/[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 77 by nigel, Sat Feb 24 21:40:45 2007 UTC revision 838 by ph10, Thu Dec 29 18:27:07 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, er, *very* 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  #define PCRE_SPY        /* For Win32 build, import data, not export */  #ifdef SUPPORT_LIBREADLINE
63    #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  /* We need the internal info for displaying the results of pcre_study() and  #else
102  other internal data; pcretest also uses some of the fixed tables, and generally  #include <sys/time.h>          /* These two includes are needed */
103  has "inside information" compared to a program that strictly follows the PCRE  #include <sys/resource.h>      /* for setrlimit(). */
104  API. */  #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"  #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 PRINTOK(c) (locale_set? isprint(c) : PRINTABLE(c))
162    
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 62  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(re, extra, tables) \
263      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(re, extra, tables) \
340      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(re, extra, tables) \
480      if (use_pcre16) \
481        PCRE_PATTERN_TO_HOST_BYTE_ORDER16(re, extra, tables); \
482      else \
483        PCRE_PATTERN_TO_HOST_BYTE_ORDER8(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 70  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 83  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    /* Another buffer is needed translation to 16-bit character strings. It will
600    obtained and extended as required. */
601    
602    #ifdef SUPPORT_PCRE16
603    static int buffer16_size = 0;
604    static pcre_uint16 *buffer16 = NULL;
605    
606    #ifdef SUPPORT_PCRE8
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    #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    #endif /* SUPPORT_PCRE8 */
629    
630    static const pcre_uint16 OP_lengths16[] = { OP_LENGTHS };
631    #endif  /* SUPPORT_PCRE16 */
632    
633    /* If we have 8-bit support, default use_pcre16 to false; if there is also
634    16-bit support, it can be changed by an option. If there is no 8-bit support,
635    there must be 16-bit support, so default it to 1. */
636    
637    #ifdef SUPPORT_PCRE8
638    static int use_pcre16 = 0;
639    #else
640    static int use_pcre16 = 1;
641    #endif
642    
643    /* Textual explanations for runtime error codes */
644    
645    static const char *errtexts[] = {
646      NULL,  /* 0 is no error */
647      NULL,  /* NOMATCH is handled specially */
648      "NULL argument passed",
649      "bad option value",
650      "magic number missing",
651      "unknown opcode - pattern overwritten?",
652      "no more memory",
653      NULL,  /* never returned by pcre_exec() or pcre_dfa_exec() */
654      "match limit exceeded",
655      "callout error code",
656      NULL,  /* BADUTF8/16 is handled specially */
657      NULL,  /* BADUTF8/16 offset is handled specially */
658      NULL,  /* PARTIAL is handled specially */
659      "not used - internal error",
660      "internal error - pattern overwritten?",
661      "bad count value",
662      "item unsupported for DFA matching",
663      "backreference condition or recursion test not supported for DFA matching",
664      "match limit not supported for DFA matching",
665      "workspace size exceeded in DFA matching",
666      "too much recursion for DFA matching",
667      "recursion limit exceeded",
668      "not used - internal error",
669      "invalid combination of newline options",
670      "bad offset value",
671      NULL,  /* SHORTUTF8/16 is handled specially */
672      "nested recursion at the same subject position",
673      "JIT stack limit reached",
674      "pattern compiled in wrong mode: 8-bit/16-bit error"
675    };
676    
677    
678  /*************************************************  /*************************************************
679  *          Read number from string               *  *         Alternate character tables             *
680  *************************************************/  *************************************************/
681    
682  /* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess  /* By default, the "tables" pointer when calling PCRE is set to NULL, thereby
683  around with conditional compilation, just do the job by hand. It is only used  using the default tables of the library. However, the T option can be used to
684  for unpicking the -o argument, so just keep it simple.  select alternate sets of tables, for different kinds of testing. Note also that
685    the L (locale) option also adjusts the tables. */
686    
687    /* This is the set of tables distributed as default with PCRE. It recognizes
688    only ASCII characters. */
689    
690    static const pcre_uint8 tables0[] = {
691    
692    /* This table is a lower casing table. */
693    
694        0,  1,  2,  3,  4,  5,  6,  7,
695        8,  9, 10, 11, 12, 13, 14, 15,
696       16, 17, 18, 19, 20, 21, 22, 23,
697       24, 25, 26, 27, 28, 29, 30, 31,
698       32, 33, 34, 35, 36, 37, 38, 39,
699       40, 41, 42, 43, 44, 45, 46, 47,
700       48, 49, 50, 51, 52, 53, 54, 55,
701       56, 57, 58, 59, 60, 61, 62, 63,
702       64, 97, 98, 99,100,101,102,103,
703      104,105,106,107,108,109,110,111,
704      112,113,114,115,116,117,118,119,
705      120,121,122, 91, 92, 93, 94, 95,
706       96, 97, 98, 99,100,101,102,103,
707      104,105,106,107,108,109,110,111,
708      112,113,114,115,116,117,118,119,
709      120,121,122,123,124,125,126,127,
710      128,129,130,131,132,133,134,135,
711      136,137,138,139,140,141,142,143,
712      144,145,146,147,148,149,150,151,
713      152,153,154,155,156,157,158,159,
714      160,161,162,163,164,165,166,167,
715      168,169,170,171,172,173,174,175,
716      176,177,178,179,180,181,182,183,
717      184,185,186,187,188,189,190,191,
718      192,193,194,195,196,197,198,199,
719      200,201,202,203,204,205,206,207,
720      208,209,210,211,212,213,214,215,
721      216,217,218,219,220,221,222,223,
722      224,225,226,227,228,229,230,231,
723      232,233,234,235,236,237,238,239,
724      240,241,242,243,244,245,246,247,
725      248,249,250,251,252,253,254,255,
726    
727    /* This table is a case flipping table. */
728    
729        0,  1,  2,  3,  4,  5,  6,  7,
730        8,  9, 10, 11, 12, 13, 14, 15,
731       16, 17, 18, 19, 20, 21, 22, 23,
732       24, 25, 26, 27, 28, 29, 30, 31,
733       32, 33, 34, 35, 36, 37, 38, 39,
734       40, 41, 42, 43, 44, 45, 46, 47,
735       48, 49, 50, 51, 52, 53, 54, 55,
736       56, 57, 58, 59, 60, 61, 62, 63,
737       64, 97, 98, 99,100,101,102,103,
738      104,105,106,107,108,109,110,111,
739      112,113,114,115,116,117,118,119,
740      120,121,122, 91, 92, 93, 94, 95,
741       96, 65, 66, 67, 68, 69, 70, 71,
742       72, 73, 74, 75, 76, 77, 78, 79,
743       80, 81, 82, 83, 84, 85, 86, 87,
744       88, 89, 90,123,124,125,126,127,
745      128,129,130,131,132,133,134,135,
746      136,137,138,139,140,141,142,143,
747      144,145,146,147,148,149,150,151,
748      152,153,154,155,156,157,158,159,
749      160,161,162,163,164,165,166,167,
750      168,169,170,171,172,173,174,175,
751      176,177,178,179,180,181,182,183,
752      184,185,186,187,188,189,190,191,
753      192,193,194,195,196,197,198,199,
754      200,201,202,203,204,205,206,207,
755      208,209,210,211,212,213,214,215,
756      216,217,218,219,220,221,222,223,
757      224,225,226,227,228,229,230,231,
758      232,233,234,235,236,237,238,239,
759      240,241,242,243,244,245,246,247,
760      248,249,250,251,252,253,254,255,
761    
762    /* This table contains bit maps for various character classes. Each map is 32
763    bytes long and the bits run from the least significant end of each byte. The
764    classes that have their own maps are: space, xdigit, digit, upper, lower, word,
765    graph, print, punct, and cntrl. Other classes are built from combinations. */
766    
767      0x00,0x3e,0x00,0x00,0x01,0x00,0x00,0x00,
768      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
769      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
770      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
771    
772      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
773      0x7e,0x00,0x00,0x00,0x7e,0x00,0x00,0x00,
774      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
775      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
776    
777      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
778      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
779      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
780      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
781    
782      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
783      0xfe,0xff,0xff,0x07,0x00,0x00,0x00,0x00,
784      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
785      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
786    
787      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
788      0x00,0x00,0x00,0x00,0xfe,0xff,0xff,0x07,
789      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
790      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
791    
792      0x00,0x00,0x00,0x00,0x00,0x00,0xff,0x03,
793      0xfe,0xff,0xff,0x87,0xfe,0xff,0xff,0x07,
794      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
795      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
796    
797      0x00,0x00,0x00,0x00,0xfe,0xff,0xff,0xff,
798      0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x7f,
799      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
800      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
801    
802      0x00,0x00,0x00,0x00,0xff,0xff,0xff,0xff,
803      0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x7f,
804      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
805      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
806    
807      0x00,0x00,0x00,0x00,0xfe,0xff,0x00,0xfc,
808      0x01,0x00,0x00,0xf8,0x01,0x00,0x00,0x78,
809      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
810      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
811    
812      0xff,0xff,0xff,0xff,0x00,0x00,0x00,0x00,
813      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x80,
814      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
815      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
816    
817    /* This table identifies various classes of character by individual bits:
818      0x01   white space character
819      0x02   letter
820      0x04   decimal digit
821      0x08   hexadecimal digit
822      0x10   alphanumeric or '_'
823      0x80   regular expression metacharacter or binary zero
824    */
825    
826  Arguments:    0x80,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   0-  7 */
827    str           string to be converted    0x00,0x01,0x01,0x00,0x01,0x01,0x00,0x00, /*   8- 15 */
828    endptr        where to put the end pointer    0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  16- 23 */
829      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*  24- 31 */
830      0x01,0x00,0x00,0x00,0x80,0x00,0x00,0x00, /*    - '  */
831      0x80,0x80,0x80,0x80,0x00,0x00,0x80,0x00, /*  ( - /  */
832      0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c, /*  0 - 7  */
833      0x1c,0x1c,0x00,0x00,0x00,0x00,0x00,0x80, /*  8 - ?  */
834      0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  @ - G  */
835      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  H - O  */
836      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  P - W  */
837      0x12,0x12,0x12,0x80,0x80,0x00,0x80,0x10, /*  X - _  */
838      0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /*  ` - g  */
839      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  h - o  */
840      0x12,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /*  p - w  */
841      0x12,0x12,0x12,0x80,0x80,0x00,0x00,0x00, /*  x -127 */
842      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 128-135 */
843      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 136-143 */
844      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144-151 */
845      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 152-159 */
846      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160-167 */
847      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 168-175 */
848      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 176-183 */
849      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
850      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 192-199 */
851      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 200-207 */
852      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 208-215 */
853      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 216-223 */
854      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 224-231 */
855      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 232-239 */
856      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 240-247 */
857      0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00};/* 248-255 */
858    
859    /* This is a set of tables that came orginally from a Windows user. It seems to
860    be at least an approximation of ISO 8859. In particular, there are characters
861    greater than 128 that are marked as spaces, letters, etc. */
862    
863    static const pcre_uint8 tables1[] = {
864    0,1,2,3,4,5,6,7,
865    8,9,10,11,12,13,14,15,
866    16,17,18,19,20,21,22,23,
867    24,25,26,27,28,29,30,31,
868    32,33,34,35,36,37,38,39,
869    40,41,42,43,44,45,46,47,
870    48,49,50,51,52,53,54,55,
871    56,57,58,59,60,61,62,63,
872    64,97,98,99,100,101,102,103,
873    104,105,106,107,108,109,110,111,
874    112,113,114,115,116,117,118,119,
875    120,121,122,91,92,93,94,95,
876    96,97,98,99,100,101,102,103,
877    104,105,106,107,108,109,110,111,
878    112,113,114,115,116,117,118,119,
879    120,121,122,123,124,125,126,127,
880    128,129,130,131,132,133,134,135,
881    136,137,138,139,140,141,142,143,
882    144,145,146,147,148,149,150,151,
883    152,153,154,155,156,157,158,159,
884    160,161,162,163,164,165,166,167,
885    168,169,170,171,172,173,174,175,
886    176,177,178,179,180,181,182,183,
887    184,185,186,187,188,189,190,191,
888    224,225,226,227,228,229,230,231,
889    232,233,234,235,236,237,238,239,
890    240,241,242,243,244,245,246,215,
891    248,249,250,251,252,253,254,223,
892    224,225,226,227,228,229,230,231,
893    232,233,234,235,236,237,238,239,
894    240,241,242,243,244,245,246,247,
895    248,249,250,251,252,253,254,255,
896    0,1,2,3,4,5,6,7,
897    8,9,10,11,12,13,14,15,
898    16,17,18,19,20,21,22,23,
899    24,25,26,27,28,29,30,31,
900    32,33,34,35,36,37,38,39,
901    40,41,42,43,44,45,46,47,
902    48,49,50,51,52,53,54,55,
903    56,57,58,59,60,61,62,63,
904    64,97,98,99,100,101,102,103,
905    104,105,106,107,108,109,110,111,
906    112,113,114,115,116,117,118,119,
907    120,121,122,91,92,93,94,95,
908    96,65,66,67,68,69,70,71,
909    72,73,74,75,76,77,78,79,
910    80,81,82,83,84,85,86,87,
911    88,89,90,123,124,125,126,127,
912    128,129,130,131,132,133,134,135,
913    136,137,138,139,140,141,142,143,
914    144,145,146,147,148,149,150,151,
915    152,153,154,155,156,157,158,159,
916    160,161,162,163,164,165,166,167,
917    168,169,170,171,172,173,174,175,
918    176,177,178,179,180,181,182,183,
919    184,185,186,187,188,189,190,191,
920    224,225,226,227,228,229,230,231,
921    232,233,234,235,236,237,238,239,
922    240,241,242,243,244,245,246,215,
923    248,249,250,251,252,253,254,223,
924    192,193,194,195,196,197,198,199,
925    200,201,202,203,204,205,206,207,
926    208,209,210,211,212,213,214,247,
927    216,217,218,219,220,221,222,255,
928    0,62,0,0,1,0,0,0,
929    0,0,0,0,0,0,0,0,
930    32,0,0,0,1,0,0,0,
931    0,0,0,0,0,0,0,0,
932    0,0,0,0,0,0,255,3,
933    126,0,0,0,126,0,0,0,
934    0,0,0,0,0,0,0,0,
935    0,0,0,0,0,0,0,0,
936    0,0,0,0,0,0,255,3,
937    0,0,0,0,0,0,0,0,
938    0,0,0,0,0,0,12,2,
939    0,0,0,0,0,0,0,0,
940    0,0,0,0,0,0,0,0,
941    254,255,255,7,0,0,0,0,
942    0,0,0,0,0,0,0,0,
943    255,255,127,127,0,0,0,0,
944    0,0,0,0,0,0,0,0,
945    0,0,0,0,254,255,255,7,
946    0,0,0,0,0,4,32,4,
947    0,0,0,128,255,255,127,255,
948    0,0,0,0,0,0,255,3,
949    254,255,255,135,254,255,255,7,
950    0,0,0,0,0,4,44,6,
951    255,255,127,255,255,255,127,255,
952    0,0,0,0,254,255,255,255,
953    255,255,255,255,255,255,255,127,
954    0,0,0,0,254,255,255,255,
955    255,255,255,255,255,255,255,255,
956    0,2,0,0,255,255,255,255,
957    255,255,255,255,255,255,255,127,
958    0,0,0,0,255,255,255,255,
959    255,255,255,255,255,255,255,255,
960    0,0,0,0,254,255,0,252,
961    1,0,0,248,1,0,0,120,
962    0,0,0,0,254,255,255,255,
963    0,0,128,0,0,0,128,0,
964    255,255,255,255,0,0,0,0,
965    0,0,0,0,0,0,0,128,
966    255,255,255,255,0,0,0,0,
967    0,0,0,0,0,0,0,0,
968    128,0,0,0,0,0,0,0,
969    0,1,1,0,1,1,0,0,
970    0,0,0,0,0,0,0,0,
971    0,0,0,0,0,0,0,0,
972    1,0,0,0,128,0,0,0,
973    128,128,128,128,0,0,128,0,
974    28,28,28,28,28,28,28,28,
975    28,28,0,0,0,0,0,128,
976    0,26,26,26,26,26,26,18,
977    18,18,18,18,18,18,18,18,
978    18,18,18,18,18,18,18,18,
979    18,18,18,128,128,0,128,16,
980    0,26,26,26,26,26,26,18,
981    18,18,18,18,18,18,18,18,
982    18,18,18,18,18,18,18,18,
983    18,18,18,128,128,0,0,0,
984    0,0,0,0,0,1,0,0,
985    0,0,0,0,0,0,0,0,
986    0,0,0,0,0,0,0,0,
987    0,0,0,0,0,0,0,0,
988    1,0,0,0,0,0,0,0,
989    0,0,18,0,0,0,0,0,
990    0,0,20,20,0,18,0,0,
991    0,20,18,0,0,0,0,0,
992    18,18,18,18,18,18,18,18,
993    18,18,18,18,18,18,18,18,
994    18,18,18,18,18,18,18,0,
995    18,18,18,18,18,18,18,18,
996    18,18,18,18,18,18,18,18,
997    18,18,18,18,18,18,18,18,
998    18,18,18,18,18,18,18,0,
999    18,18,18,18,18,18,18,18
1000    };
1001    
 Returns:        the unsigned long  
 */  
1002    
1003  static int  
1004  get_value(unsigned char *str, unsigned char **endptr)  
1005    #ifndef HAVE_STRERROR
1006    /*************************************************
1007    *     Provide strerror() for non-ANSI libraries  *
1008    *************************************************/
1009    
1010    /* Some old-fashioned systems still around (e.g. SunOS4) don't have strerror()
1011    in their libraries, but can provide the same facility by this simple
1012    alternative function. */
1013    
1014    extern int   sys_nerr;
1015    extern char *sys_errlist[];
1016    
1017    char *
1018    strerror(int n)
1019  {  {
1020  int result = 0;  if (n < 0 || n >= sys_nerr) return "unknown error number";
1021  while(*str != 0 && isspace(*str)) str++;  return sys_errlist[n];
 while (isdigit(*str)) result = result * 10 + (int)(*str++ - '0');  
 *endptr = str;  
 return(result);  
1022  }  }
1023    #endif /* HAVE_STRERROR */
1024    
1025    
1026    /*************************************************
1027    *         JIT memory callback                    *
1028    *************************************************/
1029    
1030    static pcre_jit_stack* jit_callback(void *arg)
1031    {
1032    return (pcre_jit_stack *)arg;
1033    }
1034    
1035    
1036    #if !defined NOUTF || defined SUPPORT_PCRE16
1037  /*************************************************  /*************************************************
1038  *            Convert UTF-8 string to value       *  *            Convert UTF-8 string to value       *
1039  *************************************************/  *************************************************/
# Line 128  return(result); Line 1042  return(result);
1042  and returns the value of the character.  and returns the value of the character.
1043    
1044  Argument:  Argument:
1045    buffer   a pointer to the byte vector    utf8bytes   a pointer to the byte vector
1046    vptr     a pointer to an int to receive the value    vptr        a pointer to an int to receive the value
1047    
1048  Returns:   >  0 => the number of bytes consumed  Returns:      >  0 => the number of bytes consumed
1049             -6 to 0 => malformed UTF-8 character at offset = (-return)                -6 to 0 => malformed UTF-8 character at offset = (-return)
1050  */  */
1051    
1052  static int  static int
1053  utf82ord(unsigned char *buffer, int *vptr)  utf82ord(pcre_uint8 *utf8bytes, int *vptr)
1054  {  {
1055  int c = *buffer++;  int c = *utf8bytes++;
1056  int d = c;  int d = c;
1057  int i, j, s;  int i, j, s;
1058    
# Line 154  if (i == 0 || i == 6) return 0;        / Line 1068  if (i == 0 || i == 6) return 0;        /
1068  /* i now has a value in the range 1-5 */  /* i now has a value in the range 1-5 */
1069    
1070  s = 6*i;  s = 6*i;
1071  d = (c & _pcre_utf8_table3[i]) << s;  d = (c & utf8_table3[i]) << s;
1072    
1073  for (j = 0; j < i; j++)  for (j = 0; j < i; j++)
1074    {    {
1075    c = *buffer++;    c = *utf8bytes++;
1076    if ((c & 0xc0) != 0x80) return -(j+1);    if ((c & 0xc0) != 0x80) return -(j+1);
1077    s -= 6;    s -= 6;
1078    d |= (c & 0x3f) << s;    d |= (c & 0x3f) << s;
# Line 166  for (j = 0; j < i; j++) Line 1080  for (j = 0; j < i; j++)
1080    
1081  /* Check that encoding was the correct unique one */  /* Check that encoding was the correct unique one */
1082    
1083  for (j = 0; j < _pcre_utf8_table1_size; j++)  for (j = 0; j < utf8_table1_size; j++)
1084    if (d <= _pcre_utf8_table1[j]) break;    if (d <= utf8_table1[j]) break;
1085  if (j != i) return -(i+1);  if (j != i) return -(i+1);
1086    
1087  /* Valid value */  /* Valid value */
# Line 175  if (j != i) return -(i+1); Line 1089  if (j != i) return -(i+1);
1089  *vptr = d;  *vptr = d;
1090  return i+1;  return i+1;
1091  }  }
1092    #endif /* NOUTF || SUPPORT_PCRE16 */
1093    
1094    
1095    
1096    #if !defined NOUTF || defined SUPPORT_PCRE16
1097    /*************************************************
1098    *       Convert character value to UTF-8         *
1099    *************************************************/
1100    
1101    /* This function takes an integer value in the range 0 - 0x7fffffff
1102    and encodes it as a UTF-8 character in 0 to 6 bytes.
1103    
1104    Arguments:
1105      cvalue     the character value
1106      utf8bytes  pointer to buffer for result - at least 6 bytes long
1107    
1108    Returns:     number of characters placed in the buffer
1109    */
1110    
1111    static int
1112    ord2utf8(int cvalue, pcre_uint8 *utf8bytes)
1113    {
1114    register int i, j;
1115    for (i = 0; i < utf8_table1_size; i++)
1116      if (cvalue <= utf8_table1[i]) break;
1117    utf8bytes += i;
1118    for (j = i; j > 0; j--)
1119     {
1120     *utf8bytes-- = 0x80 | (cvalue & 0x3f);
1121     cvalue >>= 6;
1122     }
1123    *utf8bytes = utf8_table2[i] | cvalue;
1124    return i + 1;
1125    }
1126    #endif /* NOUTF || SUPPORT_PCRE16 */
1127    
1128    
1129    
1130    #ifdef SUPPORT_PCRE16
1131    /*************************************************
1132    *         Convert a string to 16-bit             *
1133    *************************************************/
1134    
1135    /* In non-UTF mode, the space needed for a 16-bit string is exactly double the
1136    8-bit size. For a UTF-8 string, the size needed for UTF-16 is no more than
1137    double, because up to 0xffff uses no more than 3 bytes in UTF-8 but possibly 4
1138    in UTF-16. Higher values use 4 bytes in UTF-8 and up to 4 bytes in UTF-16. The
1139    result is always left in buffer16.
1140    
1141    Note that this function does not object to surrogate values. This is
1142    deliberate; it makes it possible to construct UTF-16 strings that are invalid,
1143    for the purpose of testing that they are correctly faulted.
1144    
1145    Patterns to be converted are either plain ASCII or UTF-8; data lines are always
1146    in UTF-8 so that values greater than 255 can be handled.
1147    
1148    Arguments:
1149      data       TRUE if converting a data line; FALSE for a regex
1150      p          points to a byte string
1151      utf        true if UTF-8 (to be converted to UTF-16)
1152      len        number of bytes in the string (excluding trailing zero)
1153    
1154    Returns:     number of 16-bit data items used (excluding trailing zero)
1155                 OR -1 if a UTF-8 string is malformed
1156                 OR -2 if a value > 0x10ffff is encountered
1157                 OR -3 if a value > 0xffff is encountered when not in UTF mode
1158    */
1159    
1160    static int
1161    to16(int data, pcre_uint8 *p, int utf, int len)
1162    {
1163    pcre_uint16 *pp;
1164    
1165    if (buffer16_size < 2*len + 2)
1166      {
1167      if (buffer16 != NULL) free(buffer16);
1168      buffer16_size = 2*len + 2;
1169      buffer16 = (pcre_uint16 *)malloc(buffer16_size);
1170      if (buffer16 == NULL)
1171        {
1172        fprintf(stderr, "pcretest: malloc(%d) failed for buffer16\n", buffer16_size);
1173        exit(1);
1174        }
1175      }
1176    
1177    pp = buffer16;
1178    
1179    if (!utf && !data)
1180      {
1181      while (len-- > 0) *pp++ = *p++;
1182      }
1183    
1184    else
1185      {
1186      int c = 0;
1187      while (len > 0)
1188        {
1189        int chlen = utf82ord(p, &c);
1190        if (chlen <= 0) return -1;
1191        if (c > 0x10ffff) return -2;
1192        p += chlen;
1193        len -= chlen;
1194        if (c < 0x10000) *pp++ = c; else
1195          {
1196          if (!utf) return -3;
1197          c -= 0x10000;
1198          *pp++ = 0xD800 | (c >> 10);
1199          *pp++ = 0xDC00 | (c & 0x3ff);
1200          }
1201        }
1202      }
1203    
1204    *pp = 0;
1205    return pp - buffer16;
1206    }
1207    #endif
1208    
1209    
1210    /*************************************************
1211    *        Read or extend an input line            *
1212    *************************************************/
1213    
1214    /* Input lines are read into buffer, but both patterns and data lines can be
1215    continued over multiple input lines. In addition, if the buffer fills up, we
1216    want to automatically expand it so as to be able to handle extremely large
1217    lines that are needed for certain stress tests. When the input buffer is
1218    expanded, the other two buffers must also be expanded likewise, and the
1219    contents of pbuffer, which are a copy of the input for callouts, must be
1220    preserved (for when expansion happens for a data line). This is not the most
1221    optimal way of handling this, but hey, this is just a test program!
1222    
1223    Arguments:
1224      f            the file to read
1225      start        where in buffer to start (this *must* be within buffer)
1226      prompt       for stdin or readline()
1227    
1228    Returns:       pointer to the start of new data
1229                   could be a copy of start, or could be moved
1230                   NULL if no data read and EOF reached
1231    */
1232    
1233    static pcre_uint8 *
1234    extend_inputline(FILE *f, pcre_uint8 *start, const char *prompt)
1235    {
1236    pcre_uint8 *here = start;
1237    
1238    for (;;)
1239      {
1240      int rlen = (int)(buffer_size - (here - buffer));
1241    
1242      if (rlen > 1000)
1243        {
1244        int dlen;
1245    
1246        /* If libreadline support is required, use readline() to read a line if the
1247        input is a terminal. Note that readline() removes the trailing newline, so
1248        we must put it back again, to be compatible with fgets(). */
1249    
1250    #ifdef SUPPORT_LIBREADLINE
1251        if (isatty(fileno(f)))
1252          {
1253          size_t len;
1254          char *s = readline(prompt);
1255          if (s == NULL) return (here == start)? NULL : start;
1256          len = strlen(s);
1257          if (len > 0) add_history(s);
1258          if (len > rlen - 1) len = rlen - 1;
1259          memcpy(here, s, len);
1260          here[len] = '\n';
1261          here[len+1] = 0;
1262          free(s);
1263          }
1264        else
1265    #endif
1266    
1267        /* Read the next line by normal means, prompting if the file is stdin. */
1268    
1269          {
1270          if (f == stdin) printf("%s", prompt);
1271          if (fgets((char *)here, rlen,  f) == NULL)
1272            return (here == start)? NULL : start;
1273          }
1274    
1275        dlen = (int)strlen((char *)here);
1276        if (dlen > 0 && here[dlen - 1] == '\n') return start;
1277        here += dlen;
1278        }
1279    
1280      else
1281        {
1282        int new_buffer_size = 2*buffer_size;
1283        pcre_uint8 *new_buffer = (pcre_uint8 *)malloc(new_buffer_size);
1284        pcre_uint8 *new_dbuffer = (pcre_uint8 *)malloc(new_buffer_size);
1285        pcre_uint8 *new_pbuffer = (pcre_uint8 *)malloc(new_buffer_size);
1286    
1287        if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
1288          {
1289          fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
1290          exit(1);
1291          }
1292    
1293        memcpy(new_buffer, buffer, buffer_size);
1294        memcpy(new_pbuffer, pbuffer, buffer_size);
1295    
1296        buffer_size = new_buffer_size;
1297    
1298        start = new_buffer + (start - buffer);
1299        here = new_buffer + (here - buffer);
1300    
1301        free(buffer);
1302        free(dbuffer);
1303        free(pbuffer);
1304    
1305        buffer = new_buffer;
1306        dbuffer = new_dbuffer;
1307        pbuffer = new_pbuffer;
1308        }
1309      }
1310    
1311    return NULL;  /* Control never gets here */
1312    }
1313    
1314    
1315    
1316    /*************************************************
1317    *          Read number from string               *
1318    *************************************************/
1319    
1320    /* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess
1321    around with conditional compilation, just do the job by hand. It is only used
1322    for unpicking arguments, so just keep it simple.
1323    
1324    Arguments:
1325      str           string to be converted
1326      endptr        where to put the end pointer
1327    
1328    Returns:        the unsigned long
1329    */
1330    
1331    static int
1332    get_value(pcre_uint8 *str, pcre_uint8 **endptr)
1333    {
1334    int result = 0;
1335    while(*str != 0 && isspace(*str)) str++;
1336    while (isdigit(*str)) result = result * 10 + (int)(*str++ - '0');
1337    *endptr = str;
1338    return(result);
1339    }
1340    
1341    
1342    
1343    /*************************************************
1344    *             Print one character                *
1345    *************************************************/
1346    
1347    /* Print a single character either literally, or as a hex escape. */
1348    
1349    static int pchar(int c, FILE *f)
1350    {
1351    if (PRINTOK(c))
1352      {
1353      if (f != NULL) fprintf(f, "%c", c);
1354      return 1;
1355      }
1356    
1357    if (c < 0x100)
1358      {
1359      if (use_utf)
1360        {
1361        if (f != NULL) fprintf(f, "\\x{%02x}", c);
1362        return 6;
1363        }
1364      else
1365        {
1366        if (f != NULL) fprintf(f, "\\x%02x", c);
1367        return 4;
1368        }
1369      }
1370    
1371    if (f != NULL) fprintf(f, "\\x{%02x}", c);
1372    return (c <= 0x000000ff)? 6 :
1373           (c <= 0x00000fff)? 7 :
1374           (c <= 0x0000ffff)? 8 :
1375           (c <= 0x000fffff)? 9 : 10;
1376    }
1377    
1378    
1379    
1380    #ifdef SUPPORT_PCRE8
1381  /*************************************************  /*************************************************
1382  *             Print character string             *  *         Print 8-bit character string           *
1383  *************************************************/  *************************************************/
1384    
1385  /* Character string printing function. Must handle UTF-8 strings in utf8  /* Must handle UTF-8 strings in utf8 mode. Yields number of characters printed.
1386  mode. Yields number of characters printed. If handed a NULL file, just counts  If handed a NULL file, just counts chars without printing. */
 chars without printing. */  
1387    
1388  static int pchars(unsigned char *p, int length, FILE *f)  static int pchars(pcre_uint8 *p, int length, FILE *f)
1389  {  {
1390  int c;  int c = 0;
1391  int yield = 0;  int yield = 0;
1392    
1393    if (length < 0)
1394      length = strlen((char *)p);
1395    
1396  while (length-- > 0)  while (length-- > 0)
1397    {    {
1398    if (use_utf8)  #if !defined NOUTF
1399      if (use_utf)
1400      {      {
1401      int rc = utf82ord(p, &c);      int rc = utf82ord(p, &c);
   
1402      if (rc > 0 && rc <= length + 1)   /* Mustn't run over the end */      if (rc > 0 && rc <= length + 1)   /* Mustn't run over the end */
1403        {        {
1404        length -= rc - 1;        length -= rc - 1;
1405        p += rc;        p += rc;
1406        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;  
         }  
1407        continue;        continue;
1408        }        }
1409      }      }
1410    #endif
1411      c = *p++;
1412      yield += pchar(c, f);
1413      }
1414    
1415    return yield;
1416    }
1417    #endif
1418    
    /* Not UTF-8, or malformed UTF-8  */  
1419    
1420    if (isprint(c = *(p++)))  
1421      {  #ifdef SUPPORT_PCRE16
1422      if (f != NULL) fprintf(f, "%c", c);  /*************************************************
1423      yield++;  *    Find length of 0-terminated 16-bit string   *
1424      }  *************************************************/
1425    else  
1426    static int strlen16(PCRE_SPTR16 p)
1427    {
1428    int len = 0;
1429    while (*p++ != 0) len++;
1430    return len;
1431    }
1432    #endif  /* SUPPORT_PCRE16 */
1433    
1434    
1435    #ifdef SUPPORT_PCRE16
1436    /*************************************************
1437    *           Print 16-bit character string        *
1438    *************************************************/
1439    
1440    /* Must handle UTF-16 strings in utf mode. Yields number of characters printed.
1441    If handed a NULL file, just counts chars without printing. */
1442    
1443    static int pchars16(PCRE_SPTR16 p, int length, FILE *f)
1444    {
1445    int yield = 0;
1446    
1447    if (length < 0)
1448      length = strlen16(p);
1449    
1450    while (length-- > 0)
1451      {
1452      int c = *p++ & 0xffff;
1453    #if !defined NOUTF
1454      if (use_utf && c >= 0xD800 && c < 0xDC00 && length > 0)
1455      {      {
1456      if (f != NULL) fprintf(f, "\\x%02x", c);      int d = *p & 0xffff;
1457      yield += 4;      if (d >= 0xDC00 && d < 0xDFFF)
1458          {
1459          c = ((c & 0x3ff) << 10) + (d & 0x3ff) + 0x10000;
1460          length--;
1461          p++;
1462          }
1463      }      }
1464    #endif
1465      yield += pchar(c, f);
1466    }    }
1467    
1468  return yield;  return yield;
1469  }  }
1470    #endif  /* SUPPORT_PCRE16 */
1471    
1472    
1473    
1474    #ifdef SUPPORT_PCRE8
1475    /*************************************************
1476    *     Read a capture name (8-bit) and check it   *
1477    *************************************************/
1478    
1479    static pcre_uint8 *
1480    read_capture_name8(pcre_uint8 *p, pcre_uint8 **pp, pcre *re)
1481    {
1482    pcre_uint8 *npp = *pp;
1483    while (isalnum(*p)) *npp++ = *p++;
1484    *npp++ = 0;
1485    *npp = 0;
1486    if (pcre_get_stringnumber(re, (char *)(*pp)) < 0)
1487      {
1488      fprintf(outfile, "no parentheses with name \"");
1489      PCHARSV(*pp, 0, -1, outfile);
1490      fprintf(outfile, "\"\n");
1491      }
1492    
1493    *pp = npp;
1494    return p;
1495    }
1496    #endif  /* SUPPORT_PCRE8 */
1497    
1498    
1499    
1500    #ifdef SUPPORT_PCRE16
1501    /*************************************************
1502    *     Read a capture name (16-bit) and check it  *
1503    *************************************************/
1504    
1505    /* Note that the text being read is 8-bit. */
1506    
1507    static pcre_uint8 *
1508    read_capture_name16(pcre_uint8 *p, pcre_uint16 **pp, pcre *re)
1509    {
1510    pcre_uint16 *npp = *pp;
1511    while (isalnum(*p)) *npp++ = *p++;
1512    *npp++ = 0;
1513    *npp = 0;
1514    if (pcre16_get_stringnumber(re, (PCRE_SPTR16)(*pp)) < 0)
1515      {
1516      fprintf(outfile, "no parentheses with name \"");
1517      PCHARSV(*pp, 0, -1, outfile);
1518      fprintf(outfile, "\"\n");
1519      }
1520    *pp = npp;
1521    return p;
1522    }
1523    #endif  /* SUPPORT_PCRE16 */
1524    
1525    
1526    
# Line 260  if (callout_extra) Line 1549  if (callout_extra)
1549      else      else
1550        {        {
1551        fprintf(f, "%2d: ", i/2);        fprintf(f, "%2d: ", i/2);
1552        (void)pchars((unsigned char *)cb->subject + cb->offset_vector[i],        PCHARSV(cb->subject, cb->offset_vector[i],
1553          cb->offset_vector[i+1] - cb->offset_vector[i], f);          cb->offset_vector[i+1] - cb->offset_vector[i], f);
1554        fprintf(f, "\n");        fprintf(f, "\n");
1555        }        }
# Line 273  printed lengths of the substrings. */ Line 1562  printed lengths of the substrings. */
1562    
1563  if (f != NULL) fprintf(f, "--->");  if (f != NULL) fprintf(f, "--->");
1564    
1565  pre_start = pchars((unsigned char *)cb->subject, cb->start_match, f);  PCHARS(pre_start, cb->subject, 0, cb->start_match, f);
1566  post_start = pchars((unsigned char *)(cb->subject + cb->start_match),  PCHARS(post_start, cb->subject, cb->start_match,
1567    cb->current_position - cb->start_match, f);    cb->current_position - cb->start_match, f);
1568    
1569  subject_length = pchars((unsigned char *)cb->subject, cb->subject_length, NULL);  PCHARS(subject_length, cb->subject, 0, cb->subject_length, NULL);
1570    
1571  (void)pchars((unsigned char *)(cb->subject + cb->current_position),  PCHARSV(cb->subject, cb->current_position,
1572    cb->subject_length - cb->current_position, f);    cb->subject_length - cb->current_position, f);
1573    
1574  if (f != NULL) fprintf(f, "\n");  if (f != NULL) fprintf(f, "\n");
# Line 298  else Line 1587  else
1587      else fprintf(outfile, "%3d ", cb->callout_number);      else fprintf(outfile, "%3d ", cb->callout_number);
1588    }    }
1589    
1590  for (i = 0; i < pre_start; i++) fprintf(outfile, " ");  for (i = 0; i < pre_start; i++) fprintf(outfile, " ");
1591  fprintf(outfile, "^");  fprintf(outfile, "^");
1592    
1593    if (post_start > 0)
1594      {
1595      for (i = 0; i < post_start - 1; i++) fprintf(outfile, " ");
1596      fprintf(outfile, "^");
1597      }
1598    
1599    for (i = 0; i < subject_length - pre_start - post_start + 4; i++)
1600      fprintf(outfile, " ");
1601    
1602    fprintf(outfile, "%.*s", (cb->next_item_length == 0)? 1 : cb->next_item_length,
1603      pbuffer + cb->pattern_position);
1604    
1605    fprintf(outfile, "\n");
1606    first_callout = 0;
1607    
1608    if (cb->mark != last_callout_mark)
1609      {
1610      if (cb->mark == NULL)
1611        fprintf(outfile, "Latest Mark: <unset>\n");
1612      else
1613        {
1614        fprintf(outfile, "Latest Mark: ");
1615        PCHARSV(cb->mark, 0, -1, outfile);
1616        putc('\n', outfile);
1617        }
1618      last_callout_mark = cb->mark;
1619      }
1620    
1621    if (cb->callout_data != NULL)
1622      {
1623      int callout_data = *((int *)(cb->callout_data));
1624      if (callout_data != 0)
1625        {
1626        fprintf(outfile, "Callout data = %d\n", callout_data);
1627        return callout_data;
1628        }
1629      }
1630    
1631    return (cb->callout_number != callout_fail_id)? 0 :
1632           (++callout_count >= callout_fail_count)? 1 : 0;
1633    }
1634    
1635    
1636    /*************************************************
1637    *            Local malloc functions              *
1638    *************************************************/
1639    
1640    /* Alternative malloc function, to test functionality and save the size of a
1641    compiled re, which is the first store request that pcre_compile() makes. The
1642    show_malloc variable is set only during matching. */
1643    
1644    static void *new_malloc(size_t size)
1645    {
1646    void *block = malloc(size);
1647    gotten_store = size;
1648    if (first_gotten_store == 0) first_gotten_store = size;
1649    if (show_malloc)
1650      fprintf(outfile, "malloc       %3d %p\n", (int)size, block);
1651    return block;
1652    }
1653    
1654    static void new_free(void *block)
1655    {
1656    if (show_malloc)
1657      fprintf(outfile, "free             %p\n", block);
1658    free(block);
1659    }
1660    
1661    /* For recursion malloc/free, to test stacking calls */
1662    
1663    static void *stack_malloc(size_t size)
1664    {
1665    void *block = malloc(size);
1666    if (show_malloc)
1667      fprintf(outfile, "stack_malloc %3d %p\n", (int)size, block);
1668    return block;
1669    }
1670    
1671    static void stack_free(void *block)
1672    {
1673    if (show_malloc)
1674      fprintf(outfile, "stack_free       %p\n", block);
1675    free(block);
1676    }
1677    
1678    
1679    /*************************************************
1680    *          Call pcre_fullinfo()                  *
1681    *************************************************/
1682    
1683    /* Get one piece of information from the pcre_fullinfo() function. When only
1684    one of 8-bit or 16-bit is supported, use_pcre16 should always have the correct
1685    value, but the code is defensive.
1686    
1687    Arguments:
1688      re        compiled regex
1689      study     study data
1690      option    PCRE_INFO_xxx option
1691      ptr       where to put the data
1692    
1693    Returns:    0 when OK, < 0 on error
1694    */
1695    
1696    static int
1697    new_info(pcre *re, pcre_extra *study, int option, void *ptr)
1698    {
1699    int rc;
1700    
1701    if (use_pcre16)
1702    #ifdef SUPPORT_PCRE16
1703      rc = pcre16_fullinfo(re, study, option, ptr);
1704    #else
1705      rc = PCRE_ERROR_BADMODE;
1706    #endif
1707    else
1708    #ifdef SUPPORT_PCRE8
1709      rc = pcre_fullinfo(re, study, option, ptr);
1710    #else
1711      rc = PCRE_ERROR_BADMODE;
1712    #endif
1713    
1714    if (rc < 0)
1715      {
1716      fprintf(outfile, "Error %d from pcre%s_fullinfo(%d)\n", rc,
1717        use_pcre16? "16" : "", option);
1718      if (rc == PCRE_ERROR_BADMODE)
1719        fprintf(outfile, "Running in %s-bit mode but pattern was compiled in "
1720          "%s-bit mode\n", use_pcre16? "16":"8", use_pcre16? "8":"16");
1721      }
1722    
1723    return rc;
1724    }
1725    
1726    
1727    
1728    /*************************************************
1729    *             Swap byte functions                *
1730    *************************************************/
1731    
1732    /* The following functions swap the bytes of a pcre_uint16 and pcre_uint32
1733    value, respectively.
1734    
1735    Arguments:
1736      value        any number
1737    
1738    Returns:       the byte swapped value
1739    */
1740    
1741    static pcre_uint32
1742    swap_uint32(pcre_uint32 value)
1743    {
1744    return ((value & 0x000000ff) << 24) |
1745           ((value & 0x0000ff00) <<  8) |
1746           ((value & 0x00ff0000) >>  8) |
1747           (value >> 24);
1748    }
1749    
1750    static pcre_uint16
1751    swap_uint16(pcre_uint16 value)
1752    {
1753    return (value >> 8) | (value << 8);
1754    }
1755    
1756    
1757    
1758    /*************************************************
1759    *        Flip bytes in a compiled pattern        *
1760    *************************************************/
1761    
1762    /* This function is called if the 'F' option was present on a pattern that is
1763    to be written to a file. We flip the bytes of all the integer fields in the
1764    regex data block and the study block. In 16-bit mode this also flips relevant
1765    bytes in the pattern itself. This is to make it possible to test PCRE's
1766    ability to reload byte-flipped patterns, e.g. those compiled on a different
1767    architecture. */
1768    
1769    static void
1770    regexflip(pcre *ere, pcre_extra *extra)
1771    {
1772    real_pcre *re = (real_pcre *)ere;
1773    #ifdef SUPPORT_PCRE16
1774    int op;
1775    pcre_uint16 *ptr = (pcre_uint16 *)re + re->name_table_offset;
1776    int length = re->name_count * re->name_entry_size;
1777    #ifdef SUPPORT_UTF
1778    BOOL utf = (re->options & PCRE_UTF16) != 0;
1779    BOOL utf16_char = FALSE;
1780    #endif /* SUPPORT_UTF */
1781    #endif /* SUPPORT_PCRE16 */
1782    
1783    /* Always flip the bytes in the main data block and study blocks. */
1784    
1785    re->magic_number = REVERSED_MAGIC_NUMBER;
1786    re->size = swap_uint32(re->size);
1787    re->options = swap_uint32(re->options);
1788    re->flags = swap_uint16(re->flags);
1789    re->top_bracket = swap_uint16(re->top_bracket);
1790    re->top_backref = swap_uint16(re->top_backref);
1791    re->first_char = swap_uint16(re->first_char);
1792    re->req_char = swap_uint16(re->req_char);
1793    re->name_table_offset = swap_uint16(re->name_table_offset);
1794    re->name_entry_size = swap_uint16(re->name_entry_size);
1795    re->name_count = swap_uint16(re->name_count);
1796    
1797    if (extra != NULL)
1798      {
1799      pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
1800      rsd->size = swap_uint32(rsd->size);
1801      rsd->flags = swap_uint32(rsd->flags);
1802      rsd->minlength = swap_uint32(rsd->minlength);
1803      }
1804    
1805    /* In 8-bit mode, that is all we need to do. In 16-bit mode we must swap bytes
1806    in the name table, if present, and then in the pattern itself. */
1807    
1808    #ifdef SUPPORT_PCRE16
1809    if (!use_pcre16) return;
1810    
1811    while(TRUE)
1812      {
1813      /* Swap previous characters. */
1814      while (length-- > 0)
1815        {
1816        *ptr = swap_uint16(*ptr);
1817        ptr++;
1818        }
1819    #ifdef SUPPORT_UTF
1820      if (utf16_char)
1821        {
1822        if ((ptr[-1] & 0xfc00) == 0xd800)
1823          {
1824          /* We know that there is only one extra character in UTF-16. */
1825          *ptr = swap_uint16(*ptr);
1826          ptr++;
1827          }
1828        }
1829      utf16_char = FALSE;
1830    #endif /* SUPPORT_UTF */
1831    
1832      /* Get next opcode. */
1833    
1834      length = 0;
1835      op = *ptr;
1836      *ptr++ = swap_uint16(op);
1837    
1838      switch (op)
1839        {
1840        case OP_END:
1841        return;
1842    
1843    #ifdef SUPPORT_UTF
1844        case OP_CHAR:
1845        case OP_CHARI:
1846        case OP_NOT:
1847        case OP_NOTI:
1848        case OP_STAR:
1849        case OP_MINSTAR:
1850        case OP_PLUS:
1851        case OP_MINPLUS:
1852        case OP_QUERY:
1853        case OP_MINQUERY:
1854        case OP_UPTO:
1855        case OP_MINUPTO:
1856        case OP_EXACT:
1857        case OP_POSSTAR:
1858        case OP_POSPLUS:
1859        case OP_POSQUERY:
1860        case OP_POSUPTO:
1861        case OP_STARI:
1862        case OP_MINSTARI:
1863        case OP_PLUSI:
1864        case OP_MINPLUSI:
1865        case OP_QUERYI:
1866        case OP_MINQUERYI:
1867        case OP_UPTOI:
1868        case OP_MINUPTOI:
1869        case OP_EXACTI:
1870        case OP_POSSTARI:
1871        case OP_POSPLUSI:
1872        case OP_POSQUERYI:
1873        case OP_POSUPTOI:
1874        case OP_NOTSTAR:
1875        case OP_NOTMINSTAR:
1876        case OP_NOTPLUS:
1877        case OP_NOTMINPLUS:
1878        case OP_NOTQUERY:
1879        case OP_NOTMINQUERY:
1880        case OP_NOTUPTO:
1881        case OP_NOTMINUPTO:
1882        case OP_NOTEXACT:
1883        case OP_NOTPOSSTAR:
1884        case OP_NOTPOSPLUS:
1885        case OP_NOTPOSQUERY:
1886        case OP_NOTPOSUPTO:
1887        case OP_NOTSTARI:
1888        case OP_NOTMINSTARI:
1889        case OP_NOTPLUSI:
1890        case OP_NOTMINPLUSI:
1891        case OP_NOTQUERYI:
1892        case OP_NOTMINQUERYI:
1893        case OP_NOTUPTOI:
1894        case OP_NOTMINUPTOI:
1895        case OP_NOTEXACTI:
1896        case OP_NOTPOSSTARI:
1897        case OP_NOTPOSPLUSI:
1898        case OP_NOTPOSQUERYI:
1899        case OP_NOTPOSUPTOI:
1900        if (utf) utf16_char = TRUE;
1901    #endif
1902        /* Fall through. */
1903    
1904        default:
1905        length = OP_lengths16[op] - 1;
1906        break;
1907    
1908        case OP_CLASS:
1909        case OP_NCLASS:
1910        /* Skip the character bit map. */
1911        ptr += 32/sizeof(pcre_uint16);
1912        length = 0;
1913        break;
1914    
1915        case OP_XCLASS:
1916        /* Reverse the size of the XCLASS instance. */
1917        ptr++;
1918        *ptr = swap_uint16(*ptr);
1919        if (LINK_SIZE > 1)
1920          {
1921          /* LINK_SIZE can be 1 or 2 in 16 bit mode. */
1922          ptr++;
1923          *ptr = swap_uint16(*ptr);
1924          }
1925        ptr++;
1926    
1927        if (LINK_SIZE > 1)
1928          length = ((ptr[-LINK_SIZE] << 16) | ptr[-LINK_SIZE + 1]) -
1929            (1 + LINK_SIZE + 1);
1930        else
1931          length = ptr[-LINK_SIZE] - (1 + LINK_SIZE + 1);
1932    
1933        op = *ptr;
1934        *ptr = swap_uint16(op);
1935        if ((op & XCL_MAP) != 0)
1936          {
1937          /* Skip the character bit map. */
1938          ptr += 32/sizeof(pcre_uint16);
1939          length -= 32/sizeof(pcre_uint16);
1940          }
1941        break;
1942        }
1943      }
1944    /* Control should never reach here in 16 bit mode. */
1945    #endif /* SUPPORT_PCRE16 */
1946    }
1947    
1948    
 if (post_start > 0)  
   {  
   for (i = 0; i < post_start - 1; i++) fprintf(outfile, " ");  
   fprintf(outfile, "^");  
   }  
1949    
1950  for (i = 0; i < subject_length - pre_start - post_start + 4; i++)  /*************************************************
1951    fprintf(outfile, " ");  *        Check match or recursion limit          *
1952    *************************************************/
1953    
1954  fprintf(outfile, "%.*s", (cb->next_item_length == 0)? 1 : cb->next_item_length,  static int
1955    pbuffer + cb->pattern_position);  check_match_limit(pcre *re, pcre_extra *extra, pcre_uint8 *bptr, int len,
1956      int start_offset, int options, int *use_offsets, int use_size_offsets,
1957      int flag, unsigned long int *limit, int errnumber, const char *msg)
1958    {
1959    int count;
1960    int min = 0;
1961    int mid = 64;
1962    int max = -1;
1963    
1964  fprintf(outfile, "\n");  extra->flags |= flag;
 first_callout = 0;  
1965    
1966  if (cb->callout_data != NULL)  for (;;)
1967    {    {
1968    int callout_data = *((int *)(cb->callout_data));    *limit = mid;
1969    if (callout_data != 0)  
1970      PCRE_EXEC(count, re, extra, bptr, len, start_offset, options,
1971        use_offsets, use_size_offsets);
1972    
1973      if (count == errnumber)
1974      {      {
1975      fprintf(outfile, "Callout data = %d\n", callout_data);      /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
1976      return callout_data;      min = mid;
1977        mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;
1978        }
1979    
1980      else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||
1981                             count == PCRE_ERROR_PARTIAL)
1982        {
1983        if (mid == min + 1)
1984          {
1985          fprintf(outfile, "Minimum %s limit = %d\n", msg, mid);
1986          break;
1987          }
1988        /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
1989        max = mid;
1990        mid = (min + mid)/2;
1991      }      }
1992      else break;    /* Some other error */
1993    }    }
1994    
1995  return (cb->callout_number != callout_fail_id)? 0 :  extra->flags &= ~flag;
1996         (++callout_count >= callout_fail_count)? 1 : 0;  return count;
1997  }  }
1998    
1999    
2000    
2001  /*************************************************  /*************************************************
2002  *            Local malloc functions              *  *         Case-independent strncmp() function    *
2003  *************************************************/  *************************************************/
2004    
2005  /* Alternative malloc function, to test functionality and show the size of the  /*
2006  compiled re. */  Arguments:
2007      s         first string
2008  static void *new_malloc(size_t size)    t         second string
2009  {    n         number of characters to compare
 void *block = malloc(size);  
 gotten_store = size;  
 if (show_malloc)  
   fprintf(outfile, "malloc       %3d %p\n", (int)size, block);  
 return block;  
 }  
   
 static void new_free(void *block)  
 {  
 if (show_malloc)  
   fprintf(outfile, "free             %p\n", block);  
 free(block);  
 }  
   
2010    
2011  /* For recursion malloc/free, to test stacking calls */  Returns:    < 0, = 0, or > 0, according to the comparison
2012    */
2013    
2014  static void *stack_malloc(size_t size)  static int
2015    strncmpic(pcre_uint8 *s, pcre_uint8 *t, int n)
2016  {  {
2017  void *block = malloc(size);  while (n--)
2018  if (show_malloc)    {
2019    fprintf(outfile, "stack_malloc %3d %p\n", (int)size, block);    int c = tolower(*s++) - tolower(*t++);
2020  return block;    if (c) return c;
2021      }
2022    return 0;
2023  }  }
2024    
 static void stack_free(void *block)  
 {  
 if (show_malloc)  
   fprintf(outfile, "stack_free       %p\n", block);  
 free(block);  
 }  
2025    
2026    
2027  /*************************************************  /*************************************************
2028  *          Call pcre_fullinfo()                  *  *         Check newline indicator                *
2029  *************************************************/  *************************************************/
2030    
2031  /* 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
2032    a message and return 0 if there is no match.
2033    
2034    Arguments:
2035      p           points after the leading '<'
2036      f           file for error message
2037    
2038    Returns:      appropriate PCRE_NEWLINE_xxx flags, or 0
2039    */
2040    
2041  static void new_info(pcre *re, pcre_extra *study, int option, void *ptr)  static int
2042    check_newline(pcre_uint8 *p, FILE *f)
2043  {  {
2044  int rc;  if (strncmpic(p, (pcre_uint8 *)"cr>", 3) == 0) return PCRE_NEWLINE_CR;
2045  if ((rc = pcre_fullinfo(re, study, option, ptr)) < 0)  if (strncmpic(p, (pcre_uint8 *)"lf>", 3) == 0) return PCRE_NEWLINE_LF;
2046    fprintf(outfile, "Error %d from pcre_fullinfo(%d)\n", rc, option);  if (strncmpic(p, (pcre_uint8 *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
2047    if (strncmpic(p, (pcre_uint8 *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF;
2048    if (strncmpic(p, (pcre_uint8 *)"any>", 4) == 0) return PCRE_NEWLINE_ANY;
2049    if (strncmpic(p, (pcre_uint8 *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF;
2050    if (strncmpic(p, (pcre_uint8 *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE;
2051    fprintf(f, "Unknown newline type at: <%s\n", p);
2052    return 0;
2053  }  }
2054    
2055    
2056    
2057  /*************************************************  /*************************************************
2058  *         Byte flipping function                 *  *             Usage function                     *
2059  *************************************************/  *************************************************/
2060    
2061  static long int  static void
2062  byteflip(long int value, int n)  usage(void)
2063  {  {
2064  if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);  printf("Usage:     pcretest [options] [<input file> [<output file>]]\n\n");
2065  return ((value & 0x000000ff) << 24) |  printf("Input and output default to stdin and stdout.\n");
2066         ((value & 0x0000ff00) <<  8) |  #ifdef SUPPORT_LIBREADLINE
2067         ((value & 0x00ff0000) >>  8) |  printf("If input is a terminal, readline() is used to read from it.\n");
2068         ((value & 0xff000000) >> 24);  #else
2069    printf("This version of pcretest is not linked with readline().\n");
2070    #endif
2071    printf("\nOptions:\n");
2072    #ifdef SUPPORT_PCRE16
2073    printf("  -16      use 16-bit interface\n");
2074    #endif
2075    printf("  -b       show compiled code (bytecode)\n");
2076    printf("  -C       show PCRE compile-time options and exit\n");
2077    printf("  -C arg   show a specific compile-time option\n");
2078    printf("           and exit with its value. The arg can be:\n");
2079    printf("     linksize     internal link size [2, 3, 4]\n");
2080    printf("     pcre8        8 bit library support enabled [0, 1]\n");
2081    printf("     pcre16       16 bit library support enabled [0, 1]\n");
2082    printf("     utf          Unicode Transformation Format supported [0, 1]\n");
2083    printf("     ucp          Unicode Properties supported [0, 1]\n");
2084    printf("     jit          Just-in-time compiler supported [0, 1]\n");
2085    printf("  -d       debug: show compiled code and information (-b and -i)\n");
2086    #if !defined NODFA
2087    printf("  -dfa     force DFA matching for all subjects\n");
2088    #endif
2089    printf("  -help    show usage information\n");
2090    printf("  -i       show information about compiled patterns\n"
2091           "  -M       find MATCH_LIMIT minimum for each subject\n"
2092           "  -m       output memory used information\n"
2093           "  -o <n>   set size of offsets vector to <n>\n");
2094    #if !defined NOPOSIX
2095    printf("  -p       use POSIX interface\n");
2096    #endif
2097    printf("  -q       quiet: do not output PCRE version number at start\n");
2098    printf("  -S <n>   set stack size to <n> megabytes\n");
2099    printf("  -s       force each pattern to be studied at basic level\n"
2100           "  -s+      force each pattern to be studied, using JIT if available\n"
2101           "  -t       time compilation and execution\n");
2102    printf("  -t <n>   time compilation and execution, repeating <n> times\n");
2103    printf("  -tm      time execution (matching) only\n");
2104    printf("  -tm <n>  time execution (matching) only, repeating <n> times\n");
2105  }  }
2106    
2107    
2108    
   
2109  /*************************************************  /*************************************************
2110  *                Main Program                    *  *                Main Program                    *
2111  *************************************************/  *************************************************/
# Line 416  options, followed by a set of test data, Line 2117  options, followed by a set of test data,
2117  int main(int argc, char **argv)  int main(int argc, char **argv)
2118  {  {
2119  FILE *infile = stdin;  FILE *infile = stdin;
2120    const char *version;
2121  int options = 0;  int options = 0;
2122  int study_options = 0;  int study_options = 0;
2123    int default_find_match_limit = FALSE;
2124  int op = 1;  int op = 1;
2125  int timeit = 0;  int timeit = 0;
2126    int timeitm = 0;
2127  int showinfo = 0;  int showinfo = 0;
2128  int showstore = 0;  int showstore = 0;
2129    int force_study = -1;
2130    int force_study_options = 0;
2131    int quiet = 0;
2132  int size_offsets = 45;  int size_offsets = 45;
2133  int size_offsets_max;  int size_offsets_max;
2134  int *offsets = NULL;  int *offsets = NULL;
# Line 432  int debug = 0; Line 2139  int debug = 0;
2139  int done = 0;  int done = 0;
2140  int all_use_dfa = 0;  int all_use_dfa = 0;
2141  int yield = 0;  int yield = 0;
2142    int stack_size;
2143    
2144    pcre_jit_stack *jit_stack = NULL;
2145    
2146    /* These vectors store, end-to-end, a list of zero-terminated captured
2147    substring names, each list itself being terminated by an empty name. Assume
2148    that 1024 is plenty long enough for the few names we'll be testing. It is
2149    easiest to keep separate 8-bit and 16-bit versions, using the 16-bit version
2150    for the actual memory, to ensure alignment. By defining these variables always
2151    (whether or not 8-bit or 16-bit is supported), we avoid too much mess with
2152    #ifdefs in the code. */
2153    
2154    pcre_uint16 copynames[1024];
2155    pcre_uint16 getnames[1024];
2156    
2157    pcre_uint16 *cn16ptr;
2158    pcre_uint16 *gn16ptr;
2159    
2160    pcre_uint8 *copynames8 = (pcre_uint8 *)copynames;
2161    pcre_uint8 *getnames8 = (pcre_uint8 *)getnames;
2162    pcre_uint8 *cn8ptr;
2163    pcre_uint8 *gn8ptr;
2164    
2165    /* Get buffers from malloc() so that valgrind will check their misuse when
2166    debugging. They grow automatically when very long lines are read. The 16-bit
2167    buffer (buffer16) is obtained only if needed. */
2168    
2169    buffer = (pcre_uint8 *)malloc(buffer_size);
2170    dbuffer = (pcre_uint8 *)malloc(buffer_size);
2171    pbuffer = (pcre_uint8 *)malloc(buffer_size);
2172    
2173  unsigned char *buffer;  /* The outfile variable is static so that new_malloc can use it. */
 unsigned char *dbuffer;  
2174    
2175  /* Get buffers from malloc() so that Electric Fence will check their misuse  outfile = stdout;
 when I am debugging. */  
2176    
2177  buffer = (unsigned char *)malloc(BUFFER_SIZE);  /* The following  _setmode() stuff is some Windows magic that tells its runtime
2178  dbuffer = (unsigned char *)malloc(DBUFFER_SIZE);  library to translate CRLF into a single LF character. At least, that's what
2179  pbuffer = (unsigned char *)malloc(PBUFFER_SIZE);  I've been told: never having used Windows I take this all on trust. Originally
2180    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.  */  
2181    
2182  #if defined(_WIN32) || defined(WIN32)  #if defined(_WIN32) || defined(WIN32)
2183  _setmode( _fileno( stdout ), 0x8000 );  _setmode( _fileno( stdout ), _O_BINARY );
2184  #endif  /* defined(_WIN32) || defined(WIN32) */  #endif
2185    
2186  outfile = stdout;  /* Get the version number: both pcre_version() and pcre16_version() give the
2187    same answer. We just need to ensure that we call one that is available. */
2188    
2189    #ifdef SUPPORT_PCRE8
2190    version = pcre_version();
2191    #else
2192    version = pcre16_version();
2193    #endif
2194    
2195  /* Scan options */  /* Scan options */
2196    
2197  while (argc > 1 && argv[op][0] == '-')  while (argc > 1 && argv[op][0] == '-')
2198    {    {
2199    unsigned char *endptr;    pcre_uint8 *endptr;
2200    
2201    if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)    if (strcmp(argv[op], "-m") == 0) showstore = 1;
2202      showstore = 1;    else if (strcmp(argv[op], "-s") == 0) force_study = 0;
2203    else if (strcmp(argv[op], "-t") == 0) timeit = 1;    else if (strcmp(argv[op], "-s+") == 0)
2204        {
2205        force_study = 1;
2206        force_study_options = PCRE_STUDY_JIT_COMPILE;
2207        }
2208      else if (strcmp(argv[op], "-16") == 0)
2209        {
2210    #ifdef SUPPORT_PCRE16
2211        use_pcre16 = 1;
2212    #else
2213        printf("** This version of PCRE was built without 16-bit support\n");
2214        exit(1);
2215    #endif
2216        }
2217      else if (strcmp(argv[op], "-q") == 0) quiet = 1;
2218      else if (strcmp(argv[op], "-b") == 0) debug = 1;
2219    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;    else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
2220    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;    else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
2221      else if (strcmp(argv[op], "-M") == 0) default_find_match_limit = TRUE;
2222    #if !defined NODFA
2223    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;    else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
2224    #endif
2225    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&    else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&
2226        ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),        ((size_offsets = get_value((pcre_uint8 *)argv[op+1], &endptr)),
2227            *endptr == 0))
2228        {
2229        op++;
2230        argc--;
2231        }
2232      else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
2233        {
2234        int both = argv[op][2] == 0;
2235        int temp;
2236        if (argc > 2 && (temp = get_value((pcre_uint8 *)argv[op+1], &endptr),
2237                         *endptr == 0))
2238          {
2239          timeitm = temp;
2240          op++;
2241          argc--;
2242          }
2243        else timeitm = LOOPREPEAT;
2244        if (both) timeit = timeitm;
2245        }
2246      else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
2247          ((stack_size = get_value((pcre_uint8 *)argv[op+1], &endptr)),
2248          *endptr == 0))          *endptr == 0))
2249      {      {
2250    #if defined(_WIN32) || defined(WIN32) || defined(__minix)
2251        printf("PCRE: -S not supported on this OS\n");
2252        exit(1);
2253    #else
2254        int rc;
2255        struct rlimit rlim;
2256        getrlimit(RLIMIT_STACK, &rlim);
2257        rlim.rlim_cur = stack_size * 1024 * 1024;
2258        rc = setrlimit(RLIMIT_STACK, &rlim);
2259        if (rc != 0)
2260          {
2261        printf("PCRE: setrlimit() failed with error %d\n", rc);
2262        exit(1);
2263          }
2264      op++;      op++;
2265      argc--;      argc--;
2266    #endif
2267      }      }
2268  #if !defined NOPOSIX  #if !defined NOPOSIX
2269    else if (strcmp(argv[op], "-p") == 0) posix = 1;    else if (strcmp(argv[op], "-p") == 0) posix = 1;
# Line 478  while (argc > 1 && argv[op][0] == '-') Line 2271  while (argc > 1 && argv[op][0] == '-')
2271    else if (strcmp(argv[op], "-C") == 0)    else if (strcmp(argv[op], "-C") == 0)
2272      {      {
2273      int rc;      int rc;
2274      printf("PCRE version %s\n", pcre_version());      unsigned long int lrc;
2275    
2276        if (argc > 2)
2277          {
2278          if (strcmp(argv[op + 1], "linksize") == 0)
2279            {
2280            (void)PCRE_CONFIG(PCRE_CONFIG_LINK_SIZE, &rc);
2281            printf("%d\n", rc);
2282            yield = rc;
2283            goto EXIT;
2284            }
2285          if (strcmp(argv[op + 1], "pcre8") == 0)
2286            {
2287    #ifdef SUPPORT_PCRE8
2288            printf("1\n");
2289            yield = 1;
2290    #else
2291            printf("0\n");
2292            yield = 0;
2293    #endif
2294            goto EXIT;
2295            }
2296          if (strcmp(argv[op + 1], "pcre16") == 0)
2297            {
2298    #ifdef SUPPORT_PCRE16
2299            printf("1\n");
2300            yield = 1;
2301    #else
2302            printf("0\n");
2303            yield = 0;
2304    #endif
2305            goto EXIT;
2306            }
2307          if (strcmp(argv[op + 1], "utf") == 0)
2308            {
2309    #ifdef SUPPORT_PCRE8
2310            (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
2311            printf("%d\n", rc);
2312            yield = rc;
2313    #else
2314            (void)pcre16_config(PCRE_CONFIG_UTF16, &rc);
2315            printf("%d\n", rc);
2316            yield = rc;
2317    #endif
2318            goto EXIT;
2319            }
2320          if (strcmp(argv[op + 1], "ucp") == 0)
2321            {
2322            (void)PCRE_CONFIG(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
2323            printf("%d\n", rc);
2324            yield = rc;
2325            goto EXIT;
2326            }
2327          if (strcmp(argv[op + 1], "jit") == 0)
2328            {
2329            (void)PCRE_CONFIG(PCRE_CONFIG_JIT, &rc);
2330            printf("%d\n", rc);
2331            yield = rc;
2332            goto EXIT;
2333            }
2334          if (strcmp(argv[op + 1], "newline") == 0)
2335            {
2336            (void)PCRE_CONFIG(PCRE_CONFIG_NEWLINE, &rc);
2337            /* Note that these values are always the ASCII values, even
2338            in EBCDIC environments. CR is 13 and NL is 10. */
2339            printf("%s\n", (rc == 13)? "CR" :
2340              (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
2341              (rc == -2)? "ANYCRLF" :
2342              (rc == -1)? "ANY" : "???");
2343            goto EXIT;
2344            }
2345          printf("Unknown -C option: %s\n", argv[op + 1]);
2346          goto EXIT;
2347          }
2348    
2349        printf("PCRE version %s\n", version);
2350      printf("Compiled with\n");      printf("Compiled with\n");
2351    
2352    /* At least one of SUPPORT_PCRE8 and SUPPORT_PCRE16 will be set. If both
2353    are set, either both UTFs are supported or both are not supported. */
2354    
2355    #if defined SUPPORT_PCRE8 && defined SUPPORT_PCRE16
2356        printf("  8-bit and 16-bit support\n");
2357        (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
2358        if (rc)
2359          printf("  UTF-8 and UTF-16 support\n");
2360        else
2361          printf("  No UTF-8 or UTF-16 support\n");
2362    #elif defined SUPPORT_PCRE8
2363        printf("  8-bit support only\n");
2364      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);      (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
2365      printf("  %sUTF-8 support\n", rc? "" : "No ");      printf("  %sUTF-8 support\n", rc? "" : "No ");
2366      (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);  #else
2367        printf("  16-bit support only\n");
2368        (void)pcre16_config(PCRE_CONFIG_UTF16, &rc);
2369        printf("  %sUTF-16 support\n", rc? "" : "No ");
2370    #endif
2371    
2372        (void)PCRE_CONFIG(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
2373      printf("  %sUnicode properties support\n", rc? "" : "No ");      printf("  %sUnicode properties support\n", rc? "" : "No ");
2374      (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);      (void)PCRE_CONFIG(PCRE_CONFIG_JIT, &rc);
2375      printf("  Newline character is %s\n", (rc == '\r')? "CR" : "LF");      if (rc)
2376      (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);        printf("  Just-in-time compiler support\n");
2377        else
2378          printf("  No just-in-time compiler support\n");
2379        (void)PCRE_CONFIG(PCRE_CONFIG_NEWLINE, &rc);
2380        /* Note that these values are always the ASCII values, even
2381        in EBCDIC environments. CR is 13 and NL is 10. */
2382        printf("  Newline sequence is %s\n", (rc == 13)? "CR" :
2383          (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
2384          (rc == -2)? "ANYCRLF" :
2385          (rc == -1)? "ANY" : "???");
2386        (void)PCRE_CONFIG(PCRE_CONFIG_BSR, &rc);
2387        printf("  \\R matches %s\n", rc? "CR, LF, or CRLF only" :
2388                                         "all Unicode newlines");
2389        (void)PCRE_CONFIG(PCRE_CONFIG_LINK_SIZE, &rc);
2390      printf("  Internal link size = %d\n", rc);      printf("  Internal link size = %d\n", rc);
2391      (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);      (void)PCRE_CONFIG(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
2392      printf("  POSIX malloc threshold = %d\n", rc);      printf("  POSIX malloc threshold = %d\n", rc);
2393      (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc);      (void)PCRE_CONFIG(PCRE_CONFIG_MATCH_LIMIT, &lrc);
2394      printf("  Default match limit = %d\n", rc);      printf("  Default match limit = %ld\n", lrc);
2395      (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);      (void)PCRE_CONFIG(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &lrc);
2396        printf("  Default recursion depth limit = %ld\n", lrc);
2397        (void)PCRE_CONFIG(PCRE_CONFIG_STACKRECURSE, &rc);
2398      printf("  Match recursion uses %s\n", rc? "stack" : "heap");      printf("  Match recursion uses %s\n", rc? "stack" : "heap");
2399      exit(0);      goto EXIT;
2400        }
2401      else if (strcmp(argv[op], "-help") == 0 ||
2402               strcmp(argv[op], "--help") == 0)
2403        {
2404        usage();
2405        goto EXIT;
2406      }      }
2407    else    else
2408      {      {
2409      printf("** Unknown or malformed option %s\n", argv[op]);      printf("** Unknown or malformed option %s\n", argv[op]);
2410      printf("Usage:   pcretest [-d] [-i] [-o <n>] [-p] [-s] [-t] [<input> [<output>]]\n");      usage();
     printf("  -C     show PCRE compile-time options and exit\n");  
     printf("  -d     debug: show compiled code; implies -i\n");  
     printf("  -dfa   force DFA matching for all subjects\n");  
     printf("  -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");  
2411      yield = 1;      yield = 1;
2412      goto EXIT;      goto EXIT;
2413      }      }
# Line 525  offsets = (int *)malloc(size_offsets_max Line 2422  offsets = (int *)malloc(size_offsets_max
2422  if (offsets == NULL)  if (offsets == NULL)
2423    {    {
2424    printf("** Failed to get %d bytes of memory for offsets vector\n",    printf("** Failed to get %d bytes of memory for offsets vector\n",
2425      size_offsets_max * sizeof(int));      (int)(size_offsets_max * sizeof(int)));
2426    yield = 1;    yield = 1;
2427    goto EXIT;    goto EXIT;
2428    }    }
# Line 534  if (offsets == NULL) Line 2431  if (offsets == NULL)
2431    
2432  if (argc > 1)  if (argc > 1)
2433    {    {
2434    infile = fopen(argv[op], "rb");    infile = fopen(argv[op], INPUT_MODE);
2435    if (infile == NULL)    if (infile == NULL)
2436      {      {
2437      printf("** Failed to open %s\n", argv[op]);      printf("** Failed to open %s\n", argv[op]);
# Line 545  if (argc > 1) Line 2442  if (argc > 1)
2442    
2443  if (argc > 2)  if (argc > 2)
2444    {    {
2445    outfile = fopen(argv[op+1], "wb");    outfile = fopen(argv[op+1], OUTPUT_MODE);
2446    if (outfile == NULL)    if (outfile == NULL)
2447      {      {
2448      printf("** Failed to open %s\n", argv[op+1]);      printf("** Failed to open %s\n", argv[op+1]);
# Line 556  if (argc > 2) Line 2453  if (argc > 2)
2453    
2454  /* Set alternative malloc function */  /* Set alternative malloc function */
2455    
2456    #ifdef SUPPORT_PCRE8
2457  pcre_malloc = new_malloc;  pcre_malloc = new_malloc;
2458  pcre_free = new_free;  pcre_free = new_free;
2459  pcre_stack_malloc = stack_malloc;  pcre_stack_malloc = stack_malloc;
2460  pcre_stack_free = stack_free;  pcre_stack_free = stack_free;
2461    #endif
2462    
2463    #ifdef SUPPORT_PCRE16
2464    pcre16_malloc = new_malloc;
2465    pcre16_free = new_free;
2466    pcre16_stack_malloc = stack_malloc;
2467    pcre16_stack_free = stack_free;
2468    #endif
2469    
2470  /* Heading line, then prompt for first regex if stdin */  /* Heading line unless quiet, then prompt for first regex if stdin */
2471    
2472  fprintf(outfile, "PCRE version %s\n\n", pcre_version());  if (!quiet) fprintf(outfile, "PCRE version %s\n\n", version);
2473    
2474  /* Main loop */  /* Main loop */
2475    
# Line 578  while (!done) Line 2484  while (!done)
2484  #endif  #endif
2485    
2486    const char *error;    const char *error;
2487    unsigned char *p, *pp, *ppp;    pcre_uint8 *markptr;
2488    unsigned char *to_file = NULL;    pcre_uint8 *p, *pp, *ppp;
2489    const unsigned char *tables = NULL;    pcre_uint8 *to_file = NULL;
2490      const pcre_uint8 *tables = NULL;
2491    unsigned long int true_size, true_study_size = 0;    unsigned long int true_size, true_study_size = 0;
2492    size_t size, regex_gotten_store;    size_t size, regex_gotten_store;
2493      int do_allcaps = 0;
2494      int do_mark = 0;
2495    int do_study = 0;    int do_study = 0;
2496      int no_force_study = 0;
2497    int do_debug = debug;    int do_debug = debug;
2498    int do_G = 0;    int do_G = 0;
2499    int do_g = 0;    int do_g = 0;
2500    int do_showinfo = showinfo;    int do_showinfo = showinfo;
2501    int do_showrest = 0;    int do_showrest = 0;
2502      int do_showcaprest = 0;
2503    int do_flip = 0;    int do_flip = 0;
2504    int erroroffset, len, delimiter;    int erroroffset, len, delimiter, poffset;
2505    
2506    use_utf8 = 0;    use_utf = 0;
2507      debug_lengths = 1;
2508    
2509    if (infile == stdin) printf("  re> ");    if (extend_inputline(infile, buffer, "  re> ") == NULL) break;
   if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break;  
2510    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);    if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
2511    fflush(outfile);    fflush(outfile);
2512    
# Line 607  while (!done) Line 2518  while (!done)
2518    
2519    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)    if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
2520      {      {
2521      unsigned long int magic;      unsigned long int magic, get_options;
2522      uschar sbuf[8];      pcre_uint8 sbuf[8];
2523      FILE *f;      FILE *f;
2524    
2525      p++;      p++;
# Line 631  while (!done) Line 2542  while (!done)
2542        (sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7];        (sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7];
2543    
2544      re = (real_pcre *)new_malloc(true_size);      re = (real_pcre *)new_malloc(true_size);
2545      regex_gotten_store = gotten_store;      regex_gotten_store = first_gotten_store;
2546    
2547      if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ;      if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ;
2548    
2549      magic = ((real_pcre *)re)->magic_number;      magic = ((real_pcre *)re)->magic_number;
2550      if (magic != MAGIC_NUMBER)      if (magic != MAGIC_NUMBER)
2551        {        {
2552        if (byteflip(magic, sizeof(magic)) == MAGIC_NUMBER)        if (swap_uint32(magic) == MAGIC_NUMBER)
2553          {          {
2554          do_flip = 1;          do_flip = 1;
2555          }          }
# Line 650  while (!done) Line 2561  while (!done)
2561          }          }
2562        }        }
2563    
2564      fprintf(outfile, "Compiled regex%s loaded from %s\n",      fprintf(outfile, "Compiled pattern%s loaded from %s\n",
2565        do_flip? " (byte-inverted)" : "", p);        do_flip? " (byte-inverted)" : "", p);
2566    
2567      /* 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 */  
2568    
2569      if (true_study_size != 0)      if (true_study_size != 0)
2570        {        {
# Line 674  while (!done) Line 2580  while (!done)
2580          {          {
2581          FAIL_READ:          FAIL_READ:
2582          fprintf(outfile, "Failed to read data from %s\n", p);          fprintf(outfile, "Failed to read data from %s\n", p);
2583          if (extra != NULL) new_free(extra);          if (extra != NULL)
2584              {
2585              PCRE_FREE_STUDY(extra);
2586              }
2587          if (re != NULL) new_free(re);          if (re != NULL) new_free(re);
2588          fclose(f);          fclose(f);
2589          continue;          continue;
# Line 684  while (!done) Line 2593  while (!done)
2593        }        }
2594      else fprintf(outfile, "No study data\n");      else fprintf(outfile, "No study data\n");
2595    
2596        /* Flip the necessary bytes. */
2597        if (do_flip)
2598          {
2599          PCRE_PATTERN_TO_HOST_BYTE_ORDER(re, extra, NULL);
2600          }
2601    
2602        /* Need to know if UTF-8 for printing data strings. */
2603    
2604        if (new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options) < 0) continue;
2605        use_utf = (get_options & PCRE_UTF8) != 0;
2606    
2607      fclose(f);      fclose(f);
2608      goto SHOW_INFO;      goto SHOW_INFO;
2609      }      }
2610    
2611    /* 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
2612    the pattern; if is isn't complete, read more. */    the pattern; if it isn't complete, read more. */
2613    
2614    delimiter = *p++;    delimiter = *p++;
2615    
2616    if (isalnum(delimiter) || delimiter == '\\')    if (isalnum(delimiter) || delimiter == '\\')
2617      {      {
2618      fprintf(outfile, "** Delimiter must not be alphameric or \\\n");      fprintf(outfile, "** Delimiter must not be alphanumeric or \\\n");
2619      goto SKIP_DATA;      goto SKIP_DATA;
2620      }      }
2621    
2622    pp = p;    pp = p;
2623      poffset = (int)(p - buffer);
2624    
2625    for(;;)    for(;;)
2626      {      {
# Line 710  while (!done) Line 2631  while (!done)
2631        pp++;        pp++;
2632        }        }
2633      if (*pp != 0) break;      if (*pp != 0) break;
2634        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)  
2635        {        {
2636        fprintf(outfile, "** Unexpected EOF\n");        fprintf(outfile, "** Unexpected EOF\n");
2637        done = 1;        done = 1;
# Line 728  while (!done) Line 2640  while (!done)
2640      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);      if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
2641      }      }
2642    
2643      /* The buffer may have moved while being extended; reset the start of data
2644      pointer to the correct relative point in the buffer. */
2645    
2646      p = buffer + poffset;
2647    
2648    /* If the first character after the delimiter is backslash, make    /* If the first character after the delimiter is backslash, make
2649    the pattern end with backslash. This is purely to provide a way    the pattern end with backslash. This is purely to provide a way
2650    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 757  while (!done) Line 2674  while (!done)
2674        case 's': options |= PCRE_DOTALL; break;        case 's': options |= PCRE_DOTALL; break;
2675        case 'x': options |= PCRE_EXTENDED; break;        case 'x': options |= PCRE_EXTENDED; break;
2676    
2677        case '+': do_showrest = 1; break;        case '+':
2678          if (do_showrest) do_showcaprest = 1; else do_showrest = 1;
2679          break;
2680    
2681          case '=': do_allcaps = 1; break;
2682        case 'A': options |= PCRE_ANCHORED; break;        case 'A': options |= PCRE_ANCHORED; break;
2683          case 'B': do_debug = 1; break;
2684        case 'C': options |= PCRE_AUTO_CALLOUT; break;        case 'C': options |= PCRE_AUTO_CALLOUT; break;
2685        case 'D': do_debug = do_showinfo = 1; break;        case 'D': do_debug = do_showinfo = 1; break;
2686        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;        case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
2687        case 'F': do_flip = 1; break;        case 'F': do_flip = 1; break;
2688        case 'G': do_G = 1; break;        case 'G': do_G = 1; break;
2689        case 'I': do_showinfo = 1; break;        case 'I': do_showinfo = 1; break;
2690          case 'J': options |= PCRE_DUPNAMES; break;
2691          case 'K': do_mark = 1; break;
2692        case 'M': log_store = 1; break;        case 'M': log_store = 1; break;
2693        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;        case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
2694    
# Line 772  while (!done) Line 2696  while (!done)
2696        case 'P': do_posix = 1; break;        case 'P': do_posix = 1; break;
2697  #endif  #endif
2698    
2699        case 'S': do_study = 1; break;        case 'S':
2700          if (do_study == 0)
2701            {
2702            do_study = 1;
2703            if (*pp == '+')
2704              {
2705              study_options |= PCRE_STUDY_JIT_COMPILE;
2706              pp++;
2707              }
2708            }
2709          else
2710            {
2711            do_study = 0;
2712            no_force_study = 1;
2713            }
2714          break;
2715    
2716        case 'U': options |= PCRE_UNGREEDY; break;        case 'U': options |= PCRE_UNGREEDY; break;
2717          case 'W': options |= PCRE_UCP; break;
2718        case 'X': options |= PCRE_EXTRA; break;        case 'X': options |= PCRE_EXTRA; break;
2719        case '8': options |= PCRE_UTF8; use_utf8 = 1; break;        case 'Y': options |= PCRE_NO_START_OPTIMISE; break;
2720          case 'Z': debug_lengths = 0; break;
2721          case '8': options |= PCRE_UTF8; use_utf = 1; break;
2722        case '?': options |= PCRE_NO_UTF8_CHECK; break;        case '?': options |= PCRE_NO_UTF8_CHECK; break;
2723    
2724          case 'T':
2725          switch (*pp++)
2726            {
2727            case '0': tables = tables0; break;
2728            case '1': tables = tables1; break;
2729    
2730            case '\r':
2731            case '\n':
2732            case ' ':
2733            case 0:
2734            fprintf(outfile, "** Missing table number after /T\n");
2735            goto SKIP_DATA;
2736    
2737            default:
2738            fprintf(outfile, "** Bad table number \"%c\" after /T\n", pp[-1]);
2739            goto SKIP_DATA;
2740            }
2741          break;
2742    
2743        case 'L':        case 'L':
2744        ppp = pp;        ppp = pp;
2745        /* The '\r' test here is so that it works on Windows */        /* The '\r' test here is so that it works on Windows. */
2746        while (*ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;        /* The '0' test is just in case this is an unterminated line. */
2747          while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
2748        *ppp = 0;        *ppp = 0;
2749        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)        if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
2750          {          {
2751          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);          fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
2752          goto SKIP_DATA;          goto SKIP_DATA;
2753          }          }
2754        tables = pcre_maketables();        locale_set = 1;
2755          tables = PCRE_MAKETABLES;
2756        pp = ppp;        pp = ppp;
2757        break;        break;
2758    
# Line 799  while (!done) Line 2763  while (!done)
2763        *pp = 0;        *pp = 0;
2764        break;        break;
2765    
2766          case '<':
2767            {
2768            if (strncmpic(pp, (pcre_uint8 *)"JS>", 3) == 0)
2769              {
2770              options |= PCRE_JAVASCRIPT_COMPAT;
2771              pp += 3;
2772              }
2773            else
2774              {
2775              int x = check_newline(pp, outfile);
2776              if (x == 0) goto SKIP_DATA;
2777              options |= x;
2778              while (*pp++ != '>');
2779              }
2780            }
2781          break;
2782    
2783        case '\r':                      /* So that it works in Windows */        case '\r':                      /* So that it works in Windows */
2784        case '\n':        case '\n':
2785        case ' ':        case ' ':
# Line 812  while (!done) Line 2793  while (!done)
2793    
2794    /* Handle compiling via the POSIX interface, which doesn't support the    /* Handle compiling via the POSIX interface, which doesn't support the
2795    timing, showing, or debugging options, nor the ability to pass over    timing, showing, or debugging options, nor the ability to pass over
2796    local character tables. */    local character tables. Neither does it have 16-bit support. */
2797    
2798  #if !defined NOPOSIX  #if !defined NOPOSIX
2799    if (posix || do_posix)    if (posix || do_posix)
# Line 823  while (!done) Line 2804  while (!done)
2804      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;      if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
2805      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;      if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
2806      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;      if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
2807        if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
2808        if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
2809        if ((options & PCRE_UCP) != 0) cflags |= REG_UCP;
2810        if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY;
2811    
2812        first_gotten_store = 0;
2813      rc = regcomp(&preg, (char *)p, cflags);      rc = regcomp(&preg, (char *)p, cflags);
2814    
2815      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
# Line 830  while (!done) Line 2817  while (!done)
2817    
2818      if (rc != 0)      if (rc != 0)
2819        {        {
2820        (void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE);        (void)regerror(rc, &preg, (char *)buffer, buffer_size);
2821        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);        fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
2822        goto SKIP_DATA;        goto SKIP_DATA;
2823        }        }
# Line 842  while (!done) Line 2829  while (!done)
2829  #endif  /* !defined NOPOSIX */  #endif  /* !defined NOPOSIX */
2830    
2831      {      {
2832      if (timeit)      unsigned long int get_options;
2833    
2834        /* In 16-bit mode, convert the input. */
2835    
2836    #ifdef SUPPORT_PCRE16
2837        if (use_pcre16)
2838          {
2839          switch(to16(FALSE, p, options & PCRE_UTF8, (int)strlen((char *)p)))
2840            {
2841            case -1:
2842            fprintf(outfile, "**Failed: invalid UTF-8 string cannot be "
2843              "converted to UTF-16\n");
2844            goto SKIP_DATA;
2845    
2846            case -2:
2847            fprintf(outfile, "**Failed: character value greater than 0x10ffff "
2848              "cannot be converted to UTF-16\n");
2849            goto SKIP_DATA;
2850    
2851            case -3: /* "Impossible error" when to16 is called arg1 FALSE */
2852            fprintf(outfile, "**Failed: character value greater than 0xffff "
2853              "cannot be converted to 16-bit in non-UTF mode\n");
2854            goto SKIP_DATA;
2855    
2856            default:
2857            break;
2858            }
2859          p = (pcre_uint8 *)buffer16;
2860          }
2861    #endif
2862    
2863        /* Compile many times when timing */
2864    
2865        if (timeit > 0)
2866        {        {
2867        register int i;        register int i;
2868        clock_t time_taken;        clock_t time_taken;
2869        clock_t start_time = clock();        clock_t start_time = clock();
2870        for (i = 0; i < LOOPREPEAT; i++)        for (i = 0; i < timeit; i++)
2871          {          {
2872          re = pcre_compile((char *)p, options, &error, &erroroffset, tables);          PCRE_COMPILE(re, p, options, &error, &erroroffset, tables);
2873          if (re != NULL) free(re);          if (re != NULL) free(re);
2874          }          }
2875        time_taken = clock() - start_time;        time_taken = clock() - start_time;
2876        fprintf(outfile, "Compile time %.3f milliseconds\n",        fprintf(outfile, "Compile time %.4f milliseconds\n",
2877          (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /          (((double)time_taken * 1000.0) / (double)timeit) /
2878            (double)CLOCKS_PER_SEC);            (double)CLOCKS_PER_SEC);
2879        }        }
2880    
2881      re = pcre_compile((char *)p, options, &error, &erroroffset, tables);      first_gotten_store = 0;
2882        PCRE_COMPILE(re, p, options, &error, &erroroffset, tables);
2883    
2884      /* Compilation failed; go back for another re, skipping to blank line      /* Compilation failed; go back for another re, skipping to blank line
2885      if non-interactive. */      if non-interactive. */
# Line 871  while (!done) Line 2892  while (!done)
2892          {          {
2893          for (;;)          for (;;)
2894            {            {
2895            if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)            if (extend_inputline(infile, buffer, NULL) == NULL)
2896              {              {
2897              done = 1;              done = 1;
2898              goto CONTINUE;              goto CONTINUE;
# Line 885  while (!done) Line 2906  while (!done)
2906        goto CONTINUE;        goto CONTINUE;
2907        }        }
2908    
2909      /* Compilation succeeded; print data if required. There are now two      /* Compilation succeeded. It is now possible to set the UTF-8 option from
2910      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
2911      returns only limited data. Check that it agrees with the newer one. */      lines. */
2912    
2913      if (log_store)      if (new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options) < 0)
2914        fprintf(outfile, "Memory allocation (code space): %d\n",        goto SKIP_DATA;
2915          (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));  
2916    
2917      /* Extract the size for possible writing before possibly flipping it,      /* Extract the size for possible writing before possibly flipping it,
2918      and remember the store that was got. */      and remember the store that was got. */
2919    
2920      true_size = ((real_pcre *)re)->size;      true_size = ((real_pcre *)re)->size;
2921      regex_gotten_store = gotten_store;      regex_gotten_store = first_gotten_store;
2922    
2923      /* If /S was present, study the regexp to generate additional info to      /* Output code size information if requested */
     help with the matching. */  
2924    
2925      if (do_study)      if (log_store)
2926          fprintf(outfile, "Memory allocation (code space): %d\n",
2927            (int)(first_gotten_store -
2928                  sizeof(real_pcre) -
2929                  ((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size));
2930    
2931        /* If -s or /S was present, study the regex to generate additional info to
2932        help with the matching, unless the pattern has the SS option, which
2933        suppresses the effect of /S (used for a few test patterns where studying is
2934        never sensible). */
2935    
2936        if (do_study || (force_study >= 0 && !no_force_study))
2937        {        {
2938        if (timeit)        if (timeit > 0)
2939          {          {
2940          register int i;          register int i;
2941          clock_t time_taken;          clock_t time_taken;
2942          clock_t start_time = clock();          clock_t start_time = clock();
2943          for (i = 0; i < LOOPREPEAT; i++)          for (i = 0; i < timeit; i++)
2944            extra = pcre_study(re, study_options, &error);            {
2945              PCRE_STUDY(extra, re, study_options | force_study_options, &error);
2946              }
2947          time_taken = clock() - start_time;          time_taken = clock() - start_time;
2948          if (extra != NULL) free(extra);          if (extra != NULL)
2949          fprintf(outfile, "  Study time %.3f milliseconds\n",            {
2950            (((double)time_taken * 1000.0) / (double)LOOPREPEAT) /            PCRE_FREE_STUDY(extra);
2951              }
2952            fprintf(outfile, "  Study time %.4f milliseconds\n",
2953              (((double)time_taken * 1000.0) / (double)timeit) /
2954              (double)CLOCKS_PER_SEC);              (double)CLOCKS_PER_SEC);
2955          }          }
2956        extra = pcre_study(re, study_options, &error);        PCRE_STUDY(extra, re, study_options | force_study_options, &error);
2957        if (error != NULL)        if (error != NULL)
2958          fprintf(outfile, "Failed to study: %s\n", error);          fprintf(outfile, "Failed to study: %s\n", error);
2959        else if (extra != NULL)        else if (extra != NULL)
2960            {
2961          true_study_size = ((pcre_study_data *)(extra->study_data))->size;          true_study_size = ((pcre_study_data *)(extra->study_data))->size;
2962            if (log_store)
2963              {
2964              size_t jitsize;
2965              if (new_info(re, extra, PCRE_INFO_JITSIZE, &jitsize) == 0 &&
2966                  jitsize != 0)
2967                fprintf(outfile, "Memory allocation (JIT code): %d\n", (int)jitsize);
2968              }
2969            }
2970        }        }
2971    
2972      /* 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. */  
2973    
2974      if (do_flip)      if (do_mark)
2975        {        {
2976        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)  
2977          {          {
2978          pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);          extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2979          rsd->size = byteflip(rsd->size, sizeof(rsd->size));          extra->flags = 0;
         rsd->options = byteflip(rsd->options, sizeof(rsd->options));  
2980          }          }
2981          extra->mark = &markptr;
2982          extra->flags |= PCRE_EXTRA_MARK;
2983        }        }
2984    
2985      /* Extract information from the compiled data if required */      /* Extract and display information from the compiled data if required. */
2986    
2987      SHOW_INFO:      SHOW_INFO:
2988    
2989        if (do_debug)
2990          {
2991          fprintf(outfile, "------------------------------------------------------------------\n");
2992          PCRE_PRINTINT(re, outfile, debug_lengths);
2993          }
2994    
2995        /* We already have the options in get_options (see above) */
2996    
2997      if (do_showinfo)      if (do_showinfo)
2998        {        {
2999        unsigned long int get_options, all_options;        unsigned long int all_options;
3000        int old_first_char, old_options, old_count;        int count, backrefmax, first_char, need_char, okpartial, jchanged,
3001        int count, backrefmax, first_char, need_char;          hascrorlf;
3002        int nameentrysize, namecount;        int nameentrysize, namecount;
3003        const uschar *nametable;        const pcre_uint8 *nametable;
   
       if (do_debug)  
         {  
         fprintf(outfile, "------------------------------------------------------------------\n");  
         _pcre_printint(re, outfile);  
         }  
3004    
3005        new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);        if (new_info(re, NULL, PCRE_INFO_SIZE, &size) +
3006        new_info(re, NULL, PCRE_INFO_SIZE, &size);            new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count) +
3007        new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);            new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax) +
3008        new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);            new_info(re, NULL, PCRE_INFO_FIRSTBYTE, &first_char) +
3009        new_info(re, NULL, PCRE_INFO_FIRSTBYTE, &first_char);            new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char) +
3010        new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char);            new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize) +
3011        new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);            new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount) +
3012        new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);            new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable) +
3013        new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);            new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial) +
3014              new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged) +
3015        old_count = pcre_info(re, &old_options, &old_first_char);            new_info(re, NULL, PCRE_INFO_HASCRORLF, &hascrorlf)
3016        if (count < 0) fprintf(outfile,            != 0)
3017          "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);  
         }  
3018    
3019        if (size != regex_gotten_store) fprintf(outfile,        if (size != regex_gotten_store) fprintf(outfile,
3020          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",          "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",
# Line 1014  while (!done) Line 3029  while (!done)
3029          fprintf(outfile, "Named capturing subpatterns:\n");          fprintf(outfile, "Named capturing subpatterns:\n");
3030          while (namecount-- > 0)          while (namecount-- > 0)
3031            {            {
3032            fprintf(outfile, "  %s %*s%3d\n", nametable + 2,  #if defined SUPPORT_PCRE8 && defined SUPPORT_PCRE16
3033              nameentrysize - 3 - (int)strlen((char *)nametable + 2), "",            int imm2_size = use_pcre16 ? 1 : 2;
3034              GET2(nametable, 0));  #else
3035              int imm2_size = IMM2_SIZE;
3036    #endif
3037              int length = (int)STRLEN(nametable + imm2_size);
3038              fprintf(outfile, "  ");
3039              PCHARSV(nametable, imm2_size, length, outfile);
3040              while (length++ < nameentrysize - imm2_size) putc(' ', outfile);
3041    #if defined SUPPORT_PCRE8 && defined SUPPORT_PCRE16
3042              fprintf(outfile, "%3d\n", use_pcre16?
3043                 (int)(((PCRE_SPTR16)nametable)[0])
3044                :((int)nametable[0] << 8) | (int)nametable[1]);
3045              nametable += nameentrysize * (use_pcre16 ? 2 : 1);
3046    #else
3047              fprintf(outfile, "%3d\n", GET2(nametable, 0));
3048    #ifdef SUPPORT_PCRE8
3049            nametable += nameentrysize;            nametable += nameentrysize;
3050    #else
3051              nametable += nameentrysize * 2;
3052    #endif
3053    #endif
3054            }            }
3055          }          }
3056    
3057        /* The NOPARTIAL bit is a private bit in the options, so we have        if (!okpartial) fprintf(outfile, "Partial matching not supported\n");
3058        to fish it out via out back door */        if (hascrorlf) fprintf(outfile, "Contains explicit CR or LF match\n");
3059    
3060        all_options = ((real_pcre *)re)->options;        all_options = ((real_pcre *)re)->options;
3061        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");  
3062    
3063        if (get_options == 0) fprintf(outfile, "No options\n");        if (get_options == 0) fprintf(outfile, "No options\n");
3064          else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s\n",          else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s\n",
3065            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",            ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
3066            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",            ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
3067            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",            ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
3068            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",            ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",
3069            ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",            ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",
3070            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",            ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",
3071              ((get_options & PCRE_BSR_ANYCRLF) != 0)? " bsr_anycrlf" : "",
3072              ((get_options & PCRE_BSR_UNICODE) != 0)? " bsr_unicode" : "",
3073            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",            ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
3074            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",            ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
3075            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",            ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
3076            ((get_options & PCRE_UTF8) != 0)? " utf8" : "",            ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
3077            ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "");            ((get_options & PCRE_UTF8) != 0)? " utf" : "",
3078              ((get_options & PCRE_UCP) != 0)? " ucp" : "",
3079              ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf_check" : "",
3080              ((get_options & PCRE_NO_START_OPTIMIZE) != 0)? " no_start_optimize" : "",
3081              ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
3082    
3083          if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
3084    
3085          switch (get_options & PCRE_NEWLINE_BITS)
3086            {
3087            case PCRE_NEWLINE_CR:
3088            fprintf(outfile, "Forced newline sequence: CR\n");
3089            break;
3090    
3091            case PCRE_NEWLINE_LF:
3092            fprintf(outfile, "Forced newline sequence: LF\n");
3093            break;
3094    
3095            case PCRE_NEWLINE_CRLF:
3096            fprintf(outfile, "Forced newline sequence: CRLF\n");
3097            break;
3098    
3099            case PCRE_NEWLINE_ANYCRLF:
3100            fprintf(outfile, "Forced newline sequence: ANYCRLF\n");
3101            break;
3102    
3103            case PCRE_NEWLINE_ANY:
3104            fprintf(outfile, "Forced newline sequence: ANY\n");
3105            break;
3106    
3107        if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0)          default:
3108          fprintf(outfile, "Case state changes\n");          break;
3109            }
3110    
3111        if (first_char == -1)        if (first_char == -1)
3112          {          {
3113          fprintf(outfile, "First char at start or follows \\n\n");          fprintf(outfile, "First char at start or follows newline\n");
3114          }          }
3115        else if (first_char < 0)        else if (first_char < 0)
3116          {          {
# Line 1060  while (!done) Line 3118  while (!done)
3118          }          }
3119        else        else
3120          {          {
3121          int ch = first_char & 255;          const char *caseless =
3122          const char *caseless = ((first_char & REQ_CASELESS) == 0)?            ((((real_pcre *)re)->flags & PCRE_FCH_CASELESS) == 0)?
3123            "" : " (caseless)";            "" : " (caseless)";
3124          if (isprint(ch))  
3125            fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);          if (PRINTOK(first_char))
3126              fprintf(outfile, "First char = \'%c\'%s\n", first_char, caseless);
3127          else          else
3128            fprintf(outfile, "First char = %d%s\n", ch, caseless);            {
3129              fprintf(outfile, "First char = ");
3130              pchar(first_char, outfile);
3131              fprintf(outfile, "%s\n", caseless);
3132              }
3133          }          }
3134    
3135        if (need_char < 0)        if (need_char < 0)
# Line 1075  while (!done) Line 3138  while (!done)
3138          }          }
3139        else        else
3140          {          {
3141          int ch = need_char & 255;          const char *caseless =
3142          const char *caseless = ((need_char & REQ_CASELESS) == 0)?            ((((real_pcre *)re)->flags & PCRE_RCH_CASELESS) == 0)?
3143            "" : " (caseless)";            "" : " (caseless)";
3144          if (isprint(ch))  
3145            fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);          if (PRINTOK(need_char))
3146              fprintf(outfile, "Need char = \'%c\'%s\n", need_char, caseless);
3147          else          else
3148            fprintf(outfile, "Need char = %d%s\n", ch, caseless);            {
3149              fprintf(outfile, "Need char = ");
3150              pchar(need_char, outfile);
3151              fprintf(outfile, "%s\n", caseless);
3152              }
3153          }          }
3154    
3155        /* 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
3156        value, but it varies, depending on the computer architecture, and        value, but it varies, depending on the computer architecture, and
3157        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
3158        flipped.) */        flipped.) If study was forced by an external -s, don't show this
3159          information unless -i or -d was also present. This means that, except
3160          when auto-callouts are involved, the output from runs with and without
3161          -s should be identical. */
3162    
3163        if (do_study)        if (do_study || (force_study >= 0 && showinfo && !no_force_study))
3164          {          {
3165          if (extra == NULL)          if (extra == NULL)
3166            fprintf(outfile, "Study returned NULL\n");            fprintf(outfile, "Study returned NULL\n");
3167          else          else
3168            {            {
3169            uschar *start_bits = NULL;            pcre_uint8 *start_bits = NULL;
3170            new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);            int minlength;
3171    
3172            if (start_bits == NULL)            if (new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength) == 0)
3173              fprintf(outfile, "No starting byte set\n");              fprintf(outfile, "Subject length lower bound = %d\n", minlength);
3174            else  
3175              if (new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits) == 0)
3176              {              {
3177              int i;              if (start_bits == NULL)
3178              int c = 24;                fprintf(outfile, "No set of starting bytes\n");
3179              fprintf(outfile, "Starting byte set: ");              else
             for (i = 0; i < 256; i++)  
3180                {                {
3181                if ((start_bits[i/8] & (1<<(i&7))) != 0)                int i;
3182                  int c = 24;
3183                  fprintf(outfile, "Starting byte set: ");
3184                  for (i = 0; i < 256; i++)
3185                  {                  {
3186                  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  
3187                    {                    {
3188                    fprintf(outfile, "\\x%02x ", i);                    if (c > 75)
3189                    c += 5;                      {
3190                        fprintf(outfile, "\n  ");
3191                        c = 2;
3192                        }
3193                      if (PRINTOK(i) && i != ' ')
3194                        {
3195                        fprintf(outfile, "%c ", i);
3196                        c += 2;
3197                        }
3198                      else
3199                        {
3200                        fprintf(outfile, "\\x%02x ", i);
3201                        c += 5;
3202                        }
3203                    }                    }
3204                  }                  }
3205                  fprintf(outfile, "\n");
3206                }                }
3207              fprintf(outfile, "\n");              }
3208              }
3209    
3210            /* Show this only if the JIT was set by /S, not by -s. */
3211    
3212            if ((study_options & PCRE_STUDY_JIT_COMPILE) != 0)
3213              {
3214              int jit;
3215              if (new_info(re, extra, PCRE_INFO_JIT, &jit) == 0)
3216                {
3217                if (jit)
3218                  fprintf(outfile, "JIT study was successful\n");
3219                else
3220    #ifdef SUPPORT_JIT
3221                  fprintf(outfile, "JIT study was not successful\n");
3222    #else
3223                  fprintf(outfile, "JIT support is not available in this version of PCRE\n");
3224    #endif
3225              }              }
3226            }            }
3227          }          }
# Line 1145  while (!done) Line 3240  while (!done)
3240          }          }
3241        else        else
3242          {          {
3243          uschar sbuf[8];          pcre_uint8 sbuf[8];
3244          sbuf[0] = (true_size >> 24)  & 255;  
3245          sbuf[1] = (true_size >> 16)  & 255;          if (do_flip) regexflip(re, extra);
3246          sbuf[2] = (true_size >>  8)  & 255;          sbuf[0] = (pcre_uint8)((true_size >> 24) & 255);
3247          sbuf[3] = (true_size)  & 255;          sbuf[1] = (pcre_uint8)((true_size >> 16) & 255);
3248            sbuf[2] = (pcre_uint8)((true_size >>  8) & 255);
3249          sbuf[4] = (true_study_size >> 24)  & 255;          sbuf[3] = (pcre_uint8)((true_size) & 255);
3250          sbuf[5] = (true_study_size >> 16)  & 255;          sbuf[4] = (pcre_uint8)((true_study_size >> 24) & 255);
3251          sbuf[6] = (true_study_size >>  8)  & 255;          sbuf[5] = (pcre_uint8)((true_study_size >> 16) & 255);
3252          sbuf[7] = (true_study_size)  & 255;          sbuf[6] = (pcre_uint8)((true_study_size >>  8) & 255);
3253            sbuf[7] = (pcre_uint8)((true_study_size) & 255);
3254    
3255          if (fwrite(sbuf, 1, 8, f) < 8 ||          if (fwrite(sbuf, 1, 8, f) < 8 ||
3256              fwrite(re, 1, true_size, f) < true_size)              fwrite(re, 1, true_size, f) < true_size)
# Line 1163  while (!done) Line 3259  while (!done)
3259            }            }
3260          else          else
3261            {            {
3262            fprintf(outfile, "Compiled regex written to %s\n", to_file);            fprintf(outfile, "Compiled pattern written to %s\n", to_file);
3263    
3264              /* If there is study data, write it. */
3265    
3266            if (extra != NULL)            if (extra != NULL)
3267              {              {
3268              if (fwrite(extra->study_data, 1, true_study_size, f) <              if (fwrite(extra->study_data, 1, true_study_size, f) <
# Line 1179  while (!done) Line 3278  while (!done)
3278          }          }
3279    
3280        new_free(re);        new_free(re);
3281        if (extra != NULL) new_free(extra);        if (extra != NULL)
3282        if (tables != NULL) new_free((void *)tables);          {
3283            PCRE_FREE_STUDY(extra);
3284            }
3285          if (locale_set)
3286            {
3287            new_free((void *)tables);
3288            setlocale(LC_CTYPE, "C");
3289            locale_set = 0;
3290            }
3291        continue;  /* With next regex */        continue;  /* With next regex */
3292        }        }
3293      }        /* End of non-POSIX compile */      }        /* End of non-POSIX compile */
# Line 1189  while (!done) Line 3296  while (!done)
3296    
3297    for (;;)    for (;;)
3298      {      {
3299      unsigned char *q;      pcre_uint8 *q;
3300      unsigned char *bptr = dbuffer;      pcre_uint8 *bptr;
3301      int *use_offsets = offsets;      int *use_offsets = offsets;
3302      int use_size_offsets = size_offsets;      int use_size_offsets = size_offsets;
3303      int callout_data = 0;      int callout_data = 0;
3304      int callout_data_set = 0;      int callout_data_set = 0;
3305      int count, c;      int count, c;
3306      int copystrings = 0;      int copystrings = 0;
3307      int find_match_limit = 0;      int find_match_limit = default_find_match_limit;
3308      int getstrings = 0;      int getstrings = 0;
3309      int getlist = 0;      int getlist = 0;
3310      int gmatched = 0;      int gmatched = 0;
3311      int start_offset = 0;      int start_offset = 0;
3312        int start_offset_sign = 1;
3313      int g_notempty = 0;      int g_notempty = 0;
3314      int use_dfa = 0;      int use_dfa = 0;
3315    
3316      options = 0;      *copynames = 0;
3317        *getnames = 0;
3318    
3319      pcre_callout = callout;      cn16ptr = copynames;
3320        gn16ptr = getnames;
3321        cn8ptr = copynames8;
3322        gn8ptr = getnames8;
3323    
3324        SET_PCRE_CALLOUT(callout);
3325      first_callout = 1;      first_callout = 1;
3326        last_callout_mark = NULL;
3327      callout_extra = 0;      callout_extra = 0;
3328      callout_count = 0;      callout_count = 0;
3329      callout_fail_count = 999999;      callout_fail_count = 999999;
3330      callout_fail_id = -1;      callout_fail_id = -1;
3331      show_malloc = 0;      show_malloc = 0;
3332        options = 0;
3333    
3334      if (infile == stdin) printf("data> ");      if (extra != NULL) extra->flags &=
3335      if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL)        ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
3336    
3337        len = 0;
3338        for (;;)
3339        {        {
3340        done = 1;        if (extend_inputline(infile, buffer + len, "data> ") == NULL)
3341        goto CONTINUE;          {
3342            if (len > 0)    /* Reached EOF without hitting a newline */
3343              {
3344              fprintf(outfile, "\n");
3345              break;
3346              }
3347            done = 1;
3348            goto CONTINUE;
3349            }
3350          if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
3351          len = (int)strlen((char *)buffer);
3352          if (buffer[len-1] == '\n') break;
3353        }        }
     if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);  
3354    
     len = (int)strlen((char *)buffer);  
3355      while (len > 0 && isspace(buffer[len-1])) len--;      while (len > 0 && isspace(buffer[len-1])) len--;
3356      buffer[len] = 0;      buffer[len] = 0;
3357      if (len == 0) break;      if (len == 0) break;
# Line 1231  while (!done) Line 3359  while (!done)
3359      p = buffer;      p = buffer;
3360      while (isspace(*p)) p++;      while (isspace(*p)) p++;
3361    
3362      q = dbuffer;      bptr = q = dbuffer;
3363      while ((c = *p++) != 0)      while ((c = *p++) != 0)
3364        {        {
3365        int i = 0;        int i = 0;
3366        int n = 0;        int n = 0;
3367    
3368        if (c == '\\') switch ((c = *p++))        /* In UTF mode, input can be UTF-8, so just copy all non-backslash bytes.
3369          In non-UTF mode, allow the value of the byte to fall through to later,
3370          where values greater than 127 are turned into UTF-8 when running in
3371          16-bit mode. */
3372    
3373          if (c != '\\')
3374            {
3375            if (use_utf)
3376              {
3377              *q++ = c;
3378              continue;
3379              }
3380            }
3381    
3382          /* Handle backslash escapes */
3383    
3384          else switch ((c = *p++))
3385          {          {
3386          case 'a': c =    7; break;          case 'a': c =    7; break;
3387          case 'b': c = '\b'; break;          case 'b': c = '\b'; break;
# Line 1256  while (!done) Line 3400  while (!done)
3400          break;          break;
3401    
3402          case 'x':          case 'x':
   
         /* Handle \x{..} specially - new Perl thing for utf8 */  
   
3403          if (*p == '{')          if (*p == '{')
3404            {            {
3405            unsigned char *pt = p;            pcre_uint8 *pt = p;
3406            c = 0;            c = 0;
3407            while (isxdigit(*(++pt)))  
3408              c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W');            /* We used to have "while (isxdigit(*(++pt)))" here, but it fails
3409              when isxdigit() is a macro that refers to its argument more than
3410              once. This is banned by the C Standard, but apparently happens in at
3411              least one MacOS environment. */
3412    
3413              for (pt++; isxdigit(*pt); pt++)
3414                c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'a' - 10);
3415            if (*pt == '}')            if (*pt == '}')
3416              {              {
             unsigned char buff8[8];  
             int ii, utn;  
             utn = _pcre_ord2utf8(c, buff8);  
             for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];  
             c = buff8[ii];   /* Last byte */  
3417              p = pt + 1;              p = pt + 1;
3418              break;              break;
3419              }              }
3420            /* Not correct form; fall through */            /* Not correct form for \x{...}; fall through */
3421            }            }
3422    
3423          /* Ordinary \x */          /* \x without {} always defines just one byte in 8-bit mode. This
3424            allows UTF-8 characters to be constructed byte by byte, and also allows
3425            invalid UTF-8 sequences to be made. Just copy the byte in UTF mode.
3426            Otherwise, pass it down to later code so that it can be turned into
3427            UTF-8 when running in 16-bit mode. */
3428    
3429          c = 0;          c = 0;
3430          while (i++ < 2 && isxdigit(*p))          while (i++ < 2 && isxdigit(*p))
3431            {            {
3432            c = c * 16 + tolower(*p) - ((isdigit(*p))? '0' : 'W');            c = c * 16 + tolower(*p) - ((isdigit(*p))? '0' : 'a' - 10);
3433            p++;            p++;
3434            }            }
3435            if (use_utf)
3436              {
3437              *q++ = c;
3438              continue;
3439              }
3440          break;          break;
3441    
3442          case 0:   /* \ followed by EOF allows for an empty line */          case 0:   /* \ followed by EOF allows for an empty line */
# Line 1293  while (!done) Line 3444  while (!done)
3444          continue;          continue;
3445    
3446          case '>':          case '>':
3447            if (*p == '-')
3448              {
3449              start_offset_sign = -1;
3450              p++;
3451              }
3452          while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';          while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';
3453            start_offset *= start_offset_sign;
3454          continue;          continue;
3455    
3456          case 'A':  /* Option setting */          case 'A':  /* Option setting */
# Line 1312  while (!done) Line 3469  while (!done)
3469            }            }
3470          else if (isalnum(*p))          else if (isalnum(*p))
3471            {            {
3472            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;  
3473            }            }
3474          else if (*p == '+')          else if (*p == '+')
3475            {            {
# Line 1328  while (!done) Line 3478  while (!done)
3478            }            }
3479          else if (*p == '-')          else if (*p == '-')
3480            {            {
3481            pcre_callout = NULL;            SET_PCRE_CALLOUT(NULL);
3482            p++;            p++;
3483            }            }
3484          else if (*p == '!')          else if (*p == '!')
# Line 1357  while (!done) Line 3507  while (!done)
3507            }            }
3508          continue;          continue;
3509    
3510    #if !defined NODFA
3511          case 'D':          case 'D':
3512    #if !defined NOPOSIX
3513          if (posix || do_posix)          if (posix || do_posix)
3514            printf("** Can't use dfa matching in POSIX mode: \\D ignored\n");            printf("** Can't use dfa matching in POSIX mode: \\D ignored\n");
3515          else          else
3516    #endif
3517            use_dfa = 1;            use_dfa = 1;
3518          continue;          continue;
3519    #endif
3520    
3521    #if !defined NODFA
3522          case 'F':          case 'F':
3523          options |= PCRE_DFA_SHORTEST;          options |= PCRE_DFA_SHORTEST;
3524          continue;          continue;
3525    #endif
3526    
3527          case 'G':          case 'G':
3528          if (isdigit(*p))          if (isdigit(*p))
# Line 1376  while (!done) Line 3532  while (!done)
3532            }            }
3533          else if (isalnum(*p))          else if (isalnum(*p))
3534            {            {
3535            uschar name[256];            READ_CAPTURE_NAME(p, &gn8ptr, &gn16ptr, re);
3536            uschar *npp = name;            }
3537            while (isalnum(*p)) *npp++ = *p++;          continue;
3538            *npp = 0;  
3539            n = pcre_get_stringnumber(re, (char *)name);          case 'J':
3540            if (n < 0)          while(isdigit(*p)) n = n * 10 + *p++ - '0';
3541              fprintf(outfile, "no parentheses with name \"%s\"\n", name);          if (extra != NULL
3542            else getstrings |= 1 << n;              && (extra->flags & PCRE_EXTRA_EXECUTABLE_JIT) != 0
3543                && extra->executable_jit != NULL)
3544              {
3545              if (jit_stack != NULL) PCRE_JIT_STACK_FREE(jit_stack);
3546              jit_stack = PCRE_JIT_STACK_ALLOC(1, n * 1024);
3547              PCRE_ASSIGN_JIT_STACK(extra, jit_callback, jit_stack);
3548            }            }
3549          continue;          continue;
3550    
# Line 1396  while (!done) Line 3557  while (!done)
3557          continue;          continue;
3558    
3559          case 'N':          case 'N':
3560          options |= PCRE_NOTEMPTY;          if ((options & PCRE_NOTEMPTY) != 0)
3561              options = (options & ~PCRE_NOTEMPTY) | PCRE_NOTEMPTY_ATSTART;
3562            else
3563              options |= PCRE_NOTEMPTY;
3564          continue;          continue;
3565    
3566          case 'O':          case 'O':
# Line 1409  while (!done) Line 3573  while (!done)
3573            if (offsets == NULL)            if (offsets == NULL)
3574              {              {
3575              printf("** Failed to get %d bytes of memory for offsets vector\n",              printf("** Failed to get %d bytes of memory for offsets vector\n",
3576                size_offsets_max * sizeof(int));                (int)(size_offsets_max * sizeof(int)));
3577              yield = 1;              yield = 1;
3578              goto EXIT;              goto EXIT;
3579              }              }
# Line 1419  while (!done) Line 3583  while (!done)
3583          continue;          continue;
3584    
3585          case 'P':          case 'P':
3586          options |= PCRE_PARTIAL;          options |= ((options & PCRE_PARTIAL_SOFT) == 0)?
3587              PCRE_PARTIAL_SOFT : PCRE_PARTIAL_HARD;
3588            continue;
3589    
3590            case 'Q':
3591            while(isdigit(*p)) n = n * 10 + *p++ - '0';
3592            if (extra == NULL)
3593              {
3594              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
3595              extra->flags = 0;
3596              }
3597            extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
3598            extra->match_limit_recursion = n;
3599            continue;
3600    
3601            case 'q':
3602            while(isdigit(*p)) n = n * 10 + *p++ - '0';
3603            if (extra == NULL)
3604              {
3605              extra = (pcre_extra *)malloc(sizeof(pcre_extra));
3606              extra->flags = 0;
3607              }
3608            extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
3609            extra->match_limit = n;
3610          continue;          continue;
3611    
3612    #if !defined NODFA
3613          case 'R':          case 'R':
3614          options |= PCRE_DFA_RESTART;          options |= PCRE_DFA_RESTART;
3615          continue;          continue;
3616    #endif
3617    
3618          case 'S':          case 'S':
3619          show_malloc = 1;          show_malloc = 1;
3620          continue;          continue;
3621    
3622            case 'Y':
3623            options |= PCRE_NO_START_OPTIMIZE;
3624            continue;
3625    
3626          case 'Z':          case 'Z':
3627          options |= PCRE_NOTEOL;          options |= PCRE_NOTEOL;
3628          continue;          continue;
# Line 1437  while (!done) Line 3630  while (!done)
3630          case '?':          case '?':
3631          options |= PCRE_NO_UTF8_CHECK;          options |= PCRE_NO_UTF8_CHECK;
3632          continue;          continue;
3633    
3634            case '<':
3635              {
3636              int x = check_newline(p, outfile);
3637              if (x == 0) goto NEXT_DATA;
3638              options |= x;
3639              while (*p++ != '>');
3640              }
3641            continue;
3642            }
3643    
3644          /* We now have a character value in c that may be greater than 255. In
3645          16-bit mode, we always convert characters to UTF-8 so that values greater
3646          than 255 can be passed to non-UTF 16-bit strings. In 8-bit mode we
3647          convert to UTF-8 if we are in UTF mode. Values greater than 127 in UTF
3648          mode must have come from \x{...} or octal constructs because values from
3649          \x.. get this far only in non-UTF mode. */
3650    
3651          if (use_pcre16 || use_utf)
3652            {
3653            pcre_uint8 buff8[8];
3654            int ii, utn;
3655            utn = ord2utf8(c, buff8);
3656            for (ii = 0; ii < utn; ii++) *q++ = buff8[ii];
3657            }
3658          else
3659            {
3660            if (c > 255)
3661              {
3662              fprintf(outfile, "** Character \\x{%x} is greater than 255 "
3663                "and UTF-8 mode is not enabled.\n", c);
3664              fprintf(outfile, "** Truncation will probably give the wrong "
3665                "result.\n");
3666              }
3667            *q++ = c;
3668          }          }
       *q++ = c;  
3669        }        }
3670    
3671        /* Reached end of subject string */
3672    
3673      *q = 0;      *q = 0;
3674      len = q - dbuffer;      len = (int)(q - dbuffer);
3675    
3676        /* Move the data to the end of the buffer so that a read over the end of
3677        the buffer will be seen by valgrind, even if it doesn't cause a crash. If