/[pcre]/code/trunk/pcretest.c
ViewVC logotype

Diff of /code/trunk/pcretest.c

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

revision 289 by ph10, Sun Dec 23 12:17:20 2007 UTC revision 773 by ph10, Wed Nov 30 18:10:27 2011 UTC
# Line 49  POSSIBILITY OF SUCH DAMAGE. Line 49  POSSIBILITY OF SUCH DAMAGE.
49  #include <errno.h>  #include <errno.h>
50    
51  #ifdef SUPPORT_LIBREADLINE  #ifdef SUPPORT_LIBREADLINE
52    #ifdef HAVE_UNISTD_H
53  #include <unistd.h>  #include <unistd.h>
54    #endif
55  #include <readline/readline.h>  #include <readline/readline.h>
56  #include <readline/history.h>  #include <readline/history.h>
57  #endif  #endif
# Line 69  input mode under Windows. */ Line 71  input mode under Windows. */
71  #define INPUT_MODE   "r"  #define INPUT_MODE   "r"
72  #define OUTPUT_MODE  "wb"  #define OUTPUT_MODE  "wb"
73    
74    #ifndef isatty
75    #define isatty _isatty         /* This is what Windows calls them, I'm told, */
76    #endif                         /* though in some environments they seem to   */
77                                   /* be already defined, hence the #ifndefs.    */
78    #ifndef fileno
79    #define fileno _fileno
80    #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 89  appropriately for an application, not fo Line 107  appropriately for an application, not fo
107  #include "pcre.h"  #include "pcre.h"
108  #include "pcre_internal.h"  #include "pcre_internal.h"
109    
110  /* We need access to the data tables that PCRE uses. So as not to have to keep  /* We need access to some of the data tables that PCRE uses. So as not to have
111  two copies, we include the source file here, changing the names of the external  to keep two copies, we include the source file here, changing the names of the
112  symbols to prevent clashes. */  external symbols to prevent clashes. */
113    
114    #define _pcre_ucp_gentype      ucp_gentype
115    #define _pcre_ucp_typerange    ucp_typerange
116  #define _pcre_utf8_table1      utf8_table1  #define _pcre_utf8_table1      utf8_table1
117  #define _pcre_utf8_table1_size utf8_table1_size  #define _pcre_utf8_table1_size utf8_table1_size
118  #define _pcre_utf8_table2      utf8_table2  #define _pcre_utf8_table2      utf8_table2
119  #define _pcre_utf8_table3      utf8_table3  #define _pcre_utf8_table3      utf8_table3
120  #define _pcre_utf8_table4      utf8_table4  #define _pcre_utf8_table4      utf8_table4
121    #define _pcre_utf8_char_sizes  utf8_char_sizes
122  #define _pcre_utt              utt  #define _pcre_utt              utt
123  #define _pcre_utt_size         utt_size  #define _pcre_utt_size         utt_size
124  #define _pcre_utt_names        utt_names  #define _pcre_utt_names        utt_names
# Line 107  symbols to prevent clashes. */ Line 128  symbols to prevent clashes. */
128    
129  /* We also need the pcre_printint() function for printing out compiled  /* We also need the pcre_printint() function for printing out compiled
130  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
131  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
132    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. */  
133    
134    #define COMPILING_PCRETEST
135  #include "pcre_printint.src"  #include "pcre_printint.src"
136    
137  #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))  /* The definition of the macro PRINTABLE, which determines whether to print an
138    output character as-is or as a hex value when showing compiled patterns, is
139    contained in the printint.src file. We uses it here also, in cases when the
140    locale has not been explicitly changed, so as to get consistent output from
141    systems that differ in their output from isprint() even in the "C" locale. */
142    
143    #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
144    
145  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
146  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 169  static int locale_set = 0; Line 191  static int locale_set = 0;
191  static int show_malloc;  static int show_malloc;
192  static int use_utf8;  static int use_utf8;
193  static size_t gotten_store;  static size_t gotten_store;
194    static size_t first_gotten_store = 0;
195    static const unsigned char *last_callout_mark = NULL;
196    
197  /* The buffers grow automatically if very long input lines are encountered. */  /* The buffers grow automatically if very long input lines are encountered. */
198    
# Line 177  static uschar *buffer = NULL; Line 201  static uschar *buffer = NULL;
201  static uschar *dbuffer = NULL;  static uschar *dbuffer = NULL;
202  static uschar *pbuffer = NULL;  static uschar *pbuffer = NULL;
203    
204    /* Textual explanations for runtime error codes */
205    
206    static const char *errtexts[] = {
207      NULL,  /* 0 is no error */
208      NULL,  /* NOMATCH is handled specially */
209      "NULL argument passed",
210      "bad option value",
211      "magic number missing",
212      "unknown opcode - pattern overwritten?",
213      "no more memory",
214      NULL,  /* never returned by pcre_exec() or pcre_dfa_exec() */
215      "match limit exceeded",
216      "callout error code",
217      NULL,  /* BADUTF8 is handled specially */
218      "bad UTF-8 offset",
219      NULL,  /* PARTIAL is handled specially */
220      "not used - internal error",
221      "internal error - pattern overwritten?",
222      "bad count value",
223      "item unsupported for DFA matching",
224      "backreference condition or recursion test not supported for DFA matching",
225      "match limit not supported for DFA matching",
226      "workspace size exceeded in DFA matching",
227      "too much recursion for DFA matching",
228      "recursion limit exceeded",
229      "not used - internal error",
230      "invalid combination of newline options",
231      "bad offset value",
232      NULL,  /* SHORTUTF8 is handled specially */
233      "nested recursion at the same subject position",
234      "JIT stack limit reached"
235    };
236    
237    
238    /*************************************************
239    *         Alternate character tables             *
240    *************************************************/
241    
242    /* By default, the "tables" pointer when calling PCRE is set to NULL, thereby
243    using the default tables of the library. However, the T option can be used to
244    select alternate sets of tables, for different kinds of testing. Note also that
245    the L (locale) option also adjusts the tables. */
246    
247    /* This is the set of tables distributed as default with PCRE. It recognizes
248    only ASCII characters. */
249    
250    static const unsigned char tables0[] = {
251    
252    /* This table is a lower casing table. */
253    
254        0,  1,  2,  3,  4,  5,  6,  7,
255        8,  9, 10, 11, 12, 13, 14, 15,
256       16, 17, 18, 19, 20, 21, 22, 23,
257       24, 25, 26, 27, 28, 29, 30, 31,
258       32, 33, 34, 35, 36, 37, 38, 39,
259       40, 41, 42, 43, 44, 45, 46, 47,
260       48, 49, 50, 51, 52, 53, 54, 55,
261       56, 57, 58, 59, 60, 61, 62, 63,
262       64, 97, 98, 99,100,101,102,103,
263      104,105,106,107,108,109,110,111,
264      112,113,114,115,116,117,118,119,
265      120,121,122, 91, 92, 93, 94, 95,
266       96, 97, 98, 99,100,101,102,103,
267      104,105,106,107,108,109,110,111,
268      112,113,114,115,116,117,118,119,
269      120,121,122,123,124,125,126,127,
270      128,129,130,131,132,133,134,135,
271      136,137,138,139,140,141,142,143,
272      144,145,146,147,148,149,150,151,
273      152,153,154,155,156,157,158,159,
274      160,161,162,163,164,165,166,167,
275      168,169,170,171,172,173,174,175,
276      176,177,178,179,180,181,182,183,
277      184,185,186,187,188,189,190,191,
278      192,193,194,195,196,197,198,199,
279      200,201,202,203,204,205,206,207,
280      208,209,210,211,212,213,214,215,
281      216,217,218,219,220,221,222,223,
282      224,225,226,227,228,229,230,231,
283      232,233,234,235,236,237,238,239,
284      240,241,242,243,244,245,246,247,
285      248,249,250,251,252,253,254,255,
286    
287    /* This table is a case flipping table. */
288    
289        0,  1,  2,  3,  4,  5,  6,  7,
290        8,  9, 10, 11, 12, 13, 14, 15,
291       16, 17, 18, 19, 20, 21, 22, 23,
292       24, 25, 26, 27, 28, 29, 30, 31,
293       32, 33, 34, 35, 36, 37, 38, 39,
294       40, 41, 42, 43, 44, 45, 46, 47,
295       48, 49, 50, 51, 52, 53, 54, 55,
296       56, 57, 58, 59, 60, 61, 62, 63,
297       64, 97, 98, 99,100,101,102,103,
298      104,105,106,107,108,109,110,111,
299      112,113,114,115,116,117,118,119,
300      120,121,122, 91, 92, 93, 94, 95,
301       96, 65, 66, 67, 68, 69, 70, 71,
302       72, 73, 74, 75, 76, 77, 78, 79,
303       80, 81, 82, 83, 84, 85, 86, 87,
304       88, 89, 90,123,124,125,126,127,
305      128,129,130,131,132,133,134,135,
306      136,137,138,139,140,141,142,143,
307      144,145,146,147,148,149,150,151,
308      152,153,154,155,156,157,158,159,
309      160,161,162,163,164,165,166,167,
310      168,169,170,171,172,173,174,175,
311      176,177,178,179,180,181,182,183,
312      184,185,186,187,188,189,190,191,
313      192,193,194,195,196,197,198,199,
314      200,201,202,203,204,205,206,207,
315      208,209,210,211,212,213,214,215,
316      216,217,218,219,220,221,222,223,
317      224,225,226,227,228,229,230,231,
318      232,233,234,235,236,237,238,239,
319      240,241,242,243,244,245,246,247,
320      248,249,250,251,252,253,254,255,
321    
322    /* This table contains bit maps for various character classes. Each map is 32
323    bytes long and the bits run from the least significant end of each byte. The
324    classes that have their own maps are: space, xdigit, digit, upper, lower, word,
325    graph, print, punct, and cntrl. Other classes are built from combinations. */
326    
327      0x00,0x3e,0x00,0x00,0x01,0x00,0x00,0x00,
328      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
329      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
330      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
331    
332      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
333      0x7e,0x00,0x00,0x00,0x7e,0x00,0x00,0x00,
334      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
335      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
336    
337      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
338      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
339      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
340      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
341    
342      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
343      0xfe,0xff,0xff,0x07,0x00,0x00,0x00,0x00,
344      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
345      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
346    
347      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
348      0x00,0x00,0x00,0x00,0xfe,0xff,0xff,0x07,
349      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
350      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
351    
352      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
353      0xfe,0xff,0xff,0x87,0xfe,0xff,0xff,0x07,
354      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
355      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
356    
357      0x00,0x00,0x00,0x00,0xfe,0xff,0xff,0xff,
358      0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x7f,
359      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
360      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
361    
362      0x00,0x00,0x00,0x00,0xff,0xff,0xff,0xff,
363      0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x7f,
364      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
365      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
366    
367      0x00,0x00,0x00,0x00,0xfe,0xff,0x00,0xfc,
368      0x01,0x00,0x00,0xf8,0x01,0x00,0x00,0x78,
369      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
370      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
371    
372      0xff,0xff,0xff,0xff,0x00,0x00,0x00,0x00,
373      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x80,
374      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
375      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
376    
377    /* This table identifies various classes of character by individual bits:
378      0x01   white space character
379      0x02   letter
380      0x04   decimal digit
381      0x08   hexadecimal digit
382      0x10   alphanumeric or '_'
383      0x80   regular expression metacharacter or binary zero
384    */
385    
386      0x80,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   0-  7 */
387      0x00,0x01,0x01,0x00,0x01,0x01,0x00,0x00, /*   8- 15 */
388      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  16- 23 */
389      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  24- 31 */
390      0x01,0x00,0x00,0x00,0x80,0x00,0x00,0x00, /*    - '  */
391      0x80,0x80,0x80,0x80,0x00,0x00,0x80,0x00, /*  ( - /  */
392      0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c, /*  0 - 7  */
393      0x1c,0x1c,0x00,0x00,0x00,0x00,0x00,0x80, /*  8 - ?  */
394      0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  @ - G  */
395      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  H - O  */
396      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  P - W  */
397      0x12,0x12,0x12,0x80,0x80,0x00,0x80,0x10, /*  X - _  */
398      0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  ` - g  */
399      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  h - o  */
400      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  p - w  */
401      0x12,0x12,0x12,0x80,0x80,0x00,0x00,0x00, /*  x -127 */
402      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 128-135 */
403      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 136-143 */
404      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144-151 */
405      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 152-159 */
406      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160-167 */
407      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 168-175 */
408      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 176-183 */
409      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
410      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 192-199 */
411      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 200-207 */
412      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 208-215 */
413      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 216-223 */
414      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 224-231 */
415      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 232-239 */
416      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 240-247 */
417      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00};/* 248-255 */
418    
419    /* This is a set of tables that came orginally from a Windows user. It seems to
420    be at least an approximation of ISO 8859. In particular, there are characters
421    greater than 128 that are marked as spaces, letters, etc. */
422    
423    static const unsigned char tables1[] = {
424    0,1,2,3,4,5,6,7,
425    8,9,10,11,12,13,14,15,
426    16,17,18,19,20,21,22,23,
427    24,25,26,27,28,29,30,31,
428    32,33,34,35,36,37,38,39,
429    40,41,42,43,44,45,46,47,
430    48,49,50,51,52,53,54,55,
431    56,57,58,59,60,61,62,63,
432    64,97,98,99,100,101,102,103,
433    104,105,106,107,108,109,110,111,
434    112,113,114,115,116,117,118,119,
435    120,121,122,91,92,93,94,95,
436    96,97,98,99,100,101,102,103,
437    104,105,106,107,108,109,110,111,
438    112,113,114,115,116,117,118,119,
439    120,121,122,123,124,125,126,127,
440    128,129,130,131,132,133,134,135,
441    136,137,138,139,140,141,142,143,
442    144,145,146,147,148,149,150,151,
443    152,153,154,155,156,157,158,159,
444    160,161,162,163,164,165,166,167,
445    168,169,170,171,172,173,174,175,
446    176,177,178,179,180,181,182,183,
447    184,185,186,187,188,189,190,191,
448    224,225,226,227,228,229,230,231,
449    232,233,234,235,236,237,238,239,
450    240,241,242,243,244,245,246,215,
451    248,249,250,251,252,253,254,223,
452    224,225,226,227,228,229,230,231,
453    232,233,234,235,236,237,238,239,
454    240,241,242,243,244,245,246,247,
455    248,249,250,251,252,253,254,255,
456    0,1,2,3,4,5,6,7,
457    8,9,10,11,12,13,14,15,
458    16,17,18,19,20,21,22,23,
459    24,25,26,27,28,29,30,31,
460    32,33,34,35,36,37,38,39,
461    40,41,42,43,44,45,46,47,
462    48,49,50,51,52,53,54,55,
463    56,57,58,59,60,61,62,63,
464    64,97,98,99,100,101,102,103,
465    104,105,106,107,108,109,110,111,
466    112,113,114,115,116,117,118,119,
467    120,121,122,91,92,93,94,95,
468    96,65,66,67,68,69,70,71,
469    72,73,74,75,76,77,78,79,
470    80,81,82,83,84,85,86,87,
471    88,89,90,123,124,125,126,127,
472    128,129,130,131,132,133,134,135,
473    136,137,138,139,140,141,142,143,
474    144,145,146,147,148,149,150,151,
475    152,153,154,155,156,157,158,159,
476    160,161,162,163,164,165,166,167,
477    168,169,170,171,172,173,174,175,
478    176,177,178,179,180,181,182,183,
479    184,185,186,187,188,189,190,191,
480    224,225,226,227,228,229,230,231,
481    232,233,234,235,236,237,238,239,
482    240,241,242,243,244,245,246,215,
483    248,249,250,251,252,253,254,223,
484    192,193,194,195,196,197,198,199,
485    200,201,202,203,204,205,206,207,
486    208,209,210,211,212,213,214,247,
487    216,217,218,219,220,221,222,255,
488    0,62,0,0,1,0,0,0,
489    0,0,0,0,0,0,0,0,
490    32,0,0,0,1,0,0,0,
491    0,0,0,0,0,0,0,0,
492    0,0,0,0,0,0,255,3,
493    126,0,0,0,126,0,0,0,
494    0,0,0,0,0,0,0,0,
495    0,0,0,0,0,0,0,0,
496    0,0,0,0,0,0,255,3,
497    0,0,0,0,0,0,0,0,
498    0,0,0,0,0,0,12,2,
499    0,0,0,0,0,0,0,0,
500    0,0,0,0,0,0,0,0,
501    254,255,255,7,0,0,0,0,
502    0,0,0,0,0,0,0,0,
503    255,255,127,127,0,0,0,0,
504    0,0,0,0,0,0,0,0,
505    0,0,0,0,254,255,255,7,
506    0,0,0,0,0,4,32,4,
507    0,0,0,128,255,255,127,255,
508    0,0,0,0,0,0,255,3,
509    254,255,255,135,254,255,255,7,
510    0,0,0,0,0,4,44,6,
511    255,255,127,255,255,255,127,255,
512    0,0,0,0,254,255,255,255,
513    255,255,255,255,255,255,255,127,
514    0,0,0,0,254,255,255,255,
515    255,255,255,255,255,255,255,255,
516    0,2,0,0,255,255,255,255,
517    255,255,255,255,255,255,255,127,
518    0,0,0,0,255,255,255,255,
519    255,255,255,255,255,255,255,255,
520    0,0,0,0,254,255,0,252,
521    1,0,0,248,1,0,0,120,
522    0,0,0,0,254,255,255,255,
523    0,0,128,0,0,0,128,0,
524    255,255,255,255,0,0,0,0,
525    0,0,0,0,0,0,0,128,
526    255,255,255,255,0,0,0,0,
527    0,0,0,0,0,0,0,0,
528    128,0,0,0,0,0,0,0,
529    0,1,1,0,1,1,0,0,
530    0,0,0,0,0,0,0,0,
531    0,0,0,0,0,0,0,0,
532    1,0,0,0,128,0,0,0,
533    128,128,128,128,0,0,128,0,
534    28,28,28,28,28,28,28,28,
535    28,28,0,0,0,0,0,128,
536    0,26,26,26,26,26,26,18,
537    18,18,18,18,18,18,18,18,
538    18,18,18,18,18,18,18,18,
539    18,18,18,128,128,0,128,16,
540    0,26,26,26,26,26,26,18,
541    18,18,18,18,18,18,18,18,
542    18,18,18,18,18,18,18,18,
543    18,18,18,128,128,0,0,0,
544    0,0,0,0,0,1,0,0,
545    0,0,0,0,0,0,0,0,
546    0,0,0,0,0,0,0,0,
547    0,0,0,0,0,0,0,0,
548    1,0,0,0,0,0,0,0,
549    0,0,18,0,0,0,0,0,
550    0,0,20,20,0,18,0,0,
551    0,20,18,0,0,0,0,0,
552    18,18,18,18,18,18,18,18,
553    18,18,18,18,18,18,18,18,
554    18,18,18,18,18,18,18,0,
555    18,18,18,18,18,18,18,18,
556    18,18,18,18,18,18,18,18,
557    18,18,18,18,18,18,18,18,
558    18,18,18,18,18,18,18,0,
559    18,18,18,18,18,18,18,18
560    };
561    
562    
563    
564    
565    #ifndef HAVE_STRERROR
566    /*************************************************
567    *     Provide strerror() for non-ANSI libraries  *
568    *************************************************/
569    
570    /* Some old-fashioned systems still around (e.g. SunOS4) don't have strerror()
571    in their libraries, but can provide the same facility by this simple
572    alternative function. */
573    
574    extern int   sys_nerr;
575    extern char *sys_errlist[];
576    
577    char *
578    strerror(int n)
579    {
580    if (n < 0 || n >= sys_nerr) return "unknown error number";
581    return sys_errlist[n];
582    }
583    #endif /* HAVE_STRERROR */
584    
585    
586    /*************************************************
587    *         JIT memory callback                    *
588    *************************************************/
589    
590    static pcre_jit_stack* jit_callback(void *arg)
591    {
592    return (pcre_jit_stack *)arg;
593    }
594    
595    
596  /*************************************************  /*************************************************
# Line 209  uschar *here = start; Line 623  uschar *here = start;
623    
624  for (;;)  for (;;)
625    {    {
626    int rlen = buffer_size - (here - buffer);    int rlen = (int)(buffer_size - (here - buffer));
627    
628    if (rlen > 1000)    if (rlen > 1000)
629      {      {
# Line 239  for (;;) Line 653  for (;;)
653      /* 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. */
654    
655        {        {
656        if (f == stdin) printf(prompt);        if (f == stdin) printf("%s", prompt);
657        if (fgets((char *)here, rlen,  f) == NULL)        if (fgets((char *)here, rlen,  f) == NULL)
658          return (here == start)? NULL : start;          return (here == start)? NULL : start;
659        }        }
# Line 559  fprintf(outfile, "%.*s", (cb->next_item_ Line 973  fprintf(outfile, "%.*s", (cb->next_item_
973  fprintf(outfile, "\n");  fprintf(outfile, "\n");
974  first_callout = 0;  first_callout = 0;
975    
976    if (cb->mark != last_callout_mark)
977      {
978      fprintf(outfile, "Latest Mark: %s\n",
979        (cb->mark == NULL)? "<unset>" : (char *)(cb->mark));
980      last_callout_mark = cb->mark;
981      }
982    
983  if (cb->callout_data != NULL)  if (cb->callout_data != NULL)
984    {    {
985    int callout_data = *((int *)(cb->callout_data));    int callout_data = *((int *)(cb->callout_data));
# Line 578  return (cb->callout_number != callout_fa Line 999  return (cb->callout_number != callout_fa
999  *            Local malloc functions              *  *            Local malloc functions              *
1000  *************************************************/  *************************************************/
1001    
1002  /* Alternative malloc function, to test functionality and show the size of the  /* Alternative malloc function, to test functionality and save the size of a
1003  compiled re. */  compiled re, which is the first store request that pcre_compile() makes. The
1004    show_malloc variable is set only during matching. */
1005    
1006  static void *new_malloc(size_t size)  static void *new_malloc(size_t size)
1007  {  {
1008  void *block = malloc(size);  void *block = malloc(size);
1009  gotten_store = size;  gotten_store = size;
1010    if (first_gotten_store == 0) first_gotten_store = size;
1011  if (show_malloc)  if (show_malloc)
1012    fprintf(outfile, "malloc       %3d %p\n", (int)size, block);    fprintf(outfile, "malloc       %3d %p\n", (int)size, block);
1013  return block;  return block;
# Line 597  if (show_malloc) Line 1020  if (show_malloc)
1020  free(block);  free(block);
1021  }  }
1022    
   
1023  /* For recursion malloc/free, to test stacking calls */  /* For recursion malloc/free, to test stacking calls */
1024    
1025  static void *stack_malloc(size_t size)  static void *stack_malloc(size_t size)
# Line 729  return 0; Line 1151  return 0;
1151  *         Check newline indicator                *  *         Check newline indicator                *
1152  *************************************************/  *************************************************/
1153    
1154  /* 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
1155  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.  
1156    
1157  Arguments:  Arguments:
1158    p           points after the leading '<'    p           points after the leading '<'
# Line 779  printf("  -dfa     force DFA matching fo Line 1200  printf("  -dfa     force DFA matching fo
1200  #endif  #endif
1201  printf("  -help    show usage information\n");  printf("  -help    show usage information\n");
1202  printf("  -i       show information about compiled patterns\n"  printf("  -i       show information about compiled patterns\n"
1203           "  -M       find MATCH_LIMIT minimum for each subject\n"
1204         "  -m       output memory used information\n"         "  -m       output memory used information\n"
1205         "  -o <n>   set size of offsets vector to <n>\n");         "  -o <n>   set size of offsets vector to <n>\n");
1206  #if !defined NOPOSIX  #if !defined NOPOSIX
# Line 786  printf("  -p       use POSIX interface\n Line 1208  printf("  -p       use POSIX interface\n
1208  #endif  #endif
1209  printf("  -q       quiet: do not output PCRE version number at start\n");  printf("  -q       quiet: do not output PCRE version number at start\n");
1210  printf("  -S <n>   set stack size to <n> megabytes\n");  printf("  -S <n>   set stack size to <n> megabytes\n");
1211  printf("  -s       output store (memory) used information\n"  printf("  -s       force each pattern to be studied at basic level\n"
1212           "  -s+      force each pattern to be studied, using JIT if available\n"
1213         "  -t       time compilation and execution\n");         "  -t       time compilation and execution\n");
1214  printf("  -t <n>   time compilation and execution, repeating <n> times\n");  printf("  -t <n>   time compilation and execution, repeating <n> times\n");
1215  printf("  -tm      time execution (matching) only\n");  printf("  -tm      time execution (matching) only\n");
# Line 808  int main(int argc, char **argv) Line 1231  int main(int argc, char **argv)
1231  FILE *infile = stdin;  FILE *infile = stdin;
1232  int options = 0;  int options = 0;
1233  int study_options = 0;  int study_options = 0;
1234    int default_find_match_limit = FALSE;
1235  int op = 1;  int op = 1;
1236  int timeit = 0;  int timeit = 0;
1237  int timeitm = 0;  int timeitm = 0;
1238  int showinfo = 0;  int showinfo = 0;
1239  int showstore = 0;  int showstore = 0;
1240    int force_study = -1;
1241    int force_study_options = 0;
1242  int quiet = 0;  int quiet = 0;
1243  int size_offsets = 45;  int size_offsets = 45;
1244  int size_offsets_max;  int size_offsets_max;
# Line 826  int all_use_dfa = 0; Line 1252  int all_use_dfa = 0;
1252  int yield = 0;  int yield = 0;
1253  int stack_size;  int stack_size;
1254    
1255    pcre_jit_stack *jit_stack = NULL;
1256    
1257    
1258  /* These vectors store, end-to-end, a list of captured substring names. Assume  /* These vectors store, end-to-end, a list of captured substring names. Assume
1259  that 1024 is plenty long enough for the few names we'll be testing. */  that 1024 is plenty long enough for the few names we'll be testing. */
1260    
# Line 861  while (argc > 1 && argv[op][0] == '-') Line 1290  while (argc > 1 && argv[op][0] == '-')
1290    {    {
1291    unsigned char *endptr;    unsigned char *endptr;
1292    
1293    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-m") == 0) showstore = 1;
1294      showstore = 1;    else if (strcmp(argv[op], "-s") == 0) force_study = 0;
1295      else if (strcmp(argv[op], "-s+") == 0)
1296        {
1297        force_study = 1;
1298        force_study_options = PCRE_STUDY_JIT_COMPILE;
1299        }
1300    else if (strcmp(argv[op], "-q") == 0) quiet = 1;    else if (strcmp(argv[op], "-q") == 0) quiet = 1;
1301    else if (strcmp(argv[op], "-b") == 0) debug = 1;    else if (strcmp(argv[op], "-b") == 0) debug = 1;
1302    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
1303    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
1304      else if (strcmp(argv[op], "-M") == 0) default_find_match_limit = TRUE;
1305  #if !defined NODFA  #if !defined NODFA
1306    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
1307  #endif  #endif
# Line 895  while (argc > 1 && argv[op][0] == '-') Line 1330  while (argc > 1 && argv[op][0] == '-')
1330        ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),        ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
1331          *endptr == 0))          *endptr == 0))
1332      {      {
1333  #if defined(_WIN32) || defined(WIN32)  #if defined(_WIN32) || defined(WIN32) || defined(__minix)
1334      printf("PCRE: -S not supported on this OS\n");      printf("PCRE: -S not supported on this OS\n");
1335      exit(1);      exit(1);
1336  #else  #else
# Line 919  while (argc > 1 && argv[op][0] == '-') Line 1354  while (argc > 1 && argv[op][0] == '-')
1354    else if (strcmp(argv[op], "-C") == 0)    else if (strcmp(argv[op], "-C") == 0)
1355      {      {
1356      int rc;      int rc;
1357        unsigned long int lrc;
1358      printf("PCRE version %s\n", pcre_version());      printf("PCRE version %s\n", pcre_version());
1359      printf("Compiled with\n");      printf("Compiled with\n");
1360      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
1361      printf("  %sUTF-8 support\n", rc? "" : "No ");      printf("  %sUTF-8 support\n", rc? "" : "No ");
1362      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
1363      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
1364        (void)pcre_config(PCRE_CONFIG_JIT, &rc);
1365        if (rc)
1366          printf("  Just-in-time compiler support\n");
1367        else
1368          printf("  No just-in-time compiler support\n");
1369      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
1370      printf("  Newline sequence is %s\n", (rc == '\r')? "CR" :      /* Note that these values are always the ASCII values, even
1371        (rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" :      in EBCDIC environments. CR is 13 and NL is 10. */
1372        printf("  Newline sequence is %s\n", (rc == 13)? "CR" :
1373          (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
1374        (rc == -2)? "ANYCRLF" :        (rc == -2)? "ANYCRLF" :
1375        (rc == -1)? "ANY" : "???");        (rc == -1)? "ANY" : "???");
1376      (void)pcre_config(PCRE_CONFIG_BSR, &rc);      (void)pcre_config(PCRE_CONFIG_BSR, &rc);
# Line 937  while (argc > 1 && argv[op][0] == '-') Line 1380  while (argc > 1 && argv[op][0] == '-')
1380      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
1381      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
1382      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
1383      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &lrc);
1384      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %ld\n", lrc);
1385      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &rc);      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &lrc);
1386      printf("  Default recursion depth limit = %d\n", rc);      printf("  Default recursion depth limit = %ld\n", lrc);
1387      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
1388      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
1389      goto EXIT;      goto EXIT;
# Line 1022  while (!done) Line 1465  while (!done)
1465  #endif  #endif
1466    
1467    const char *error;    const char *error;
1468      unsigned char *markptr;
1469    unsigned char *p, *pp, *ppp;    unsigned char *p, *pp, *ppp;
1470    unsigned char *to_file = NULL;    unsigned char *to_file = NULL;
1471    const unsigned char *tables = NULL;    const unsigned char *tables = NULL;
1472    unsigned long int true_size, true_study_size = 0;    unsigned long int true_size, true_study_size = 0;
1473    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
1474      int do_allcaps = 0;
1475      int do_mark = 0;
1476    int do_study = 0;    int do_study = 0;
1477      int no_force_study = 0;
1478    int do_debug = debug;    int do_debug = debug;
1479    int do_G = 0;    int do_G = 0;
1480    int do_g = 0;    int do_g = 0;
1481    int do_showinfo = showinfo;    int do_showinfo = showinfo;
1482    int do_showrest = 0;    int do_showrest = 0;
1483      int do_showcaprest = 0;
1484    int do_flip = 0;    int do_flip = 0;
1485    int erroroffset, len, delimiter, poffset;    int erroroffset, len, delimiter, poffset;
1486    
# Line 1075  while (!done) Line 1523  while (!done)
1523        (sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7];        (sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7];
1524    
1525      re = (real_pcre *)new_malloc(true_size);      re = (real_pcre *)new_malloc(true_size);
1526      regex_gotten_store = gotten_store;      regex_gotten_store = first_gotten_store;
1527    
1528      if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ;      if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ;
1529    
# Line 1094  while (!done) Line 1542  while (!done)
1542          }          }
1543        }        }
1544    
1545      fprintf(outfile, "Compiled regex%s loaded from %s\n",      fprintf(outfile, "Compiled pattern%s loaded from %s\n",
1546        do_flip? " (byte-inverted)" : "", p);        do_flip? " (byte-inverted)" : "", p);
1547    
1548      /* Need to know if UTF-8 for printing data strings */      /* Need to know if UTF-8 for printing data strings */
# Line 1102  while (!done) Line 1550  while (!done)
1550      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);      new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1551      use_utf8 = (get_options & PCRE_UTF8) != 0;      use_utf8 = (get_options & PCRE_UTF8) != 0;
1552    
1553      /* Now see if there is any following study data */      /* Now see if there is any following study data. */
1554    
1555      if (true_study_size != 0)      if (true_study_size != 0)
1556        {        {
# Line 1118  while (!done) Line 1566  while (!done)
1566          {          {
1567          FAIL_READ:          FAIL_READ:
1568          fprintf(outfile, "Failed to read data from %s\n", p);          fprintf(outfile, "Failed to read data from %s\n", p);
1569          if (extra != NULL) new_free(extra);          if (extra != NULL) pcre_free_study(extra);
1570          if (re != NULL) new_free(re);          if (re != NULL) new_free(re);
1571          fclose(f);          fclose(f);
1572          continue;          continue;
# Line 1144  while (!done) Line 1592  while (!done)
1592      }      }
1593    
1594    pp = p;    pp = p;
1595    poffset = p - buffer;    poffset = (int)(p - buffer);
1596    
1597    for(;;)    for(;;)
1598      {      {
# Line 1184  while (!done) Line 1632  while (!done)
1632    /* Look for options after final delimiter */    /* Look for options after final delimiter */
1633    
1634    options = 0;    options = 0;
   study_options = 0;  
1635    log_store = showstore;  /* default from command line */    log_store = showstore;  /* default from command line */
1636    
1637    while (*pp != 0)    while (*pp != 0)
# Line 1198  while (!done) Line 1645  while (!done)
1645        case 's': options |= PCRE_DOTALL; break;        case 's': options |= PCRE_DOTALL; break;
1646        case 'x': options |= PCRE_EXTENDED; break;        case 'x': options |= PCRE_EXTENDED; break;
1647    
1648        case '+': do_showrest = 1; break;        case '+':
1649          if (do_showrest) do_showcaprest = 1; else do_showrest = 1;
1650          break;
1651    
1652          case '=': do_allcaps = 1; break;
1653        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
1654        case 'B': do_debug = 1; break;        case 'B': do_debug = 1; break;
1655        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
# Line 1208  while (!done) Line 1659  while (!done)
1659        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
1660        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
1661        case 'J': options |= PCRE_DUPNAMES; break;        case 'J': options |= PCRE_DUPNAMES; break;
1662          case 'K': do_mark = 1; break;
1663        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
1664        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1665    
# Line 1215  while (!done) Line 1667  while (!done)
1667        case 'P': do_posix = 1; break;        case 'P': do_posix = 1; break;
1668  #endif  #endif
1669    
1670        case 'S': do_study = 1; break;        case 'S':
1671          if (do_study == 0)
1672            {
1673            do_study = 1;
1674            if (*pp == '+')
1675              {
1676              study_options |= PCRE_STUDY_JIT_COMPILE;
1677              pp++;
1678              }
1679            }
1680          else
1681            {
1682            do_study = 0;
1683            no_force_study = 1;
1684            }
1685          break;
1686    
1687        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
1688          case 'W': options |= PCRE_UCP; break;
1689        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
1690          case 'Y': options |= PCRE_NO_START_OPTIMISE; break;
1691        case 'Z': debug_lengths = 0; break;        case 'Z': debug_lengths = 0; break;
1692        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1693        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
1694    
1695          case 'T':
1696          switch (*pp++)
1697            {
1698            case '0': tables = tables0; break;
1699            case '1': tables = tables1; break;
1700    
1701            case '\r':
1702            case '\n':
1703            case ' ':
1704            case 0:
1705            fprintf(outfile, "** Missing table number after /T\n");
1706            goto SKIP_DATA;
1707    
1708            default:
1709            fprintf(outfile, "** Bad table number \"%c\" after /T\n", pp[-1]);
1710            goto SKIP_DATA;
1711            }
1712          break;
1713    
1714        case 'L':        case 'L':
1715        ppp = pp;        ppp = pp;
1716        /* The '\r' test here is so that it works on Windows. */        /* The '\r' test here is so that it works on Windows. */
# Line 1247  while (!done) Line 1736  while (!done)
1736    
1737        case '<':        case '<':
1738          {          {
1739          int x = check_newline(pp, outfile);          if (strncmpic(pp, (uschar *)"JS>", 3) == 0)
1740          if (x == 0) goto SKIP_DATA;            {
1741          options |= x;            options |= PCRE_JAVASCRIPT_COMPAT;
1742          while (*pp++ != '>');            pp += 3;
1743              }
1744            else
1745              {
1746              int x = check_newline(pp, outfile);
1747              if (x == 0) goto SKIP_DATA;
1748              options |= x;
1749              while (*pp++ != '>');
1750              }
1751          }          }
1752        break;        break;
1753    
# Line 1280  while (!done) Line 1777  while (!done)
1777      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1778      if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;      if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1779      if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;      if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1780        if ((options & PCRE_UCP) != 0) cflags |= REG_UCP;
1781        if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY;
1782    
1783        first_gotten_store = 0;
1784      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
1785    
1786      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
# Line 1300  while (!done) Line 1800  while (!done)
1800  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
1801    
1802      {      {
1803        unsigned long int get_options;
1804    
1805      if (timeit > 0)      if (timeit > 0)
1806        {        {
1807        register int i;        register int i;
# Line 1316  while (!done) Line 1818  while (!done)
1818            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
1819        }        }
1820    
1821        first_gotten_store = 0;
1822      re = pcre_compile((char *)p, options, &error, &erroroffset, tables);      re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1823    
1824      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
# Line 1343  while (!done) Line 1846  while (!done)
1846        goto CONTINUE;        goto CONTINUE;
1847        }        }
1848    
1849      /* Compilation succeeded; print data if required. There are now two      /* Compilation succeeded. It is now possible to set the UTF-8 option from
1850      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
1851      returns only limited data. Check that it agrees with the newer one. */      lines. */
1852    
1853        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1854        if ((get_options & PCRE_UTF8) != 0) use_utf8 = 1;
1855    
1856        /* Print information if required. There are now two info-returning
1857        functions. The old one has a limited interface and returns only limited
1858        data. Check that it agrees with the newer one. */
1859    
1860      if (log_store)      if (log_store)
1861        fprintf(outfile, "Memory allocation (code space): %d\n",        fprintf(outfile, "Memory allocation (code space): %d\n",
1862          (int)(gotten_store -          (int)(first_gotten_store -
1863                sizeof(real_pcre) -                sizeof(real_pcre) -
1864                ((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size));                ((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size));
1865    
# Line 1357  while (!done) Line 1867  while (!done)
1867      and remember the store that was got. */      and remember the store that was got. */
1868    
1869      true_size = ((real_pcre *)re)->size;      true_size = ((real_pcre *)re)->size;
1870      regex_gotten_store = gotten_store;      regex_gotten_store = first_gotten_store;
1871    
1872      /* If /S was present, study the regexp to generate additional info to      /* If -s or /S was present, study the regex to generate additional info to
1873      help with the matching. */      help with the matching, unless the pattern has the SS option, which
1874        suppresses the effect of /S (used for a few test patterns where studying is
1875        never sensible). */
1876    
1877      if (do_study)      if (do_study || (force_study >= 0 && !no_force_study))
1878        {        {
1879        if (timeit > 0)        if (timeit > 0)
1880          {          {
# Line 1370  while (!done) Line 1882  while (!done)
1882          clock_t time_taken;          clock_t time_taken;
1883          clock_t start_time = clock();          clock_t start_time = clock();
1884          for (i = 0; i < timeit; i++)          for (i = 0; i < timeit; i++)
1885            extra = pcre_study(re, study_options, &error);            extra = pcre_study(re, study_options | force_study_options, &error);
1886          time_taken = clock() - start_time;          time_taken = clock() - start_time;
1887          if (extra != NULL) free(extra);          if (extra != NULL) pcre_free_study(extra);
1888          fprintf(outfile, "  Study time %.4f milliseconds\n",          fprintf(outfile, "  Study time %.4f milliseconds\n",
1889            (((double)time_taken * 1000.0) / (double)timeit) /            (((double)time_taken * 1000.0) / (double)timeit) /
1890              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
1891          }          }
1892        extra = pcre_study(re, study_options, &error);        extra = pcre_study(re, study_options | force_study_options, &error);
1893        if (error != NULL)        if (error != NULL)
1894          fprintf(outfile, "Failed to study: %s\n", error);          fprintf(outfile, "Failed to study: %s\n", error);
1895        else if (extra != NULL)        else if (extra != NULL)
1896          true_study_size = ((pcre_study_data *)(extra->study_data))->size;          true_study_size = ((pcre_study_data *)(extra->study_data))->size;
1897        }        }
1898    
1899        /* If /K was present, we set up for handling MARK data. */
1900    
1901        if (do_mark)
1902          {
1903          if (extra == NULL)
1904            {
1905            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
1906            extra->flags = 0;
1907            }
1908          extra->mark = &markptr;
1909          extra->flags |= PCRE_EXTRA_MARK;
1910          }
1911    
1912      /* 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
1913      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
1914      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 1416  while (!done) Line 1941  while (!done)
1941          {          {
1942          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
1943          rsd->size = byteflip(rsd->size, sizeof(rsd->size));          rsd->size = byteflip(rsd->size, sizeof(rsd->size));
1944          rsd->options = byteflip(rsd->options, sizeof(rsd->options));          rsd->flags = byteflip(rsd->flags, sizeof(rsd->flags));
1945            rsd->minlength = byteflip(rsd->minlength, sizeof(rsd->minlength));
1946          }          }
1947        }        }
1948    
# Line 1430  while (!done) Line 1956  while (!done)
1956        pcre_printint(re, outfile, debug_lengths);        pcre_printint(re, outfile, debug_lengths);
1957        }        }
1958    
1959        /* We already have the options in get_options (see above) */
1960    
1961      if (do_showinfo)      if (do_showinfo)
1962        {        {
1963        unsigned long int get_options, all_options;        unsigned long int all_options;
1964  #if !defined NOINFOCHECK  #if !defined NOINFOCHECK
1965        int old_first_char, old_options, old_count;        int old_first_char, old_options, old_count;
1966  #endif  #endif
# Line 1441  while (!done) Line 1969  while (!done)
1969        int nameentrysize, namecount;        int nameentrysize, namecount;
1970        const uschar *nametable;        const uschar *nametable;
1971    
       new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);  
1972        new_info(re, NULL, PCRE_INFO_SIZE, &size);        new_info(re, NULL, PCRE_INFO_SIZE, &size);
1973        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
1974        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);
# Line 1501  while (!done) Line 2028  while (!done)
2028        if (do_flip) all_options = byteflip(all_options, sizeof(all_options));        if (do_flip) all_options = byteflip(all_options, sizeof(all_options));
2029    
2030        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
2031          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",
2032            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
2033            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
2034            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
# Line 1515  while (!done) Line 2042  while (!done)
2042            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
2043            ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",            ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
2044            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
2045              ((get_options & PCRE_UCP) != 0)? " ucp" : "",
2046            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
2047              ((get_options & PCRE_NO_START_OPTIMIZE) != 0)? " no_start_optimize" : "",
2048            ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");            ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
2049    
2050        if (jchanged) fprintf(outfile, "Duplicate name status changes\n");        if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
# Line 1583  while (!done) Line 2112  while (!done)
2112        /* Don't output study size; at present it is in any case a fixed        /* Don't output study size; at present it is in any case a fixed
2113        value, but it varies, depending on the computer architecture, and        value, but it varies, depending on the computer architecture, and
2114        so messes up the test suite. (And with the /F option, it might be        so messes up the test suite. (And with the /F option, it might be
2115        flipped.) */        flipped.) If study was forced by an external -s, don't show this
2116          information unless -i or -d was also present. This means that, except
2117          when auto-callouts are involved, the output from runs with and without
2118          -s should be identical. */
2119    
2120        if (do_study)        if (do_study || (force_study >= 0 && showinfo && !no_force_study))
2121          {          {
2122          if (extra == NULL)          if (extra == NULL)
2123            fprintf(outfile, "Study returned NULL\n");            fprintf(outfile, "Study returned NULL\n");
2124          else          else
2125            {            {
2126            uschar *start_bits = NULL;            uschar *start_bits = NULL;
2127            new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);            int minlength;
2128    
2129              new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength);
2130              fprintf(outfile, "Subject length lower bound = %d\n", minlength);
2131    
2132              new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);
2133            if (start_bits == NULL)            if (start_bits == NULL)
2134              fprintf(outfile, "No starting byte set\n");              fprintf(outfile, "No set of starting bytes\n");
2135            else            else
2136              {              {
2137              int i;              int i;
# Line 1625  while (!done) Line 2161  while (!done)
2161              fprintf(outfile, "\n");              fprintf(outfile, "\n");
2162              }              }
2163            }            }
2164    
2165            /* Show this only if the JIT was set by /S, not by -s. */
2166    
2167            if ((study_options & PCRE_STUDY_JIT_COMPILE) != 0)
2168              {
2169              int jit;
2170              new_info(re, extra, PCRE_INFO_JIT, &jit);
2171              if (jit)
2172                fprintf(outfile, "JIT study was successful\n");
2173              else
2174    #ifdef SUPPORT_JIT
2175                fprintf(outfile, "JIT study was not successful\n");
2176    #else
2177                fprintf(outfile, "JIT support is not available in this version of PCRE\n");
2178    #endif
2179              }
2180          }          }
2181        }        }
2182    
# Line 1659  while (!done) Line 2211  while (!done)
2211            }            }
2212          else          else
2213            {            {
2214            fprintf(outfile, "Compiled regex written to %s\n", to_file);            fprintf(outfile, "Compiled pattern written to %s\n", to_file);
2215    
2216              /* If there is study data, write it. */
2217    
2218            if (extra != NULL)            if (extra != NULL)
2219              {              {
2220              if (fwrite(extra->study_data, 1, true_study_size, f) <              if (fwrite(extra->study_data, 1, true_study_size, f) <
# Line 1669  while (!done) Line 2224  while (!done)
2224                  strerror(errno));                  strerror(errno));
2225                }                }
2226              else fprintf(outfile, "Study data written to %s\n", to_file);              else fprintf(outfile, "Study data written to %s\n", to_file);
   
2227              }              }
2228            }            }
2229          fclose(f);          fclose(f);
2230          }          }
2231    
2232        new_free(re);        new_free(re);
2233        if (extra != NULL) new_free(extra);        if (extra != NULL) pcre_free_study(extra);
2234        if (tables != NULL) new_free((void *)tables);        if (locale_set)
2235            {
2236            new_free((void *)tables);
2237            setlocale(LC_CTYPE, "C");
2238            locale_set = 0;
2239            }
2240        continue;  /* With next regex */        continue;  /* With next regex */
2241        }        }
2242      }        /* End of non-POSIX compile */      }        /* End of non-POSIX compile */
# Line 1694  while (!done) Line 2253  while (!done)
2253      int callout_data_set = 0;      int callout_data_set = 0;
2254      int count, c;      int count, c;
2255      int copystrings = 0;      int copystrings = 0;
2256      int find_match_limit = 0;      int find_match_limit = default_find_match_limit;
2257      int getstrings = 0;      int getstrings = 0;
2258      int getlist = 0;      int getlist = 0;
2259      int gmatched = 0;      int gmatched = 0;
2260      int start_offset = 0;      int start_offset = 0;
2261        int start_offset_sign = 1;
2262      int g_notempty = 0;      int g_notempty = 0;
2263      int use_dfa = 0;      int use_dfa = 0;
2264    
# Line 1712  while (!done) Line 2272  while (!done)
2272    
2273      pcre_callout = callout;      pcre_callout = callout;
2274      first_callout = 1;      first_callout = 1;
2275        last_callout_mark = NULL;
2276      callout_extra = 0;      callout_extra = 0;
2277      callout_count = 0;      callout_count = 0;
2278      callout_fail_count = 999999;      callout_fail_count = 999999;
# Line 1726  while (!done) Line 2287  while (!done)
2287        {        {
2288        if (extend_inputline(infile, buffer + len, "data> ") == NULL)        if (extend_inputline(infile, buffer + len, "data> ") == NULL)
2289          {          {
2290          if (len > 0) break;          if (len > 0)    /* Reached EOF without hitting a newline */
2291              {
2292              fprintf(outfile, "\n");
2293              break;
2294              }
2295          done = 1;          done = 1;
2296          goto CONTINUE;          goto CONTINUE;
2297          }          }
# Line 1786  while (!done) Line 2351  while (!done)
2351            {            {
2352            unsigned char *pt = p;            unsigned char *pt = p;
2353            c = 0;            c = 0;
2354            while (isxdigit(*(++pt)))  
2355              c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W');            /* We used to have "while (isxdigit(*(++pt)))" here, but it fails
2356              when isxdigit() is a macro that refers to its argument more than
2357              once. This is banned by the C Standard, but apparently happens in at
2358              least one MacOS environment. */
2359    
2360              for (pt++; isxdigit(*pt); pt++)
2361                c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'a' - 10);
2362            if (*pt == '}')            if (*pt == '}')
2363              {              {
2364              unsigned char buff8[8];              unsigned char buff8[8];
2365              int ii, utn;              int ii, utn;
2366              utn = ord2utf8(c, buff8);              if (use_utf8)
2367              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];                {
2368              c = buff8[ii];   /* Last byte */                utn = ord2utf8(c, buff8);
2369                  for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
2370                  c = buff8[ii];   /* Last byte */
2371                  }
2372                else
2373                 {
2374                 if (c > 255)
2375                   fprintf(outfile, "** Character \\x{%x} is greater than 255 and "
2376                     "UTF-8 mode is not enabled.\n"
2377                     "** Truncation will probably give the wrong result.\n", c);
2378                 }
2379              p = pt + 1;              p = pt + 1;
2380              break;              break;
2381              }              }
# Line 1807  while (!done) Line 2388  while (!done)
2388          c = 0;          c = 0;
2389          while (i++ < 2 && isxdigit(*p))          while (i++ < 2 && isxdigit(*p))
2390            {            {
2391            c = c * 16 + tolower(*p) - ((isdigit(*p))? '0' : 'W');            c = c * 16 + tolower(*p) - ((isdigit(*p))? '0' : 'a' - 10);
2392            p++;            p++;
2393            }            }
2394          break;          break;
# Line 1817  while (!done) Line 2398  while (!done)
2398          continue;          continue;
2399    
2400          case '>':          case '>':
2401            if (*p == '-')
2402              {
2403              start_offset_sign = -1;
2404              p++;
2405              }
2406          while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';          while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';
2407            start_offset *= start_offset_sign;
2408          continue;          continue;
2409    
2410          case 'A':  /* Option setting */          case 'A':  /* Option setting */
# Line 1890  while (!done) Line 2477  while (!done)
2477  #endif  #endif
2478            use_dfa = 1;            use_dfa = 1;
2479          continue;          continue;
2480    #endif
2481    
2482    #if !defined NODFA
2483          case 'F':          case 'F':
2484          options |= PCRE_DFA_SHORTEST;          options |= PCRE_DFA_SHORTEST;
2485          continue;          continue;
# Line 1915  while (!done) Line 2504  while (!done)
2504            }            }
2505          continue;          continue;
2506    
2507            case 'J':
2508            while(isdigit(*p)) n = n * 10 + *p++ - '0';
2509            if (extra != NULL
2510                && (extra->flags & PCRE_EXTRA_EXECUTABLE_JIT) != 0
2511                && extra->executable_jit != NULL)
2512              {
2513              if (jit_stack != NULL) pcre_jit_stack_free(jit_stack);
2514              jit_stack = pcre_jit_stack_alloc(1, n * 1024);
2515              pcre_assign_jit_stack(extra, jit_callback, jit_stack);
2516              }
2517            continue;
2518    
2519          case 'L':          case 'L':
2520          getlist = 1;          getlist = 1;
2521          continue;          continue;
# Line 1924  while (!done) Line 2525  while (!done)
2525          continue;          continue;
2526    
2527          case 'N':          case 'N':
2528          options |= PCRE_NOTEMPTY;          if ((options & PCRE_NOTEMPTY) != 0)
2529              options = (options & ~PCRE_NOTEMPTY) | PCRE_NOTEMPTY_ATSTART;
2530            else
2531              options |= PCRE_NOTEMPTY;
2532          continue;          continue;
2533    
2534          case 'O':          case 'O':
# Line 1947  while (!done) Line 2551  while (!done)
2551          continue;          continue;
2552    
2553          case 'P':          case 'P':
2554          options |= PCRE_PARTIAL;          options |= ((options & PCRE_PARTIAL_SOFT) == 0)?
2555              PCRE_PARTIAL_SOFT : PCRE_PARTIAL_HARD;
2556          continue;          continue;
2557    
2558          case 'Q':          case 'Q':
# Line 1982  while (!done) Line 2587  while (!done)
2587          show_malloc = 1;          show_malloc = 1;
2588          continue;          continue;
2589    
2590            case 'Y':
2591            options |= PCRE_NO_START_OPTIMIZE;
2592            continue;
2593    
2594          case 'Z':          case 'Z':
2595          options |= PCRE_NOTEOL;          options |= PCRE_NOTEOL;
2596          continue;          continue;
# Line 2002  while (!done) Line 2611  while (!done)
2611        *q++ = c;        *q++ = c;
2612        }        }
2613      *q = 0;      *q = 0;
2614      len = q - dbuffer;      len = (int)(q - dbuffer);
2615    
2616        /* Move the data to the end of the buffer so that a read over the end of
2617        the buffer will be seen by valgrind, even if it doesn't cause a crash. If
2618        we are using the POSIX interface, we must include the terminating zero. */
2619    
2620    #if !defined NOPOSIX
2621        if (posix || do_posix)
2622          {
2623          memmove(bptr + buffer_size - len - 1, bptr, len + 1);
2624          bptr += buffer_size - len - 1;
2625          }
2626        else
2627    #endif
2628          {
2629          memmove(bptr + buffer_size - len, bptr, len);
2630          bptr += buffer_size - len;
2631          }
2632    
2633      if ((all_use_dfa || use_dfa) && find_match_limit)      if ((all_use_dfa || use_dfa) && find_match_limit)
2634        {        {
# Line 2023  while (!done) Line 2649  while (!done)
2649          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);
2650        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;
2651        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;
2652          if ((options & PCRE_NOTEMPTY) != 0) eflags |= REG_NOTEMPTY;
2653    
2654        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);
2655    
# Line 2047  while (!done) Line 2674  while (!done)
2674              (void)pchars(dbuffer + pmatch[i].rm_so,              (void)pchars(dbuffer + pmatch[i].rm_so,
2675                pmatch[i].rm_eo - pmatch[i].rm_so, outfile);                pmatch[i].rm_eo - pmatch[i].rm_so, outfile);
2676              fprintf(outfile, "\n");              fprintf(outfile, "\n");
2677              if (i == 0 && do_showrest)              if (do_showcaprest || (i == 0 && do_showrest))
2678                {                {
2679                fprintf(outfile, " 0+ ");                fprintf(outfile, "%2d+ ", (int)i);
2680                (void)pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo,                (void)pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo,
2681                  outfile);                  outfile);
2682                fprintf(outfile, "\n");                fprintf(outfile, "\n");
# Line 2067  while (!done) Line 2694  while (!done)
2694    
2695      for (;; gmatched++)    /* Loop for /g or /G */      for (;; gmatched++)    /* Loop for /g or /G */
2696        {        {
2697          markptr = NULL;
2698    
2699        if (timeitm > 0)        if (timeitm > 0)
2700          {          {
2701          register int i;          register int i;
# Line 2078  while (!done) Line 2707  while (!done)
2707            {            {
2708            int workspace[1000];            int workspace[1000];
2709            for (i = 0; i < timeitm; i++)            for (i = 0; i < timeitm; i++)
2710              count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,              count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2711                options | g_notempty, use_offsets, use_size_offsets, workspace,                options | g_notempty, use_offsets, use_size_offsets, workspace,
2712                sizeof(workspace)/sizeof(int));                sizeof(workspace)/sizeof(int));
2713            }            }
# Line 2097  while (!done) Line 2726  while (!done)
2726    
2727        /* If find_match_limit is set, we want to do repeated matches with        /* If find_match_limit is set, we want to do repeated matches with
2728        varying limits in order to find the minimum value for the match limit and        varying limits in order to find the minimum value for the match limit and
2729        for the recursion limit. */        for the recursion limit. The match limits are relevant only to the normal
2730          running of pcre_exec(), so disable the JIT optimization. This makes it
2731          possible to run the same set of tests with and without JIT externally
2732          requested. */
2733    
2734        if (find_match_limit)        if (find_match_limit)
2735          {          {
# Line 2106  while (!done) Line 2738  while (!done)
2738            extra = (pcre_extra *)malloc(sizeof(pcre_extra));            extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2739            extra->flags = 0;            extra->flags = 0;
2740            }            }
2741            else extra->flags &= ~PCRE_EXTRA_EXECUTABLE_JIT;
2742    
2743          (void)check_match_limit(re, extra, bptr, len, start_offset,          (void)check_match_limit(re, extra, bptr, len, start_offset,
2744            options|g_notempty, use_offsets, use_size_offsets,            options|g_notempty, use_offsets, use_size_offsets,
# Line 2141  while (!done) Line 2774  while (!done)
2774        else if (all_use_dfa || use_dfa)        else if (all_use_dfa || use_dfa)
2775          {          {
2776          int workspace[1000];          int workspace[1000];
2777          count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,          count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset,
2778            options | g_notempty, use_offsets, use_size_offsets, workspace,            options | g_notempty, use_offsets, use_size_offsets, workspace,
2779            sizeof(workspace)/sizeof(int));            sizeof(workspace)/sizeof(int));
2780          if (count == 0)          if (count == 0)
# Line 2189  while (!done) Line 2822  while (!done)
2822              }              }
2823            }            }
2824    
2825            /* do_allcaps requests showing of all captures in the pattern, to check
2826            unset ones at the end. */
2827    
2828            if (do_allcaps)
2829              {
2830              new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
2831              count++;   /* Allow for full match */
2832              if (count * 2 > use_size_offsets) count = use_size_offsets/2;
2833              }
2834    
2835            /* Output the captured substrings */
2836    
2837          for (i = 0; i < count * 2; i += 2)          for (i = 0; i < count * 2; i += 2)
2838            {            {
2839            if (use_offsets[i] < 0)            if (use_offsets[i] < 0)
2840                {
2841                if (use_offsets[i] != -1)
2842                  fprintf(outfile, "ERROR: bad negative value %d for offset %d\n",
2843                    use_offsets[i], i);
2844                if (use_offsets[i+1] != -1)
2845                  fprintf(outfile, "ERROR: bad negative value %d for offset %d\n",
2846                    use_offsets[i+1], i+1);
2847              fprintf(outfile, "%2d: <unset>\n", i/2);              fprintf(outfile, "%2d: <unset>\n", i/2);
2848                }
2849            else            else
2850              {              {
2851              fprintf(outfile, "%2d: ", i/2);              fprintf(outfile, "%2d: ", i/2);
2852              (void)pchars(bptr + use_offsets[i],              (void)pchars(bptr + use_offsets[i],
2853                use_offsets[i+1] - use_offsets[i], outfile);                use_offsets[i+1] - use_offsets[i], outfile);
2854              fprintf(outfile, "\n");              fprintf(outfile, "\n");
2855              if (i == 0)              if (do_showcaprest || (i == 0 && do_showrest))
2856                {                {
2857                if (do_showrest)                fprintf(outfile, "%2d+ ", i/2);
2858                  {                (void)pchars(bptr + use_offsets[i+1], len - use_offsets[i+1],
2859                  fprintf(outfile, " 0+ ");                  outfile);
2860                  (void)pchars(bptr + use_offsets[i+1], len - use_offsets[i+1],                fprintf(outfile, "\n");
                   outfile);  
                 fprintf(outfile, "\n");  
                 }  
2861                }                }
2862              }              }
2863            }            }
2864    
2865            if (markptr != NULL) fprintf(outfile, "MK: %s\n", markptr);
2866    
2867          for (i = 0; i < 32; i++)          for (i = 0; i < 32; i++)
2868            {            {
2869            if ((copystrings & (1 << i)) != 0)            if ((copystrings & (1 << i)) != 0)
# Line 2285  while (!done) Line 2937  while (!done)
2937                fprintf(outfile, "%2dL %s\n", i, stringlist[i]);                fprintf(outfile, "%2dL %s\n", i, stringlist[i]);
2938              if (stringlist[i] != NULL)              if (stringlist[i] != NULL)
2939                fprintf(outfile, "string list not terminated by NULL\n");                fprintf(outfile, "string list not terminated by NULL\n");
             /* free((void *)stringlist); */  
2940              pcre_free_substring_list(stringlist);              pcre_free_substring_list(stringlist);
2941              }              }
2942            }            }
# Line 2295  while (!done) Line 2946  while (!done)
2946    
2947        else if (count == PCRE_ERROR_PARTIAL)        else if (count == PCRE_ERROR_PARTIAL)
2948          {          {
2949          fprintf(outfile, "Partial match");          if (markptr == NULL) fprintf(outfile, "Partial match");
2950  #if !defined NODFA            else fprintf(outfile, "Partial match, mark=%s", markptr);
2951          if ((all_use_dfa || use_dfa) && use_size_offsets > 2)          if (use_size_offsets > 1)
2952            fprintf(outfile, ": %.*s", use_offsets[1] - use_offsets[0],            {
2953              bptr + use_offsets[0]);            fprintf(outfile, ": ");
2954  #endif            pchars(bptr + use_offsets[0], use_offsets[1] - use_offsets[0],
2955                outfile);
2956              }
2957          fprintf(outfile, "\n");          fprintf(outfile, "\n");
2958          break;  /* Out of the /g loop */          break;  /* Out of the /g loop */
2959          }          }
# Line 2310  while (!done) Line 2963  while (!done)
2963        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
2964        string - that was checked before setting g_notempty.        string - that was checked before setting g_notempty.
2965    
2966        Complication arises in the case when the newline option is "any" or        Complication arises in the case when the newline convention is "any",
2967        "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
2968        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,
2969        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
2970        Fudge the offset value to achieve this.        pcre_exec(). Fudge the offset value to achieve this. We check for a
2971          newline setting in the pattern; if none was set, use pcre_config() to
2972          find the default.
2973    
2974        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
2975        character, not one byte. */        character, not one byte. */
# Line 2330  while (!done) Line 2985  while (!done)
2985              {              {
2986              int d;              int d;
2987              (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);              (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);
2988              obits = (d == '\r')? PCRE_NEWLINE_CR :              /* Note that these values are always the ASCII ones, even in
2989                      (d == '\n')? PCRE_NEWLINE_LF :              EBCDIC environments. CR = 13, NL = 10. */
2990                      (d == ('\r'<<8 | '\n'))? PCRE_NEWLINE_CRLF :              obits = (d == 13)? PCRE_NEWLINE_CR :
2991                        (d == 10)? PCRE_NEWLINE_LF :
2992                        (d == (13<<8 | 10))? PCRE_NEWLINE_CRLF :
2993                      (d == -2)? PCRE_NEWLINE_ANYCRLF :                      (d == -2)? PCRE_NEWLINE_ANYCRLF :
2994                      (d == -1)? PCRE_NEWLINE_ANY : 0;                      (d == -1)? PCRE_NEWLINE_ANY : 0;
2995              }              }
2996            if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||            if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2997                   (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_CRLF ||
2998                 (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)                 (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2999                &&                &&
3000                start_offset < len - 1 &&                start_offset < len - 1 &&
# Line 2347  while (!done) Line 3005  while (!done)
3005              {              {
3006              while (start_offset + onechar < len)              while (start_offset + onechar < len)
3007                {                {
3008                int tb = bptr[start_offset+onechar];                if ((bptr[start_offset+onechar] & 0xc0) != 0x80) break;
3009                if (tb <= 127) break;                onechar++;
               tb &= 0xc0;  
               if (tb != 0 && tb != 0xc0) onechar++;  
3010                }                }
3011              }              }
3012            use_offsets[1] = start_offset + onechar;            use_offsets[1] = start_offset + onechar;
3013            }            }
3014          else          else
3015            {            {
3016            if (count == PCRE_ERROR_NOMATCH)            switch(count)
3017              {              {
3018              if (gmatched == 0) fprintf(outfile, "No match\n");              case PCRE_ERROR_NOMATCH:
3019                if (gmatched == 0)
3020                  {
3021                  if (markptr == NULL) fprintf(outfile, "No match\n");
3022                    else fprintf(outfile, "No match, mark = %s\n", markptr);
3023                  }
3024                break;
3025    
3026                case PCRE_ERROR_BADUTF8:
3027                case PCRE_ERROR_SHORTUTF8:
3028                fprintf(outfile, "Error %d (%s UTF-8 string)", count,
3029                  (count == PCRE_ERROR_BADUTF8)? "bad" : "short");
3030                if (use_size_offsets >= 2)
3031                  fprintf(outfile, " offset=%d reason=%d", use_offsets[0],
3032                    use_offsets[1]);
3033                fprintf(outfile, "\n");
3034                break;
3035    
3036                default:
3037                if (count < 0 && (-count) < sizeof(errtexts)/sizeof(const char *))
3038                  fprintf(outfile, "Error %d (%s)\n", count, errtexts[-count]);
3039                else
3040                  fprintf(outfile, "Error %d (Unexpected value)\n", count);
3041                break;
3042              }              }
3043            else fprintf(outfile, "Error %d\n", count);  
3044            break;  /* Out of the /g loop */            break;  /* Out of the /g loop */
3045            }            }
3046          }          }
# Line 2371  while (!done) Line 3050  while (!done)
3050        if (!do_g && !do_G) break;        if (!do_g && !do_G) break;
3051    
3052        /* 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
3053        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
3054        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
3055        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
3056        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
3057        character. */        character. */
3058    
# Line 2382  while (!done) Line 3061  while (!done)
3061        if (use_offsets[0] == use_offsets[1])        if (use_offsets[0] == use_offsets[1])
3062          {          {
3063          if (use_offsets[0] == len) break;          if (use_offsets[0] == len) break;
3064          g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED;          g_notempty = PCRE_NOTEMPTY_ATSTART | PCRE_ANCHORED;
3065          }          }
3066    
3067        /* For /g, update the start offset, leaving the rest alone */        /* For /g, update the start offset, leaving the rest alone */
# Line 2408  while (!done) Line 3087  while (!done)
3087  #endif  #endif
3088    
3089    if (re != NULL) new_free(re);    if (re != NULL) new_free(re);
3090    if (extra != NULL) new_free(extra);    if (extra != NULL) pcre_free_study(extra);
3091    if (tables != NULL)    if (locale_set)
3092      {      {
3093      new_free((void *)tables);      new_free((void *)tables);
3094      setlocale(LC_CTYPE, "C");      setlocale(LC_CTYPE, "C");
3095      locale_set = 0;      locale_set = 0;
3096      }      }
3097      if (jit_stack != NULL)
3098        {
3099        pcre_jit_stack_free(jit_stack);
3100        jit_stack = NULL;
3101        }
3102    }    }
3103    
3104  if (infile == stdin) fprintf(outfile, "\n");  if (infile == stdin) fprintf(outfile, "\n");

Legend:
Removed from v.289  
changed lines
  Added in v.773

  ViewVC Help
Powered by ViewVC 1.1.5