/[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 75 by nigel, Sat Feb 24 21:40:37 2007 UTC revision 841 by zherczeg, Sat Dec 31 07:04:43 2011 UTC
# Line 4  Line 4 
4    
5  /* This program was hacked up as a tester for PCRE. I really should have  /* This program was hacked up as a tester for PCRE. I really should have
6  written it more tidily in the first place. Will I ever learn? It has grown and  written it more tidily in the first place. Will I ever learn? It has grown and
7  been extended and consequently is now rather untidy in places.  been extended and consequently is now rather, er, *very* untidy in places. The
8    addition of 16-bit support has made it even worse. :-(
9    
10  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
11  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 35  POSSIBILITY OF SUCH DAMAGE. Line 36  POSSIBILITY OF SUCH DAMAGE.
36  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
37  */  */
38    
39    /* This program now supports the testing of both the 8-bit and 16-bit PCRE
40    libraries in a single program. This is different from the modules such as
41    pcre_compile.c in the library itself, which are compiled separately for each
42    mode. If both modes are enabled, for example, pcre_compile.c is compiled twice
43    (the second time with COMPILE_PCRE16 defined). By contrast, pcretest.c is
44    compiled only once. Therefore, it must not make use of any of the macros from
45    pcre_internal.h that depend on COMPILE_PCRE8 or COMPILE_PCRE16. It does,
46    however, make use of SUPPORT_PCRE8 and SUPPORT_PCRE16 to ensure that it calls
47    only supported library functions. */
48    
49    
50    #ifdef HAVE_CONFIG_H
51    #include "config.h"
52    #endif
53    
54  #include <ctype.h>  #include <ctype.h>
55  #include <stdio.h>  #include <stdio.h>
# Line 44  POSSIBILITY OF SUCH DAMAGE. Line 59  POSSIBILITY OF SUCH DAMAGE.
59  #include <locale.h>  #include <locale.h>
60  #include <errno.h>  #include <errno.h>
61    
62  /* We need the internal info for displaying the results of pcre_study(). Also  #ifdef SUPPORT_LIBREADLINE
63  for getting the opcodes for showing compiled code. */  #ifdef HAVE_UNISTD_H
64    #include <unistd.h>
65    #endif
66    #include <readline/readline.h>
67    #include <readline/history.h>
68    #endif
69    
70    
71    /* A number of things vary for Windows builds. Originally, pcretest opened its
72    input and output without "b"; then I was told that "b" was needed in some
73    environments, so it was added for release 5.0 to both the input and output. (It
74    makes no difference on Unix-like systems.) Later I was told that it is wrong
75    for the input on Windows. I've now abstracted the modes into two macros that
76    are set here, to make it easier to fiddle with them, and removed "b" from the
77    input mode under Windows. */
78    
79    #if defined(_WIN32) || defined(WIN32)
80    #include <io.h>                /* For _setmode() */
81    #include <fcntl.h>             /* For _O_BINARY */
82    #define INPUT_MODE   "r"
83    #define OUTPUT_MODE  "wb"
84    
85    #ifndef isatty
86    #define isatty _isatty         /* This is what Windows calls them, I'm told, */
87    #endif                         /* though in some environments they seem to   */
88                                   /* be already defined, hence the #ifndefs.    */
89    #ifndef fileno
90    #define fileno _fileno
91    #endif
92    
93    /* A user sent this fix for Borland Builder 5 under Windows. */
94    
95    #ifdef __BORLANDC__
96    #define _setmode(handle, mode) setmode(handle, mode)
97    #endif
98    
99    /* Not Windows */
100    
101    #else
102    #include <sys/time.h>          /* These two includes are needed */
103    #include <sys/resource.h>      /* for setrlimit(). */
104    #define INPUT_MODE   "rb"
105    #define OUTPUT_MODE  "wb"
106    #endif
107    
108    
109    /* We have to include pcre_internal.h because we need the internal info for
110    displaying the results of pcre_study() and we also need to know about the
111    internal macros, structures, and other internal data values; pcretest has
112    "inside information" compared to a program that strictly follows the PCRE API.
113    
114    Although pcre_internal.h does itself include pcre.h, we explicitly include it
115    here before pcre_internal.h so that the PCRE_EXP_xxx macros get set
116    appropriately for an application, not for building PCRE. */
117    
118    #include "pcre.h"
119    
120    #if defined SUPPORT_PCRE16 && !defined SUPPORT_PCRE8
121    /* Configure internal macros to 16 bit mode. */
122    #define COMPILE_PCRE16
123    #endif
124    
125    #include "pcre_internal.h"
126    
127    /* The pcre_printint() function, which prints the internal form of a compiled
128    regex, is held in a separate file so that (a) it can be compiled in either
129    8-bit or 16-bit mode, and (b) it can be #included directly in pcre_compile.c
130    when that is compiled in debug mode. */
131    
132    #ifdef SUPPORT_PCRE8
133    void pcre_printint(pcre *external_re, FILE *f, BOOL print_lengths);
134    #endif
135    #ifdef SUPPORT_PCRE16
136    void pcre16_printint(pcre *external_re, FILE *f, BOOL print_lengths);
137    #endif
138    
139    /* We need access to some of the data tables that PCRE uses. So as not to have
140    to keep two copies, we include the source file here, changing the names of the
141    external symbols to prevent clashes. */
142    
143    #define PCRE_INCLUDED
144    #undef PRIV
145    #define PRIV(name) name
146    
147    #include "pcre_tables.c"
148    
149    /* The definition of the macro PRINTABLE, which determines whether to print an
150    output character as-is or as a hex value when showing compiled patterns, is
151    the same as in the printint.src file. We uses it here in cases when the locale
152    has not been explicitly changed, so as to get consistent output from systems
153    that differ in their output from isprint() even in the "C" locale. */
154    
155    #ifdef EBCDIC
156    #define PRINTABLE(c) ((c) >= 64 && (c) < 255)
157    #else
158    #define PRINTABLE(c) ((c) >= 32 && (c) < 127)
159    #endif
160    
161  #define PCRE_SPY        /* For Win32 build, import data, not export */  #define PRINTOK(c) (locale_set? isprint(c) : PRINTABLE(c))
162  #include "internal.h"  
163    /* Posix support is disabled in 16 bit only mode. */
164    #if defined SUPPORT_PCRE16 && !defined SUPPORT_PCRE8 && !defined NOPOSIX
165    #define NOPOSIX
166    #endif
167    
168  /* It is possible to compile this test program without including support for  /* It is possible to compile this test program without including support for
169  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 58  Makefile. */ Line 173  Makefile. */
173  #include "pcreposix.h"  #include "pcreposix.h"
174  #endif  #endif
175    
176    /* It is also possible, originally for the benefit of a version that was
177    imported into Exim, to build pcretest without support for UTF8 or UTF16 (define
178    NOUTF), without the interface to the DFA matcher (NODFA). In fact, we
179    automatically cut out the UTF support if PCRE is built without it. */
180    
181    #ifndef SUPPORT_UTF
182    #ifndef NOUTF
183    #define NOUTF
184    #endif
185    #endif
186    
187    /* To make the code a bit tidier for 8-bit and 16-bit support, we define macros
188    for all the pcre[16]_xxx functions (except pcre16_fullinfo, which is called
189    only from one place and is handled differently). I couldn't dream up any way of
190    using a single macro to do this in a generic way, because of the many different
191    argument requirements. We know that at least one of SUPPORT_PCRE8 and
192    SUPPORT_PCRE16 must be set. First define macros for each individual mode; then
193    use these in the definitions of generic macros.
194    
195    **** Special note about the PCHARSxxx macros: the address of the string to be
196    printed is always given as two arguments: a base address followed by an offset.
197    The base address is cast to the correct data size for 8 or 16 bit data; the
198    offset is in units of this size. If the string were given as base+offset in one
199    argument, the casting might be incorrectly applied. */
200    
201    #ifdef SUPPORT_PCRE8
202    
203    #define PCHARS8(lv, p, offset, len, f) \
204      lv = pchars((pcre_uint8 *)(p) + offset, len, f)
205    
206    #define PCHARSV8(p, offset, len, f) \
207      (void)pchars((pcre_uint8 *)(p) + offset, len, f)
208    
209    #define READ_CAPTURE_NAME8(p, cn8, cn16, re) \
210      p = read_capture_name8(p, cn8, re)
211    
212    #define SET_PCRE_CALLOUT8(callout) \
213      pcre_callout = callout
214    
215    #define STRLEN8(p) ((int)strlen((char *)p))
216    
217    
218    #define PCRE_COMPILE8(re, pat, options, error, erroffset, tables) \
219      re = pcre_compile((char *)pat, options, error, erroffset, tables)
220    
221    #define PCRE_COPY_NAMED_SUBSTRING8(rc, re, bptr, offsets, count, \
222        namesptr, cbuffer, size) \
223      rc = pcre_copy_named_substring(re, (char *)bptr, offsets, count, \
224        (char *)namesptr, cbuffer, size)
225    
226    #define PCRE_COPY_SUBSTRING8(rc, bptr, offsets, count, i, cbuffer, size) \
227      rc = pcre_copy_substring((char *)bptr, offsets, count, i, cbuffer, size)
228    
229    #define PCRE_DFA_EXEC8(count, re, extra, bptr, len, start_offset, options, \
230        offsets, size_offsets, workspace, size_workspace) \
231      count = pcre_dfa_exec(re, extra, (char *)bptr, len, start_offset, options, \
232        offsets, size_offsets, workspace, size_workspace)
233    
234    #define PCRE_EXEC8(count, re, extra, bptr, len, start_offset, options, \
235        offsets, size_offsets) \
236      count = pcre_exec(re, extra, (char *)bptr, len, start_offset, options, \
237        offsets, size_offsets)
238    
239    #define PCRE_FREE_STUDY8(extra) \
240      pcre_free_study(extra)
241    
242    #define PCRE_FREE_SUBSTRING8(substring) \
243      pcre_free_substring(substring)
244    
245    #define PCRE_FREE_SUBSTRING_LIST8(listptr) \
246      pcre_free_substring_list(listptr)
247    
248    #define PCRE_GET_NAMED_SUBSTRING8(rc, re, bptr, offsets, count, \
249        getnamesptr, subsptr) \
250      rc = pcre_get_named_substring(re, (char *)bptr, offsets, count, \
251        (char *)getnamesptr, subsptr)
252    
253    #define PCRE_GET_STRINGNUMBER8(n, rc, ptr) \
254      n = pcre_get_stringnumber(re, (char *)ptr)
255    
256    #define PCRE_GET_SUBSTRING8(rc, bptr, offsets, count, i, subsptr) \
257      rc = pcre_get_substring((char *)bptr, offsets, count, i, subsptr)
258    
259    #define PCRE_GET_SUBSTRING_LIST8(rc, bptr, offsets, count, listptr) \
260      rc = pcre_get_substring_list((const char *)bptr, offsets, count, listptr)
261    
262    #define PCRE_PATTERN_TO_HOST_BYTE_ORDER8(rc, re, extra, tables) \
263      rc = pcre_pattern_to_host_byte_order(re, extra, tables)
264    
265    #define PCRE_PRINTINT8(re, outfile, debug_lengths) \
266      pcre_printint(re, outfile, debug_lengths)
267    
268    #define PCRE_STUDY8(extra, re, options, error) \
269      extra = pcre_study(re, options, error)
270    
271    #endif /* SUPPORT_PCRE8 */
272    
273    /* -----------------------------------------------------------*/
274    
275    #ifdef SUPPORT_PCRE16
276    
277    #define PCHARS16(lv, p, offset, len, f) \
278      lv = pchars16((PCRE_SPTR16)(p) + offset, len, f)
279    
280    #define PCHARSV16(p, offset, len, f) \
281      (void)pchars16((PCRE_SPTR16)(p) + offset, len, f)
282    
283    #define READ_CAPTURE_NAME16(p, cn8, cn16, re) \
284      p = read_capture_name16(p, cn16, re)
285    
286    #define STRLEN16(p) ((int)strlen16((PCRE_SPTR16)p))
287    
288    #define SET_PCRE_CALLOUT16(callout) \
289      pcre16_callout = callout
290    
291    
292    #define PCRE_COMPILE16(re, pat, options, error, erroffset, tables) \
293      re = pcre16_compile((PCRE_SPTR16)pat, options, error, erroffset, tables)
294    
295    #define PCRE_COPY_NAMED_SUBSTRING16(rc, re, bptr, offsets, count, \
296        namesptr, cbuffer, size) \
297      rc = pcre16_copy_named_substring(re, (PCRE_SPTR16)bptr, offsets, count, \
298        (PCRE_SPTR16)namesptr, (PCRE_SCHAR16 *)cbuffer, size/2)
299    
300    #define PCRE_COPY_SUBSTRING16(rc, bptr, offsets, count, i, cbuffer, size) \
301      rc = pcre16_copy_substring((PCRE_SPTR16)bptr, offsets, count, i, \
302        (PCRE_SCHAR16 *)cbuffer, size/2)
303    
304    #define PCRE_DFA_EXEC16(count, re, extra, bptr, len, start_offset, options, \
305        offsets, size_offsets, workspace, size_workspace) \
306      count = pcre16_dfa_exec(re, extra, (PCRE_SPTR16)bptr, len, start_offset, \
307        options, offsets, size_offsets, workspace, size_workspace)
308    
309    #define PCRE_EXEC16(count, re, extra, bptr, len, start_offset, options, \
310        offsets, size_offsets) \
311      count = pcre16_exec(re, extra, (PCRE_SPTR16)bptr, len, start_offset, \
312        options, offsets, size_offsets)
313    
314    #define PCRE_FREE_STUDY16(extra) \
315      pcre16_free_study(extra)
316    
317    #define PCRE_FREE_SUBSTRING16(substring) \
318      pcre16_free_substring((PCRE_SPTR16)substring)
319    
320    #define PCRE_FREE_SUBSTRING_LIST16(listptr) \
321      pcre16_free_substring_list((PCRE_SPTR16 *)listptr)
322    
323    #define PCRE_GET_NAMED_SUBSTRING16(rc, re, bptr, offsets, count, \
324        getnamesptr, subsptr) \
325      rc = pcre16_get_named_substring(re, (PCRE_SPTR16)bptr, offsets, count, \
326        (PCRE_SPTR16)getnamesptr, (PCRE_SPTR16 *)(void*)subsptr)
327    
328    #define PCRE_GET_STRINGNUMBER16(n, rc, ptr) \
329      n = pcre16_get_stringnumber(re, (PCRE_SPTR16)ptr)
330    
331    #define PCRE_GET_SUBSTRING16(rc, bptr, offsets, count, i, subsptr) \
332      rc = pcre16_get_substring((PCRE_SPTR16)bptr, offsets, count, i, \
333        (PCRE_SPTR16 *)(void*)subsptr)
334    
335    #define PCRE_GET_SUBSTRING_LIST16(rc, bptr, offsets, count, listptr) \
336      rc = pcre16_get_substring_list((PCRE_SPTR16)bptr, offsets, count, \
337        (PCRE_SPTR16 **)(void*)listptr)
338    
339    #define PCRE_PATTERN_TO_HOST_BYTE_ORDER16(rc, re, extra, tables) \
340      rc = pcre16_pattern_to_host_byte_order(re, extra, tables)
341    
342    #define PCRE_PRINTINT16(re, outfile, debug_lengths) \
343      pcre16_printint(re, outfile, debug_lengths)
344    
345    #define PCRE_STUDY16(extra, re, options, error) \
346      extra = pcre16_study(re, options, error)
347    
348    #endif /* SUPPORT_PCRE16 */
349    
350    
351    /* ----- Both modes are supported; a runtime test is needed, except for
352    pcre_config(), and the JIT stack functions, when it doesn't matter which
353    version is called. ----- */
354    
355    #if defined SUPPORT_PCRE8 && defined SUPPORT_PCRE16
356    
357    #define CHAR_SIZE (use_pcre16? 2:1)
358    
359    #define PCHARS(lv, p, offset, len, f) \
360      if (use_pcre16) \
361        PCHARS16(lv, p, offset, len, f); \
362      else \
363        PCHARS8(lv, p, offset, len, f)
364    
365    #define PCHARSV(p, offset, len, f) \
366      if (use_pcre16) \
367        PCHARSV16(p, offset, len, f); \
368      else \
369        PCHARSV8(p, offset, len, f)
370    
371    #define READ_CAPTURE_NAME(p, cn8, cn16, re) \
372      if (use_pcre16) \
373        READ_CAPTURE_NAME16(p, cn8, cn16, re); \
374      else \
375        READ_CAPTURE_NAME8(p, cn8, cn16, re)
376    
377    #define SET_PCRE_CALLOUT(callout) \
378      if (use_pcre16) \
379        SET_PCRE_CALLOUT16(callout); \
380      else \
381        SET_PCRE_CALLOUT8(callout)
382    
383    #define STRLEN(p) (use_pcre16? STRLEN16(p) : STRLEN8(p))
384    
385    #define PCRE_ASSIGN_JIT_STACK pcre_assign_jit_stack
386    
387    #define PCRE_COMPILE(re, pat, options, error, erroffset, tables) \
388      if (use_pcre16) \
389        PCRE_COMPILE16(re, pat, options, error, erroffset, tables); \
390      else \
391        PCRE_COMPILE8(re, pat, options, error, erroffset, tables)
392    
393    #define PCRE_CONFIG pcre_config
394    
395    #define PCRE_COPY_NAMED_SUBSTRING(rc, re, bptr, offsets, count, \
396        namesptr, cbuffer, size) \
397      if (use_pcre16) \
398        PCRE_COPY_NAMED_SUBSTRING16(rc, re, bptr, offsets, count, \
399          namesptr, cbuffer, size); \
400      else \
401        PCRE_COPY_NAMED_SUBSTRING8(rc, re, bptr, offsets, count, \
402          namesptr, cbuffer, size)
403    
404    #define PCRE_COPY_SUBSTRING(rc, bptr, offsets, count, i, cbuffer, size) \
405      if (use_pcre16) \
406        PCRE_COPY_SUBSTRING16(rc, bptr, offsets, count, i, cbuffer, size); \
407      else \
408        PCRE_COPY_SUBSTRING8(rc, bptr, offsets, count, i, cbuffer, size)
409    
410    #define PCRE_DFA_EXEC(count, re, extra, bptr, len, start_offset, options, \
411        offsets, size_offsets, workspace, size_workspace) \
412      if (use_pcre16) \
413        PCRE_DFA_EXEC16(count, re, extra, bptr, len, start_offset, options, \
414          offsets, size_offsets, workspace, size_workspace); \
415      else \
416        PCRE_DFA_EXEC8(count, re, extra, bptr, len, start_offset, options, \
417          offsets, size_offsets, workspace, size_workspace)
418    
419    #define PCRE_EXEC(count, re, extra, bptr, len, start_offset, options, \
420        offsets, size_offsets) \
421      if (use_pcre16) \
422        PCRE_EXEC16(count, re, extra, bptr, len, start_offset, options, \
423          offsets, size_offsets); \
424      else \
425        PCRE_EXEC8(count, re, extra, bptr, len, start_offset, options, \
426          offsets, size_offsets)
427    
428    #define PCRE_FREE_STUDY(extra) \
429      if (use_pcre16) \
430        PCRE_FREE_STUDY16(extra); \
431      else \
432        PCRE_FREE_STUDY8(extra)
433    
434    #define PCRE_FREE_SUBSTRING(substring) \
435      if (use_pcre16) \
436        PCRE_FREE_SUBSTRING16(substring); \
437      else \
438        PCRE_FREE_SUBSTRING8(substring)
439    
440    #define PCRE_FREE_SUBSTRING_LIST(listptr) \
441      if (use_pcre16) \
442        PCRE_FREE_SUBSTRING_LIST16(listptr); \
443      else \
444        PCRE_FREE_SUBSTRING_LIST8(listptr)
445    
446    #define PCRE_GET_NAMED_SUBSTRING(rc, re, bptr, offsets, count, \
447        getnamesptr, subsptr) \
448      if (use_pcre16) \
449        PCRE_GET_NAMED_SUBSTRING16(rc, re, bptr, offsets, count, \
450          getnamesptr, subsptr); \
451      else \
452        PCRE_GET_NAMED_SUBSTRING8(rc, re, bptr, offsets, count, \
453          getnamesptr, subsptr)
454    
455    #define PCRE_GET_STRINGNUMBER(n, rc, ptr) \
456      if (use_pcre16) \
457        PCRE_GET_STRINGNUMBER16(n, rc, ptr); \
458      else \
459        PCRE_GET_STRINGNUMBER8(n, rc, ptr)
460    
461    #define PCRE_GET_SUBSTRING(rc, bptr, use_offsets, count, i, subsptr) \
462      if (use_pcre16) \
463        PCRE_GET_SUBSTRING16(rc, bptr, use_offsets, count, i, subsptr); \
464      else \
465        PCRE_GET_SUBSTRING8(rc, bptr, use_offsets, count, i, subsptr)
466    
467    #define PCRE_GET_SUBSTRING_LIST(rc, bptr, offsets, count, listptr) \
468      if (use_pcre16) \
469        PCRE_GET_SUBSTRING_LIST16(rc, bptr, offsets, count, listptr); \
470      else \
471        PCRE_GET_SUBSTRING_LIST8(rc, bptr, offsets, count, listptr)
472    
473    #define PCRE_JIT_STACK_ALLOC pcre_jit_stack_alloc
474    #define PCRE_JIT_STACK_FREE pcre_jit_stack_free
475    
476    #define PCRE_MAKETABLES \
477      (use_pcre16? pcre16_maketables() : pcre_maketables())
478    
479    #define PCRE_PATTERN_TO_HOST_BYTE_ORDER(rc, re, extra, tables) \
480      if (use_pcre16) \
481        PCRE_PATTERN_TO_HOST_BYTE_ORDER16(rc, re, extra, tables); \
482      else \
483        PCRE_PATTERN_TO_HOST_BYTE_ORDER8(rc, re, extra, tables)
484    
485    #define PCRE_PRINTINT(re, outfile, debug_lengths) \
486      if (use_pcre16) \
487        PCRE_PRINTINT16(re, outfile, debug_lengths); \
488      else \
489        PCRE_PRINTINT8(re, outfile, debug_lengths)
490    
491    #define PCRE_STUDY(extra, re, options, error) \
492      if (use_pcre16) \
493        PCRE_STUDY16(extra, re, options, error); \
494      else \
495        PCRE_STUDY8(extra, re, options, error)
496    
497    /* ----- Only 8-bit mode is supported ----- */
498    
499    #elif defined SUPPORT_PCRE8
500    #define CHAR_SIZE                 1
501    #define PCHARS                    PCHARS8
502    #define PCHARSV                   PCHARSV8
503    #define READ_CAPTURE_NAME         READ_CAPTURE_NAME8
504    #define SET_PCRE_CALLOUT          SET_PCRE_CALLOUT8
505    #define STRLEN                    STRLEN8
506    #define PCRE_ASSIGN_JIT_STACK     pcre_assign_jit_stack
507    #define PCRE_COMPILE              PCRE_COMPILE8
508    #define PCRE_CONFIG               pcre_config
509    #define PCRE_COPY_NAMED_SUBSTRING PCRE_COPY_NAMED_SUBSTRING8
510    #define PCRE_COPY_SUBSTRING       PCRE_COPY_SUBSTRING8
511    #define PCRE_DFA_EXEC             PCRE_DFA_EXEC8
512    #define PCRE_EXEC                 PCRE_EXEC8
513    #define PCRE_FREE_STUDY           PCRE_FREE_STUDY8
514    #define PCRE_FREE_SUBSTRING       PCRE_FREE_SUBSTRING8
515    #define PCRE_FREE_SUBSTRING_LIST  PCRE_FREE_SUBSTRING_LIST8
516    #define PCRE_GET_NAMED_SUBSTRING  PCRE_GET_NAMED_SUBSTRING8
517    #define PCRE_GET_STRINGNUMBER     PCRE_GET_STRINGNUMBER8
518    #define PCRE_GET_SUBSTRING        PCRE_GET_SUBSTRING8
519    #define PCRE_GET_SUBSTRING_LIST   PCRE_GET_SUBSTRING_LIST8
520    #define PCRE_JIT_STACK_ALLOC      pcre_jit_stack_alloc
521    #define PCRE_JIT_STACK_FREE       pcre_jit_stack_free
522    #define PCRE_MAKETABLES           pcre_maketables()
523    #define PCRE_PATTERN_TO_HOST_BYTE_ORDER PCRE_PATTERN_TO_HOST_BYTE_ORDER8
524    #define PCRE_PRINTINT             PCRE_PRINTINT8
525    #define PCRE_STUDY                PCRE_STUDY8
526    
527    /* ----- Only 16-bit mode is supported ----- */
528    
529    #else
530    #define CHAR_SIZE                 2
531    #define PCHARS                    PCHARS16
532    #define PCHARSV                   PCHARSV16
533    #define READ_CAPTURE_NAME         READ_CAPTURE_NAME16
534    #define SET_PCRE_CALLOUT          SET_PCRE_CALLOUT16
535    #define STRLEN                    STRLEN16
536    #define PCRE_ASSIGN_JIT_STACK     pcre16_assign_jit_stack
537    #define PCRE_COMPILE              PCRE_COMPILE16
538    #define PCRE_CONFIG               pcre16_config
539    #define PCRE_COPY_NAMED_SUBSTRING PCRE_COPY_NAMED_SUBSTRING16
540    #define PCRE_COPY_SUBSTRING       PCRE_COPY_SUBSTRING16
541    #define PCRE_DFA_EXEC             PCRE_DFA_EXEC16
542    #define PCRE_EXEC                 PCRE_EXEC16
543    #define PCRE_FREE_STUDY           PCRE_FREE_STUDY16
544    #define PCRE_FREE_SUBSTRING       PCRE_FREE_SUBSTRING16
545    #define PCRE_FREE_SUBSTRING_LIST  PCRE_FREE_SUBSTRING_LIST16
546    #define PCRE_GET_NAMED_SUBSTRING  PCRE_GET_NAMED_SUBSTRING16
547    #define PCRE_GET_STRINGNUMBER     PCRE_GET_STRINGNUMBER16
548    #define PCRE_GET_SUBSTRING        PCRE_GET_SUBSTRING16
549    #define PCRE_GET_SUBSTRING_LIST   PCRE_GET_SUBSTRING_LIST16
550    #define PCRE_JIT_STACK_ALLOC      pcre16_jit_stack_alloc
551    #define PCRE_JIT_STACK_FREE       pcre16_jit_stack_free
552    #define PCRE_MAKETABLES           pcre16_maketables()
553    #define PCRE_PATTERN_TO_HOST_BYTE_ORDER PCRE_PATTERN_TO_HOST_BYTE_ORDER16
554    #define PCRE_PRINTINT             PCRE_PRINTINT16
555    #define PCRE_STUDY                PCRE_STUDY16
556    #endif
557    
558    /* ----- End of mode-specific function call macros ----- */
559    
560    
561    /* Other parameters */
562    
563  #ifndef CLOCKS_PER_SEC  #ifndef CLOCKS_PER_SEC
564  #ifdef CLK_TCK  #ifdef CLK_TCK
565  #define CLOCKS_PER_SEC CLK_TCK  #define CLOCKS_PER_SEC CLK_TCK
# Line 66  Makefile. */ Line 568  Makefile. */
568  #endif  #endif
569  #endif  #endif
570    
571  #define LOOPREPEAT 500000  /* This is the default loop count for timing. */
572    
573  #define BUFFER_SIZE 30000  #define LOOPREPEAT 500000
 #define PBUFFER_SIZE BUFFER_SIZE  
 #define DBUFFER_SIZE BUFFER_SIZE  
574    
575    /* Static variables */
576    
577  static FILE *outfile;  static FILE *outfile;
578  static int log_store = 0;  static int log_store = 0;
# Line 79  static int callout_count; Line 580  static int callout_count;
580  static int callout_extra;  static int callout_extra;
581  static int callout_fail_count;  static int callout_fail_count;
582  static int callout_fail_id;  static int callout_fail_id;
583    static int debug_lengths;
584  static int first_callout;  static int first_callout;
585    static int locale_set = 0;
586  static int show_malloc;  static int show_malloc;
587  static int use_utf8;  static int use_utf;
588  static size_t gotten_store;  static size_t gotten_store;
589    static size_t first_gotten_store = 0;
590    static const unsigned char *last_callout_mark = NULL;
591    
592  static uschar *pbuffer = NULL;  /* The buffers grow automatically if very long input lines are encountered. */
593    
594    static int buffer_size = 50000;
595    static pcre_uint8 *buffer = NULL;
596    static pcre_uint8 *dbuffer = NULL;
597    static pcre_uint8 *pbuffer = NULL;
598    
599  static const int utf8_table1[] = {  /* Another buffer is needed translation to 16-bit character strings. It will
600    0x0000007f, 0x000007ff, 0x0000ffff, 0x001fffff, 0x03ffffff, 0x7fffffff};  obtained and extended as required. */
601    
602  static const int utf8_table2[] = {  #ifdef SUPPORT_PCRE16
603    0, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc};  static int buffer16_size = 0;
604    static pcre_uint16 *buffer16 = NULL;
605    
606  static const int utf8_table3[] = {  #ifdef SUPPORT_PCRE8
   0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01};  
607    
608    /* We need the table of operator lengths that is used for 16-bit compiling, in
609    order to swap bytes in a pattern for saving/reloading testing. Luckily, the
610    data is defined as a macro. However, we must ensure that LINK_SIZE is adjusted
611    appropriately for the 16-bit world. Just as a safety check, make sure that
612    COMPILE_PCRE16 is *not* set. */
613    
614    #ifdef COMPILE_PCRE16
615    #error COMPILE_PCRE16 must not be set when compiling pcretest.c
616    #endif
617    
618  /*************************************************  #if LINK_SIZE == 2
619  *         Print compiled regex                   *  #undef LINK_SIZE
620  *************************************************/  #define LINK_SIZE 1
621    #elif LINK_SIZE == 3 || LINK_SIZE == 4
622    #undef LINK_SIZE
623    #define LINK_SIZE 2
624    #else
625    #error LINK_SIZE must be either 2, 3, or 4
626    #endif
627    
628  /* The code for doing this is held in a separate file that is also included in  #undef IMM2_SIZE
629  pcre.c when it is compiled with the debug switch. It defines a function called  #define IMM2_SIZE 1
 print_internals(), which uses a table of opcode lengths defined by the macro  
 OP_LENGTHS, whose name must be OP_lengths. It also uses a table that translates  
 Unicode property names to numbers; this is kept in a separate file. */  
630    
631  static uschar OP_lengths[] = { OP_LENGTHS };  #endif /* SUPPORT_PCRE8 */
632    
633  #include "ucp.h"  static const pcre_uint16 OP_lengths16[] = { OP_LENGTHS };
634  #include "ucptypetable.c"  #endif  /* SUPPORT_PCRE16 */
 #include "printint.c"  
635    
636    /* If we have 8-bit support, default use_pcre16 to false; if there is also
637    16-bit support, it can be changed by an option. If there is no 8-bit support,
638    there must be 16-bit support, so default it to 1. */
639    
640    #ifdef SUPPORT_PCRE8
641    static int use_pcre16 = 0;
642    #else
643    static int use_pcre16 = 1;
644    #endif
645    
646  /*************************************************  /* Textual explanations for runtime error codes */
 *          Read number from string               *  
 *************************************************/  
647    
648  /* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess  static const char *errtexts[] = {
649  around with conditional compilation, just do the job by hand. It is only used    NULL,  /* 0 is no error */
650  for unpicking the -o argument, so just keep it simple.    NULL,  /* NOMATCH is handled specially */
651      "NULL argument passed",
652      "bad option value",
653      "magic number missing",
654      "unknown opcode - pattern overwritten?",
655      "no more memory",
656      NULL,  /* never returned by pcre_exec() or pcre_dfa_exec() */
657      "match limit exceeded",
658      "callout error code",
659      NULL,  /* BADUTF8/16 is handled specially */
660      NULL,  /* BADUTF8/16 offset is handled specially */
661      NULL,  /* PARTIAL is handled specially */
662      "not used - internal error",
663      "internal error - pattern overwritten?",
664      "bad count value",
665      "item unsupported for DFA matching",
666      "backreference condition or recursion test not supported for DFA matching",
667      "match limit not supported for DFA matching",
668      "workspace size exceeded in DFA matching",
669      "too much recursion for DFA matching",
670      "recursion limit exceeded",
671      "not used - internal error",
672      "invalid combination of newline options",
673      "bad offset value",
674      NULL,  /* SHORTUTF8/16 is handled specially */
675      "nested recursion at the same subject position",
676      "JIT stack limit reached",
677      "pattern compiled in wrong mode: 8-bit/16-bit error"
678    };
679    
 Arguments:  
   str           string to be converted  
   endptr        where to put the end pointer  
680    
681  Returns:        the unsigned long  /*************************************************
682    *         Alternate character tables             *
683    *************************************************/
684    
685    /* By default, the "tables" pointer when calling PCRE is set to NULL, thereby
686    using the default tables of the library. However, the T option can be used to
687    select alternate sets of tables, for different kinds of testing. Note also that
688    the L (locale) option also adjusts the tables. */
689    
690    /* This is the set of tables distributed as default with PCRE. It recognizes
691    only ASCII characters. */
692    
693    static const pcre_uint8 tables0[] = {
694    
695    /* This table is a lower casing table. */
696    
697        0,  1,  2,  3,  4,  5,  6,  7,
698        8,  9, 10, 11, 12, 13, 14, 15,
699       16, 17, 18, 19, 20, 21, 22, 23,
700       24, 25, 26, 27, 28, 29, 30, 31,
701       32, 33, 34, 35, 36, 37, 38, 39,
702       40, 41, 42, 43, 44, 45, 46, 47,
703       48, 49, 50, 51, 52, 53, 54, 55,
704       56, 57, 58, 59, 60, 61, 62, 63,
705       64, 97, 98, 99,100,101,102,103,
706      104,105,106,107,108,109,110,111,
707      112,113,114,115,116,117,118,119,
708      120,121,122, 91, 92, 93, 94, 95,
709       96, 97, 98, 99,100,101,102,103,
710      104,105,106,107,108,109,110,111,
711      112,113,114,115,116,117,118,119,
712      120,121,122,123,124,125,126,127,
713      128,129,130,131,132,133,134,135,
714      136,137,138,139,140,141,142,143,
715      144,145,146,147,148,149,150,151,
716      152,153,154,155,156,157,158,159,
717      160,161,162,163,164,165,166,167,
718      168,169,170,171,172,173,174,175,
719      176,177,178,179,180,181,182,183,
720      184,185,186,187,188,189,190,191,
721      192,193,194,195,196,197,198,199,
722      200,201,202,203,204,205,206,207,
723      208,209,210,211,212,213,214,215,
724      216,217,218,219,220,221,222,223,
725      224,225,226,227,228,229,230,231,
726      232,233,234,235,236,237,238,239,
727      240,241,242,243,244,245,246,247,
728      248,249,250,251,252,253,254,255,
729    
730    /* This table is a case flipping table. */
731    
732        0,  1,  2,  3,  4,  5,  6,  7,
733        8,  9, 10, 11, 12, 13, 14, 15,
734       16, 17, 18, 19, 20, 21, 22, 23,
735       24, 25, 26, 27, 28, 29, 30, 31,
736       32, 33, 34, 35, 36, 37, 38, 39,
737       40, 41, 42, 43, 44, 45, 46, 47,
738       48, 49, 50, 51, 52, 53, 54, 55,
739       56, 57, 58, 59, 60, 61, 62, 63,
740       64, 97, 98, 99,100,101,102,103,
741      104,105,106,107,108,109,110,111,
742      112,113,114,115,116,117,118,119,
743      120,121,122, 91, 92, 93, 94, 95,
744       96, 65, 66, 67, 68, 69, 70, 71,
745       72, 73, 74, 75, 76, 77, 78, 79,
746       80, 81, 82, 83, 84, 85, 86, 87,
747       88, 89, 90,123,124,125,126,127,
748      128,129,130,131,132,133,134,135,
749      136,137,138,139,140,141,142,143,
750      144,145,146,147,148,149,150,151,
751      152,153,154,155,156,157,158,159,
752      160,161,162,163,164,165,166,167,
753      168,169,170,171,172,173,174,175,
754      176,177,178,179,180,181,182,183,
755      184,185,186,187,188,189,190,191,
756      192,193,194,195,196,197,198,199,
757      200,201,202,203,204,205,206,207,
758      208,209,210,211,212,213,214,215,
759      216,217,218,219,220,221,222,223,
760      224,225,226,227,228,229,230,231,
761      232,233,234,235,236,237,238,239,
762      240,241,242,243,244,245,246,247,
763      248,249,250,251,252,253,254,255,
764    
765    /* This table contains bit maps for various character classes. Each map is 32
766    bytes long and the bits run from the least significant end of each byte. The
767    classes that have their own maps are: space, xdigit, digit, upper, lower, word,
768    graph, print, punct, and cntrl. Other classes are built from combinations. */
769    
770      0x00,0x3e,0x00,0x00,0x01,0x00,0x00,0x00,
771      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
772      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
773      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
774    
775      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
776      0x7e,0x00,0x00,0x00,0x7e,0x00,0x00,0x00,
777      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
778      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
779    
780      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
781      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
782      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
783      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
784    
785      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
786      0xfe,0xff,0xff,0x07,0x00,0x00,0x00,0x00,
787      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
788      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
789    
790      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
791      0x00,0x00,0x00,0x00,0xfe,0xff,0xff,0x07,
792      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
793      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
794    
795      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
796      0xfe,0xff,0xff,0x87,0xfe,0xff,0xff,0x07,
797      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
798      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
799    
800      0x00,0x00,0x00,0x00,0xfe,0xff,0xff,0xff,
801      0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x7f,
802      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
803      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
804    
805      0x00,0x00,0x00,0x00,0xff,0xff,0xff,0xff,
806      0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x7f,
807      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
808      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
809    
810      0x00,0x00,0x00,0x00,0xfe,0xff,0x00,0xfc,
811      0x01,0x00,0x00,0xf8,0x01,0x00,0x00,0x78,
812      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
813      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
814    
815      0xff,0xff,0xff,0xff,0x00,0x00,0x00,0x00,
816      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x80,
817      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
818      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
819    
820    /* This table identifies various classes of character by individual bits:
821      0x01   white space character
822      0x02   letter
823      0x04   decimal digit
824      0x08   hexadecimal digit
825      0x10   alphanumeric or '_'
826      0x80   regular expression metacharacter or binary zero
827  */  */
828    
829  static int    0x80,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   0-  7 */
830  get_value(unsigned char *str, unsigned char **endptr)    0x00,0x01,0x01,0x00,0x01,0x01,0x00,0x00, /*   8- 15 */
831  {    0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  16- 23 */
832  int result = 0;    0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  24- 31 */
833  while(*str != 0 && isspace(*str)) str++;    0x01,0x00,0x00,0x00,0x80,0x00,0x00,0x00, /*    - '  */
834  while (isdigit(*str)) result = result * 10 + (int)(*str++ - '0');    0x80,0x80,0x80,0x80,0x00,0x00,0x80,0x00, /*  ( - /  */
835  *endptr = str;    0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c, /*  0 - 7  */
836  return(result);    0x1c,0x1c,0x00,0x00,0x00,0x00,0x00,0x80, /*  8 - ?  */
837  }    0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  @ - G  */
838      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  H - O  */
839      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  P - W  */
840      0x12,0x12,0x12,0x80,0x80,0x00,0x80,0x10, /*  X - _  */
841      0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  ` - g  */
842      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  h - o  */
843      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  p - w  */
844      0x12,0x12,0x12,0x80,0x80,0x00,0x00,0x00, /*  x -127 */
845      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 128-135 */
846      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 136-143 */
847      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144-151 */
848      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 152-159 */
849      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160-167 */
850      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 168-175 */
851      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 176-183 */
852      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
853      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 192-199 */
854      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 200-207 */
855      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 208-215 */
856      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 216-223 */
857      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 224-231 */
858      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 232-239 */
859      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 240-247 */
860      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00};/* 248-255 */
861    
862    /* This is a set of tables that came orginally from a Windows user. It seems to
863    be at least an approximation of ISO 8859. In particular, there are characters
864    greater than 128 that are marked as spaces, letters, etc. */
865    
866    static const pcre_uint8 tables1[] = {
867    0,1,2,3,4,5,6,7,
868    8,9,10,11,12,13,14,15,
869    16,17,18,19,20,21,22,23,
870    24,25,26,27,28,29,30,31,
871    32,33,34,35,36,37,38,39,
872    40,41,42,43,44,45,46,47,
873    48,49,50,51,52,53,54,55,
874    56,57,58,59,60,61,62,63,
875    64,97,98,99,100,101,102,103,
876    104,105,106,107,108,109,110,111,
877    112,113,114,115,116,117,118,119,
878    120,121,122,91,92,93,94,95,
879    96,97,98,99,100,101,102,103,
880    104,105,106,107,108,109,110,111,
881    112,113,114,115,116,117,118,119,
882    120,121,122,123,124,125,126,127,
883    128,129,130,131,132,133,134,135,
884    136,137,138,139,140,141,142,143,
885    144,145,146,147,148,149,150,151,
886    152,153,154,155,156,157,158,159,
887    160,161,162,163,164,165,166,167,
888    168,169,170,171,172,173,174,175,
889    176,177,178,179,180,181,182,183,
890    184,185,186,187,188,189,190,191,
891    224,225,226,227,228,229,230,231,
892    232,233,234,235,236,237,238,239,
893    240,241,242,243,244,245,246,215,
894    248,249,250,251,252,253,254,223,
895    224,225,226,227,228,229,230,231,
896    232,233,234,235,236,237,238,239,
897    240,241,242,243,244,245,246,247,
898    248,249,250,251,252,253,254,255,
899    0,1,2,3,4,5,6,7,
900    8,9,10,11,12,13,14,15,
901    16,17,18,19,20,21,22,23,
902    24,25,26,27,28,29,30,31,
903    32,33,34,35,36,37,38,39,
904    40,41,42,43,44,45,46,47,
905    48,49,50,51,52,53,54,55,
906    56,57,58,59,60,61,62,63,
907    64,97,98,99,100,101,102,103,
908    104,105,106,107,108,109,110,111,
909    112,113,114,115,116,117,118,119,
910    120,121,122,91,92,93,94,95,
911    96,65,66,67,68,69,70,71,
912    72,73,74,75,76,77,78,79,
913    80,81,82,83,84,85,86,87,
914    88,89,90,123,124,125,126,127,
915    128,129,130,131,132,133,134,135,
916    136,137,138,139,140,141,142,143,
917    144,145,146,147,148,149,150,151,
918    152,153,154,155,156,157,158,159,
919    160,161,162,163,164,165,166,167,
920    168,169,170,171,172,173,174,175,
921    176,177,178,179,180,181,182,183,
922    184,185,186,187,188,189,190,191,
923    224,225,226,227,228,229,230,231,
924    232,233,234,235,236,237,238,239,
925    240,241,242,243,244,245,246,215,
926    248,249,250,251,252,253,254,223,
927    192,193,194,195,196,197,198,199,
928    200,201,202,203,204,205,206,207,
929    208,209,210,211,212,213,214,247,
930    216,217,218,219,220,221,222,255,
931    0,62,0,0,1,0,0,0,
932    0,0,0,0,0,0,0,0,
933    32,0,0,0,1,0,0,0,
934    0,0,0,0,0,0,0,0,
935    0,0,0,0,0,0,255,3,
936    126,0,0,0,126,0,0,0,
937    0,0,0,0,0,0,0,0,
938    0,0,0,0,0,0,0,0,
939    0,0,0,0,0,0,255,3,
940    0,0,0,0,0,0,0,0,
941    0,0,0,0,0,0,12,2,
942    0,0,0,0,0,0,0,0,
943    0,0,0,0,0,0,0,0,
944    254,255,255,7,0,0,0,0,
945    0,0,0,0,0,0,0,0,
946    255,255,127,127,0,0,0,0,
947    0,0,0,0,0,0,0,0,
948    0,0,0,0,254,255,255,7,
949    0,0,0,0,0,4,32,4,
950    0,0,0,128,255,255,127,255,
951    0,0,0,0,0,0,255,3,
952    254,255,255,135,254,255,255,7,
953    0,0,0,0,0,4,44,6,
954    255,255,127,255,255,255,127,255,
955    0,0,0,0,254,255,255,255,
956    255,255,255,255,255,255,255,127,
957    0,0,0,0,254,255,255,255,
958    255,255,255,255,255,255,255,255,
959    0,2,0,0,255,255,255,255,
960    255,255,255,255,255,255,255,127,
961    0,0,0,0,255,255,255,255,
962    255,255,255,255,255,255,255,255,
963    0,0,0,0,254,255,0,252,
964    1,0,0,248,1,0,0,120,
965    0,0,0,0,254,255,255,255,
966    0,0,128,0,0,0,128,0,
967    255,255,255,255,0,0,0,0,
968    0,0,0,0,0,0,0,128,
969    255,255,255,255,0,0,0,0,
970    0,0,0,0,0,0,0,0,
971    128,0,0,0,0,0,0,0,
972    0,1,1,0,1,1,0,0,
973    0,0,0,0,0,0,0,0,
974    0,0,0,0,0,0,0,0,
975    1,0,0,0,128,0,0,0,
976    128,128,128,128,0,0,128,0,
977    28,28,28,28,28,28,28,28,
978    28,28,0,0,0,0,0,128,
979    0,26,26,26,26,26,26,18,
980    18,18,18,18,18,18,18,18,
981    18,18,18,18,18,18,18,18,
982    18,18,18,128,128,0,128,16,
983    0,26,26,26,26,26,26,18,
984    18,18,18,18,18,18,18,18,
985    18,18,18,18,18,18,18,18,
986    18,18,18,128,128,0,0,0,
987    0,0,0,0,0,1,0,0,
988    0,0,0,0,0,0,0,0,
989    0,0,0,0,0,0,0,0,
990    0,0,0,0,0,0,0,0,
991    1,0,0,0,0,0,0,0,
992    0,0,18,0,0,0,0,0,
993    0,0,20,20,0,18,0,0,
994    0,20,18,0,0,0,0,0,
995    18,18,18,18,18,18,18,18,
996    18,18,18,18,18,18,18,18,
997    18,18,18,18,18,18,18,0,
998    18,18,18,18,18,18,18,18,
999    18,18,18,18,18,18,18,18,
1000    18,18,18,18,18,18,18,18,
1001    18,18,18,18,18,18,18,0,
1002    18,18,18,18,18,18,18,18
1003    };
1004    
1005    
1006    
1007    
1008    #ifndef HAVE_STRERROR
1009  /*************************************************  /*************************************************
1010  *       Convert character value to UTF-8         *  *     Provide strerror() for non-ANSI libraries  *
1011  *************************************************/  *************************************************/
1012    
1013  /* This function takes an integer value in the range 0 - 0x7fffffff  /* Some old-fashioned systems still around (e.g. SunOS4) don't have strerror()
1014  and encodes it as a UTF-8 character in 0 to 6 bytes.  in their libraries, but can provide the same facility by this simple
1015    alternative function. */
 Arguments:  
   cvalue     the character value  
   buffer     pointer to buffer for result - at least 6 bytes long  
1016    
1017  Returns:     number of characters placed in the buffer  extern int   sys_nerr;
1018               -1 if input character is negative  extern char *sys_errlist[];
              0 if input character is positive but too big (only when  
              int is longer than 32 bits)  
 */  
1019    
1020  static int  char *
1021  ord2utf8(int cvalue, unsigned char *buffer)  strerror(int n)
1022  {  {
1023  register int i, j;  if (n < 0 || n >= sys_nerr) return "unknown error number";
1024  for (i = 0; i < sizeof(utf8_table1)/sizeof(int); i++)  return sys_errlist[n];
1025    if (cvalue <= utf8_table1[i]) break;  }
1026  if (i >= sizeof(utf8_table1)/sizeof(int)) return 0;  #endif /* HAVE_STRERROR */
 if (cvalue < 0) return -1;  
1027    
1028  buffer += i;  
1029  for (j = i; j > 0; j--)  /*************************************************
1030   {  *         JIT memory callback                    *
1031   *buffer-- = 0x80 | (cvalue & 0x3f);  *************************************************/
1032   cvalue >>= 6;  
1033   }  static pcre_jit_stack* jit_callback(void *arg)
1034  *buffer = utf8_table2[i] | cvalue;  {
1035  return i + 1;  return (pcre_jit_stack *)arg;
1036  }  }
1037    
1038    
1039    #if !defined NOUTF || defined SUPPORT_PCRE16
1040  /*************************************************  /*************************************************
1041  *            Convert UTF-8 string to value       *  *            Convert UTF-8 string to value       *
1042  *************************************************/  *************************************************/
# Line 188  return i + 1; Line 1045  return i + 1;
1045  and returns the value of the character.  and returns the value of the character.
1046    
1047  Argument:  Argument:
1048    buffer   a pointer to the byte vector    utf8bytes   a pointer to the byte vector
1049    vptr     a pointer to an int to receive the value    vptr        a pointer to an int to receive the value
1050    
1051  Returns:   >  0 => the number of bytes consumed  Returns:      >  0 => the number of bytes consumed
1052             -6 to 0 => malformed UTF-8 character at offset = (-return)                -6 to 0 => malformed UTF-8 character at offset = (-return)
1053  */  */
1054    
1055  static int  static int
1056  utf82ord(unsigned char *buffer, int *vptr)  utf82ord(pcre_uint8 *utf8bytes, int *vptr)
1057  {  {
1058  int c = *buffer++;  int c = *utf8bytes++;
1059  int d = c;  int d = c;
1060  int i, j, s;  int i, j, s;
1061    
# Line 218  d = (c & utf8_table3[i]) << s; Line 1075  d = (c & utf8_table3[i]) << s;
1075    
1076  for (j = 0; j < i; j++)  for (j = 0; j < i; j++)
1077    {    {
1078    c = *buffer++;    c = *utf8bytes++;
1079    if ((c & 0xc0) != 0x80) return -(j+1);    if ((c & 0xc0) != 0x80) return -(j+1);
1080    s -= 6;    s -= 6;
1081    d |= (c & 0x3f) << s;    d |= (c & 0x3f) << s;
# Line 226  for (j = 0; j < i; j++) Line 1083  for (j = 0; j < i; j++)
1083    
1084  /* Check that encoding was the correct unique one */  /* Check that encoding was the correct unique one */
1085    
1086  for (j = 0; j < sizeof(utf8_table1)/sizeof(int); j++)  for (j = 0; j < utf8_table1_size; j++)
1087    if (d <= utf8_table1[j]) break;    if (d <= utf8_table1[j]) break;
1088  if (j != i) return -(i+1);  if (j != i) return -(i+1);
1089    
# Line 235  if (j != i) return -(i+1); Line 1092  if (j != i) return -(i+1);
1092  *vptr = d;  *vptr = d;
1093  return i+1;  return i+1;
1094  }  }
1095    #endif /* NOUTF || SUPPORT_PCRE16 */
1096    
1097    
1098    
1099    #if !defined NOUTF || defined SUPPORT_PCRE16
1100    /*************************************************
1101    *       Convert character value to UTF-8         *
1102    *************************************************/
1103    
1104    /* This function takes an integer value in the range 0 - 0x7fffffff
1105    and encodes it as a UTF-8 character in 0 to 6 bytes.
1106    
1107    Arguments:
1108      cvalue     the character value
1109      utf8bytes  pointer to buffer for result - at least 6 bytes long
1110    
1111    Returns:     number of characters placed in the buffer
1112    */
1113    
1114    static int
1115    ord2utf8(int cvalue, pcre_uint8 *utf8bytes)
1116    {
1117    register int i, j;
1118    for (i = 0; i < utf8_table1_size; i++)
1119      if (cvalue <= utf8_table1[i]) break;
1120    utf8bytes += i;
1121    for (j = i; j > 0; j--)
1122     {
1123     *utf8bytes-- = 0x80 | (cvalue & 0x3f);
1124     cvalue >>= 6;
1125     }
1126    *utf8bytes = utf8_table2[i] | cvalue;
1127    return i + 1;
1128    }
1129    #endif /* NOUTF || SUPPORT_PCRE16 */
1130    
1131    
1132    
1133    #ifdef SUPPORT_PCRE16
1134  /*************************************************  /*************************************************
1135  *             Print character string             *  *         Convert a string to 16-bit             *
1136  *************************************************/  *************************************************/
1137    
1138  /* Character string printing function. Must handle UTF-8 strings in utf8  /* In non-UTF mode, the space needed for a 16-bit string is exactly double the
1139  mode. Yields number of characters printed. If handed a NULL file, just counts  8-bit size. For a UTF-8 string, the size needed for UTF-16 is no more than
1140  chars without printing. */  double, because up to 0xffff uses no more than 3 bytes in UTF-8 but possibly 4
1141    in UTF-16. Higher values use 4 bytes in UTF-8 and up to 4 bytes in UTF-16. The
1142    result is always left in buffer16.
1143    
1144    Note that this function does not object to surrogate values. This is
1145    deliberate; it makes it possible to construct UTF-16 strings that are invalid,
1146    for the purpose of testing that they are correctly faulted.
1147    
1148    Patterns to be converted are either plain ASCII or UTF-8; data lines are always
1149    in UTF-8 so that values greater than 255 can be handled.
1150    
1151  static int pchars(unsigned char *p, int length, FILE *f)  Arguments:
1152      data       TRUE if converting a data line; FALSE for a regex
1153      p          points to a byte string
1154      utf        true if UTF-8 (to be converted to UTF-16)
1155      len        number of bytes in the string (excluding trailing zero)
1156    
1157    Returns:     number of 16-bit data items used (excluding trailing zero)
1158                 OR -1 if a UTF-8 string is malformed
1159                 OR -2 if a value > 0x10ffff is encountered
1160                 OR -3 if a value > 0xffff is encountered when not in UTF mode
1161    */
1162    
1163    static int
1164    to16(int data, pcre_uint8 *p, int utf, int len)
1165  {  {
1166  int c;  pcre_uint16 *pp;
1167    
1168    if (buffer16_size < 2*len + 2)
1169      {
1170      if (buffer16 != NULL) free(buffer16);
1171      buffer16_size = 2*len + 2;
1172      buffer16 = (pcre_uint16 *)malloc(buffer16_size);
1173      if (buffer16 == NULL)
1174        {
1175        fprintf(stderr, "pcretest: malloc(%d) failed for buffer16\n", buffer16_size);
1176        exit(1);
1177        }
1178      }
1179    
1180    pp = buffer16;
1181    
1182    if (!utf && !data)
1183      {
1184      while (len-- > 0) *pp++ = *p++;
1185      }
1186    
1187    else
1188      {
1189      int c = 0;
1190      while (len > 0)
1191        {
1192        int chlen = utf82ord(p, &c);
1193        if (chlen <= 0) return -1;
1194        if (c > 0x10ffff) return -2;
1195        p += chlen;
1196        len -= chlen;
1197        if (c < 0x10000) *pp++ = c; else
1198          {
1199          if (!utf) return -3;
1200          c -= 0x10000;
1201          *pp++ = 0xD800 | (c >> 10);
1202          *pp++ = 0xDC00 | (c & 0x3ff);
1203          }
1204        }
1205      }
1206    
1207    *pp = 0;
1208    return pp - buffer16;
1209    }
1210    #endif
1211    
1212    
1213    /*************************************************
1214    *        Read or extend an input line            *
1215    *************************************************/
1216    
1217    /* Input lines are read into buffer, but both patterns and data lines can be
1218    continued over multiple input lines. In addition, if the buffer fills up, we
1219    want to automatically expand it so as to be able to handle extremely large
1220    lines that are needed for certain stress tests. When the input buffer is
1221    expanded, the other two buffers must also be expanded likewise, and the
1222    contents of pbuffer, which are a copy of the input for callouts, must be
1223    preserved (for when expansion happens for a data line). This is not the most
1224    optimal way of handling this, but hey, this is just a test program!
1225    
1226    Arguments:
1227      f            the file to read
1228      start        where in buffer to start (this *must* be within buffer)
1229      prompt       for stdin or readline()
1230    
1231    Returns:       pointer to the start of new data
1232                   could be a copy of start, or could be moved
1233                   NULL if no data read and EOF reached
1234    */
1235    
1236    static pcre_uint8 *
1237    extend_inputline(FILE *f, pcre_uint8 *start, const char *prompt)
1238    {
1239    pcre_uint8 *here = start;
1240    
1241    for (;;)
1242      {
1243      int rlen = (int)(buffer_size - (here - buffer));
1244    
1245      if (rlen > 1000)
1246        {
1247        int dlen;
1248    
1249        /* If libreadline support is required, use readline() to read a line if the
1250        input is a terminal. Note that readline() removes the trailing newline, so
1251        we must put it back again, to be compatible with fgets(). */
1252    
1253    #ifdef SUPPORT_LIBREADLINE
1254        if (isatty(fileno(f)))
1255          {
1256          size_t len;
1257          char *s = readline(prompt);
1258          if (s == NULL) return (here == start)? NULL : start;
1259          len = strlen(s);
1260          if (len > 0) add_history(s);
1261          if (len > rlen - 1) len = rlen - 1;
1262          memcpy(here, s, len);
1263          here[len] = '\n';
1264          here[len+1] = 0;
1265          free(s);
1266          }
1267        else
1268    #endif
1269    
1270        /* Read the next line by normal means, prompting if the file is stdin. */
1271    
1272          {
1273          if (f == stdin) printf("%s", prompt);
1274          if (fgets((char *)here, rlen,  f) == NULL)
1275            return (here == start)? NULL : start;
1276          }
1277    
1278        dlen = (int)strlen((char *)here);
1279        if (dlen > 0 && here[dlen - 1] == '\n') return start;
1280        here += dlen;
1281        }
1282    
1283      else
1284        {
1285        int new_buffer_size = 2*buffer_size;
1286        pcre_uint8 *new_buffer = (pcre_uint8 *)malloc(new_buffer_size);
1287        pcre_uint8 *new_dbuffer = (pcre_uint8 *)malloc(new_buffer_size);
1288        pcre_uint8 *new_pbuffer = (pcre_uint8 *)malloc(new_buffer_size);
1289    
1290        if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
1291          {
1292          fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
1293          exit(1);
1294          }
1295    
1296        memcpy(new_buffer, buffer, buffer_size);
1297        memcpy(new_pbuffer, pbuffer, buffer_size);
1298    
1299        buffer_size = new_buffer_size;
1300    
1301        start = new_buffer + (start - buffer);
1302        here = new_buffer + (here - buffer);
1303    
1304        free(buffer);
1305        free(dbuffer);
1306        free(pbuffer);
1307    
1308        buffer = new_buffer;
1309        dbuffer = new_dbuffer;
1310        pbuffer = new_pbuffer;
1311        }
1312      }
1313    
1314    return NULL;  /* Control never gets here */
1315    }
1316    
1317    
1318    
1319    /*************************************************
1320    *          Read number from string               *
1321    *************************************************/
1322    
1323    /* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess
1324    around with conditional compilation, just do the job by hand. It is only used
1325    for unpicking arguments, so just keep it simple.
1326    
1327    Arguments:
1328      str           string to be converted
1329      endptr        where to put the end pointer
1330    
1331    Returns:        the unsigned long
1332    */
1333    
1334    static int
1335    get_value(pcre_uint8 *str, pcre_uint8 **endptr)
1336    {
1337    int result = 0;
1338    while(*str != 0 && isspace(*str)) str++;
1339    while (isdigit(*str)) result = result * 10 + (int)(*str++ - '0');
1340    *endptr = str;
1341    return(result);
1342    }
1343    
1344    
1345    
1346    /*************************************************
1347    *             Print one character                *
1348    *************************************************/
1349    
1350    /* Print a single character either literally, or as a hex escape. */
1351    
1352    static int pchar(int c, FILE *f)
1353    {
1354    if (PRINTOK(c))
1355      {
1356      if (f != NULL) fprintf(f, "%c", c);
1357      return 1;
1358      }
1359    
1360    if (c < 0x100)
1361      {
1362      if (use_utf)
1363        {
1364        if (f != NULL) fprintf(f, "\\x{%02x}", c);
1365        return 6;
1366        }
1367      else
1368        {
1369        if (f != NULL) fprintf(f, "\\x%02x", c);
1370        return 4;
1371        }
1372      }
1373    
1374    if (f != NULL) fprintf(f, "\\x{%02x}", c);
1375    return (c <= 0x000000ff)? 6 :
1376           (c <= 0x00000fff)? 7 :
1377           (c <= 0x0000ffff)? 8 :
1378           (c <= 0x000fffff)? 9 : 10;
1379    }
1380    
1381    
1382    
1383    #ifdef SUPPORT_PCRE8
1384    /*************************************************
1385    *         Print 8-bit character string           *
1386    *************************************************/
1387    
1388    /* Must handle UTF-8 strings in utf8 mode. Yields number of characters printed.
1389    If handed a NULL file, just counts chars without printing. */
1390    
1391    static int pchars(pcre_uint8 *p, int length, FILE *f)
1392    {
1393    int c = 0;
1394  int yield = 0;  int yield = 0;
1395    
1396    if (length < 0)
1397      length = strlen((char *)p);
1398    
1399  while (length-- > 0)  while (length-- > 0)
1400    {    {
1401    if (use_utf8)  #if !defined NOUTF
1402      if (use_utf)
1403      {      {
1404      int rc = utf82ord(p, &c);      int rc = utf82ord(p, &c);
   
1405      if (rc > 0 && rc <= length + 1)   /* Mustn't run over the end */      if (rc > 0 && rc <= length + 1)   /* Mustn't run over the end */
1406        {        {
1407        length -= rc - 1;        length -= rc - 1;
1408        p += rc;        p += rc;
1409        if (c < 256 && isprint(c))        yield += pchar(c, f);
         {  
         if (f != NULL) fprintf(f, "%c", c);  
         yield++;  
         }  
       else  
         {  
         int n;  
         if (f != NULL) fprintf(f, "\\x{%02x}%n", c, &n);  
         yield += n;  
         }  
1410        continue;        continue;
1411        }        }
1412      }      }
1413    #endif
1414      c = *p++;
1415      yield += pchar(c, f);
1416      }
1417    
1418     /* Not UTF-8, or malformed UTF-8  */  return yield;
1419    }
1420    #endif
1421    
1422    if (isprint(c = *(p++)))  
1423      {  
1424      if (f != NULL) fprintf(f, "%c", c);  #ifdef SUPPORT_PCRE16
1425      yield++;  /*************************************************
1426      }  *    Find length of 0-terminated 16-bit string   *
1427    else  *************************************************/
1428    
1429    static int strlen16(PCRE_SPTR16 p)
1430    {
1431    int len = 0;
1432    while (*p++ != 0) len++;
1433    return len;
1434    }
1435    #endif  /* SUPPORT_PCRE16 */
1436    
1437    
1438    #ifdef SUPPORT_PCRE16
1439    /*************************************************
1440    *           Print 16-bit character string        *
1441    *************************************************/
1442    
1443    /* Must handle UTF-16 strings in utf mode. Yields number of characters printed.
1444    If handed a NULL file, just counts chars without printing. */
1445    
1446    static int pchars16(PCRE_SPTR16 p, int length, FILE *f)
1447    {
1448    int yield = 0;
1449    
1450    if (length < 0)
1451      length = strlen16(p);
1452    
1453    while (length-- > 0)
1454      {
1455      int c = *p++ & 0xffff;
1456    #if !defined NOUTF
1457      if (use_utf && c >= 0xD800 && c < 0xDC00 && length > 0)
1458      {      {
1459      if (f != NULL) fprintf(f, "\\x%02x", c);      int d = *p & 0xffff;
1460      yield += 4;      if (d >= 0xDC00 && d < 0xDFFF)
1461          {
1462          c = ((c & 0x3ff) << 10) + (d & 0x3ff) + 0x10000;
1463          length--;
1464          p++;
1465          }
1466      }      }
1467    #endif
1468      yield += pchar(c, f);
1469    }    }
1470    
1471  return yield;  return yield;
1472  }  }
1473    #endif  /* SUPPORT_PCRE16 */
1474    
1475    
1476    
1477    #ifdef SUPPORT_PCRE8
1478    /*************************************************
1479    *     Read a capture name (8-bit) and check it   *
1480    *************************************************/
1481    
1482    static pcre_uint8 *
1483    read_capture_name8(pcre_uint8 *p, pcre_uint8 **pp, pcre *re)
1484    {
1485    pcre_uint8 *npp = *pp;
1486    while (isalnum(*p)) *npp++ = *p++;
1487    *npp++ = 0;
1488    *npp = 0;
1489    if (pcre_get_stringnumber(re, (char *)(*pp)) < 0)
1490      {
1491      fprintf(outfile, "no parentheses with name \"");
1492      PCHARSV(*pp, 0, -1, outfile);
1493      fprintf(outfile, "\"\n");
1494      }
1495    
1496    *pp = npp;
1497    return p;
1498    }
1499    #endif  /* SUPPORT_PCRE8 */
1500    
1501    
1502    
1503    #ifdef SUPPORT_PCRE16
1504    /*************************************************
1505    *     Read a capture name (16-bit) and check it  *
1506    *************************************************/
1507    
1508    /* Note that the text being read is 8-bit. */
1509    
1510    static pcre_uint8 *
1511    read_capture_name16(pcre_uint8 *p, pcre_uint16 **pp, pcre *re)
1512    {
1513    pcre_uint16 *npp = *pp;
1514    while (isalnum(*p)) *npp++ = *p++;
1515    *npp++ = 0;
1516    *npp = 0;
1517    if (pcre16_get_stringnumber(re, (PCRE_SPTR16)(*pp)) < 0)
1518      {
1519      fprintf(outfile, "no parentheses with name \"");
1520      PCHARSV(*pp, 0, -1, outfile);
1521      fprintf(outfile, "\"\n");
1522      }
1523    *pp = npp;
1524    return p;
1525    }
1526    #endif  /* SUPPORT_PCRE16 */
1527    
1528    
1529    
# Line 320  if (callout_extra) Line 1552  if (callout_extra)
1552      else      else
1553        {        {
1554        fprintf(f, "%2d: ", i/2);        fprintf(f, "%2d: ", i/2);
1555        (void)pchars((unsigned char *)cb->subject + cb->offset_vector[i],        PCHARSV(cb->subject, cb->offset_vector[i],
1556          cb->offset_vector[i+1] - cb->offset_vector[i], f);          cb->offset_vector[i+1] - cb->offset_vector[i], f);
1557        fprintf(f, "\n");        fprintf(f, "\n");
1558        }        }
# Line 333  printed lengths of the substrings. */ Line 1565  printed lengths of the substrings. */
1565    
1566  if (f != NULL) fprintf(f, "--->");  if (f != NULL) fprintf(f, "--->");
1567    
1568  pre_start = pchars((unsigned char *)cb->subject, cb->start_match, f);  PCHARS(pre_start, cb->subject, 0, cb->start_match, f);
1569  post_start = pchars((unsigned char *)(cb->subject + cb->start_match),  PCHARS(post_start, cb->subject, cb->start_match,
1570    cb->current_position - cb->start_match, f);    cb->current_position - cb->start_match, f);
1571    
1572  subject_length = pchars((unsigned char *)cb->subject, cb->subject_length, NULL);  PCHARS(subject_length, cb->subject, 0, cb->subject_length, NULL);
1573    
1574  (void)pchars((unsigned char *)(cb->subject + cb->current_position),  PCHARSV(cb->subject, cb->current_position,
1575    cb->subject_length - cb->current_position, f);    cb->subject_length - cb->current_position, f);
1576    
1577  if (f != NULL) fprintf(f, "\n");  if (f != NULL) fprintf(f, "\n");
# Line 376  fprintf(outfile, "%.*s", (cb->next_item_ Line 1608  fprintf(outfile, "%.*s", (cb->next_item_
1608  fprintf(outfile, "\n");  fprintf(outfile, "\n");
1609  first_callout = 0;  first_callout = 0;
1610    
1611    if (cb->mark != last_callout_mark)
1612      {
1613      if (cb->mark == NULL)
1614        fprintf(outfile, "Latest Mark: <unset>\n");
1615      else
1616        {
1617        fprintf(outfile, "Latest Mark: ");
1618        PCHARSV(cb->mark, 0, -1, outfile);
1619        putc('\n', outfile);
1620        }
1621      last_callout_mark = cb->mark;
1622      }
1623    
1624  if (cb->callout_data != NULL)  if (cb->callout_data != NULL)
1625    {    {
1626    int callout_data = *((int *)(cb->callout_data));    int callout_data = *((int *)(cb->callout_data));
# Line 386  if (cb->callout_data != NULL) Line 1631  if (cb->callout_data != NULL)
1631      }      }
1632    }    }
1633    
1634  return (cb->callout_number != callout_fail_id)? 0 :  return (cb->callout_number != callout_fail_id)? 0 :
1635         (++callout_count >= callout_fail_count)? 1 : 0;         (++callout_count >= callout_fail_count)? 1 : 0;
1636    }
1637    
1638    
1639    /*************************************************
1640    *            Local malloc functions              *
1641    *************************************************/
1642    
1643    /* Alternative malloc function, to test functionality and save the size of a
1644    compiled re, which is the first store request that pcre_compile() makes. The
1645    show_malloc variable is set only during matching. */
1646    
1647    static void *new_malloc(size_t size)
1648    {
1649    void *block = malloc(size);
1650    gotten_store = size;
1651    if (first_gotten_store == 0) first_gotten_store = size;
1652    if (show_malloc)
1653      fprintf(outfile, "malloc       %3d %p\n", (int)size, block);
1654    return block;
1655    }
1656    
1657    static void new_free(void *block)
1658    {
1659    if (show_malloc)
1660      fprintf(outfile, "free             %p\n", block);
1661    free(block);
1662    }
1663    
1664    /* For recursion malloc/free, to test stacking calls */
1665    
1666    static void *stack_malloc(size_t size)
1667    {
1668    void *block = malloc(size);
1669    if (show_malloc)
1670      fprintf(outfile, "stack_malloc %3d %p\n", (int)size, block);
1671    return block;
1672    }
1673    
1674    static void stack_free(void *block)
1675    {
1676    if (show_malloc)
1677      fprintf(outfile, "stack_free       %p\n", block);
1678    free(block);
1679    }
1680    
1681    
1682    /*************************************************
1683    *          Call pcre_fullinfo()                  *
1684    *************************************************/
1685    
1686    /* Get one piece of information from the pcre_fullinfo() function. When only
1687    one of 8-bit or 16-bit is supported, use_pcre16 should always have the correct
1688    value, but the code is defensive.
1689    
1690    Arguments:
1691      re        compiled regex
1692      study     study data
1693      option    PCRE_INFO_xxx option
1694      ptr       where to put the data
1695    
1696    Returns:    0 when OK, < 0 on error
1697    */
1698    
1699    static int
1700    new_info(pcre *re, pcre_extra *study, int option, void *ptr)
1701    {
1702    int rc;
1703    
1704    if (use_pcre16)
1705    #ifdef SUPPORT_PCRE16
1706      rc = pcre16_fullinfo(re, study, option, ptr);
1707    #else
1708      rc = PCRE_ERROR_BADMODE;
1709    #endif
1710    else
1711    #ifdef SUPPORT_PCRE8
1712      rc = pcre_fullinfo(re, study, option, ptr);
1713    #else
1714      rc = PCRE_ERROR_BADMODE;
1715    #endif
1716    
1717    if (rc < 0)
1718      {
1719      fprintf(outfile, "Error %d from pcre%s_fullinfo(%d)\n", rc,
1720        use_pcre16? "16" : "", option);
1721      if (rc == PCRE_ERROR_BADMODE)
1722        fprintf(outfile, "Running in %s-bit mode but pattern was compiled in "
1723          "%s-bit mode\n", use_pcre16? "16":"8", use_pcre16? "8":"16");
1724      }
1725    
1726    return rc;
1727    }
1728    
1729    
1730    
1731    /*************************************************
1732    *             Swap byte functions                *
1733    *************************************************/
1734    
1735    /* The following functions swap the bytes of a pcre_uint16 and pcre_uint32
1736    value, respectively.
1737    
1738    Arguments:
1739      value        any number
1740    
1741    Returns:       the byte swapped value
1742    */
1743    
1744    static pcre_uint32
1745    swap_uint32(pcre_uint32 value)
1746    {
1747    return ((value & 0x000000ff) << 24) |
1748           ((value & 0x0000ff00) <<  8) |
1749           ((value & 0x00ff0000) >>  8) |
1750           (value >> 24);
1751    }
1752    
1753    static pcre_uint16
1754    swap_uint16(pcre_uint16 value)
1755    {
1756    return (value >> 8) | (value << 8);
1757    }
1758    
1759    
1760    
1761    /*************************************************
1762    *        Flip bytes in a compiled pattern        *
1763    *************************************************/
1764    
1765    /* This function is called if the 'F' option was present on a pattern that is
1766    to be written to a file. We flip the bytes of all the integer fields in the
1767    regex data block and the study block. In 16-bit mode this also flips relevant
1768    bytes in the pattern itself. This is to make it possible to test PCRE's
1769    ability to reload byte-flipped patterns, e.g. those compiled on a different
1770    architecture. */
1771    
1772    static void
1773    regexflip(pcre *ere, pcre_extra *extra)
1774    {
1775    real_pcre *re = (real_pcre *)ere;
1776    #ifdef SUPPORT_PCRE16
1777    int op;
1778    pcre_uint16 *ptr = (pcre_uint16 *)re + re->name_table_offset;
1779    int length = re->name_count * re->name_entry_size;
1780    #ifdef SUPPORT_UTF
1781    BOOL utf = (re->options & PCRE_UTF16) != 0;
1782    BOOL utf16_char = FALSE;
1783    #endif /* SUPPORT_UTF */
1784    #endif /* SUPPORT_PCRE16 */
1785    
1786    /* Always flip the bytes in the main data block and study blocks. */
1787    
1788    re->magic_number = REVERSED_MAGIC_NUMBER;
1789    re->size = swap_uint32(re->size);
1790    re->options = swap_uint32(re->options);
1791    re->flags = swap_uint16(re->flags);
1792    re->top_bracket = swap_uint16(re->top_bracket);
1793    re->top_backref = swap_uint16(re->top_backref);
1794    re->first_char = swap_uint16(re->first_char);
1795    re->req_char = swap_uint16(re->req_char);
1796    re->name_table_offset = swap_uint16(re->name_table_offset);
1797    re->name_entry_size = swap_uint16(re->name_entry_size);
1798    re->name_count = swap_uint16(re->name_count);
1799    
1800    if (extra != NULL)
1801      {
1802      pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
1803      rsd->size = swap_uint32(rsd->size);
1804      rsd->flags = swap_uint32(rsd->flags);
1805      rsd->minlength = swap_uint32(rsd->minlength);
1806      }
1807    
1808    /* In 8-bit mode, that is all we need to do. In 16-bit mode we must swap bytes
1809    in the name table, if present, and then in the pattern itself. */
1810    
1811    #ifdef SUPPORT_PCRE16
1812    if (!use_pcre16) return;
1813    
1814    while(TRUE)
1815      {
1816      /* Swap previous characters. */
1817      while (length-- > 0)
1818        {
1819        *ptr = swap_uint16(*ptr);
1820        ptr++;
1821        }
1822    #ifdef SUPPORT_UTF
1823      if (utf16_char)
1824        {
1825        if ((ptr[-1] & 0xfc00) == 0xd800)
1826          {
1827          /* We know that there is only one extra character in UTF-16. */
1828          *ptr = swap_uint16(*ptr);
1829          ptr++;
1830          }
1831        }
1832      utf16_char = FALSE;
1833    #endif /* SUPPORT_UTF */
1834    
1835      /* Get next opcode. */
1836    
1837      length = 0;
1838      op = *ptr;
1839      *ptr++ = swap_uint16(op);
1840    
1841      switch (op)
1842        {
1843        case OP_END:
1844        return;
1845    
1846    #ifdef SUPPORT_UTF
1847        case OP_CHAR:
1848        case OP_CHARI:
1849        case OP_NOT:
1850        case OP_NOTI:
1851        case OP_STAR:
1852        case OP_MINSTAR:
1853        case OP_PLUS:
1854        case OP_MINPLUS:
1855        case OP_QUERY:
1856        case OP_MINQUERY:
1857        case OP_UPTO:
1858        case OP_MINUPTO:
1859        case OP_EXACT:
1860        case OP_POSSTAR:
1861        case OP_POSPLUS:
1862        case OP_POSQUERY:
1863        case OP_POSUPTO:
1864        case OP_STARI:
1865        case OP_MINSTARI:
1866        case OP_PLUSI:
1867        case OP_MINPLUSI:
1868        case OP_QUERYI:
1869        case OP_MINQUERYI:
1870        case OP_UPTOI:
1871        case OP_MINUPTOI:
1872        case OP_EXACTI:
1873        case OP_POSSTARI:
1874        case OP_POSPLUSI:
1875        case OP_POSQUERYI:
1876        case OP_POSUPTOI:
1877        case OP_NOTSTAR:
1878        case OP_NOTMINSTAR:
1879        case OP_NOTPLUS:
1880        case OP_NOTMINPLUS:
1881        case OP_NOTQUERY:
1882        case OP_NOTMINQUERY:
1883        case OP_NOTUPTO:
1884        case OP_NOTMINUPTO:
1885        case OP_NOTEXACT:
1886        case OP_NOTPOSSTAR:
1887        case OP_NOTPOSPLUS:
1888        case OP_NOTPOSQUERY:
1889        case OP_NOTPOSUPTO:
1890        case OP_NOTSTARI:
1891        case OP_NOTMINSTARI:
1892        case OP_NOTPLUSI:
1893        case OP_NOTMINPLUSI:
1894        case OP_NOTQUERYI:
1895        case OP_NOTMINQUERYI:
1896        case OP_NOTUPTOI:
1897        case OP_NOTMINUPTOI:
1898        case OP_NOTEXACTI:
1899        case OP_NOTPOSSTARI:
1900        case OP_NOTPOSPLUSI:
1901        case OP_NOTPOSQUERYI:
1902        case OP_NOTPOSUPTOI:
1903        if (utf) utf16_char = TRUE;
1904    #endif
1905        /* Fall through. */
1906    
1907        default:
1908        length = OP_lengths16[op] - 1;
1909        break;
1910    
1911        case OP_CLASS:
1912        case OP_NCLASS:
1913        /* Skip the character bit map. */
1914        ptr += 32/sizeof(pcre_uint16);
1915        length = 0;
1916        break;
1917    
1918        case OP_XCLASS:
1919        /* LINK_SIZE can be 1 or 2 in 16 bit mode. */
1920        if (LINK_SIZE > 1)
1921          length = (int)((((unsigned int)(ptr[0]) << 16) | (unsigned int)(ptr[1]))
1922            - (1 + LINK_SIZE + 1));
1923        else
1924          length = (int)((unsigned int)(ptr[0]) - (1 + LINK_SIZE + 1));
1925    
1926        /* Reverse the size of the XCLASS instance. */
1927        *ptr = swap_uint16(*ptr);
1928        ptr++;
1929        if (LINK_SIZE > 1)
1930          {
1931          *ptr = swap_uint16(*ptr);
1932          ptr++;
1933          }
1934    
1935        op = *ptr;
1936        *ptr = swap_uint16(op);
1937        ptr++;
1938        if ((op & XCL_MAP) != 0)
1939          {
1940          /* Skip the character bit map. */
1941          ptr += 32/sizeof(pcre_uint16);
1942          length -= 32/sizeof(pcre_uint16);
1943          }
1944        break;
1945        }
1946      }
1947    /* Control should never reach here in 16 bit mode. */
1948    #endif /* SUPPORT_PCRE16 */
1949    }
1950    
1951    
1952    
1953    /*************************************************
1954    *        Check match or recursion limit          *
1955    *************************************************/
1956    
1957    static int
1958    check_match_limit(pcre *re, pcre_extra *extra, pcre_uint8 *bptr, int len,
1959      int start_offset, int options, int *use_offsets, int use_size_offsets,
1960      int flag, unsigned long int *limit, int errnumber, const char *msg)
1961    {
1962    int count;
1963    int min = 0;
1964    int mid = 64;
1965    int max = -1;
1966    
1967    extra->flags |= flag;
1968    
1969    for (;;)
1970      {
1971      *limit = mid;
1972    
1973      PCRE_EXEC(count, re, extra, bptr, len, start_offset, options,
1974        use_offsets, use_size_offsets);
1975    
1976      if (count == errnumber)
1977        {
1978        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
1979        min = mid;
1980        mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;
1981        }
1982    
1983      else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||
1984                             count == PCRE_ERROR_PARTIAL)
1985        {
1986        if (mid == min + 1)
1987          {
1988          fprintf(outfile, "Minimum %s limit = %d\n", msg, mid);
1989          break;
1990          }
1991        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
1992        max = mid;
1993        mid = (min + mid)/2;
1994        }
1995      else break;    /* Some other error */
1996      }
1997    
1998    extra->flags &= ~flag;
1999    return count;
2000  }  }
2001    
2002    
2003    
2004  /*************************************************  /*************************************************
2005  *            Local malloc functions              *  *         Case-independent strncmp() function    *
2006  *************************************************/  *************************************************/
2007    
2008  /* Alternative malloc function, to test functionality and show the size of the  /*
2009  compiled re. */  Arguments:
2010      s         first string
2011  static void *new_malloc(size_t size)    t         second string
2012  {    n         number of characters to compare
 void *block = malloc(size);  
 gotten_store = size;  
 if (show_malloc)  
   fprintf(outfile, "malloc       %3d %p\n", size, block);  
 return block;  
 }  
   
 static void new_free(void *block)  
 {  
 if (show_malloc)  
   fprintf(outfile, "free             %p\n", block);  
 free(block);  
 }  
   
2013    
2014  /* For recursion malloc/free, to test stacking calls */  Returns:    < 0, = 0, or > 0, according to the comparison
2015    */
2016    
2017  static void *stack_malloc(size_t size)  static int
2018    strncmpic(pcre_uint8 *s, pcre_uint8 *t, int n)
2019  {  {
2020  void *block = malloc(size);  while (n--)
2021  if (show_malloc)    {
2022    fprintf(outfile, "stack_malloc %3d %p\n", size, block);    int c = tolower(*s++) - tolower(*t++);
2023  return block;    if (c) return c;
2024      }
2025    return 0;
2026  }  }
2027    
 static void stack_free(void *block)  
 {  
 if (show_malloc)  
   fprintf(outfile, "stack_free       %p\n", block);  
 free(block);  
 }  
2028    
2029    
2030  /*************************************************  /*************************************************
2031  *          Call pcre_fullinfo()                  *  *         Check newline indicator                *
2032  *************************************************/  *************************************************/
2033    
2034  /* Get one piece of information from the pcre_fullinfo() function */  /* This is used both at compile and run-time to check for <xxx> escapes. Print
2035    a message and return 0 if there is no match.
2036    
2037    Arguments:
2038      p           points after the leading '<'
2039      f           file for error message
2040    
2041    Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
2042    */
2043    
2044  static void new_info(pcre *re, pcre_extra *study, int option, void *ptr)  static int
2045    check_newline(pcre_uint8 *p, FILE *f)
2046  {  {
2047  int rc;  if (strncmpic(p, (pcre_uint8 *)"cr>", 3) == 0) return PCRE_NEWLINE_CR;
2048  if ((rc = pcre_fullinfo(re, study, option, ptr)) < 0)  if (strncmpic(p, (pcre_uint8 *)"lf>", 3) == 0) return PCRE_NEWLINE_LF;
2049    fprintf(outfile, "Error %d from pcre_fullinfo(%d)\n", rc, option);  if (strncmpic(p, (pcre_uint8 *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
2050    if (strncmpic(p, (pcre_uint8 *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF;
2051    if (strncmpic(p, (pcre_uint8 *)"any>", 4) == 0) return PCRE_NEWLINE_ANY;
2052    if (strncmpic(p, (pcre_uint8 *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF;
2053    if (strncmpic(p, (pcre_uint8 *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE;
2054    fprintf(f, "Unknown newline type at: <%s\n", p);
2055    return 0;
2056  }  }
2057    
2058    
2059    
2060  /*************************************************  /*************************************************
2061  *         Byte flipping function                 *  *             Usage function                     *
2062  *************************************************/  *************************************************/
2063    
2064  static long int  static void
2065  byteflip(long int value, int n)  usage(void)
2066  {  {
2067  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);  printf("Usage:     pcretest [options] [<input file> [<output file>]]\n\n");
2068  return ((value & 0x000000ff) << 24) |  printf("Input and output default to stdin and stdout.\n");
2069         ((value & 0x0000ff00) <<  8) |  #ifdef SUPPORT_LIBREADLINE
2070         ((value & 0x00ff0000) >>  8) |  printf("If input is a terminal, readline() is used to read from it.\n");
2071         ((value & 0xff000000) >> 24);  #else
2072    printf("This version of pcretest is not linked with readline().\n");
2073    #endif
2074    printf("\nOptions:\n");
2075    #ifdef SUPPORT_PCRE16
2076    printf("  -16      use 16-bit interface\n");
2077    #endif
2078    printf("  -b       show compiled code (bytecode)\n");
2079    printf("  -C       show PCRE compile-time options and exit\n");
2080    printf("  -C arg   show a specific compile-time option\n");
2081    printf("           and exit with its value. The arg can be:\n");
2082    printf("     linksize     internal link size [2, 3, 4]\n");
2083    printf("     pcre8        8 bit library support enabled [0, 1]\n");
2084    printf("     pcre16       16 bit library support enabled [0, 1]\n");
2085    printf("     utf          Unicode Transformation Format supported [0, 1]\n");
2086    printf("     ucp          Unicode Properties supported [0, 1]\n");
2087    printf("     jit          Just-in-time compiler supported [0, 1]\n");
2088    printf("     newline      Newline type [CR, LF, CRLF, ANYCRLF, ANY, ???]\n");
2089    printf("  -d       debug: show compiled code and information (-b and -i)\n");
2090    #if !defined NODFA
2091    printf("  -dfa     force DFA matching for all subjects\n");
2092    #endif
2093    printf("  -help    show usage information\n");
2094    printf("  -i       show information about compiled patterns\n"
2095           "  -M       find MATCH_LIMIT minimum for each subject\n"
2096           "  -m       output memory used information\n"
2097           "  -o <n>   set size of offsets vector to <n>\n");
2098    #if !defined NOPOSIX
2099    printf("  -p       use POSIX interface\n");
2100    #endif
2101    printf("  -q       quiet: do not output PCRE version number at start\n");
2102    printf("  -S <n>   set stack size to <n> megabytes\n");
2103    printf("  -s       force each pattern to be studied at basic level\n"
2104           "  -s+      force each pattern to be studied, using JIT if available\n"
2105           "  -t       time compilation and execution\n");
2106    printf("  -t <n>   time compilation and execution, repeating <n> times\n");
2107    printf("  -tm      time execution (matching) only\n");
2108    printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
2109  }  }
2110    
2111    
2112    
   
2113  /*************************************************  /*************************************************
2114  *                Main Program                    *  *                Main Program                    *
2115  *************************************************/  *************************************************/
# Line 476  options, followed by a set of test data, Line 2121  options, followed by a set of test data,
2121  int main(int argc, char **argv)  int main(int argc, char **argv)
2122  {  {
2123  FILE *infile = stdin;  FILE *infile = stdin;
2124    const char *version;
2125  int options = 0;  int options = 0;
2126  int study_options = 0;  int study_options = 0;
2127    int default_find_match_limit = FALSE;
2128  int op = 1;  int op = 1;
2129  int timeit = 0;  int timeit = 0;
2130    int timeitm = 0;
2131  int showinfo = 0;  int showinfo = 0;
2132  int showstore = 0;  int showstore = 0;
2133    int force_study = -1;
2134    int force_study_options = 0;
2135    int quiet = 0;
2136  int size_offsets = 45;  int size_offsets = 45;
2137  int size_offsets_max;  int size_offsets_max;
2138  int *offsets;  int *offsets = NULL;
2139  #if !defined NOPOSIX  #if !defined NOPOSIX
2140  int posix = 0;  int posix = 0;
2141  #endif  #endif
2142  int debug = 0;  int debug = 0;
2143  int done = 0;  int done = 0;
2144    int all_use_dfa = 0;
2145    int yield = 0;
2146    int stack_size;
2147    
2148    pcre_jit_stack *jit_stack = NULL;
2149    
2150  unsigned char *buffer;  /* These vectors store, end-to-end, a list of zero-terminated captured
2151  unsigned char *dbuffer;  substring names, each list itself being terminated by an empty name. Assume
2152    that 1024 is plenty long enough for the few names we'll be testing. It is
2153    easiest to keep separate 8-bit and 16-bit versions, using the 16-bit version
2154    for the actual memory, to ensure alignment. By defining these variables always
2155    (whether or not 8-bit or 16-bit is supported), we avoid too much mess with
2156    #ifdefs in the code. */
2157    
2158    pcre_uint16 copynames[1024];
2159    pcre_uint16 getnames[1024];
2160    
2161    pcre_uint16 *cn16ptr;
2162    pcre_uint16 *gn16ptr;
2163    
2164    pcre_uint8 *copynames8 = (pcre_uint8 *)copynames;
2165    pcre_uint8 *getnames8 = (pcre_uint8 *)getnames;
2166    pcre_uint8 *cn8ptr;
2167    pcre_uint8 *gn8ptr;
2168    
2169    /* Get buffers from malloc() so that valgrind will check their misuse when
2170    debugging. They grow automatically when very long lines are read. The 16-bit
2171    buffer (buffer16) is obtained only if needed. */
2172    
2173    buffer = (pcre_uint8 *)malloc(buffer_size);
2174    dbuffer = (pcre_uint8 *)malloc(buffer_size);
2175    pbuffer = (pcre_uint8 *)malloc(buffer_size);
2176    
2177  /* Get buffers from malloc() so that Electric Fence will check their misuse  /* The outfile variable is static so that new_malloc can use it. */
2178  when I am debugging. */  
2179    outfile = stdout;
2180    
2181  buffer = (unsigned char *)malloc(BUFFER_SIZE);  /* The following  _setmode() stuff is some Windows magic that tells its runtime
2182  dbuffer = (unsigned char *)malloc(DBUFFER_SIZE);  library to translate CRLF into a single LF character. At least, that's what
2183  pbuffer = (unsigned char *)malloc(PBUFFER_SIZE);  I've been told: never having used Windows I take this all on trust. Originally
2184    it set 0x8000, but then I was advised that _O_BINARY was better. */
 /* The outfile variable is static so that new_malloc can use it. The _setmode()  
 stuff is some magic that I don't understand, but which apparently does good  
 things in Windows. It's related to line terminations.  */  
2185    
2186  #if defined(_WIN32) || defined(WIN32)  #if defined(_WIN32) || defined(WIN32)
2187  _setmode( _fileno( stdout ), 0x8000 );  _setmode( _fileno( stdout ), _O_BINARY );
2188  #endif  /* defined(_WIN32) || defined(WIN32) */  #endif
2189    
2190  outfile = stdout;  /* Get the version number: both pcre_version() and pcre16_version() give the
2191    same answer. We just need to ensure that we call one that is available. */
2192    
2193    #ifdef SUPPORT_PCRE8
2194    version = pcre_version();
2195    #else
2196    version = pcre16_version();
2197    #endif
2198    
2199  /* Scan options */  /* Scan options */
2200    
2201  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
2202    {    {
2203    unsigned char *endptr;    pcre_uint8 *endptr;
2204    
2205    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-m") == 0) showstore = 1;
2206      showstore = 1;    else if (strcmp(argv[op], "-s") == 0) force_study = 0;
2207    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-s+") == 0)
2208        {
2209        force_study = 1;
2210        force_study_options = PCRE_STUDY_JIT_COMPILE;
2211        }
2212      else if (strcmp(argv[op], "-16") == 0)
2213        {
2214    #ifdef SUPPORT_PCRE16
2215        use_pcre16 = 1;
2216    #else
2217        printf("** This version of PCRE was built without 16-bit support\n");
2218        exit(1);
2219    #endif
2220        }
2221      else if (strcmp(argv[op], "-q") == 0) quiet = 1;
2222      else if (strcmp(argv[op], "-b") == 0) debug = 1;
2223    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
2224    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
2225      else if (strcmp(argv[op], "-M") == 0) default_find_match_limit = TRUE;
2226    #if !defined NODFA
2227      else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
2228    #endif
2229    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&
2230        ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),        ((size_offsets = get_value((pcre_uint8 *)argv[op+1], &endptr)),
2231            *endptr == 0))
2232        {
2233        op++;
2234        argc--;
2235        }
2236      else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
2237        {
2238        int both = argv[op][2] == 0;
2239        int temp;
2240        if (argc > 2 && (temp = get_value((pcre_uint8 *)argv[op+1], &endptr),
2241                         *endptr == 0))
2242          {
2243          timeitm = temp;
2244          op++;
2245          argc--;
2246          }
2247        else timeitm = LOOPREPEAT;
2248        if (both) timeit = timeitm;
2249        }
2250      else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
2251          ((stack_size = get_value((pcre_uint8 *)argv[op+1], &endptr)),
2252          *endptr == 0))          *endptr == 0))
2253      {      {
2254    #if defined(_WIN32) || defined(WIN32) || defined(__minix)
2255        printf("PCRE: -S not supported on this OS\n");
2256        exit(1);
2257    #else
2258        int rc;
2259        struct rlimit rlim;
2260        getrlimit(RLIMIT_STACK, &rlim);
2261        rlim.rlim_cur = stack_size * 1024 * 1024;
2262        rc = setrlimit(RLIMIT_STACK, &rlim);
2263        if (rc != 0)
2264          {
2265        printf("PCRE: setrlimit() failed with error %d\n", rc);
2266        exit(1);
2267          }
2268      op++;      op++;
2269      argc--;      argc--;
2270    #endif
2271      }      }
2272  #if !defined NOPOSIX  #if !defined NOPOSIX
2273    else if (strcmp(argv[op], "-p") == 0) posix = 1;    else if (strcmp(argv[op], "-p") == 0) posix = 1;
# Line 535  while (argc > 1 && argv[op][0] == '-') Line 2275  while (argc > 1 && argv[op][0] == '-')
2275    else if (strcmp(argv[op], "-C") == 0)    else if (strcmp(argv[op], "-C") == 0)
2276      {      {
2277      int rc;      int rc;
2278      printf("PCRE version %s\n", pcre_version());      unsigned long int lrc;
2279    
2280        if (argc > 2)
2281          {
2282          if (strcmp(argv[op + 1], "linksize") == 0)
2283            {
2284            (void)PCRE_CONFIG(PCRE_CONFIG_LINK_SIZE, &rc);
2285            printf("%d\n", rc);
2286            yield = rc;
2287            goto EXIT;
2288            }
2289          if (strcmp(argv[op + 1], "pcre8") == 0)
2290            {
2291    #ifdef SUPPORT_PCRE8
2292            printf("1\n");
2293            yield = 1;
2294    #else
2295            printf("0\n");
2296            yield = 0;
2297    #endif
2298            goto EXIT;
2299            }
2300          if (strcmp(argv[op + 1], "pcre16") == 0)
2301            {
2302    #ifdef SUPPORT_PCRE16
2303            printf("1\n");
2304            yield = 1;
2305    #else
2306            printf("0\n");
2307            yield = 0;
2308    #endif
2309            goto EXIT;
2310            }
2311          if (strcmp(argv[op + 1], "utf") == 0)
2312            {
2313    #ifdef SUPPORT_PCRE8
2314            (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
2315            printf("%d\n", rc);
2316            yield = rc;
2317    #else
2318            (void)pcre16_config(PCRE_CONFIG_UTF16, &rc);
2319            printf("%d\n", rc);
2320            yield = rc;
2321    #endif
2322            goto EXIT;
2323            }
2324          if (strcmp(argv[op + 1], "ucp") == 0)
2325            {
2326            (void)PCRE_CONFIG(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
2327            printf("%d\n", rc);
2328            yield = rc;
2329            goto EXIT;
2330            }
2331          if (strcmp(argv[op + 1], "jit") == 0)
2332            {
2333            (void)PCRE_CONFIG(PCRE_CONFIG_JIT, &rc);
2334            printf("%d\n", rc);
2335            yield = rc;
2336            goto EXIT;
2337            }
2338          if (strcmp(argv[op + 1], "newline") == 0)
2339            {
2340            (void)PCRE_CONFIG(PCRE_CONFIG_NEWLINE, &rc);
2341            /* Note that these values are always the ASCII values, even
2342            in EBCDIC environments. CR is 13 and NL is 10. */
2343            printf("%s\n", (rc == 13)? "CR" :
2344              (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
2345              (rc == -2)? "ANYCRLF" :
2346              (rc == -1)? "ANY" : "???");
2347            goto EXIT;
2348            }
2349          printf("Unknown -C option: %s\n", argv[op + 1]);
2350          goto EXIT;
2351          }
2352    
2353        printf("PCRE version %s\n", version);
2354      printf("Compiled with\n");      printf("Compiled with\n");
2355    
2356    /* At least one of SUPPORT_PCRE8 and SUPPORT_PCRE16 will be set. If both
2357    are set, either both UTFs are supported or both are not supported. */
2358    
2359    #if defined SUPPORT_PCRE8 && defined SUPPORT_PCRE16
2360        printf("  8-bit and 16-bit support\n");
2361        (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
2362        if (rc)
2363          printf("  UTF-8 and UTF-16 support\n");
2364        else
2365          printf("  No UTF-8 or UTF-16 support\n");
2366    #elif defined SUPPORT_PCRE8
2367        printf("  8-bit support only\n");
2368      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
2369      printf("  %sUTF-8 support\n", rc? "" : "No ");      printf("  %sUTF-8 support\n", rc? "" : "No ");
2370      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);  #else
2371        printf("  16-bit support only\n");
2372        (void)pcre16_config(PCRE_CONFIG_UTF16, &rc);
2373        printf("  %sUTF-16 support\n", rc? "" : "No ");
2374    #endif
2375    
2376        (void)PCRE_CONFIG(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
2377      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
2378      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)PCRE_CONFIG(PCRE_CONFIG_JIT, &rc);
2379      printf("  Newline character is %s\n", (rc == '\r')? "CR" : "LF");      if (rc)
2380      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);        printf("  Just-in-time compiler support\n");
2381        else
2382          printf("  No just-in-time compiler support\n");
2383        (void)PCRE_CONFIG(PCRE_CONFIG_NEWLINE, &rc);
2384        /* Note that these values are always the ASCII values, even
2385        in EBCDIC environments. CR is 13 and NL is 10. */
2386        printf("  Newline sequence is %s\n", (rc == 13)? "CR" :
2387          (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
2388          (rc == -2)? "ANYCRLF" :
2389          (rc == -1)? "ANY" : "???");
2390        (void)PCRE_CONFIG(PCRE_CONFIG_BSR, &rc);
2391        printf("  \\R matches %s\n", rc? "CR, LF, or CRLF only" :
2392                                         "all Unicode newlines");
2393        (void)PCRE_CONFIG(PCRE_CONFIG_LINK_SIZE, &rc);
2394      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
2395      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)PCRE_CONFIG(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
2396      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
2397      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)PCRE_CONFIG(PCRE_CONFIG_MATCH_LIMIT, &lrc);
2398      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %ld\n", lrc);
2399      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)PCRE_CONFIG(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &lrc);
2400        printf("  Default recursion depth limit = %ld\n", lrc);
2401        (void)PCRE_CONFIG(PCRE_CONFIG_STACKRECURSE, &rc);
2402      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
2403      exit(0);      goto EXIT;
2404        }
2405      else if (strcmp(argv[op], "-help") == 0 ||
2406               strcmp(argv[op], "--help") == 0)
2407        {
2408        usage();
2409        goto EXIT;
2410      }      }
2411    else    else
2412      {      {
2413      printf("** Unknown or malformed option %s\n", argv[op]);      printf("** Unknown or malformed option %s\n", argv[op]);
2414      printf("Usage:   pcretest [-d] [-i] [-o <n>] [-p] [-s] [-t] [<input> [<output>]]\n");      usage();
2415      printf("  -C     show PCRE compile-time options and exit\n");      yield = 1;
2416      printf("  -d     debug: show compiled code; implies -i\n"      goto EXIT;
            "  -i     show information about compiled pattern\n"  
            "  -m     output memory used information\n"  
            "  -o <n> set size of offsets vector to <n>\n");  
 #if !defined NOPOSIX  
     printf("  -p     use POSIX interface\n");  
 #endif  
     printf("  -s     output store (memory) used information\n"  
            "  -t     time compilation and execution\n");  
     return 1;  
2417      }      }
2418    op++;    op++;
2419    argc--;    argc--;
# Line 580  offsets = (int *)malloc(size_offsets_max Line 2426  offsets = (int *)malloc(size_offsets_max
2426  if (offsets == NULL)  if (offsets == NULL)
2427    {    {
2428    printf("** Failed to get %d bytes of memory for offsets vector\n",    printf("** Failed to get %d bytes of memory for offsets vector\n",
2429      size_offsets_max * sizeof(int));      (int)(size_offsets_max * sizeof(int)));
2430    return 1;    yield = 1;
2431      goto EXIT;
2432    }    }
2433    
2434  /* Sort out the input and output files */  /* Sort out the input and output files */
2435    
2436  if (argc > 1)  if (argc > 1)
2437    {    {
2438    infile = fopen(argv[op], "rb");    infile = fopen(argv[op], INPUT_MODE);
2439    if (infile == NULL)    if (infile == NULL)
2440      {      {
2441      printf("** Failed to open %s\n", argv[op]);      printf("** Failed to open %s\n", argv[op]);
2442      return 1;      yield = 1;
2443        goto EXIT;
2444      }      }
2445    }    }
2446    
2447  if (argc > 2)  if (argc > 2)
2448    {    {
2449    outfile = fopen(argv[op+1], "wb");    outfile = fopen(argv[op+1], OUTPUT_MODE);
2450    if (outfile == NULL)    if (outfile == NULL)
2451      {      {
2452      printf("** Failed to open %s\n", argv[op+1]);      printf("** Failed to open %s\n", argv[op+1]);
2453      return 1;      yield = 1;
2454        goto EXIT;
2455      }      }
2456    }    }
2457    
2458  /* Set alternative malloc function */  /* Set alternative malloc function */
2459    
2460    #ifdef SUPPORT_PCRE8
2461  pcre_malloc = new_malloc;  pcre_malloc = new_malloc;
2462  pcre_free = new_free;  pcre_free = new_free;
2463  pcre_stack_malloc = stack_malloc;  pcre_stack_malloc = stack_malloc;
2464  pcre_stack_free = stack_free;  pcre_stack_free = stack_free;
2465    #endif
2466    
2467    #ifdef SUPPORT_PCRE16
2468    pcre16_malloc = new_malloc;
2469    pcre16_free = new_free;
2470    pcre16_stack_malloc = stack_malloc;
2471    pcre16_stack_free = stack_free;
2472    #endif
2473    
2474  /* Heading line, then prompt for first regex if stdin */  /* Heading line unless quiet, then prompt for first regex if stdin */
2475    
2476  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  if (!quiet) fprintf(outfile, "PCRE version %s\n\n", version);
2477    
2478  /* Main loop */  /* Main loop */
2479    
# Line 630  while (!done) Line 2488  while (!done)
2488  #endif  #endif
2489    
2490    const char *error;    const char *error;
2491    unsigned char *p, *pp, *ppp;    pcre_uint8 *markptr;
2492    unsigned char *to_file = NULL;    pcre_uint8 *p, *pp, *ppp;
2493    const unsigned char *tables = NULL;    pcre_uint8 *to_file = NULL;
2494      const pcre_uint8 *tables = NULL;
2495      pcre_uint32 get_options;
2496    unsigned long int true_size, true_study_size = 0;    unsigned long int true_size, true_study_size = 0;
2497    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
2498      int do_allcaps = 0;
2499      int do_mark = 0;
2500    int do_study = 0;    int do_study = 0;
2501      int no_force_study = 0;
2502    int do_debug = debug;    int do_debug = debug;
2503    int do_G = 0;    int do_G = 0;
2504    int do_g = 0;    int do_g = 0;
2505    int do_showinfo = showinfo;    int do_showinfo = showinfo;
2506    int do_showrest = 0;    int do_showrest = 0;
2507      int do_showcaprest = 0;
2508    int do_flip = 0;    int do_flip = 0;
2509    int erroroffset, len, delimiter;    int erroroffset, len, delimiter, poffset;
2510    
2511    use_utf8 = 0;    use_utf = 0;
2512      debug_lengths = 1;
2513    
2514    if (infile == stdin) printf("  re> ");    if (extend_inputline(infile, buffer, "  re> ") == NULL) break;
   if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break;  
2515    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
2516    fflush(outfile);    fflush(outfile);
2517    
# Line 659  while (!done) Line 2523  while (!done)
2523    
2524    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
2525      {      {
2526      unsigned long int magic;      pcre_uint32 magic;
2527      uschar sbuf[8];      pcre_uint8 sbuf[8];
2528      FILE *f;      FILE *f;
2529    
2530      p++;      p++;
2531        if (*p == '!')
2532          {
2533          do_debug = TRUE;
2534          do_showinfo = TRUE;
2535          p++;
2536          }
2537    
2538      pp = p + (int)strlen((char *)p);      pp = p + (int)strlen((char *)p);
2539      while (isspace(pp[-1])) pp--;      while (isspace(pp[-1])) pp--;
2540      *pp = 0;      *pp = 0;
# Line 675  while (!done) Line 2546  while (!done)
2546        continue;        continue;
2547        }        }
2548    
2549        first_gotten_store = 0;
2550      if (fread(sbuf, 1, 8, f) != 8) goto FAIL_READ;      if (fread(sbuf, 1, 8, f) != 8) goto FAIL_READ;
2551    
2552      true_size =      true_size =
# Line 683  while (!done) Line 2555  while (!done)
2555        (sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7];        (sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7];
2556    
2557      re = (real_pcre *)new_malloc(true_size);      re = (real_pcre *)new_malloc(true_size);
2558      regex_gotten_store = gotten_store;      regex_gotten_store = first_gotten_store;
2559    
2560      if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ;      if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ;
2561    
2562      magic = ((real_pcre *)re)->magic_number;      magic = ((real_pcre *)re)->magic_number;
2563      if (magic != MAGIC_NUMBER)      if (magic != MAGIC_NUMBER)
2564        {        {
2565        if (byteflip(magic, sizeof(magic)) == MAGIC_NUMBER)        if (swap_uint32(magic) == MAGIC_NUMBER)
2566          {          {
2567          do_flip = 1;          do_flip = 1;
2568          }          }
# Line 702  while (!done) Line 2574  while (!done)
2574          }          }
2575        }        }
2576    
2577      fprintf(outfile, "Compiled regex%s loaded from %s\n",      /* We hide the byte-invert info for little and big endian tests. */
2578        do_flip? " (byte-inverted)" : "", p);      fprintf(outfile, "Compiled pattern%s loaded from %s\n",
2579          do_flip && (p[-1] == '<') ? " (byte-inverted)" : "", p);
2580    
2581      /* Need to know if UTF-8 for printing data strings */      /* Now see if there is any following study data. */
   
     new_info(re, NULL, PCRE_INFO_OPTIONS, &options);  
     use_utf8 = (options & PCRE_UTF8) != 0;  
   
     /* Now see if there is any following study data */  
2582    
2583      if (true_study_size != 0)      if (true_study_size != 0)
2584        {        {
# Line 726  while (!done) Line 2594  while (!done)
2594          {          {
2595          FAIL_READ:          FAIL_READ:
2596          fprintf(outfile, "Failed to read data from %s\n", p);          fprintf(outfile, "Failed to read data from %s\n", p);
2597          if (extra != NULL) new_free(extra);          if (extra != NULL)
2598              {
2599              PCRE_FREE_STUDY(extra);
2600              }
2601          if (re != NULL) new_free(re);          if (re != NULL) new_free(re);
2602          fclose(f);          fclose(f);
2603          continue;          continue;
# Line 736  while (!done) Line 2607  while (!done)
2607        }        }
2608      else fprintf(outfile, "No study data\n");      else fprintf(outfile, "No study data\n");
2609    
2610        /* Flip the necessary bytes. */
2611        if (do_flip)
2612          {
2613          int rc;
2614          PCRE_PATTERN_TO_HOST_BYTE_ORDER(rc, re, extra, NULL);
2615          if (rc == PCRE_ERROR_BADMODE)
2616            {
2617            /* Simulate the result of the function call below. */
2618            fprintf(outfile, "Error %d from pcre%s_fullinfo(%d)\n", rc,
2619              use_pcre16? "16" : "", PCRE_INFO_OPTIONS);
2620            fprintf(outfile, "Running in %s-bit mode but pattern was compiled in "
2621              "%s-bit mode\n", use_pcre16? "16":"8", use_pcre16? "8":"16");
2622            continue;
2623            }
2624          }
2625    
2626        /* Need to know if UTF-8 for printing data strings. */
2627    
2628        if (new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options) < 0) continue;
2629        use_utf = (get_options & PCRE_UTF8) != 0;
2630    
2631      fclose(f);      fclose(f);
2632      goto SHOW_INFO;      goto SHOW_INFO;
2633      }      }
2634    
2635    /* In-line pattern (the usual case). Get the delimiter and seek the end of    /* In-line pattern (the usual case). Get the delimiter and seek the end of
2636    the pattern; if is isn't complete, read more. */    the pattern; if it isn't complete, read more. */
2637    
2638    delimiter = *p++;    delimiter = *p++;
2639    
2640    if (isalnum(delimiter) || delimiter == '\\')    if (isalnum(delimiter) || delimiter == '\\')
2641      {      {
2642      fprintf(outfile, "** Delimiter must not be alphameric or \\\n");      fprintf(outfile, "** Delimiter must not be alphanumeric or \\\n");
2643      goto SKIP_DATA;      goto SKIP_DATA;
2644      }      }
2645    
2646    pp = p;    pp = p;
2647      poffset = (int)(p - buffer);
2648    
2649    for(;;)    for(;;)
2650      {      {
# Line 762  while (!done) Line 2655  while (!done)
2655        pp++;        pp++;
2656        }        }
2657      if (*pp != 0) break;      if (*pp != 0) break;
2658        if ((pp = extend_inputline(infile, pp, "    > ")) == NULL)
     len = BUFFER_SIZE - (pp - buffer);  
     if (len < 256)  
       {  
       fprintf(outfile, "** Expression too long - missing delimiter?\n");  
       goto SKIP_DATA;  
       }  
   
     if (infile == stdin) printf("    > ");  
     if (fgets((char *)pp, len, infile) == NULL)  
2659        {        {
2660        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
2661        done = 1;        done = 1;
# Line 780  while (!done) Line 2664  while (!done)
2664      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
2665      }      }
2666    
2667      /* The buffer may have moved while being extended; reset the start of data
2668      pointer to the correct relative point in the buffer. */
2669    
2670      p = buffer + poffset;
2671    
2672    /* If the first character after the delimiter is backslash, make    /* If the first character after the delimiter is backslash, make
2673    the pattern end with backslash. This is purely to provide a way    the pattern end with backslash. This is purely to provide a way
2674    of testing for the error message when a pattern ends with backslash. */    of testing for the error message when a pattern ends with backslash. */
# Line 802  while (!done) Line 2691  while (!done)
2691      {      {
2692      switch (*pp++)      switch (*pp++)
2693        {        {
2694          case 'f': options |= PCRE_FIRSTLINE; break;
2695        case 'g': do_g = 1; break;        case 'g': do_g = 1; break;
2696        case 'i': options |= PCRE_CASELESS; break;        case 'i': options |= PCRE_CASELESS; break;
2697        case 'm': options |= PCRE_MULTILINE; break;        case 'm': options |= PCRE_MULTILINE; break;
2698        case 's': options |= PCRE_DOTALL; break;        case 's': options |= PCRE_DOTALL; break;
2699        case 'x': options |= PCRE_EXTENDED; break;        case 'x': options |= PCRE_EXTENDED; break;
2700    
2701        case '+': do_showrest = 1; break;        case '+':
2702          if (do_showrest) do_showcaprest = 1; else do_showrest = 1;
2703          break;
2704    
2705          case '=': do_allcaps = 1; break;
2706        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
2707          case 'B': do_debug = 1; break;
2708        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
2709        case 'D': do_debug = do_showinfo = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
2710        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
2711        case 'F': do_flip = 1; break;        case 'F': do_flip = 1; break;
2712        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
2713        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
2714          case 'J': options |= PCRE_DUPNAMES; break;
2715          case 'K': do_mark = 1; break;
2716        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
2717        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
2718    
# Line 823  while (!done) Line 2720  while (!done)
2720        case 'P': do_posix = 1; break;        case 'P': do_posix = 1; break;
2721  #endif  #endif
2722    
2723        case 'S': do_study = 1; break;        case 'S':
2724          if (do_study == 0)
2725            {
2726            do_study = 1;
2727            if (*pp == '+')
2728              {
2729              study_options |= PCRE_STUDY_JIT_COMPILE;
2730              pp++;
2731              }
2732            }
2733          else
2734            {
2735            do_study = 0;
2736            no_force_study = 1;
2737            }
2738          break;
2739    
2740        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
2741          case 'W': options |= PCRE_UCP; break;
2742        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
2743        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case 'Y': options |= PCRE_NO_START_OPTIMISE; break;
2744          case 'Z': debug_lengths = 0; break;
2745          case '8': options |= PCRE_UTF8; use_utf = 1; break;
2746        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
2747    
2748          case 'T':
2749          switch (*pp++)
2750            {
2751            case '0': tables = tables0; break;
2752            case '1': tables = tables1; break;
2753    
2754            case '\r':
2755            case '\n':
2756            case ' ':
2757            case 0:
2758            fprintf(outfile, "** Missing table number after /T\n");
2759            goto SKIP_DATA;
2760    
2761            default:
2762            fprintf(outfile, "** Bad table number \"%c\" after /T\n", pp[-1]);
2763            goto SKIP_DATA;
2764            }
2765          break;
2766    
2767        case 'L':        case 'L':
2768        ppp = pp;        ppp = pp;
2769        while (*ppp != '\n' && *ppp != ' ') ppp++;        /* The '\r' test here is so that it works on Windows. */
2770          /* The '0' test is just in case this is an unterminated line. */
2771          while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
2772        *ppp = 0;        *ppp = 0;
2773        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
2774          {          {
2775          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
2776          goto SKIP_DATA;          goto SKIP_DATA;
2777          }          }
2778        tables = pcre_maketables();        locale_set = 1;
2779          tables = PCRE_MAKETABLES;
2780        pp = ppp;        pp = ppp;
2781        break;        break;
2782    
# Line 849  while (!done) Line 2787  while (!done)
2787        *pp = 0;        *pp = 0;
2788        break;        break;
2789    
2790        case '\n': case ' ': break;        case '<':
2791            {
2792            if (strncmpic(pp, (pcre_uint8 *)"JS>", 3) == 0)
2793              {
2794              options |= PCRE_JAVASCRIPT_COMPAT;
2795              pp += 3;
2796              }
2797            else
2798              {
2799              int x = check_newline(pp, outfile);
2800              if (x == 0) goto SKIP_DATA;
2801              options |= x;
2802              while (*pp++ != '>');
2803              }
2804            }
2805          break;
2806    
2807          case '\r':                      /* So that it works in Windows */
2808          case '\n':
2809          case ' ':
2810          break;
2811    
2812        default:        default:
2813        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);        fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);
# Line 859  while (!done) Line 2817  while (!done)
2817    
2818    /* Handle compiling via the POSIX interface, which doesn't support the    /* Handle compiling via the POSIX interface, which doesn't support the
2819    timing, showing, or debugging options, nor the ability to pass over    timing, showing, or debugging options, nor the ability to pass over
2820    local character tables. */    local character tables. Neither does it have 16-bit support. */
2821    
2822  #if !defined NOPOSIX  #if !defined NOPOSIX
2823    if (posix || do_posix)    if (posix || do_posix)
# Line 869  while (!done) Line 2827  while (!done)
2827    
2828      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
2829      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
2830        if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
2831        if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
2832        if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
2833        if ((options & PCRE_UCP) != 0) cflags |= REG_UCP;
2834        if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY;
2835    
2836        first_gotten_store = 0;
2837      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
2838    
2839      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
# Line 876  while (!done) Line 2841  while (!done)
2841    
2842      if (rc != 0)      if (rc != 0)
2843        {        {
2844        (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);        (void)regerror(rc, &preg, (char *)buffer, buffer_size);
2845        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
2846        goto SKIP_DATA;        goto SKIP_DATA;
2847        }        }
# Line 888  while (!done) Line 2853  while (!done)
2853  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
2854    
2855      {      {
2856      if (timeit)      /* In 16-bit mode, convert the input. */
2857    
2858    #ifdef SUPPORT_PCRE16
2859        if (use_pcre16)
2860          {
2861          switch(to16(FALSE, p, options & PCRE_UTF8, (int)strlen((char *)p)))
2862            {
2863            case -1:
2864            fprintf(outfile, "**Failed: invalid UTF-8 string cannot be "
2865              "converted to UTF-16\n");
2866            goto SKIP_DATA;
2867    
2868            case -2:
2869            fprintf(outfile, "**Failed: character value greater than 0x10ffff "
2870              "cannot be converted to UTF-16\n");
2871            goto SKIP_DATA;
2872    
2873            case -3: /* "Impossible error" when to16 is called arg1 FALSE */
2874            fprintf(outfile, "**Failed: character value greater than 0xffff "
2875              "cannot be converted to 16-bit in non-UTF mode\n");
2876            goto SKIP_DATA;
2877    
2878            default:
2879            break;
2880            }
2881          p = (pcre_uint8 *)buffer16;
2882          }
2883    #endif
2884    
2885        /* Compile many times when timing */
2886    
2887        if (timeit > 0)
2888        {        {
2889        register int i;        register int i;
2890        clock_t time_taken;        clock_t time_taken;
2891        clock_t start_time = clock();        clock_t start_time = clock();
2892        for (i = 0; i < LOOPREPEAT; i++)        for (i = 0; i < timeit; i++)
2893          {          {
2894          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);          PCRE_COMPILE(re, p, options, &error, &erroroffset, tables);
2895          if (re != NULL) free(re);          if (re != NULL) free(re);
2896          }          }
2897        time_taken = clock() - start_time;        time_taken = clock() - start_time;
2898        fprintf(outfile, "Compile time %.3f milliseconds\n",        fprintf(outfile, "Compile time %.4f milliseconds\n",
2899          (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /          (((double)time_taken * 1000.0) / (double)timeit) /
2900            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
2901        }        }
2902    
2903      re = pcre_compile((char *)p, options, &error, &erroroffset, tables);      first_gotten_store = 0;
2904        PCRE_COMPILE(re, p, options, &error, &erroroffset, tables);
2905    
2906      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
2907      if non-interactive. */      if non-interactive. */
# Line 917  while (!done) Line 2914  while (!done)
2914          {          {
2915          for (;;)          for (;;)
2916            {            {
2917            if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)            if (extend_inputline(infile, buffer, NULL) == NULL)
2918              {              {
2919              done = 1;              done = 1;
2920              goto CONTINUE;              goto CONTINUE;
# Line 931  while (!done) Line 2928  while (!done)
2928        goto CONTINUE;        goto CONTINUE;
2929        }        }
2930    
2931      /* Compilation succeeded; print data if required. There are now two      /* Compilation succeeded. It is now possible to set the UTF-8 option from
2932      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
2933      returns only limited data. Check that it agrees with the newer one. */      lines. */
2934    
2935      if (log_store)      if (new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options) < 0)
2936        fprintf(outfile, "Memory allocation (code space): %d\n",        goto SKIP_DATA;
2937          (int)(gotten_store -      if ((get_options & PCRE_UTF8) != 0) use_utf = 1;
               sizeof(real_pcre) -  
               ((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size));  
2938    
2939      /* Extract the size for possible writing before possibly flipping it,      /* Extract the size for possible writing before possibly flipping it,
2940      and remember the store that was got. */      and remember the store that was got. */
2941    
2942      true_size = ((real_pcre *)re)->size;      true_size = ((real_pcre *)re)->size;
2943      regex_gotten_store = gotten_store;      regex_gotten_store = first_gotten_store;
2944    
2945        /* Output code size information if requested */
2946    
2947        if (log_store)
2948          fprintf(outfile, "Memory allocation (code space): %d\n",
2949            (int)(first_gotten_store -
2950                  sizeof(real_pcre) -
2951                  ((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size));
2952    
2953      /* 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
2954      help with the matching. */      help with the matching, unless the pattern has the SS option, which
2955        suppresses the effect of /S (used for a few test patterns where studying is
2956        never sensible). */
2957    
2958      if (do_study)      if (do_study || (force_study >= 0 && !no_force_study))
2959        {        {
2960        if (timeit)        if (timeit > 0)
2961          {          {
2962          register int i;          register int i;
2963          clock_t time_taken;          clock_t time_taken;
2964          clock_t start_time = clock();          clock_t start_time = clock();
2965          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeit; i++)
2966            extra = pcre_study(re, study_options, &error);            {
2967              PCRE_STUDY(extra, re, study_options | force_study_options, &error);
2968              }
2969          time_taken = clock() - start_time;          time_taken = clock() - start_time;
2970          if (extra != NULL) free(extra);          if (extra != NULL)
2971          fprintf(outfile, "  Study time %.3f milliseconds\n",            {
2972            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            PCRE_FREE_STUDY(extra);
2973              }
2974            fprintf(outfile, "  Study time %.4f milliseconds\n",
2975              (((double)time_taken * 1000.0) / (double)timeit) /
2976              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
2977          }          }
2978        extra = pcre_study(re, study_options, &error);        PCRE_STUDY(extra, re, study_options | force_study_options, &error);
2979        if (error != NULL)        if (error != NULL)
2980          fprintf(outfile, "Failed to study: %s\n", error);          fprintf(outfile, "Failed to study: %s\n", error);
2981        else if (extra != NULL)        else if (extra != NULL)
2982            {
2983          true_study_size = ((pcre_study_data *)(extra->study_data))->size;          true_study_size = ((pcre_study_data *)(extra->study_data))->size;
2984            if (log_store)
2985              {
2986              size_t jitsize;
2987              if (new_info(re, extra, PCRE_INFO_JITSIZE, &jitsize) == 0 &&
2988                  jitsize != 0)
2989                fprintf(outfile, "Memory allocation (JIT code): %d\n", (int)jitsize);
2990              }
2991            }
2992        }        }
2993    
2994      /* If the 'F' option was present, we flip the bytes of all the integer      /* If /K was present, we set up for handling MARK data. */
     fields in the regex data block and the study block. This is to make it  
     possible to test PCRE's handling of byte-flipped patterns, e.g. those  
     compiled on a different architecture. */  
2995    
2996      if (do_flip)      if (do_mark)
2997        {        {
2998        real_pcre *rre = (real_pcre *)re;        if (extra == NULL)
       rre->magic_number = byteflip(rre->magic_number, sizeof(rre->magic_number));  
       rre->size = byteflip(rre->size, sizeof(rre->size));  
       rre->options = byteflip(rre->options, sizeof(rre->options));  
       rre->top_bracket = byteflip(rre->top_bracket, sizeof(rre->top_bracket));  
       rre->top_backref = byteflip(rre->top_backref, sizeof(rre->top_backref));  
       rre->first_byte = byteflip(rre->first_byte, sizeof(rre->first_byte));  
       rre->req_byte = byteflip(rre->req_byte, sizeof(rre->req_byte));  
       rre->name_table_offset = byteflip(rre->name_table_offset,  
         sizeof(rre->name_table_offset));  
       rre->name_entry_size = byteflip(rre->name_entry_size,  
         sizeof(rre->name_entry_size));  
       rre->name_count = byteflip(rre->name_count, sizeof(rre->name_count));  
   
       if (extra != NULL)  
2999          {          {
3000          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);          extra = (pcre_extra *)malloc(sizeof(pcre_extra));
3001          rsd->size = byteflip(rsd->size, sizeof(rsd->size));          extra->flags = 0;
         rsd->options = byteflip(rsd->options, sizeof(rsd->options));  
3002          }          }
3003          extra->mark = &markptr;
3004          extra->flags |= PCRE_EXTRA_MARK;
3005        }        }
3006    
3007      /* Extract information from the compiled data if required */      /* Extract and display information from the compiled data if required. */
3008    
3009      SHOW_INFO:      SHOW_INFO:
3010    
3011        if (do_debug)
3012          {
3013          fprintf(outfile, "------------------------------------------------------------------\n");
3014          PCRE_PRINTINT(re, outfile, debug_lengths);
3015          }
3016    
3017        /* We already have the options in get_options (see above) */
3018    
3019      if (do_showinfo)      if (do_showinfo)
3020        {        {
3021        unsigned long int get_options, all_options;        unsigned long int all_options;
3022        int old_first_char, old_options, old_count;        int count, backrefmax, first_char, need_char, okpartial, jchanged,
3023        int count, backrefmax, first_char, need_char;          hascrorlf;
3024        int nameentrysize, namecount;        int nameentrysize, namecount;
3025        const uschar *nametable;        const pcre_uint8 *nametable;
   
       if (do_debug)  
         {  
         fprintf(outfile, "------------------------------------------------------------------\n");  
         print_internals(re, outfile);  
         }  
3026    
3027        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);        if (new_info(re, NULL, PCRE_INFO_SIZE, &size) +
3028        new_info(re, NULL, PCRE_INFO_SIZE, &size);            new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count) +
3029        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);            new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax) +
3030        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);            new_info(re, NULL, PCRE_INFO_FIRSTBYTE, &first_char) +
3031        new_info(re, NULL, PCRE_INFO_FIRSTBYTE, &first_char);            new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char) +
3032        new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char);            new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize) +
3033        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);            new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount) +
3034        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);            new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable) +
3035        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);            new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial) +
3036              new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged) +
3037        old_count = pcre_info(re, &old_options, &old_first_char);            new_info(re, NULL, PCRE_INFO_HASCRORLF, &hascrorlf)
3038        if (count < 0) fprintf(outfile,            != 0)
3039          "Error %d from pcre_info()\n", count);          goto SKIP_DATA;
       else  
         {  
         if (old_count != count) fprintf(outfile,  
           "Count disagreement: pcre_fullinfo=%d pcre_info=%d\n", count,  
             old_count);  
   
         if (old_first_char != first_char) fprintf(outfile,  
           "First char disagreement: pcre_fullinfo=%d pcre_info=%d\n",  
             first_char, old_first_char);  
   
         if (old_options != (int)get_options) fprintf(outfile,  
           "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",  
             get_options, old_options);  
         }  
3040    
3041        if (size != regex_gotten_store) fprintf(outfile,        if (size != regex_gotten_store) fprintf(outfile,
3042          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",
3043          size, regex_gotten_store);          (int)size, (int)regex_gotten_store);
3044    
3045        fprintf(outfile, "Capturing subpattern count = %d\n", count);        fprintf(outfile, "Capturing subpattern count = %d\n", count);
3046        if (backrefmax > 0)        if (backrefmax > 0)
# Line 1060  while (!done) Line 3051  while (!done)
3051          fprintf(outfile, "Named capturing subpatterns:\n");          fprintf(outfile, "Named capturing subpatterns:\n");
3052          while (namecount-- > 0)          while (namecount-- > 0)
3053            {            {
3054            fprintf(outfile, "  %s %*s%3d\n", nametable + 2,  #if defined SUPPORT_PCRE8 && defined SUPPORT_PCRE16
3055              nameentrysize - 3 - (int)strlen((char *)nametable + 2), "",            int imm2_size = use_pcre16 ? 1 : 2;
3056              GET2(nametable, 0));  #else
3057              int imm2_size = IMM2_SIZE;
3058    #endif
3059              int length = (int)STRLEN(nametable + imm2_size);
3060              fprintf(outfile, "  ");
3061              PCHARSV(nametable, imm2_size, length, outfile);
3062              while (length++ < nameentrysize - imm2_size) putc(' ', outfile);
3063    #if defined SUPPORT_PCRE8 && defined SUPPORT_PCRE16
3064              fprintf(outfile, "%3d\n", use_pcre16?
3065                 (int)(((PCRE_SPTR16)nametable)[0])
3066                :((int)nametable[0] << 8) | (int)nametable[1]);
3067              nametable += nameentrysize * (use_pcre16 ? 2 : 1);
3068    #else
3069              fprintf(outfile, "%3d\n", GET2(nametable, 0));
3070    #ifdef SUPPORT_PCRE8
3071            nametable += nameentrysize;            nametable += nameentrysize;
3072    #else
3073              nametable += nameentrysize * 2;
3074    #endif
3075    #endif
3076            }            }
3077          }          }
3078    
3079        /* The NOPARTIAL bit is a private bit in the options, so we have        if (!okpartial) fprintf(outfile, "Partial matching not supported\n");
3080        to fish it out via out back door */        if (hascrorlf) fprintf(outfile, "Contains explicit CR or LF match\n");
3081    
3082        all_options = ((real_pcre *)re)->options;        all_options = ((real_pcre *)re)->options;
3083        if (do_flip)        if (do_flip) all_options = swap_uint32(all_options);
         {  
         all_options = byteflip(all_options, sizeof(all_options));  
         }  
   
       if ((all_options & PCRE_NOPARTIAL) != 0)  
         fprintf(outfile, "Partial matching not supported\n");  
3084    
3085        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
3086          else fprintf(outfile, "Options:%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",
3087            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
3088            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
3089            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
3090            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",
3091              ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",
3092            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",
3093              ((get_options & PCRE_BSR_ANYCRLF) != 0)? " bsr_anycrlf" : "",
3094              ((get_options & PCRE_BSR_UNICODE) != 0)? " bsr_unicode" : "",
3095            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
3096            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
3097            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
3098            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
3099            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "");            ((get_options & PCRE_UTF8) != 0)? " utf" : "",
3100              ((get_options & PCRE_UCP) != 0)? " ucp" : "",
3101              ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf_check" : "",
3102              ((get_options & PCRE_NO_START_OPTIMIZE) != 0)? " no_start_optimize" : "",
3103              ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
3104    
3105          if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
3106    
3107          switch (get_options & PCRE_NEWLINE_BITS)
3108            {
3109            case PCRE_NEWLINE_CR:
3110            fprintf(outfile, "Forced newline sequence: CR\n");
3111            break;
3112    
3113            case PCRE_NEWLINE_LF:
3114            fprintf(outfile, "Forced newline sequence: LF\n");
3115            break;
3116    
3117        if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0)          case PCRE_NEWLINE_CRLF:
3118          fprintf(outfile, "Case state changes\n");          fprintf(outfile, "Forced newline sequence: CRLF\n");
3119            break;
3120    
3121            case PCRE_NEWLINE_ANYCRLF:
3122            fprintf(outfile, "Forced newline sequence: ANYCRLF\n");
3123            break;
3124    
3125            case PCRE_NEWLINE_ANY:
3126            fprintf(outfile, "Forced newline sequence: ANY\n");
3127            break;
3128    
3129            default:
3130            break;
3131            }
3132    
3133        if (first_char == -1)        if (first_char == -1)
3134          {          {
3135          fprintf(outfile, "First char at start or follows \\n\n");          fprintf(outfile, "First char at start or follows newline\n");
3136          }          }
3137        else if (first_char < 0)        else if (first_char < 0)
3138          {          {
# Line 1105  while (!done) Line 3140  while (!done)
3140          }          }
3141        else        else
3142          {          {
3143          int ch = first_char & 255;          const char *caseless =
3144          const char *caseless = ((first_char & REQ_CASELESS) == 0)?            ((((real_pcre *)re)->flags & PCRE_FCH_CASELESS) == 0)?
3145            "" : " (caseless)";            "" : " (caseless)";
3146          if (isprint(ch))  
3147            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);          if (PRINTOK(first_char))
3148              fprintf(outfile, "First char = \'%c\'%s\n", first_char, caseless);
3149          else          else
3150            fprintf(outfile, "First char = %d%s\n", ch, caseless);            {
3151              fprintf(outfile, "First char = ");
3152              pchar(first_char, outfile);
3153              fprintf(outfile, "%s\n", caseless);
3154              }
3155          }          }
3156    
3157        if (need_char < 0)        if (need_char < 0)
# Line 1120  while (!done) Line 3160  while (!done)
3160          }          }
3161        else        else
3162          {          {
3163          int ch = need_char & 255;          const char *caseless =
3164          const char *caseless = ((need_char & REQ_CASELESS) == 0)?            ((((real_pcre *)re)->flags & PCRE_RCH_CASELESS) == 0)?
3165            "" : " (caseless)";            "" : " (caseless)";
3166          if (isprint(ch))  
3167            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);          if (PRINTOK(need_char))
3168              fprintf(outfile, "Need char = \'%c\'%s\n", need_char, caseless);
3169          else          else
3170            fprintf(outfile, "Need char = %d%s\n", ch, caseless);            {
3171              fprintf(outfile, "Need char = ");
3172              pchar(need_char, outfile);
3173              fprintf(outfile, "%s\n", caseless);
3174              }
3175          }          }
3176    
3177        /* 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
3178        value, but it varies, depending on the computer architecture, and        value, but it varies, depending on the computer architecture, and
3179        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
3180        flipped.) */        flipped.) If study was forced by an external -s, don't show this
3181          information unless -i or -d was also present. This means that, except
3182          when auto-callouts are involved, the output from runs with and without
3183          -s should be identical. */
3184    
3185        if (do_study)        if (do_study || (force_study >= 0 && showinfo && !no_force_study))
3186          {          {
3187          if (extra == NULL)          if (extra == NULL)
3188            fprintf(outfile, "Study returned NULL\n");            fprintf(outfile, "Study returned NULL\n");
3189          else          else
3190            {            {
3191            uschar *start_bits = NULL;            pcre_uint8 *start_bits = NULL;
3192            new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);            int minlength;
3193    
3194            if (start_bits == NULL)            if (new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength) == 0)
3195              fprintf(outfile, "No starting byte set\n");              fprintf(outfile, "Subject length lower bound = %d\n", minlength);
3196            else  
3197              if (new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits) == 0)
3198              {              {
3199              int i;              if (start_bits == NULL)
3200              int c = 24;                fprintf(outfile, "No set of starting bytes\n");
3201              fprintf(outfile, "Starting byte set: ");              else
             for (i = 0; i < 256; i++)  
3202                {                {
3203                if ((start_bits[i/8] & (1<<(i&7))) != 0)                int i;
3204                  int c = 24;
3205                  fprintf(outfile, "Starting byte set: ");
3206                  for (i = 0; i < 256; i++)
3207                  {                  {
3208                  if (c > 75)                  if ((start_bits[i/8] & (1<<(i&7))) != 0)
                   {  
                   fprintf(outfile, "\n  ");  
                   c = 2;  
                   }  
                 if (isprint(i) && i != ' ')  
                   {  
                   fprintf(outfile, "%c ", i);  
                   c += 2;  
                   }  
                 else  
3209                    {                    {
3210                    fprintf(outfile, "\\x%02x ", i);                    if (c > 75)
3211                    c += 5;                      {
3212                        fprintf(outfile, "\n  ");
3213                        c = 2;
3214                        }
3215                      if (PRINTOK(i) && i != ' ')
3216                        {
3217                        fprintf(outfile, "%c ", i);
3218                        c += 2;
3219                        }
3220                      else
3221                        {
3222                        fprintf(outfile, "\\x%02x ", i);
3223                        c += 5;
3224                        }
3225                    }                    }
3226                  }                  }
3227                  fprintf(outfile, "\n");
3228                }                }
3229              fprintf(outfile, "\n");              }
3230              }
3231    
3232            /* Show this only if the JIT was set by /S, not by -s. */
3233    
3234            if ((study_options & PCRE_STUDY_JIT_COMPILE) != 0)
3235              {
3236              int jit;
3237              if (new_info(re, extra, PCRE_INFO_JIT, &jit) == 0)
3238                {
3239                if (jit)
3240                  fprintf(outfile, "JIT study was successful\n");
3241                else
3242    #ifdef SUPPORT_JIT
3243                  fprintf(outfile, "JIT study was not successful\n");
3244    #else
3245                  fprintf(outfile, "JIT support is not available in this version of PCRE\n");
3246    #endif
3247              }              }
3248            }            }
3249          }          }
# Line 1190  while (!done) Line 3262  while (!done)
3262          }          }
3263        else        else
3264          {          {
3265          uschar sbuf[8];          pcre_uint8 sbuf[8];
3266          sbuf[0] = (true_size >> 24)  & 255;  
3267          sbuf[1] = (true_size >> 16)  & 255;          if (do_flip) regexflip(re, extra);
3268          sbuf[2] = (true_size >>  8)  & 255;          sbuf[0] = (pcre_uint8)((true_size >> 24) & 255);
3269          sbuf[3] = (true_size)  & 255;          sbuf[1] = (pcre_uint8)((true_size >> 16) & 255);
3270            sbuf[2] = (pcre_uint8)((true_size >>  8) & 255);
3271          sbuf[4] = (true_study_size >> 24)  & 255;          sbuf[3] = (pcre_uint8)((true_size) & 255);
3272          sbuf[5] = (true_study_size >> 16)  & 255;          sbuf[4] = (pcre_uint8)((true_study_size >> 24) & 255);
3273          sbuf[6] = (true_study_size >>  8)  & 255;          sbuf[5] = (pcre_uint8)((true_study_size >> 16) & 255);
3274          sbuf[7] = (true_study_size)  & 255;          sbuf[6] = (pcre_uint8)((true_study_size >>  8) & 255);
3275            sbuf[7] = (pcre_uint8)((true_study_size) & 255);
3276    
3277          if (fwrite(sbuf, 1, 8, f) < 8 ||          if (fwrite(sbuf, 1, 8, f) < 8 ||
3278              fwrite(re, 1, true_size, f) < true_size)              fwrite(re, 1, true_size, f) < true_size)
# Line 1208  while (!done) Line 3281  while (!done)
3281            }            }
3282          else          else
3283            {            {
3284            fprintf(outfile, "Compiled regex written to %s\n", to_file);            fprintf(outfile, "Compiled pattern written to %s\n", to_file);
3285    
3286              /* If there is study data, write it. */
3287    
3288            if (extra != NULL)            if (extra != NULL)
3289              {              {
3290              if (fwrite(extra->study_data, 1, true_study_size, f) <              if (fwrite(extra->study_data, 1, true_study_size, f) <
# Line 1222  while (!done) Line 3298  while (!done)
3298            }            }
3299          fclose(f);          fclose(f);
3300          }          }
3301    
3302          new_free(re);
3303          if (extra != NULL)
3304            {
3305            PCRE_FREE_STUDY(extra);
3306            }
3307          if (locale_set)
3308            {
3309            new_free((void *)tables);
3310            setlocale(LC_CTYPE, "C");
3311            locale_set = 0;
3312            }
3313        continue;  /* With next regex */        continue;  /* With next regex */
3314        }        }
3315      }        /* End of non-POSIX compile */      }        /* End of non-POSIX compile */
# Line 1230  while (!done) Line 3318  while (!done)
3318    
3319    for (;;)    for (;;)
3320      {      {
3321      unsigned char *q;      pcre_uint8 *q;
3322      unsigned char *bptr = dbuffer;      pcre_uint8 *bptr;
3323      int *use_offsets = offsets;      int *use_offsets = offsets;
3324      int use_size_offsets = size_offsets;      int use_size_offsets = size_offsets;
3325      int callout_data = 0;      int callout_data = 0;
3326      int callout_data_set = 0;      int callout_data_set = 0;
3327      int count, c;      int count, c;
3328      int copystrings = 0;      int copystrings = 0;
3329      int find_match_limit = 0;      int find_match_limit = default_find_match_limit;
3330      int getstrings = 0;      int getstrings = 0;
3331      int getlist = 0;      int getlist = 0;
3332      int gmatched = 0;      int gmatched = 0;
3333      int start_offset = 0;      int start_offset = 0;
3334        int start_offset_sign = 1;
3335      int g_notempty = 0;      int g_notempty = 0;
3336        int use_dfa = 0;
3337    
3338      options = 0;      *copynames = 0;
3339        *getnames = 0;
3340    
3341      pcre_callout = callout;      cn16ptr = copynames;
3342        gn16ptr = getnames;
3343        cn8ptr = copynames8;
3344        gn8ptr = getnames8;
3345    
3346        SET_PCRE_CALLOUT(callout);
3347      first_callout = 1;      first_callout = 1;
3348        last_callout_mark = NULL;
3349      callout_extra = 0;      callout_extra = 0;
3350      callout_count = 0;      callout_count = 0;
3351      callout_fail_count = 999999;      callout_fail_count = 999999;
3352      callout_fail_id = -1;      callout_fail_id = -1;
3353      show_malloc = 0;      show_malloc = 0;
3354        options = 0;
3355    
3356      if (infile == stdin) printf("data> ");      if (extra != NULL) extra->flags &=
3357      if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)        ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
3358    
3359        len = 0;
3360        for (;;)
3361        {        {
3362        done = 1;        if (extend_inputline(infile, buffer + len, "data> ") == NULL)
3363        goto CONTINUE;          {
3364            if (len > 0)    /* Reached EOF without hitting a newline */
3365              {
3366              fprintf(outfile, "\n");
3367              break;
3368              }
3369            done = 1;
3370            goto CONTINUE;
3371            }
3372          if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
3373          len = (int)strlen((char *)buffer);
3374          if (buffer[len-1] == '\n') break;
3375        }        }
     if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);  
3376    
     len = (int)strlen((char *)buffer);  
3377      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
3378      buffer[len] = 0;      buffer[len] = 0;
3379      if (len == 0) break;      if (len == 0) break;
# Line 1271  while (!done) Line 3381  while (!done)
3381      p = buffer;      p = buffer;
3382      while (isspace(*p)) p++;      while (isspace(*p)) p++;
3383    
3384      q = dbuffer;      bptr = q = dbuffer;
3385      while ((c = *p++) != 0)      while ((c = *p++) != 0)
3386        {        {
3387        int i = 0;        int i = 0;
3388        int n = 0;        int n = 0;
3389    
3390        if (c == '\\') switch ((c = *p++))        /* In UTF mode, input can be UTF-8, so just copy all non-backslash bytes.
3391          In non-UTF mode, allow the value of the byte to fall through to later,
3392          where values greater than 127 are turned into UTF-8 when running in
3393          16-bit mode. */
3394    
3395          if (c != '\\')
3396            {
3397            if (use_utf)
3398              {
3399              *q++ = c;
3400              continue;
3401              }
3402            }
3403    
3404          /* Handle backslash escapes */
3405    
3406          else switch ((c = *p++))
3407          {          {
3408          case 'a': c =    7; break;          case 'a': c =    7; break;
3409          case 'b': c = '\b'; break;          case 'b': c = '\b'; break;
# Line 1296  while (!done) Line 3422  while (!done)
3422          break;          break;
3423    
3424          case 'x':          case 'x':
   
         /* Handle \x{..} specially - new Perl thing for utf8 */  
   
3425          if (*p == '{')          if (*p == '{')
3426            {            {
3427            unsigned char *pt = p;            pcre_uint8 *pt = p;
3428            c = 0;            c = 0;
3429            while (isxdigit(*(++pt)))  
3430              c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W');            /* We used to have "while (isxdigit(*(++pt)))" here, but it fails
3431              when isxdigit() is a macro that refers to its argument more than
3432              once. This is banned by the C Standard, but apparently happens in at
3433              least one MacOS environment. */
3434    
3435              for (pt++; isxdigit(*pt); pt++)
3436                c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'a' - 10);
3437            if (*pt == '}')            if (*pt == '}')
3438              {              {
             unsigned char buff8[8];  
             int ii, utn;  
             utn = ord2utf8(c, buff8);  
             for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];  
             c = buff8[ii];   /* Last byte */  
3439              p = pt + 1;              p = pt + 1;
3440              break;              break;
3441              }              }
3442            /* Not correct form; fall through */            /* Not correct form for \x{...}; fall through */
3443            }            }
3444    
3445          /* Ordinary \x */          /* \x without {} always defines just one byte in 8-bit mode. This
3446            allows UTF-8 characters to be constructed byte by byte, and also allows
3447            invalid UTF-8 sequences to be made. Just copy the byte in UTF mode.
3448            Otherwise, pass it down to later code so that it can be turned into
3449            UTF-8 when running in 16-bit mode. */
3450    
3451          c = 0;          c = 0;
3452          while (i++ < 2 && isxdigit(*p))          while (i++ < 2 && isxdigit(*p))
3453            {            {
3454            c = c * 16 + tolower(*p) - ((isdigit(*p))? '0' : 'W');            c = c * 16 + tolower(*p) - ((isdigit(*p))? '0' : 'a' - 10);
3455            p++;            p++;
3456            }            }
3457            if (use_utf)
3458              {
3459              *q++ = c;
3460              continue;
3461              }
3462          break;          break;
3463    
3464          case 0:   /* \ followed by EOF allows for an empty line */          case 0:   /* \ followed by EOF allows for an empty line */
# Line 1333  while (!done) Line 3466  while (!done)
3466          continue;          continue;
3467    
3468          case '>':          case '>':
3469            if (*p == '-')
3470              {
3471              start_offset_sign = -1;
3472              p++;
3473              }
3474          while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';          while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';
3475            start_offset *= start_offset_sign;
3476          continue;          continue;
3477    
3478          case 'A':  /* Option setting */          case 'A':  /* Option setting */
# Line 1352  while (!done) Line 3491  while (!done)
3491            }            }
3492          else if (isalnum(*p))          else if (isalnum(*p))
3493            {            {
3494            uschar name[256];            READ_CAPTURE_NAME(p, &cn8ptr, &cn16ptr, re);
           uschar *npp = name;  
           while (isalnum(*p)) *npp++ = *p++;  
           *npp = 0;  
           n = pcre_get_stringnumber(re, (char *)name);  
           if (n < 0)  
             fprintf(outfile, "no parentheses with name \"%s\"\n", name);  
           else copystrings |= 1 << n;  
3495            }            }
3496          else if (*p == '+')          else if (*p == '+')
3497            {            {
# Line 1368  while (!done) Line 3500  while (!done)
3500            }            }
3501          else if (*p == '-')          else if (*p == '-')
3502            {            {
3503            pcre_callout = NULL;            SET_PCRE_CALLOUT(NULL);
3504            p++;            p++;
3505            }            }
3506          else if (*p == '!')          else if (*p == '!')
# Line 1397  while (!done) Line 3529  while (!done)
3529            }            }
3530          continue;          continue;
3531    
3532    #if !defined NODFA
3533            case 'D':
3534    #if !defined NOPOSIX
3535            if (posix || do_posix)
3536              printf("** Can't use dfa matching in POSIX mode: \\D ignored\n");
3537            else
3538    #endif
3539              use_dfa = 1;
3540            continue;
3541    #endif
3542    
3543    #if !defined NODFA
3544            case 'F':
3545            options |= PCRE_DFA_SHORTEST;
3546            continue;
3547    #endif
3548    
3549          case 'G':          case 'G':
3550          if (isdigit(*p))          if (isdigit(*p))
3551            {            {
# Line 1405  while (!done) Line 3554  while (!done)
3554            }            }
3555          else if (isalnum(*p))          else if (isalnum(*p))
3556            {            {
3557            uschar name[256];            READ_CAPTURE_NAME(p, &gn8ptr, &gn16ptr, re);
3558            uschar *npp = name;            }
3559            while (isalnum(*p)) *npp++ = *p++;          continue;
3560            *npp = 0;  
3561            n = pcre_get_stringnumber(re, (char *)name);          case 'J':
3562            if (n < 0)          while(isdigit(*p)) n = n * 10 + *p++ - '0';
3563              fprintf(outfile, "no parentheses with name \"%s\"\n", name);          if (extra != NULL
3564            else getstrings |= 1 << n;              && (extra->flags & PCRE_EXTRA_EXECUTABLE_JIT) != 0
3565                && extra->executable_jit != NULL)
3566              {
3567              if (jit_stack != NULL) PCRE_JIT_STACK_FREE(jit_stack);
3568              jit_stack = PCRE_JIT_STACK_ALLOC(1, n * 1024);
3569              PCRE_ASSIGN_JIT_STACK(extra, jit_callback, jit_stack);
3570            }            }
3571          continue;          continue;
3572    
# Line 1425  while (!done) Line 3579  while (!done)
3579          continue;          continue;
3580    
3581          case 'N':          case 'N':
3582          options |= PCRE_NOTEMPTY;          if ((options & PCRE_NOTEMPTY) != 0)
3583              options = (options & ~PCRE_NOTEMPTY) | PCRE_NOTEMPTY_ATSTART;
3584            else
3585              options |= PCRE_NOTEMPTY;
3586          continue;          continue;
3587    
3588          case 'O':          case 'O':
# Line 1438  while (!done) Line 3595  while (!done)
3595            if (offsets == NULL)            if (offsets == NULL)
3596              {              {
3597              printf("** Failed to get %d bytes of memory for offsets vector\n",              printf("** Failed to get %d bytes of memory for offsets vector\n",
3598                size_offsets_max * sizeof(int));                (int)(size_offsets_max * sizeof(int)));
3599              return 1;              yield = 1;
3600                goto EXIT;
3601              }              }
3602            }            }
3603          use_size_offsets = n;          use_size_offsets = n;
# Line 1447  while (!done) Line 3605  while (!done)
3605          continue;          continue;
3606    
3607          case 'P':          case 'P':
3608          options |= PCRE_PARTIAL;          options |= ((options & PCRE_PARTIAL_SOFT) == 0)?
3609              PCRE_PARTIAL_SOFT : PCRE_PARTIAL_HARD;
3610            continue;
3611    
3612            case 'Q':
3613            while(isdigit(*p)) n = n * 10 + *p++ - '0';
3614            if (extra == NULL)
3615              {
3616              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
3617              extra->flags = 0;
3618              }
3619            extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
3620            extra->match_limit_recursion = n;
3621            continue;
3622    
3623            case 'q':
3624            while(isdigit(*p)) n = n * 10 + *p++ - '0';
3625            if (extra == NULL)
3626              {
3627              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
3628              extra->flags = 0;
3629              }
3630            extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
3631            extra->match_limit = n;
3632            continue;
3633    
3634    #if !defined NODFA
3635            case 'R':
3636            options |= PCRE_DFA_RESTART;
3637          continue;          continue;
3638    #endif
3639    
3640          case 'S':          case 'S':
3641          show_malloc = 1;          show_malloc = 1;
3642          continue;          continue;
3643    
3644            case 'Y':
3645            options |= PCRE_NO_START_OPTIMIZE;
3646            continue;
3647    
3648          case 'Z':          case 'Z':
3649          options |= PCRE_NOTEOL;          options |= PCRE_NOTEOL;
3650          continue;          continue;
# Line 1461  while (!done) Line 3652  while (!done)
3652          case '?':          case '?':
3653          options |= PCRE_NO_UTF8_CHECK;          options |= PCRE_NO_UTF8_CHECK;
3654          continue;          continue;
3655    
3656            case '<':
3657              {
3658              int x = check_newline(p, outfile);
3659              if (x == 0) goto NEXT_DATA;
3660              options |= x;
3661              while (*p++ != '>');
3662              }
3663            continue;
3664            }
3665    
3666          /* We now have a character value in c that may be greater than 255. In
3667          16-bit mode, we always convert characters to UTF-8 so that values greater
3668          than 255 can be passed to non-UTF 16-bit strings. In 8-bit mode we
3669          convert to UTF-8 if we are in UTF mode. Values greater than 127 in UTF
3670          mode must have come from \x{...} or octal constructs because values from
3671          \x.. get this far only in non-UTF mode. */
3672    
3673          if (use_pcre16 || use_utf)
3674            {
3675            pcre_uint8 buff8[8];
3676            int ii, utn;
3677            utn = ord2utf8(c, buff8);
3678            for (ii = 0; ii < utn; ii++) *q++ = buff8[ii];
3679            }
3680          else
3681            {
3682            if (c > 255)
3683              {
3684              fprintf(outfile, "** Character \\x{%x} is greater than 255 "
3685                "and UTF-8 mode is not enabled.\n", c);
3686              fprintf(outfile, "** Truncation will probably give the wrong "
3687                "result.\n");
3688              }
3689            *q++ = c;
3690          }          }
       *q++ = c;  
3691        }        }
3692    
3693        /* Reached end of subject string */
3694    
3695      *q = 0;      *q = 0;
3696      len = q - dbuffer;      len = (int)(q - dbuffer);
3697    
3698        /* Move the data to the end of the buffer so that a read over the end of
3699        the buffer will be seen by valgrind, even if it doesn't cause a crash. If
3700        we are using the POSIX interface, we must include the terminating zero. */
3701    
3702    #if !defined NOPOSIX
3703        if (posix || do_posix)
3704          {
3705          memmove(bptr + buffer_size - len - 1, bptr, len + 1);
3706          bptr += buffer_size - len - 1;
3707          }
3708        else
3709    #endif
3710          {
3711          memmove(bptr + buffer_size - len, bptr, len);
3712          bptr += buffer_size - len;
3713          }
3714    
3715        if ((all_use_dfa || use_dfa) && find_match_limit)
3716          {
3717          printf("**Match limit not relevant for DFA matching: ignored\n");
3718          find_match_limit = 0;
3719          }
3720    
3721      /* Handle matching via the POSIX interface, which does not      /* Handle matching via the POSIX interface, which does not
3722      support timing or playing with the match limit or callout data. */      support timing or playing with the match limit or callout data. */
# Line 1480  while (!done) Line 3731  while (!done)
3731          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);          pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);
3732        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;        if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;
3733        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;        if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;
3734          if ((options & PCRE_NOTEMPTY) != 0) eflags |= REG_NOTEMPTY;
3735    
3736        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);        rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);
3737    
3738        if (rc != 0)        if (rc != 0)
3739          {          {
3740          (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);          (void)regerror(rc, &preg, (char *)buffer, buffer_size);
3741          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);          fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
3742          }          }
3743          else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
3744                  != 0)
3745            {
3746            fprintf(outfile, "Matched with REG_NOSUB\n");
3747            }