373  while (!done)
2373          continue;          continue;
2374    
2375          case 'N':          case 'N':
2376          options |= PCRE_NOTEMPTY;          if ((options & PCRE_NOTEMPTY) != 0)
2377              options = (options & ~PCRE_NOTEMPTY) | PCRE_NOTEMPTY_ATSTART;
2378            else
2379              options |= PCRE_NOTEMPTY;
2380          continue;          continue;
2381    
2382          case 'O':          case 'O':
# Line 1992  while (!done) Line 2399  while (!done)
2399          continue;          continue;
2400    
2401          case 'P':          case 'P':
2402          options |= PCRE_PARTIAL;          options |= ((options & PCRE_PARTIAL_SOFT) == 0)?
2403              PCRE_PARTIAL_SOFT : PCRE_PARTIAL_HARD;
2404          continue;          continue;
2405    
2406          case 'Q':          case 'Q':
# Line 2051  while (!done) Line 2459  while (!done)
2459        *q++ = c;        *q++ = c;
2460        }        }
2461      *q = 0;      *q = 0;
2462      len = q - dbuffer;      len = (int)(q - dbuffer);
2463    
2464      /* 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
2465      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 2134  while (!done) Line 2542  while (!done)
2542    
2543      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
2544        {        {
2545          markptr = NULL;
2546    
2547        if (timeitm > 0)        if (timeitm > 0)
2548          {          {
2549          register int i;          register int i;
# Line 2145  while (!done) Line 2555  while (!done)
2555            {            {
2556            int workspace[1000];            int workspace[1000];
2557            for (i = 0; i < timeitm; i++)            for (i = 0; i < timeitm; i++)
2558              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,              count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2559                options | g_notempty, use_offsets, use_size_offsets, workspace,                options | g_notempty, use_offsets, use_size_offsets, workspace,
2560                sizeof(workspace)/sizeof(int));                sizeof(workspace)/sizeof(int));
2561            }            }
# Line 2208  while (!done) Line 2618  while (!done)
2618        else if (all_use_dfa || use_dfa)        else if (all_use_dfa || use_dfa)
2619          {          {
2620          int workspace[1000];          int workspace[1000];
2621          count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,          count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2622            options | g_notempty, use_offsets, use_size_offsets, workspace,            options | g_notempty, use_offsets, use_size_offsets, workspace,
2623            sizeof(workspace)/sizeof(int));            sizeof(workspace)/sizeof(int));
2624          if (count == 0)          if (count == 0)
# Line 2279  while (!done) Line 2689  while (!done)
2689              }              }
2690            }            }
2691    
2692            if (markptr != NULL) fprintf(outfile, "MK: %s\n", markptr);
2693    
2694          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2695            {            {
2696            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
# Line 2362  while (!done) Line 2774  while (!done)
2774    
2775        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2776          {          {
2777          fprintf(outfile, "Partial match");          if (markptr == NULL) fprintf(outfile, "Partial match");
2778  #if !defined NODFA            else fprintf(outfile, "Partial match, mark=%s", markptr);
2779          if ((all_use_dfa || use_dfa) && use_size_offsets > 2)          if (use_size_offsets > 1)
2780            fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],            {
2781              bptr + use_offsets[0]);            fprintf(outfile, ": ");
2782  #endif            pchars(bptr + use_offsets[0], use_offsets[1] - use_offsets[0],
2783                outfile);
2784              }
2785          fprintf(outfile, "\n");          fprintf(outfile, "\n");
2786          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2787          }          }
# Line 2377  while (!done) Line 2791  while (!done)
2791        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
2792        string - that was checked before setting g_notempty.        string - that was checked before setting g_notempty.
2793    
2794        Complication arises in the case when the newline option is "any" or        Complication arises in the case when the newline convention is "any",
2795        "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
2796        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,
2797        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
2798        Fudge the offset value to achieve this.        pcre_exec(). Fudge the offset value to achieve this. We check for a
2799          newline setting in the pattern; if none was set, use pcre_config() to
2800          find the default.
2801    
2802        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
2803        character, not one byte. */        character, not one byte. */
# Line 2406  while (!done) Line 2822  while (!done)
2822                      (d == -1)? PCRE_NEWLINE_ANY : 0;                      (d == -1)? PCRE_NEWLINE_ANY : 0;
2823              }              }
2824            if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||            if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2825                   (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_CRLF ||
2826                 (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)                 (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2827                &&                &&
2828                start_offset < len - 1 &&                start_offset < len - 1 &&
# Line 2416  while (!done) Line 2833  while (!done)
2833              {              {
2834              while (start_offset + onechar < len)              while (start_offset + onechar < len)
2835                {                {
2836                int tb = bptr[start_offset+onechar];                if ((bptr[start_offset+onechar] & 0xc0) != 0x80) break;
2837                if (tb <= 127) break;                onechar++;
               tb &= 0xc0;  
               if (tb != 0 && tb != 0xc0) onechar++;  
2838                }                }
2839              }              }
2840            use_offsets[1] = start_offset + onechar;            use_offsets[1] = start_offset + onechar;
# Line 2428  while (!done) Line 2843  while (!done)
2843            {            {
2844            if (count == PCRE_ERROR_NOMATCH)            if (count == PCRE_ERROR_NOMATCH)
2845              {              {
2846              if (gmatched == 0) fprintf(outfile, "No match\n");              if (gmatched == 0)
2847                  {
2848                  if (markptr == NULL) fprintf(outfile, "No match\n");
2849                    else fprintf(outfile, "No match, mark = %s\n", markptr);
2850                  }
2851              }              }
2852            else fprintf(outfile, "Error %d\n", count);            else fprintf(outfile, "Error %d\n", count);
2853            break;  /* Out of the /g loop */            break;  /* Out of the /g loop */
# Line 2440  while (!done) Line 2859  while (!done)
2859        if (!do_g && !do_G) break;        if (!do_g && !do_G) break;
2860    
2861        /* 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
2862        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
2863        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
2864        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
2865        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
2866        character. */        character. */
2867    
# Line 2451  while (!done) Line 2870  while (!done)
2870        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
2871          {          {
2872          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
2873          g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED;          g_notempty = PCRE_NOTEMPTY_ATSTART | PCRE_ANCHORED;
2874          }          }
2875    
2876        /* For /g, update the start offset, leaving the rest alone */        /* For /g, update the start offset, leaving the rest alone */
# Line 2478  while (!done) Line 2897  while (!done)
2897    
2898    if (re != NULL) new_free(re);    if (re != NULL) new_free(re);
2899    if (extra != NULL) new_free(extra);    if (extra != NULL) new_free(extra);
2900    if (tables != NULL)    if (locale_set)
2901      {      {
2902      new_free((void *)tables);      new_free((void *)tables);
2903      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");

Legend:
Removed from v.416  
changed lines
  Added in v.566

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12