1 |
/*************************************************
|
2 |
* Perl-Compatible Regular Expressions *
|
3 |
*************************************************/
|
4 |
|
5 |
/* PCRE is a library of functions to support regular expressions whose syntax
|
6 |
and semantics are as close as possible to those of the Perl 5 language.
|
7 |
|
8 |
Written by Philip Hazel
|
9 |
Copyright (c) 1997-2012 University of Cambridge
|
10 |
|
11 |
-----------------------------------------------------------------------------
|
12 |
Redistribution and use in source and binary forms, with or without
|
13 |
modification, are permitted provided that the following conditions are met:
|
14 |
|
15 |
* Redistributions of source code must retain the above copyright notice,
|
16 |
this list of conditions and the following disclaimer.
|
17 |
|
18 |
* Redistributions in binary form must reproduce the above copyright
|
19 |
notice, this list of conditions and the following disclaimer in the
|
20 |
documentation and/or other materials provided with the distribution.
|
21 |
|
22 |
* Neither the name of the University of Cambridge nor the names of its
|
23 |
contributors may be used to endorse or promote products derived from
|
24 |
this software without specific prior written permission.
|
25 |
|
26 |
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
|
27 |
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
28 |
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
29 |
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
|
30 |
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
|
31 |
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
|
32 |
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
33 |
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
|
34 |
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
|
35 |
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
36 |
POSSIBILITY OF SUCH DAMAGE.
|
37 |
-----------------------------------------------------------------------------
|
38 |
*/
|
39 |
|
40 |
|
41 |
/* This module contains a PCRE private debugging function for printing out the
|
42 |
internal form of a compiled regular expression, along with some supporting
|
43 |
local functions. This source file is used in two places:
|
44 |
|
45 |
(1) It is #included by pcre_compile.c when it is compiled in debugging mode
|
46 |
(PCRE_DEBUG defined in pcre_internal.h). It is not included in production
|
47 |
compiles. In this case PCRE_INCLUDED is defined.
|
48 |
|
49 |
(2) It is also compiled separately and linked with pcretest.c, which can be
|
50 |
asked to print out a compiled regex for debugging purposes. */
|
51 |
|
52 |
#ifndef PCRE_INCLUDED
|
53 |
|
54 |
#ifdef HAVE_CONFIG_H
|
55 |
#include "config.h"
|
56 |
#endif
|
57 |
|
58 |
/* For pcretest program. */
|
59 |
#define PRIV(name) name
|
60 |
|
61 |
/* We have to include pcre_internal.h because we need the internal info for
|
62 |
displaying the results of pcre_study() and we also need to know about the
|
63 |
internal macros, structures, and other internal data values; pcretest has
|
64 |
"inside information" compared to a program that strictly follows the PCRE API.
|
65 |
|
66 |
Although pcre_internal.h does itself include pcre.h, we explicitly include it
|
67 |
here before pcre_internal.h so that the PCRE_EXP_xxx macros get set
|
68 |
appropriately for an application, not for building PCRE. */
|
69 |
|
70 |
#include "pcre.h"
|
71 |
#include "pcre_internal.h"
|
72 |
|
73 |
/* These are the funtions that are contained within. It doesn't seem worth
|
74 |
having a separate .h file just for this. */
|
75 |
|
76 |
#endif /* PCRE_INCLUDED */
|
77 |
|
78 |
#ifdef PCRE_INCLUDED
|
79 |
static /* Keep the following function as private. */
|
80 |
#endif
|
81 |
|
82 |
#if defined COMPILE_PCRE8
|
83 |
void pcre_printint(pcre *external_re, FILE *f, BOOL print_lengths);
|
84 |
#elif defined COMPILE_PCRE16
|
85 |
void pcre16_printint(pcre *external_re, FILE *f, BOOL print_lengths);
|
86 |
#elif defined COMPILE_PCRE32
|
87 |
void pcre32_printint(pcre *external_re, FILE *f, BOOL print_lengths);
|
88 |
#endif
|
89 |
|
90 |
/* Macro that decides whether a character should be output as a literal or in
|
91 |
hexadecimal. We don't use isprint() because that can vary from system to system
|
92 |
(even without the use of locales) and we want the output always to be the same,
|
93 |
for testing purposes. */
|
94 |
|
95 |
#ifdef EBCDIC
|
96 |
#define PRINTABLE(c) ((c) >= 64 && (c) < 255)
|
97 |
#else
|
98 |
#define PRINTABLE(c) ((c) >= 32 && (c) < 127)
|
99 |
#endif
|
100 |
|
101 |
/* The table of operator names. */
|
102 |
|
103 |
static const char *priv_OP_names[] = { OP_NAME_LIST };
|
104 |
|
105 |
/* This table of operator lengths is not actually used by the working code,
|
106 |
but its size is needed for a check that ensures it is the correct size for the
|
107 |
number of opcodes (thus catching update omissions). */
|
108 |
|
109 |
static const pcre_uint8 priv_OP_lengths[] = { OP_LENGTHS };
|
110 |
|
111 |
|
112 |
|
113 |
/*************************************************
|
114 |
* Print single- or multi-byte character *
|
115 |
*************************************************/
|
116 |
|
117 |
static unsigned int
|
118 |
print_char(FILE *f, pcre_uchar *ptr, BOOL utf)
|
119 |
{
|
120 |
pcre_uint32 c = *ptr;
|
121 |
|
122 |
#ifndef SUPPORT_UTF
|
123 |
|
124 |
(void)utf; /* Avoid compiler warning */
|
125 |
if (PRINTABLE(c)) fprintf(f, "%c", (char)c);
|
126 |
else if (c <= 0x80) fprintf(f, "\\x%02x", c);
|
127 |
else fprintf(f, "\\x{%x}", c);
|
128 |
return 0;
|
129 |
|
130 |
#else
|
131 |
|
132 |
#if defined COMPILE_PCRE8
|
133 |
|
134 |
if (!utf || (c & 0xc0) != 0xc0)
|
135 |
{
|
136 |
if (PRINTABLE(c)) fprintf(f, "%c", (char)c);
|
137 |
else if (c < 0x80) fprintf(f, "\\x%02x", c);
|
138 |
else fprintf(f, "\\x{%02x}", c);
|
139 |
return 0;
|
140 |
}
|
141 |
else
|
142 |
{
|
143 |
int i;
|
144 |
int a = PRIV(utf8_table4)[c & 0x3f]; /* Number of additional bytes */
|
145 |
int s = 6*a;
|
146 |
c = (c & PRIV(utf8_table3)[a]) << s;
|
147 |
for (i = 1; i <= a; i++)
|
148 |
{
|
149 |
/* This is a check for malformed UTF-8; it should only occur if the sanity
|
150 |
check has been turned off. Rather than swallow random bytes, just stop if
|
151 |
we hit a bad one. Print it with \X instead of \x as an indication. */
|
152 |
|
153 |
if ((ptr[i] & 0xc0) != 0x80)
|
154 |
{
|
155 |
fprintf(f, "\\X{%x}", c);
|
156 |
return i - 1;
|
157 |
}
|
158 |
|
159 |
/* The byte is OK */
|
160 |
|
161 |
s -= 6;
|
162 |
c |= (ptr[i] & 0x3f) << s;
|
163 |
}
|
164 |
fprintf(f, "\\x{%x}", c);
|
165 |
return a;
|
166 |
}
|
167 |
|
168 |
#elif defined COMPILE_PCRE16
|
169 |
|
170 |
if (!utf || (c & 0xfc00) != 0xd800)
|
171 |
{
|
172 |
if (PRINTABLE(c)) fprintf(f, "%c", (char)c);
|
173 |
else if (c <= 0x80) fprintf(f, "\\x%02x", c);
|
174 |
else fprintf(f, "\\x{%02x}", c);
|
175 |
return 0;
|
176 |
}
|
177 |
else
|
178 |
{
|
179 |
/* This is a check for malformed UTF-16; it should only occur if the sanity
|
180 |
check has been turned off. Rather than swallow a low surrogate, just stop if
|
181 |
we hit a bad one. Print it with \X instead of \x as an indication. */
|
182 |
|
183 |
if ((ptr[1] & 0xfc00) != 0xdc00)
|
184 |
{
|
185 |
fprintf(f, "\\X{%x}", c);
|
186 |
return 0;
|
187 |
}
|
188 |
|
189 |
c = (((c & 0x3ff) << 10) | (ptr[1] & 0x3ff)) + 0x10000;
|
190 |
fprintf(f, "\\x{%x}", c);
|
191 |
return 1;
|
192 |
}
|
193 |
|
194 |
#elif defined COMPILE_PCRE32
|
195 |
|
196 |
if (!utf || (c & 0xfffff800u) != 0xd800u)
|
197 |
{
|
198 |
if (PRINTABLE(c)) fprintf(f, "%c", (char)c);
|
199 |
else if (c <= 0x80) fprintf(f, "\\x%02x", c);
|
200 |
else fprintf(f, "\\x{%x}", c);
|
201 |
return 0;
|
202 |
}
|
203 |
else
|
204 |
{
|
205 |
/* This is a check for malformed UTF-32; it should only occur if the sanity
|
206 |
check has been turned off. Rather than swallow a surrogate, just stop if
|
207 |
we hit one. Print it with \X instead of \x as an indication. */
|
208 |
fprintf(f, "\\X{%x}", c);
|
209 |
return 0;
|
210 |
}
|
211 |
|
212 |
#endif /* COMPILE_PCRE[8|16|32] */
|
213 |
|
214 |
#endif /* SUPPORT_UTF */
|
215 |
}
|
216 |
|
217 |
/*************************************************
|
218 |
* Print uchar string (regardless of utf) *
|
219 |
*************************************************/
|
220 |
|
221 |
static void
|
222 |
print_puchar(FILE *f, PCRE_PUCHAR ptr)
|
223 |
{
|
224 |
while (*ptr != '\0')
|
225 |
{
|
226 |
register pcre_uint32 c = *ptr++;
|
227 |
if (PRINTABLE(c)) fprintf(f, "%c", c); else fprintf(f, "\\x{%x}", c);
|
228 |
}
|
229 |
}
|
230 |
|
231 |
/*************************************************
|
232 |
* Find Unicode property name *
|
233 |
*************************************************/
|
234 |
|
235 |
static const char *
|
236 |
get_ucpname(unsigned int ptype, unsigned int pvalue)
|
237 |
{
|
238 |
#ifdef SUPPORT_UCP
|
239 |
int i;
|
240 |
for (i = PRIV(utt_size) - 1; i >= 0; i--)
|
241 |
{
|
242 |
if (ptype == PRIV(utt)[i].type && pvalue == PRIV(utt)[i].value) break;
|
243 |
}
|
244 |
return (i >= 0)? PRIV(utt_names) + PRIV(utt)[i].name_offset : "??";
|
245 |
#else
|
246 |
/* It gets harder and harder to shut off unwanted compiler warnings. */
|
247 |
ptype = ptype * pvalue;
|
248 |
return (ptype == pvalue)? "??" : "??";
|
249 |
#endif
|
250 |
}
|
251 |
|
252 |
|
253 |
/*************************************************
|
254 |
* Print Unicode property value *
|
255 |
*************************************************/
|
256 |
|
257 |
/* "Normal" properties can be printed from tables. The PT_CLIST property is a
|
258 |
pseudo-property that contains a pointer to a list of case-equivalent
|
259 |
characters. This is used only when UCP support is available and UTF mode is
|
260 |
selected. It should never occur otherwise, but just in case it does, have
|
261 |
something ready to print. */
|
262 |
|
263 |
static void
|
264 |
print_prop(FILE *f, pcre_uchar *code, const char *before, const char *after)
|
265 |
{
|
266 |
if (code[1] != PT_CLIST)
|
267 |
{
|
268 |
fprintf(f, "%s%s %s%s", before, priv_OP_names[*code], get_ucpname(code[1],
|
269 |
code[2]), after);
|
270 |
}
|
271 |
else
|
272 |
{
|
273 |
const char *not = (*code == OP_PROP)? "" : "not ";
|
274 |
#ifndef SUPPORT_UCP
|
275 |
fprintf(f, "%s%sclist %d%s", before, not, code[2], after);
|
276 |
#else
|
277 |
const pcre_uint32 *p = PRIV(ucd_caseless_sets) + code[2];
|
278 |
fprintf (f, "%s%sclist", before, not);
|
279 |
while (*p < NOTACHAR) fprintf(f, " %04x", *p++);
|
280 |
fprintf(f, "%s", after);
|
281 |
#endif
|
282 |
}
|
283 |
}
|
284 |
|
285 |
|
286 |
|
287 |
|
288 |
/*************************************************
|
289 |
* Print compiled regex *
|
290 |
*************************************************/
|
291 |
|
292 |
/* Make this function work for a regex with integers either byte order.
|
293 |
However, we assume that what we are passed is a compiled regex. The
|
294 |
print_lengths flag controls whether offsets and lengths of items are printed.
|
295 |
They can be turned off from pcretest so that automatic tests on bytecode can be
|
296 |
written that do not depend on the value of LINK_SIZE. */
|
297 |
|
298 |
#ifdef PCRE_INCLUDED
|
299 |
static /* Keep the following function as private. */
|
300 |
#endif
|
301 |
#if defined COMPILE_PCRE8
|
302 |
void
|
303 |
pcre_printint(pcre *external_re, FILE *f, BOOL print_lengths)
|
304 |
#elif defined COMPILE_PCRE16
|
305 |
void
|
306 |
pcre16_printint(pcre *external_re, FILE *f, BOOL print_lengths)
|
307 |
#elif defined COMPILE_PCRE32
|
308 |
void
|
309 |
pcre32_printint(pcre *external_re, FILE *f, BOOL print_lengths)
|
310 |
#endif
|
311 |
{
|
312 |
REAL_PCRE *re = (REAL_PCRE *)external_re;
|
313 |
pcre_uchar *codestart, *code;
|
314 |
BOOL utf;
|
315 |
|
316 |
unsigned int options = re->options;
|
317 |
int offset = re->name_table_offset;
|
318 |
int count = re->name_count;
|
319 |
int size = re->name_entry_size;
|
320 |
|
321 |
if (re->magic_number != MAGIC_NUMBER)
|
322 |
{
|
323 |
offset = ((offset << 8) & 0xff00) | ((offset >> 8) & 0xff);
|
324 |
count = ((count << 8) & 0xff00) | ((count >> 8) & 0xff);
|
325 |
size = ((size << 8) & 0xff00) | ((size >> 8) & 0xff);
|
326 |
options = ((options << 24) & 0xff000000) |
|
327 |
((options << 8) & 0x00ff0000) |
|
328 |
((options >> 8) & 0x0000ff00) |
|
329 |
((options >> 24) & 0x000000ff);
|
330 |
}
|
331 |
|
332 |
code = codestart = (pcre_uchar *)re + offset + count * size;
|
333 |
/* PCRE_UTF(16|32) have the same value as PCRE_UTF8. */
|
334 |
utf = (options & PCRE_UTF8) != 0;
|
335 |
|
336 |
for(;;)
|
337 |
{
|
338 |
pcre_uchar *ccode;
|
339 |
const char *flag = " ";
|
340 |
pcre_uint32 c;
|
341 |
unsigned int extra = 0;
|
342 |
|
343 |
if (print_lengths)
|
344 |
fprintf(f, "%3d ", (int)(code - codestart));
|
345 |
else
|
346 |
fprintf(f, " ");
|
347 |
|
348 |
switch(*code)
|
349 |
{
|
350 |
/* ========================================================================== */
|
351 |
/* These cases are never obeyed. This is a fudge that causes a compile-
|
352 |
time error if the vectors OP_names or OP_lengths, which are indexed
|
353 |
by opcode, are not the correct length. It seems to be the only way to do
|
354 |
such a check at compile time, as the sizeof() operator does not work in
|
355 |
the C preprocessor. */
|
356 |
|
357 |
case OP_TABLE_LENGTH:
|
358 |
case OP_TABLE_LENGTH +
|
359 |
((sizeof(priv_OP_names)/sizeof(const char *) == OP_TABLE_LENGTH) &&
|
360 |
(sizeof(priv_OP_lengths) == OP_TABLE_LENGTH)):
|
361 |
break;
|
362 |
/* ========================================================================== */
|
363 |
|
364 |
case OP_END:
|
365 |
fprintf(f, " %s\n", priv_OP_names[*code]);
|
366 |
fprintf(f, "------------------------------------------------------------------\n");
|
367 |
return;
|
368 |
|
369 |
case OP_CHAR:
|
370 |
fprintf(f, " ");
|
371 |
do
|
372 |
{
|
373 |
code++;
|
374 |
code += 1 + print_char(f, code, utf);
|
375 |
}
|
376 |
while (*code == OP_CHAR);
|
377 |
fprintf(f, "\n");
|
378 |
continue;
|
379 |
|
380 |
case OP_CHARI:
|
381 |
fprintf(f, " /i ");
|
382 |
do
|
383 |
{
|
384 |
code++;
|
385 |
code += 1 + print_char(f, code, utf);
|
386 |
}
|
387 |
while (*code == OP_CHARI);
|
388 |
fprintf(f, "\n");
|
389 |
continue;
|
390 |
|
391 |
case OP_CBRA:
|
392 |
case OP_CBRAPOS:
|
393 |
case OP_SCBRA:
|
394 |
case OP_SCBRAPOS:
|
395 |
if (print_lengths) fprintf(f, "%3d ", GET(code, 1));
|
396 |
else fprintf(f, " ");
|
397 |
fprintf(f, "%s %d", priv_OP_names[*code], GET2(code, 1+LINK_SIZE));
|
398 |
break;
|
399 |
|
400 |
case OP_BRA:
|
401 |
case OP_BRAPOS:
|
402 |
case OP_SBRA:
|
403 |
case OP_SBRAPOS:
|
404 |
case OP_KETRMAX:
|
405 |
case OP_KETRMIN:
|
406 |
case OP_KETRPOS:
|
407 |
case OP_ALT:
|
408 |
case OP_KET:
|
409 |
case OP_ASSERT:
|
410 |
case OP_ASSERT_NOT:
|
411 |
case OP_ASSERTBACK:
|
412 |
case OP_ASSERTBACK_NOT:
|
413 |
case OP_ONCE:
|
414 |
case OP_ONCE_NC:
|
415 |
case OP_COND:
|
416 |
case OP_SCOND:
|
417 |
case OP_REVERSE:
|
418 |
if (print_lengths) fprintf(f, "%3d ", GET(code, 1));
|
419 |
else fprintf(f, " ");
|
420 |
fprintf(f, "%s", priv_OP_names[*code]);
|
421 |
break;
|
422 |
|
423 |
case OP_CLOSE:
|
424 |
fprintf(f, " %s %d", priv_OP_names[*code], GET2(code, 1));
|
425 |
break;
|
426 |
|
427 |
case OP_CREF:
|
428 |
fprintf(f, "%3d %s", GET2(code,1), priv_OP_names[*code]);
|
429 |
break;
|
430 |
|
431 |
case OP_DNCREF:
|
432 |
{
|
433 |
pcre_uchar *entry = (pcre_uchar *)re + offset + (GET2(code, 1) * size) +
|
434 |
IMM2_SIZE;
|
435 |
fprintf(f, " %s Cond ref <", flag);
|
436 |
print_puchar(f, entry);
|
437 |
fprintf(f, ">%d", GET2(code, 1 + IMM2_SIZE));
|
438 |
}
|
439 |
break;
|
440 |
|
441 |
case OP_RREF:
|
442 |
c = GET2(code, 1);
|
443 |
if (c == RREF_ANY)
|
444 |
fprintf(f, " Cond recurse any");
|
445 |
else
|
446 |
fprintf(f, " Cond recurse %d", c);
|
447 |
break;
|
448 |
|
449 |
case OP_DNRREF:
|
450 |
{
|
451 |
pcre_uchar *entry = (pcre_uchar *)re + offset + (GET2(code, 1) * size) +
|
452 |
IMM2_SIZE;
|
453 |
fprintf(f, " %s Cond recurse <", flag);
|
454 |
print_puchar(f, entry);
|
455 |
fprintf(f, ">%d", GET2(code, 1 + IMM2_SIZE));
|
456 |
}
|
457 |
break;
|
458 |
|
459 |
case OP_DEF:
|
460 |
fprintf(f, " Cond def");
|
461 |
break;
|
462 |
|
463 |
case OP_STARI:
|
464 |
case OP_MINSTARI:
|
465 |
case OP_POSSTARI:
|
466 |
case OP_PLUSI:
|
467 |
case OP_MINPLUSI:
|
468 |
case OP_POSPLUSI:
|
469 |
case OP_QUERYI:
|
470 |
case OP_MINQUERYI:
|
471 |
case OP_POSQUERYI:
|
472 |
flag = "/i";
|
473 |
/* Fall through */
|
474 |
case OP_STAR:
|
475 |
case OP_MINSTAR:
|
476 |
case OP_POSSTAR:
|
477 |
case OP_PLUS:
|
478 |
case OP_MINPLUS:
|
479 |
case OP_POSPLUS:
|
480 |
case OP_QUERY:
|
481 |
case OP_MINQUERY:
|
482 |
case OP_POSQUERY:
|
483 |
case OP_TYPESTAR:
|
484 |
case OP_TYPEMINSTAR:
|
485 |
case OP_TYPEPOSSTAR:
|
486 |
case OP_TYPEPLUS:
|
487 |
case OP_TYPEMINPLUS:
|
488 |
case OP_TYPEPOSPLUS:
|
489 |
case OP_TYPEQUERY:
|
490 |
case OP_TYPEMINQUERY:
|
491 |
case OP_TYPEPOSQUERY:
|
492 |
fprintf(f, " %s ", flag);
|
493 |
if (*code >= OP_TYPESTAR)
|
494 |
{
|
495 |
if (code[1] == OP_PROP || code[1] == OP_NOTPROP)
|
496 |
{
|
497 |
print_prop(f, code + 1, "", " ");
|
498 |
extra = 2;
|
499 |
}
|
500 |
else fprintf(f, "%s", priv_OP_names[code[1]]);
|
501 |
}
|
502 |
else extra = print_char(f, code+1, utf);
|
503 |
fprintf(f, "%s", priv_OP_names[*code]);
|
504 |
break;
|
505 |
|
506 |
case OP_EXACTI:
|
507 |
case OP_UPTOI:
|
508 |
case OP_MINUPTOI:
|
509 |
case OP_POSUPTOI:
|
510 |
flag = "/i";
|
511 |
/* Fall through */
|
512 |
case OP_EXACT:
|
513 |
case OP_UPTO:
|
514 |
case OP_MINUPTO:
|
515 |
case OP_POSUPTO:
|
516 |
fprintf(f, " %s ", flag);
|
517 |
extra = print_char(f, code + 1 + IMM2_SIZE, utf);
|
518 |
fprintf(f, "{");
|
519 |
if (*code != OP_EXACT && *code != OP_EXACTI) fprintf(f, "0,");
|
520 |
fprintf(f, "%d}", GET2(code,1));
|
521 |
if (*code == OP_MINUPTO || *code == OP_MINUPTOI) fprintf(f, "?");
|
522 |
else if (*code == OP_POSUPTO || *code == OP_POSUPTOI) fprintf(f, "+");
|
523 |
break;
|
524 |
|
525 |
case OP_TYPEEXACT:
|
526 |
case OP_TYPEUPTO:
|
527 |
case OP_TYPEMINUPTO:
|
528 |
case OP_TYPEPOSUPTO:
|
529 |
if (code[1 + IMM2_SIZE] == OP_PROP || code[1 + IMM2_SIZE] == OP_NOTPROP)
|
530 |
{
|
531 |
print_prop(f, code + IMM2_SIZE + 1, " ", " ");
|
532 |
extra = 2;
|
533 |
}
|
534 |
else fprintf(f, " %s", priv_OP_names[code[1 + IMM2_SIZE]]);
|
535 |
fprintf(f, "{");
|
536 |
if (*code != OP_TYPEEXACT) fprintf(f, "0,");
|
537 |
fprintf(f, "%d}", GET2(code,1));
|
538 |
if (*code == OP_TYPEMINUPTO) fprintf(f, "?");
|
539 |
else if (*code == OP_TYPEPOSUPTO) fprintf(f, "+");
|
540 |
break;
|
541 |
|
542 |
case OP_NOTI:
|
543 |
flag = "/i";
|
544 |
/* Fall through */
|
545 |
case OP_NOT:
|
546 |
fprintf(f, " %s [^", flag);
|
547 |
extra = print_char(f, code + 1, utf);
|
548 |
fprintf(f, "]");
|
549 |
break;
|
550 |
|
551 |
case OP_NOTSTARI:
|
552 |
case OP_NOTMINSTARI:
|
553 |
case OP_NOTPOSSTARI:
|
554 |
case OP_NOTPLUSI:
|
555 |
case OP_NOTMINPLUSI:
|
556 |
case OP_NOTPOSPLUSI:
|
557 |
case OP_NOTQUERYI:
|
558 |
case OP_NOTMINQUERYI:
|
559 |
case OP_NOTPOSQUERYI:
|
560 |
flag = "/i";
|
561 |
/* Fall through */
|
562 |
|
563 |
case OP_NOTSTAR:
|
564 |
case OP_NOTMINSTAR:
|
565 |
case OP_NOTPOSSTAR:
|
566 |
case OP_NOTPLUS:
|
567 |
case OP_NOTMINPLUS:
|
568 |
case OP_NOTPOSPLUS:
|
569 |
case OP_NOTQUERY:
|
570 |
case OP_NOTMINQUERY:
|
571 |
case OP_NOTPOSQUERY:
|
572 |
fprintf(f, " %s [^", flag);
|
573 |
extra = print_char(f, code + 1, utf);
|
574 |
fprintf(f, "]%s", priv_OP_names[*code]);
|
575 |
break;
|
576 |
|
577 |
case OP_NOTEXACTI:
|
578 |
case OP_NOTUPTOI:
|
579 |
case OP_NOTMINUPTOI:
|
580 |
case OP_NOTPOSUPTOI:
|
581 |
flag = "/i";
|
582 |
/* Fall through */
|
583 |
|
584 |
case OP_NOTEXACT:
|
585 |
case OP_NOTUPTO:
|
586 |
case OP_NOTMINUPTO:
|
587 |
case OP_NOTPOSUPTO:
|
588 |
fprintf(f, " %s [^", flag);
|
589 |
extra = print_char(f, code + 1 + IMM2_SIZE, utf);
|
590 |
fprintf(f, "]{");
|
591 |
if (*code != OP_NOTEXACT && *code != OP_NOTEXACTI) fprintf(f, "0,");
|
592 |
fprintf(f, "%d}", GET2(code,1));
|
593 |
if (*code == OP_NOTMINUPTO || *code == OP_NOTMINUPTOI) fprintf(f, "?");
|
594 |
else
|
595 |
if (*code == OP_NOTPOSUPTO || *code == OP_NOTPOSUPTOI) fprintf(f, "+");
|
596 |
break;
|
597 |
|
598 |
case OP_RECURSE:
|
599 |
if (print_lengths) fprintf(f, "%3d ", GET(code, 1));
|
600 |
else fprintf(f, " ");
|
601 |
fprintf(f, "%s", priv_OP_names[*code]);
|
602 |
break;
|
603 |
|
604 |
case OP_REFI:
|
605 |
flag = "/i";
|
606 |
/* Fall through */
|
607 |
case OP_REF:
|
608 |
fprintf(f, " %s \\%d", flag, GET2(code,1));
|
609 |
ccode = code + priv_OP_lengths[*code];
|
610 |
goto CLASS_REF_REPEAT;
|
611 |
|
612 |
case OP_DNREFI:
|
613 |
flag = "/i";
|
614 |
/* Fall through */
|
615 |
case OP_DNREF:
|
616 |
{
|
617 |
pcre_uchar *entry = (pcre_uchar *)re + offset + (GET2(code, 1) * size) +
|
618 |
IMM2_SIZE;
|
619 |
fprintf(f, " %s \\k<", flag);
|
620 |
print_puchar(f, entry);
|
621 |
fprintf(f, ">%d", GET2(code, 1 + IMM2_SIZE));
|
622 |
}
|
623 |
ccode = code + priv_OP_lengths[*code];
|
624 |
goto CLASS_REF_REPEAT;
|
625 |
|
626 |
case OP_CALLOUT:
|
627 |
fprintf(f, " %s %d %d %d", priv_OP_names[*code], code[1], GET(code,2),
|
628 |
GET(code, 2 + LINK_SIZE));
|
629 |
break;
|
630 |
|
631 |
case OP_PROP:
|
632 |
case OP_NOTPROP:
|
633 |
print_prop(f, code, " ", "");
|
634 |
break;
|
635 |
|
636 |
/* OP_XCLASS cannot occur in 8-bit, non-UTF mode. However, there's no harm
|
637 |
in having this code always here, and it makes it less messy without all
|
638 |
those #ifdefs. */
|
639 |
|
640 |
case OP_CLASS:
|
641 |
case OP_NCLASS:
|
642 |
case OP_XCLASS:
|
643 |
{
|
644 |
int i;
|
645 |
unsigned int min, max;
|
646 |
BOOL printmap;
|
647 |
pcre_uint8 *map;
|
648 |
|
649 |
fprintf(f, " [");
|
650 |
|
651 |
if (*code == OP_XCLASS)
|
652 |
{
|
653 |
extra = GET(code, 1);
|
654 |
ccode = code + LINK_SIZE + 1;
|
655 |
printmap = (*ccode & XCL_MAP) != 0;
|
656 |
if ((*ccode++ & XCL_NOT) != 0) fprintf(f, "^");
|
657 |
}
|
658 |
else
|
659 |
{
|
660 |
printmap = TRUE;
|
661 |
ccode = code + 1;
|
662 |
}
|
663 |
|
664 |
/* Print a bit map */
|
665 |
|
666 |
if (printmap)
|
667 |
{
|
668 |
map = (pcre_uint8 *)ccode;
|
669 |
for (i = 0; i < 256; i++)
|
670 |
{
|
671 |
if ((map[i/8] & (1 << (i&7))) != 0)
|
672 |
{
|
673 |
int j;
|
674 |
for (j = i+1; j < 256; j++)
|
675 |
if ((map[j/8] & (1 << (j&7))) == 0) break;
|
676 |
if (i == '-' || i == ']') fprintf(f, "\\");
|
677 |
if (PRINTABLE(i)) fprintf(f, "%c", i);
|
678 |
else fprintf(f, "\\x%02x", i);
|
679 |
if (--j > i)
|
680 |
{
|
681 |
if (j != i + 1) fprintf(f, "-");
|
682 |
if (j == '-' || j == ']') fprintf(f, "\\");
|
683 |
if (PRINTABLE(j)) fprintf(f, "%c", j);
|
684 |
else fprintf(f, "\\x%02x", j);
|
685 |
}
|
686 |
i = j;
|
687 |
}
|
688 |
}
|
689 |
ccode += 32 / sizeof(pcre_uchar);
|
690 |
}
|
691 |
|
692 |
/* For an XCLASS there is always some additional data */
|
693 |
|
694 |
if (*code == OP_XCLASS)
|
695 |
{
|
696 |
pcre_uchar ch;
|
697 |
while ((ch = *ccode++) != XCL_END)
|
698 |
{
|
699 |
BOOL not = FALSE;
|
700 |
const char *notch = "";
|
701 |
|
702 |
switch(ch)
|
703 |
{
|
704 |
case XCL_NOTPROP:
|
705 |
not = TRUE;
|
706 |
notch = "^";
|
707 |
/* Fall through */
|
708 |
|
709 |
case XCL_PROP:
|
710 |
{
|
711 |
unsigned int ptype = *ccode++;
|
712 |
unsigned int pvalue = *ccode++;
|
713 |
|
714 |
switch(ptype)
|
715 |
{
|
716 |
case PT_PXGRAPH:
|
717 |
fprintf(f, "[:%sgraph:]", notch);
|
718 |
break;
|
719 |
|
720 |
case PT_PXPRINT:
|
721 |
fprintf(f, "[:%sprint:]", notch);
|
722 |
break;
|
723 |
|
724 |
case PT_PXPUNCT:
|
725 |
fprintf(f, "[:%spunct:]", notch);
|
726 |
break;
|
727 |
|
728 |
default:
|
729 |
fprintf(f, "\\%c{%s}", (not? 'P':'p'),
|
730 |
get_ucpname(ptype, pvalue));
|
731 |
break;
|
732 |
}
|
733 |
}
|
734 |
break;
|
735 |
|
736 |
default:
|
737 |
ccode += 1 + print_char(f, ccode, utf);
|
738 |
if (ch == XCL_RANGE)
|
739 |
{
|
740 |
fprintf(f, "-");
|
741 |
ccode += 1 + print_char(f, ccode, utf);
|
742 |
}
|
743 |
break;
|
744 |
}
|
745 |
}
|
746 |
}
|
747 |
|
748 |
/* Indicate a non-UTF class which was created by negation */
|
749 |
|
750 |
fprintf(f, "]%s", (*code == OP_NCLASS)? " (neg)" : "");
|
751 |
|
752 |
/* Handle repeats after a class or a back reference */
|
753 |
|
754 |
CLASS_REF_REPEAT:
|
755 |
switch(*ccode)
|
756 |
{
|
757 |
case OP_CRSTAR:
|
758 |
case OP_CRMINSTAR:
|
759 |
case OP_CRPLUS:
|
760 |
case OP_CRMINPLUS:
|
761 |
case OP_CRQUERY:
|
762 |
case OP_CRMINQUERY:
|
763 |
case OP_CRPOSSTAR:
|
764 |
case OP_CRPOSPLUS:
|
765 |
case OP_CRPOSQUERY:
|
766 |
fprintf(f, "%s", priv_OP_names[*ccode]);
|
767 |
extra += priv_OP_lengths[*ccode];
|
768 |
break;
|
769 |
|
770 |
case OP_CRRANGE:
|
771 |
case OP_CRMINRANGE:
|
772 |
case OP_CRPOSRANGE:
|
773 |
min = GET2(ccode,1);
|
774 |
max = GET2(ccode,1 + IMM2_SIZE);
|
775 |
if (max == 0) fprintf(f, "{%u,}", min);
|
776 |
else fprintf(f, "{%u,%u}", min, max);
|
777 |
if (*ccode == OP_CRMINRANGE) fprintf(f, "?");
|
778 |
else if (*ccode == OP_CRPOSRANGE) fprintf(f, "+");
|
779 |
extra += priv_OP_lengths[*ccode];
|
780 |
break;
|
781 |
|
782 |
/* Do nothing if it's not a repeat; this code stops picky compilers
|
783 |
warning about the lack of a default code path. */
|
784 |
|
785 |
default:
|
786 |
break;
|
787 |
}
|
788 |
}
|
789 |
break;
|
790 |
|
791 |
case OP_MARK:
|
792 |
case OP_PRUNE_ARG:
|
793 |
case OP_SKIP_ARG:
|
794 |
case OP_THEN_ARG:
|
795 |
fprintf(f, " %s ", priv_OP_names[*code]);
|
796 |
print_puchar(f, code + 2);
|
797 |
extra += code[1];
|
798 |
break;
|
799 |
|
800 |
case OP_THEN:
|
801 |
fprintf(f, " %s", priv_OP_names[*code]);
|
802 |
break;
|
803 |
|
804 |
case OP_CIRCM:
|
805 |
case OP_DOLLM:
|
806 |
flag = "/m";
|
807 |
/* Fall through */
|
808 |
|
809 |
/* Anything else is just an item with no data, but possibly a flag. */
|
810 |
|
811 |
default:
|
812 |
fprintf(f, " %s %s", flag, priv_OP_names[*code]);
|
813 |
break;
|
814 |
}
|
815 |
|
816 |
code += priv_OP_lengths[*code] + extra;
|
817 |
fprintf(f, "\n");
|
818 |
}
|
819 |
}
|
820 |
|
821 |
/* End of pcre_printint.src */
|