2 |
* PCRE testing program * |
* PCRE testing program * |
3 |
*************************************************/ |
*************************************************/ |
4 |
|
|
5 |
|
/* 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 |
7 |
|
been extended and consequently is now rather untidy in places. */ |
8 |
|
|
9 |
#include <ctype.h> |
#include <ctype.h> |
10 |
#include <stdio.h> |
#include <stdio.h> |
11 |
#include <string.h> |
#include <string.h> |
12 |
#include <stdlib.h> |
#include <stdlib.h> |
13 |
#include <time.h> |
#include <time.h> |
14 |
|
#include <locale.h> |
15 |
|
|
16 |
/* Use the internal info for displaying the results of pcre_study(). */ |
/* We need the internal info for displaying the results of pcre_study(). Also |
17 |
|
for getting the opcodes for showing compiled code. */ |
18 |
|
|
19 |
|
#define PCRE_SPY /* For Win32 build, import data, not export */ |
20 |
#include "internal.h" |
#include "internal.h" |
21 |
|
|
22 |
|
/* It is possible to compile this test program without including support for |
23 |
|
testing the POSIX interface, though this is not available via the standard |
24 |
|
Makefile. */ |
25 |
|
|
26 |
|
#if !defined NOPOSIX |
27 |
#include "pcreposix.h" |
#include "pcreposix.h" |
28 |
|
#endif |
29 |
|
|
30 |
#ifndef CLOCKS_PER_SEC |
#ifndef CLOCKS_PER_SEC |
31 |
#ifdef CLK_TCK |
#ifdef CLK_TCK |
35 |
#endif |
#endif |
36 |
#endif |
#endif |
37 |
|
|
38 |
#define LOOPREPEAT 10000 |
#define LOOPREPEAT 50000 |
39 |
|
|
40 |
|
|
41 |
static FILE *outfile; |
static FILE *outfile; |
42 |
static int log_store = 0; |
static int log_store = 0; |
43 |
|
static int callout_count; |
44 |
|
static int callout_extra; |
45 |
|
static int callout_fail_count; |
46 |
|
static int callout_fail_id; |
47 |
|
static int first_callout; |
48 |
|
static int use_utf8; |
49 |
|
static size_t gotten_store; |
50 |
|
|
51 |
|
|
52 |
|
|
53 |
|
static int utf8_table1[] = { |
54 |
|
0x0000007f, 0x000007ff, 0x0000ffff, 0x001fffff, 0x03ffffff, 0x7fffffff}; |
55 |
|
|
56 |
|
static int utf8_table2[] = { |
57 |
|
0, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc}; |
58 |
|
|
59 |
|
static int utf8_table3[] = { |
60 |
|
0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01}; |
61 |
|
|
62 |
|
|
63 |
|
|
64 |
|
/************************************************* |
65 |
|
* Print compiled regex * |
66 |
|
*************************************************/ |
67 |
|
|
68 |
|
/* The code for doing this is held in a separate file that is also included in |
69 |
|
pcre.c when it is compiled with the debug switch. It defines a function called |
70 |
|
print_internals(), which uses a table of opcode lengths defined by the macro |
71 |
|
OP_LENGTHS, whose name must be OP_lengths. */ |
72 |
|
|
73 |
|
static uschar OP_lengths[] = { OP_LENGTHS }; |
74 |
|
|
75 |
|
#include "printint.c" |
76 |
|
|
77 |
|
|
78 |
|
|
79 |
|
/************************************************* |
80 |
|
* Read number from string * |
81 |
|
*************************************************/ |
82 |
|
|
83 |
|
/* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess |
84 |
|
around with conditional compilation, just do the job by hand. It is only used |
85 |
|
for unpicking the -o argument, so just keep it simple. |
86 |
|
|
87 |
|
Arguments: |
88 |
|
str string to be converted |
89 |
|
endptr where to put the end pointer |
90 |
|
|
91 |
|
Returns: the unsigned long |
92 |
|
*/ |
93 |
|
|
94 |
|
static int |
95 |
|
get_value(unsigned char *str, unsigned char **endptr) |
96 |
|
{ |
97 |
|
int result = 0; |
98 |
|
while(*str != 0 && isspace(*str)) str++; |
99 |
|
while (isdigit(*str)) result = result * 10 + (int)(*str++ - '0'); |
100 |
|
*endptr = str; |
101 |
|
return(result); |
102 |
|
} |
103 |
|
|
104 |
|
|
105 |
|
|
106 |
/* Debugging function to print the internal form of the regex. This is the same |
/************************************************* |
107 |
code as contained in pcre.c under the DEBUG macro. */ |
* Convert character value to UTF-8 * |
108 |
|
*************************************************/ |
109 |
|
|
110 |
static const char *OP_names[] = { |
/* This function takes an integer value in the range 0 - 0x7fffffff |
111 |
"End", "\\A", "\\B", "\\b", "\\D", "\\d", |
and encodes it as a UTF-8 character in 0 to 6 bytes. |
|
"\\S", "\\s", "\\W", "\\w", "\\Z", "\\z", |
|
|
"Opt", "^", "$", "Any", "chars", "not", |
|
|
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
|
|
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
|
|
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
|
|
"*", "*?", "+", "+?", "?", "??", "{", "{", |
|
|
"class", "Ref", |
|
|
"Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", |
|
|
"AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cref", |
|
|
"Brazero", "Braminzero", "Bra" |
|
|
}; |
|
112 |
|
|
113 |
|
Arguments: |
114 |
|
cvalue the character value |
115 |
|
buffer pointer to buffer for result - at least 6 bytes long |
116 |
|
|
117 |
|
Returns: number of characters placed in the buffer |
118 |
|
-1 if input character is negative |
119 |
|
0 if input character is positive but too big (only when |
120 |
|
int is longer than 32 bits) |
121 |
|
*/ |
122 |
|
|
123 |
static void print_internals(pcre *re, FILE *outfile) |
static int |
124 |
|
ord2utf8(int cvalue, unsigned char *buffer) |
125 |
{ |
{ |
126 |
unsigned char *code = ((real_pcre *)re)->code; |
register int i, j; |
127 |
|
for (i = 0; i < sizeof(utf8_table1)/sizeof(int); i++) |
128 |
|
if (cvalue <= utf8_table1[i]) break; |
129 |
|
if (i >= sizeof(utf8_table1)/sizeof(int)) return 0; |
130 |
|
if (cvalue < 0) return -1; |
131 |
|
|
132 |
|
buffer += i; |
133 |
|
for (j = i; j > 0; j--) |
134 |
|
{ |
135 |
|
*buffer-- = 0x80 | (cvalue & 0x3f); |
136 |
|
cvalue >>= 6; |
137 |
|
} |
138 |
|
*buffer = utf8_table2[i] | cvalue; |
139 |
|
return i + 1; |
140 |
|
} |
141 |
|
|
|
fprintf(outfile, "------------------------------------------------------------------\n"); |
|
142 |
|
|
143 |
for(;;) |
/************************************************* |
144 |
|
* Convert UTF-8 string to value * |
145 |
|
*************************************************/ |
146 |
|
|
147 |
|
/* This function takes one or more bytes that represents a UTF-8 character, |
148 |
|
and returns the value of the character. |
149 |
|
|
150 |
|
Argument: |
151 |
|
buffer a pointer to the byte vector |
152 |
|
vptr a pointer to an int to receive the value |
153 |
|
|
154 |
|
Returns: > 0 => the number of bytes consumed |
155 |
|
-6 to 0 => malformed UTF-8 character at offset = (-return) |
156 |
|
*/ |
157 |
|
|
158 |
|
static int |
159 |
|
utf82ord(unsigned char *buffer, int *vptr) |
160 |
|
{ |
161 |
|
int c = *buffer++; |
162 |
|
int d = c; |
163 |
|
int i, j, s; |
164 |
|
|
165 |
|
for (i = -1; i < 6; i++) /* i is number of additional bytes */ |
166 |
{ |
{ |
167 |
int c; |
if ((d & 0x80) == 0) break; |
168 |
int charlength; |
d <<= 1; |
169 |
|
} |
170 |
|
|
171 |
fprintf(outfile, "%3d ", (int)(code - ((real_pcre *)re)->code)); |
if (i == -1) { *vptr = c; return 1; } /* ascii character */ |
172 |
|
if (i == 0 || i == 6) return 0; /* invalid UTF-8 */ |
|
if (*code >= OP_BRA) |
|
|
{ |
|
|
fprintf(outfile, "%3d Bra %d", (code[1] << 8) + code[2], *code - OP_BRA); |
|
|
code += 2; |
|
|
} |
|
|
|
|
|
else switch(*code) |
|
|
{ |
|
|
case OP_END: |
|
|
fprintf(outfile, " %s\n", OP_names[*code]); |
|
|
fprintf(outfile, "------------------------------------------------------------------\n"); |
|
|
return; |
|
|
|
|
|
case OP_OPT: |
|
|
fprintf(outfile, " %.2x %s", code[1], OP_names[*code]); |
|
|
code++; |
|
|
break; |
|
|
|
|
|
case OP_COND: |
|
|
fprintf(outfile, "%3d Cond", (code[1] << 8) + code[2]); |
|
|
code += 2; |
|
|
break; |
|
|
|
|
|
case OP_CREF: |
|
|
fprintf(outfile, " %.2d %s", code[1], OP_names[*code]); |
|
|
code++; |
|
|
break; |
|
|
|
|
|
case OP_CHARS: |
|
|
charlength = *(++code); |
|
|
fprintf(outfile, "%3d ", charlength); |
|
|
while (charlength-- > 0) |
|
|
if (isprint(c = *(++code))) fprintf(outfile, "%c", c); |
|
|
else fprintf(outfile, "\\x%02x", c); |
|
|
break; |
|
|
|
|
|
case OP_KETRMAX: |
|
|
case OP_KETRMIN: |
|
|
case OP_ALT: |
|
|
case OP_KET: |
|
|
case OP_ASSERT: |
|
|
case OP_ASSERT_NOT: |
|
|
case OP_ASSERTBACK: |
|
|
case OP_ASSERTBACK_NOT: |
|
|
case OP_ONCE: |
|
|
fprintf(outfile, "%3d %s", (code[1] << 8) + code[2], OP_names[*code]); |
|
|
code += 2; |
|
|
break; |
|
|
|
|
|
case OP_REVERSE: |
|
|
fprintf(outfile, "%3d %s", (code[1] << 8) + code[2], OP_names[*code]); |
|
|
code += 2; |
|
|
break; |
|
|
|
|
|
case OP_STAR: |
|
|
case OP_MINSTAR: |
|
|
case OP_PLUS: |
|
|
case OP_MINPLUS: |
|
|
case OP_QUERY: |
|
|
case OP_MINQUERY: |
|
|
case OP_TYPESTAR: |
|
|
case OP_TYPEMINSTAR: |
|
|
case OP_TYPEPLUS: |
|
|
case OP_TYPEMINPLUS: |
|
|
case OP_TYPEQUERY: |
|
|
case OP_TYPEMINQUERY: |
|
|
if (*code >= OP_TYPESTAR) |
|
|
fprintf(outfile, " %s", OP_names[code[1]]); |
|
|
else if (isprint(c = code[1])) fprintf(outfile, " %c", c); |
|
|
else fprintf(outfile, " \\x%02x", c); |
|
|
fprintf(outfile, "%s", OP_names[*code++]); |
|
|
break; |
|
|
|
|
|
case OP_EXACT: |
|
|
case OP_UPTO: |
|
|
case OP_MINUPTO: |
|
|
if (isprint(c = code[3])) fprintf(outfile, " %c{", c); |
|
|
else fprintf(outfile, " \\x%02x{", c); |
|
|
if (*code != OP_EXACT) fprintf(outfile, ","); |
|
|
fprintf(outfile, "%d}", (code[1] << 8) + code[2]); |
|
|
if (*code == OP_MINUPTO) fprintf(outfile, "?"); |
|
|
code += 3; |
|
|
break; |
|
|
|
|
|
case OP_TYPEEXACT: |
|
|
case OP_TYPEUPTO: |
|
|
case OP_TYPEMINUPTO: |
|
|
fprintf(outfile, " %s{", OP_names[code[3]]); |
|
|
if (*code != OP_TYPEEXACT) fprintf(outfile, "0,"); |
|
|
fprintf(outfile, "%d}", (code[1] << 8) + code[2]); |
|
|
if (*code == OP_TYPEMINUPTO) fprintf(outfile, "?"); |
|
|
code += 3; |
|
|
break; |
|
|
|
|
|
case OP_NOT: |
|
|
if (isprint(c = *(++code))) fprintf(outfile, " [^%c]", c); |
|
|
else fprintf(outfile, " [^\\x%02x]", c); |
|
|
break; |
|
|
|
|
|
case OP_NOTSTAR: |
|
|
case OP_NOTMINSTAR: |
|
|
case OP_NOTPLUS: |
|
|
case OP_NOTMINPLUS: |
|
|
case OP_NOTQUERY: |
|
|
case OP_NOTMINQUERY: |
|
|
if (isprint(c = code[1])) fprintf(outfile, " [^%c]", c); |
|
|
else fprintf(outfile, " [^\\x%02x]", c); |
|
|
fprintf(outfile, "%s", OP_names[*code++]); |
|
|
break; |
|
|
|
|
|
case OP_NOTEXACT: |
|
|
case OP_NOTUPTO: |
|
|
case OP_NOTMINUPTO: |
|
|
if (isprint(c = code[3])) fprintf(outfile, " [^%c]{", c); |
|
|
else fprintf(outfile, " [^\\x%02x]{", c); |
|
|
if (*code != OP_NOTEXACT) fprintf(outfile, ","); |
|
|
fprintf(outfile, "%d}", (code[1] << 8) + code[2]); |
|
|
if (*code == OP_NOTMINUPTO) fprintf(outfile, "?"); |
|
|
code += 3; |
|
|
break; |
|
|
|
|
|
case OP_REF: |
|
|
fprintf(outfile, " \\%d", *(++code)); |
|
|
code++; |
|
|
goto CLASS_REF_REPEAT; |
|
|
|
|
|
case OP_CLASS: |
|
|
{ |
|
|
int i, min, max; |
|
|
code++; |
|
|
fprintf(outfile, " ["); |
|
|
|
|
|
for (i = 0; i < 256; i++) |
|
|
{ |
|
|
if ((code[i/8] & (1 << (i&7))) != 0) |
|
|
{ |
|
|
int j; |
|
|
for (j = i+1; j < 256; j++) |
|
|
if ((code[j/8] & (1 << (j&7))) == 0) break; |
|
|
if (i == '-' || i == ']') fprintf(outfile, "\\"); |
|
|
if (isprint(i)) fprintf(outfile, "%c", i); else fprintf(outfile, "\\x%02x", i); |
|
|
if (--j > i) |
|
|
{ |
|
|
fprintf(outfile, "-"); |
|
|
if (j == '-' || j == ']') fprintf(outfile, "\\"); |
|
|
if (isprint(j)) fprintf(outfile, "%c", j); else fprintf(outfile, "\\x%02x", j); |
|
|
} |
|
|
i = j; |
|
|
} |
|
|
} |
|
|
fprintf(outfile, "]"); |
|
|
code += 32; |
|
|
|
|
|
CLASS_REF_REPEAT: |
|
|
|
|
|
switch(*code) |
|
|
{ |
|
|
case OP_CRSTAR: |
|
|
case OP_CRMINSTAR: |
|
|
case OP_CRPLUS: |
|
|
case OP_CRMINPLUS: |
|
|
case OP_CRQUERY: |
|
|
case OP_CRMINQUERY: |
|
|
fprintf(outfile, "%s", OP_names[*code]); |
|
|
break; |
|
173 |
|
|
174 |
case OP_CRRANGE: |
/* i now has a value in the range 1-5 */ |
175 |
case OP_CRMINRANGE: |
|
176 |
min = (code[1] << 8) + code[2]; |
s = 6*i; |
177 |
max = (code[3] << 8) + code[4]; |
d = (c & utf8_table3[i]) << s; |
178 |
if (max == 0) fprintf(outfile, "{%d,}", min); |
|
179 |
else fprintf(outfile, "{%d,%d}", min, max); |
for (j = 0; j < i; j++) |
180 |
if (*code == OP_CRMINRANGE) fprintf(outfile, "?"); |
{ |
181 |
code += 4; |
c = *buffer++; |
182 |
break; |
if ((c & 0xc0) != 0x80) return -(j+1); |
183 |
|
s -= 6; |
184 |
|
d |= (c & 0x3f) << s; |
185 |
|
} |
186 |
|
|
187 |
|
/* Check that encoding was the correct unique one */ |
188 |
|
|
189 |
|
for (j = 0; j < sizeof(utf8_table1)/sizeof(int); j++) |
190 |
|
if (d <= utf8_table1[j]) break; |
191 |
|
if (j != i) return -(i+1); |
192 |
|
|
193 |
|
/* Valid value */ |
194 |
|
|
195 |
|
*vptr = d; |
196 |
|
return i+1; |
197 |
|
} |
198 |
|
|
199 |
|
|
200 |
|
|
201 |
|
/************************************************* |
202 |
|
* Print character string * |
203 |
|
*************************************************/ |
204 |
|
|
205 |
|
/* Character string printing function. Must handle UTF-8 strings in utf8 |
206 |
|
mode. Yields number of characters printed. If handed a NULL file, just counts |
207 |
|
chars without printing. */ |
208 |
|
|
209 |
|
static int pchars(unsigned char *p, int length, FILE *f) |
210 |
|
{ |
211 |
|
int c; |
212 |
|
int yield = 0; |
213 |
|
|
214 |
|
while (length-- > 0) |
215 |
|
{ |
216 |
|
if (use_utf8) |
217 |
|
{ |
218 |
|
int rc = utf82ord(p, &c); |
219 |
|
|
220 |
default: |
if (rc > 0 && rc <= length + 1) /* Mustn't run over the end */ |
221 |
code--; |
{ |
222 |
|
length -= rc - 1; |
223 |
|
p += rc; |
224 |
|
if (c < 256 && isprint(c)) |
225 |
|
{ |
226 |
|
if (f != NULL) fprintf(f, "%c", c); |
227 |
|
yield++; |
228 |
|
} |
229 |
|
else |
230 |
|
{ |
231 |
|
int n; |
232 |
|
if (f != NULL) fprintf(f, "\\x{%02x}%n", c, &n); |
233 |
|
yield += n; |
234 |
} |
} |
235 |
|
continue; |
236 |
} |
} |
237 |
break; |
} |
238 |
|
|
239 |
/* Anything else is just a one-node item */ |
/* Not UTF-8, or malformed UTF-8 */ |
240 |
|
|
241 |
default: |
if (isprint(c = *(p++))) |
242 |
fprintf(outfile, " %s", OP_names[*code]); |
{ |
243 |
break; |
if (f != NULL) fprintf(f, "%c", c); |
244 |
|
yield++; |
245 |
|
} |
246 |
|
else |
247 |
|
{ |
248 |
|
if (f != NULL) fprintf(f, "\\x%02x", c); |
249 |
|
yield += 4; |
250 |
} |
} |
|
|
|
|
code++; |
|
|
fprintf(outfile, "\n"); |
|
251 |
} |
} |
252 |
|
|
253 |
|
return yield; |
254 |
} |
} |
255 |
|
|
256 |
|
|
257 |
|
|
258 |
/* Character string printing function. */ |
/************************************************* |
259 |
|
* Callout function * |
260 |
|
*************************************************/ |
261 |
|
|
262 |
|
/* Called from PCRE as a result of the (?C) item. We print out where we are in |
263 |
|
the match. Yield zero unless more callouts than the fail count, or the callout |
264 |
|
data is not zero. */ |
265 |
|
|
266 |
static void pchars(unsigned char *p, int length) |
static int callout(pcre_callout_block *cb) |
267 |
{ |
{ |
268 |
int c; |
FILE *f = (first_callout | callout_extra)? outfile : NULL; |
269 |
while (length-- > 0) |
int i, pre_start, post_start; |
270 |
if (isprint(c = *(p++))) fprintf(outfile, "%c", c); |
|
271 |
else fprintf(outfile, "\\x%02x", c); |
if (callout_extra) |
272 |
|
{ |
273 |
|
fprintf(f, "Callout %d: last capture = %d\n", |
274 |
|
cb->callout_number, cb->capture_last); |
275 |
|
|
276 |
|
for (i = 0; i < cb->capture_top * 2; i += 2) |
277 |
|
{ |
278 |
|
if (cb->offset_vector[i] < 0) |
279 |
|
fprintf(f, "%2d: <unset>\n", i/2); |
280 |
|
else |
281 |
|
{ |
282 |
|
fprintf(f, "%2d: ", i/2); |
283 |
|
(void)pchars((unsigned char *)cb->subject + cb->offset_vector[i], |
284 |
|
cb->offset_vector[i+1] - cb->offset_vector[i], f); |
285 |
|
fprintf(f, "\n"); |
286 |
|
} |
287 |
|
} |
288 |
|
} |
289 |
|
|
290 |
|
/* Re-print the subject in canonical form, the first time or if giving full |
291 |
|
datails. On subsequent calls in the same match, we use pchars just to find the |
292 |
|
printed lengths of the substrings. */ |
293 |
|
|
294 |
|
if (f != NULL) fprintf(f, "--->"); |
295 |
|
|
296 |
|
pre_start = pchars((unsigned char *)cb->subject, cb->start_match, f); |
297 |
|
post_start = pchars((unsigned char *)(cb->subject + cb->start_match), |
298 |
|
cb->current_position - cb->start_match, f); |
299 |
|
|
300 |
|
(void)pchars((unsigned char *)(cb->subject + cb->current_position), |
301 |
|
cb->subject_length - cb->current_position, f); |
302 |
|
|
303 |
|
if (f != NULL) fprintf(f, "\n"); |
304 |
|
|
305 |
|
/* Always print appropriate indicators, with callout number if not already |
306 |
|
shown */ |
307 |
|
|
308 |
|
if (callout_extra) fprintf(outfile, " "); |
309 |
|
else fprintf(outfile, "%3d ", cb->callout_number); |
310 |
|
|
311 |
|
for (i = 0; i < pre_start; i++) fprintf(outfile, " "); |
312 |
|
fprintf(outfile, "^"); |
313 |
|
|
314 |
|
if (post_start > 0) |
315 |
|
{ |
316 |
|
for (i = 0; i < post_start - 1; i++) fprintf(outfile, " "); |
317 |
|
fprintf(outfile, "^"); |
318 |
|
} |
319 |
|
|
320 |
|
fprintf(outfile, "\n"); |
321 |
|
|
322 |
|
first_callout = 0; |
323 |
|
|
324 |
|
if ((int)(cb->callout_data) != 0) |
325 |
|
{ |
326 |
|
fprintf(outfile, "Callout data = %d\n", (int)(cb->callout_data)); |
327 |
|
return (int)(cb->callout_data); |
328 |
|
} |
329 |
|
|
330 |
|
return (cb->callout_number != callout_fail_id)? 0 : |
331 |
|
(++callout_count >= callout_fail_count)? 1 : 0; |
332 |
} |
} |
333 |
|
|
334 |
|
|
335 |
|
/************************************************* |
336 |
|
* Local malloc function * |
337 |
|
*************************************************/ |
338 |
|
|
339 |
/* Alternative malloc function, to test functionality and show the size of the |
/* Alternative malloc function, to test functionality and show the size of the |
340 |
compiled re. */ |
compiled re. */ |
341 |
|
|
342 |
static void *new_malloc(size_t size) |
static void *new_malloc(size_t size) |
343 |
{ |
{ |
344 |
if (log_store) fprintf(outfile, "Store size request: %d\n", (int)size); |
gotten_store = size; |
345 |
return malloc(size); |
return malloc(size); |
346 |
} |
} |
347 |
|
|
348 |
|
|
349 |
|
|
350 |
|
/************************************************* |
351 |
|
* Call pcre_fullinfo() * |
352 |
|
*************************************************/ |
353 |
|
|
354 |
|
/* Get one piece of information from the pcre_fullinfo() function */ |
355 |
|
|
356 |
|
static void new_info(pcre *re, pcre_extra *study, int option, void *ptr) |
357 |
|
{ |
358 |
|
int rc; |
359 |
|
if ((rc = pcre_fullinfo(re, study, option, ptr)) < 0) |
360 |
|
fprintf(outfile, "Error %d from pcre_fullinfo(%d)\n", rc, option); |
361 |
|
} |
362 |
|
|
363 |
|
|
364 |
|
|
365 |
|
/************************************************* |
366 |
|
* Main Program * |
367 |
|
*************************************************/ |
368 |
|
|
369 |
/* Read lines from named file or stdin and write to named file or stdout; lines |
/* Read lines from named file or stdin and write to named file or stdout; lines |
370 |
consist of a regular expression, in delimiters and optionally followed by |
consist of a regular expression, in delimiters and optionally followed by |
371 |
options, followed by a set of test data, terminated by an empty line. */ |
options, followed by a set of test data, terminated by an empty line. */ |
378 |
int op = 1; |
int op = 1; |
379 |
int timeit = 0; |
int timeit = 0; |
380 |
int showinfo = 0; |
int showinfo = 0; |
381 |
|
int showstore = 0; |
382 |
|
int size_offsets = 45; |
383 |
|
int size_offsets_max; |
384 |
|
int *offsets; |
385 |
|
#if !defined NOPOSIX |
386 |
int posix = 0; |
int posix = 0; |
387 |
|
#endif |
388 |
int debug = 0; |
int debug = 0; |
389 |
int done = 0; |
int done = 0; |
390 |
unsigned char buffer[30000]; |
unsigned char buffer[30000]; |
398 |
|
|
399 |
while (argc > 1 && argv[op][0] == '-') |
while (argc > 1 && argv[op][0] == '-') |
400 |
{ |
{ |
401 |
if (strcmp(argv[op], "-s") == 0) log_store = 1; |
unsigned char *endptr; |
402 |
|
|
403 |
|
if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0) |
404 |
|
showstore = 1; |
405 |
else if (strcmp(argv[op], "-t") == 0) timeit = 1; |
else if (strcmp(argv[op], "-t") == 0) timeit = 1; |
406 |
else if (strcmp(argv[op], "-i") == 0) showinfo = 1; |
else if (strcmp(argv[op], "-i") == 0) showinfo = 1; |
407 |
else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1; |
else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1; |
408 |
|
else if (strcmp(argv[op], "-o") == 0 && argc > 2 && |
409 |
|
((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)), |
410 |
|
*endptr == 0)) |
411 |
|
{ |
412 |
|
op++; |
413 |
|
argc--; |
414 |
|
} |
415 |
|
#if !defined NOPOSIX |
416 |
else if (strcmp(argv[op], "-p") == 0) posix = 1; |
else if (strcmp(argv[op], "-p") == 0) posix = 1; |
417 |
|
#endif |
418 |
|
else if (strcmp(argv[op], "-C") == 0) |
419 |
|
{ |
420 |
|
int rc; |
421 |
|
printf("PCRE version %s\n", pcre_version()); |
422 |
|
printf("Compiled with\n"); |
423 |
|
(void)pcre_config(PCRE_CONFIG_UTF8, &rc); |
424 |
|
printf(" %sUTF-8 support\n", rc? "" : "No "); |
425 |
|
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
426 |
|
printf(" Newline character is %s\n", (rc == '\r')? "CR" : "LF"); |
427 |
|
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
428 |
|
printf(" Internal link size = %d\n", rc); |
429 |
|
(void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc); |
430 |
|
printf(" POSIX malloc threshold = %d\n", rc); |
431 |
|
(void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc); |
432 |
|
printf(" Default match limit = %d\n", rc); |
433 |
|
exit(0); |
434 |
|
} |
435 |
else |
else |
436 |
{ |
{ |
437 |
printf("*** Unknown option %s\n", argv[op]); |
printf("** Unknown or malformed option %s\n", argv[op]); |
438 |
|
printf("Usage: pcretest [-d] [-i] [-o <n>] [-p] [-s] [-t] [<input> [<output>]]\n"); |
439 |
|
printf(" -C show PCRE compile-time options and exit\n"); |
440 |
|
printf(" -d debug: show compiled code; implies -i\n" |
441 |
|
" -i show information about compiled pattern\n" |
442 |
|
" -o <n> set size of offsets vector to <n>\n"); |
443 |
|
#if !defined NOPOSIX |
444 |
|
printf(" -p use POSIX interface\n"); |
445 |
|
#endif |
446 |
|
printf(" -s output store information\n" |
447 |
|
" -t time compilation and execution\n"); |
448 |
return 1; |
return 1; |
449 |
} |
} |
450 |
op++; |
op++; |
451 |
argc--; |
argc--; |
452 |
} |
} |
453 |
|
|
454 |
|
/* Get the store for the offsets vector, and remember what it was */ |
455 |
|
|
456 |
|
size_offsets_max = size_offsets; |
457 |
|
offsets = malloc(size_offsets_max * sizeof(int)); |
458 |
|
if (offsets == NULL) |
459 |
|
{ |
460 |
|
printf("** Failed to get %d bytes of memory for offsets vector\n", |
461 |
|
size_offsets_max * sizeof(int)); |
462 |
|
return 1; |
463 |
|
} |
464 |
|
|
465 |
/* Sort out the input and output files */ |
/* Sort out the input and output files */ |
466 |
|
|
467 |
if (argc > 1) |
if (argc > 1) |
498 |
{ |
{ |
499 |
pcre *re = NULL; |
pcre *re = NULL; |
500 |
pcre_extra *extra = NULL; |
pcre_extra *extra = NULL; |
501 |
|
|
502 |
|
#if !defined NOPOSIX /* There are still compilers that require no indent */ |
503 |
regex_t preg; |
regex_t preg; |
504 |
|
int do_posix = 0; |
505 |
|
#endif |
506 |
|
|
507 |
const char *error; |
const char *error; |
508 |
unsigned char *p, *pp; |
unsigned char *p, *pp, *ppp; |
509 |
|
const unsigned char *tables = NULL; |
510 |
int do_study = 0; |
int do_study = 0; |
511 |
int do_debug = 0; |
int do_debug = debug; |
512 |
int do_posix = 0; |
int do_G = 0; |
513 |
|
int do_g = 0; |
514 |
|
int do_showinfo = showinfo; |
515 |
|
int do_showrest = 0; |
516 |
int erroroffset, len, delimiter; |
int erroroffset, len, delimiter; |
517 |
|
|
518 |
|
use_utf8 = 0; |
519 |
|
|
520 |
if (infile == stdin) printf(" re> "); |
if (infile == stdin) printf(" re> "); |
521 |
if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) break; |
if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) break; |
522 |
if (infile != stdin) fprintf(outfile, "%s", (char *)buffer); |
if (infile != stdin) fprintf(outfile, "%s", (char *)buffer); |
523 |
|
fflush(outfile); |
524 |
|
|
525 |
p = buffer; |
p = buffer; |
526 |
while (isspace(*p)) p++; |
while (isspace(*p)) p++; |
531 |
|
|
532 |
delimiter = *p++; |
delimiter = *p++; |
533 |
|
|
534 |
if (isalnum(delimiter)) |
if (isalnum(delimiter) || delimiter == '\\') |
535 |
{ |
{ |
536 |
fprintf(outfile, "** Delimiter must not be alphameric\n"); |
fprintf(outfile, "** Delimiter must not be alphameric or \\\n"); |
537 |
goto SKIP_DATA; |
goto SKIP_DATA; |
538 |
} |
} |
539 |
|
|
541 |
|
|
542 |
for(;;) |
for(;;) |
543 |
{ |
{ |
544 |
while (*pp != 0 && *pp != delimiter) pp++; |
while (*pp != 0) |
545 |
|
{ |
546 |
|
if (*pp == '\\' && pp[1] != 0) pp++; |
547 |
|
else if (*pp == delimiter) break; |
548 |
|
pp++; |
549 |
|
} |
550 |
if (*pp != 0) break; |
if (*pp != 0) break; |
551 |
|
|
552 |
len = sizeof(buffer) - (pp - buffer); |
len = sizeof(buffer) - (pp - buffer); |
566 |
if (infile != stdin) fprintf(outfile, "%s", (char *)pp); |
if (infile != stdin) fprintf(outfile, "%s", (char *)pp); |
567 |
} |
} |
568 |
|
|
569 |
|
/* If the first character after the delimiter is backslash, make |
570 |
|
the pattern end with backslash. This is purely to provide a way |
571 |
|
of testing for the error message when a pattern ends with backslash. */ |
572 |
|
|
573 |
|
if (pp[1] == '\\') *pp++ = '\\'; |
574 |
|
|
575 |
/* Terminate the pattern at the delimiter */ |
/* Terminate the pattern at the delimiter */ |
576 |
|
|
577 |
*pp++ = 0; |
*pp++ = 0; |
580 |
|
|
581 |
options = 0; |
options = 0; |
582 |
study_options = 0; |
study_options = 0; |
583 |
|
log_store = showstore; /* default from command line */ |
584 |
|
|
585 |
while (*pp != 0) |
while (*pp != 0) |
586 |
{ |
{ |
587 |
switch (*pp++) |
switch (*pp++) |
588 |
{ |
{ |
589 |
|
case 'g': do_g = 1; break; |
590 |
case 'i': options |= PCRE_CASELESS; break; |
case 'i': options |= PCRE_CASELESS; break; |
591 |
case 'm': options |= PCRE_MULTILINE; break; |
case 'm': options |= PCRE_MULTILINE; break; |
592 |
case 's': options |= PCRE_DOTALL; break; |
case 's': options |= PCRE_DOTALL; break; |
593 |
case 'x': options |= PCRE_EXTENDED; break; |
case 'x': options |= PCRE_EXTENDED; break; |
594 |
|
|
595 |
|
case '+': do_showrest = 1; break; |
596 |
case 'A': options |= PCRE_ANCHORED; break; |
case 'A': options |= PCRE_ANCHORED; break; |
597 |
case 'D': do_debug = 1; break; |
case 'D': do_debug = do_showinfo = 1; break; |
598 |
case 'E': options |= PCRE_DOLLAR_ENDONLY; break; |
case 'E': options |= PCRE_DOLLAR_ENDONLY; break; |
599 |
|
case 'G': do_G = 1; break; |
600 |
|
case 'I': do_showinfo = 1; break; |
601 |
|
case 'M': log_store = 1; break; |
602 |
|
case 'N': options |= PCRE_NO_AUTO_CAPTURE; break; |
603 |
|
|
604 |
|
#if !defined NOPOSIX |
605 |
case 'P': do_posix = 1; break; |
case 'P': do_posix = 1; break; |
606 |
|
#endif |
607 |
|
|
608 |
case 'S': do_study = 1; break; |
case 'S': do_study = 1; break; |
609 |
case 'U': options |= PCRE_UNGREEDY; break; |
case 'U': options |= PCRE_UNGREEDY; break; |
610 |
case 'X': options |= PCRE_EXTRA; break; |
case 'X': options |= PCRE_EXTRA; break; |
611 |
|
case '8': options |= PCRE_UTF8; use_utf8 = 1; break; |
612 |
|
|
613 |
|
case 'L': |
614 |
|
ppp = pp; |
615 |
|
while (*ppp != '\n' && *ppp != ' ') ppp++; |
616 |
|
*ppp = 0; |
617 |
|
if (setlocale(LC_CTYPE, (const char *)pp) == NULL) |
618 |
|
{ |
619 |
|
fprintf(outfile, "** Failed to set locale \"%s\"\n", pp); |
620 |
|
goto SKIP_DATA; |
621 |
|
} |
622 |
|
tables = pcre_maketables(); |
623 |
|
pp = ppp; |
624 |
|
break; |
625 |
|
|
626 |
case '\n': case ' ': break; |
case '\n': case ' ': break; |
627 |
default: |
default: |
628 |
fprintf(outfile, "** Unknown option '%c'\n", pp[-1]); |
fprintf(outfile, "** Unknown option '%c'\n", pp[-1]); |
631 |
} |
} |
632 |
|
|
633 |
/* Handle compiling via the POSIX interface, which doesn't support the |
/* Handle compiling via the POSIX interface, which doesn't support the |
634 |
timing, showing, or debugging options. */ |
timing, showing, or debugging options, nor the ability to pass over |
635 |
|
local character tables. */ |
636 |
|
|
637 |
|
#if !defined NOPOSIX |
638 |
if (posix || do_posix) |
if (posix || do_posix) |
639 |
{ |
{ |
640 |
int rc; |
int rc; |
657 |
/* Handle compiling via the native interface */ |
/* Handle compiling via the native interface */ |
658 |
|
|
659 |
else |
else |
660 |
|
#endif /* !defined NOPOSIX */ |
661 |
|
|
662 |
{ |
{ |
663 |
if (timeit) |
if (timeit) |
664 |
{ |
{ |
667 |
clock_t start_time = clock(); |
clock_t start_time = clock(); |
668 |
for (i = 0; i < LOOPREPEAT; i++) |
for (i = 0; i < LOOPREPEAT; i++) |
669 |
{ |
{ |
670 |
re = pcre_compile((char *)p, options, &error, &erroroffset); |
re = pcre_compile((char *)p, options, &error, &erroroffset, tables); |
671 |
if (re != NULL) free(re); |
if (re != NULL) free(re); |
672 |
} |
} |
673 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
674 |
fprintf(outfile, "Compile time %.2f milliseconds\n", |
fprintf(outfile, "Compile time %.3f milliseconds\n", |
675 |
((double)time_taken)/(4 * CLOCKS_PER_SEC)); |
(((double)time_taken * 1000.0) / (double)LOOPREPEAT) / |
676 |
|
(double)CLOCKS_PER_SEC); |
677 |
} |
} |
678 |
|
|
679 |
re = pcre_compile((char *)p, options, &error, &erroroffset); |
re = pcre_compile((char *)p, options, &error, &erroroffset, tables); |
680 |
|
|
681 |
/* Compilation failed; go back for another re, skipping to blank line |
/* Compilation failed; go back for another re, skipping to blank line |
682 |
if non-interactive. */ |
if non-interactive. */ |
700 |
} |
} |
701 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
702 |
} |
} |
703 |
continue; |
goto CONTINUE; |
704 |
} |
} |
705 |
|
|
706 |
/* Compilation succeeded; print data if required */ |
/* Compilation succeeded; print data if required. There are now two |
707 |
|
info-returning functions. The old one has a limited interface and |
708 |
|
returns only limited data. Check that it agrees with the newer one. */ |
709 |
|
|
710 |
|
if (log_store) |
711 |
|
fprintf(outfile, "Memory allocation (code space): %d\n", |
712 |
|
(int)(gotten_store - |
713 |
|
sizeof(real_pcre) - |
714 |
|
((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size)); |
715 |
|
|
716 |
if (showinfo || do_debug) |
if (do_showinfo) |
717 |
{ |
{ |
718 |
int first_char, count; |
unsigned long int get_options; |
719 |
|
int old_first_char, old_options, old_count; |
720 |
|
int count, backrefmax, first_char, need_char; |
721 |
|
int nameentrysize, namecount; |
722 |
|
const uschar *nametable; |
723 |
|
size_t size; |
724 |
|
|
725 |
if (debug || do_debug) print_internals(re, outfile); |
if (do_debug) |
726 |
|
{ |
727 |
|
fprintf(outfile, "------------------------------------------------------------------\n"); |
728 |
|
print_internals(re, outfile); |
729 |
|
} |
730 |
|
|
731 |
|
new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options); |
732 |
|
new_info(re, NULL, PCRE_INFO_SIZE, &size); |
733 |
|
new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count); |
734 |
|
new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax); |
735 |
|
new_info(re, NULL, PCRE_INFO_FIRSTBYTE, &first_char); |
736 |
|
new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char); |
737 |
|
new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize); |
738 |
|
new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount); |
739 |
|
new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable); |
740 |
|
|
741 |
count = pcre_info(re, &options, &first_char); |
old_count = pcre_info(re, &old_options, &old_first_char); |
742 |
if (count < 0) fprintf(outfile, |
if (count < 0) fprintf(outfile, |
743 |
"Error %d while reading info\n", count); |
"Error %d from pcre_info()\n", count); |
744 |
else |
else |
745 |
{ |
{ |
746 |
fprintf(outfile, "Identifying subpattern count = %d\n", count); |
if (old_count != count) fprintf(outfile, |
747 |
if (options == 0) fprintf(outfile, "No options\n"); |
"Count disagreement: pcre_fullinfo=%d pcre_info=%d\n", count, |
748 |
else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s\n", |
old_count); |
749 |
((options & PCRE_ANCHORED) != 0)? " anchored" : "", |
|
750 |
((options & PCRE_CASELESS) != 0)? " caseless" : "", |
if (old_first_char != first_char) fprintf(outfile, |
751 |
((options & PCRE_EXTENDED) != 0)? " extended" : "", |
"First char disagreement: pcre_fullinfo=%d pcre_info=%d\n", |
752 |
((options & PCRE_MULTILINE) != 0)? " multiline" : "", |
first_char, old_first_char); |
753 |
((options & PCRE_DOTALL) != 0)? " dotall" : "", |
|
754 |
((options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
if (old_options != (int)get_options) fprintf(outfile, |
755 |
((options & PCRE_EXTRA) != 0)? " extra" : "", |
"Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n", |
756 |
((options & PCRE_UNGREEDY) != 0)? " ungreedy" : ""); |
get_options, old_options); |
757 |
if (first_char == -1) |
} |
758 |
{ |
|
759 |
fprintf(outfile, "First char at start or follows \\n\n"); |
if (size != gotten_store) fprintf(outfile, |
760 |
} |
"Size disagreement: pcre_fullinfo=%d call to malloc for %d\n", |
761 |
else if (first_char < 0) |
size, gotten_store); |
762 |
|
|
763 |
|
fprintf(outfile, "Capturing subpattern count = %d\n", count); |
764 |
|
if (backrefmax > 0) |
765 |
|
fprintf(outfile, "Max back reference = %d\n", backrefmax); |
766 |
|
|
767 |
|
if (namecount > 0) |
768 |
|
{ |
769 |
|
fprintf(outfile, "Named capturing subpatterns:\n"); |
770 |
|
while (namecount-- > 0) |
771 |
{ |
{ |
772 |
fprintf(outfile, "No first char\n"); |
fprintf(outfile, " %s %*s%3d\n", nametable + 2, |
773 |
|
nameentrysize - 3 - (int)strlen((char *)nametable + 2), "", |
774 |
|
GET2(nametable, 0)); |
775 |
|
nametable += nameentrysize; |
776 |
} |
} |
777 |
|
} |
778 |
|
|
779 |
|
if (get_options == 0) fprintf(outfile, "No options\n"); |
780 |
|
else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s\n", |
781 |
|
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
782 |
|
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
783 |
|
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
784 |
|
((get_options & PCRE_MULTILINE) != 0)? " multiline" : "", |
785 |
|
((get_options & PCRE_DOTALL) != 0)? " dotall" : "", |
786 |
|
((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
787 |
|
((get_options & PCRE_EXTRA) != 0)? " extra" : "", |
788 |
|
((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "", |
789 |
|
((get_options & PCRE_UTF8) != 0)? " utf8" : ""); |
790 |
|
|
791 |
|
if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0) |
792 |
|
fprintf(outfile, "Case state changes\n"); |
793 |
|
|
794 |
|
if (first_char == -1) |
795 |
|
{ |
796 |
|
fprintf(outfile, "First char at start or follows \\n\n"); |
797 |
|
} |
798 |
|
else if (first_char < 0) |
799 |
|
{ |
800 |
|
fprintf(outfile, "No first char\n"); |
801 |
|
} |
802 |
|
else |
803 |
|
{ |
804 |
|
int ch = first_char & 255; |
805 |
|
const char *caseless = ((first_char & REQ_CASELESS) == 0)? |
806 |
|
"" : " (caseless)"; |
807 |
|
if (isprint(ch)) |
808 |
|
fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless); |
809 |
else |
else |
810 |
{ |
fprintf(outfile, "First char = %d%s\n", ch, caseless); |
811 |
if (isprint(first_char)) |
} |
812 |
fprintf(outfile, "First char = \'%c\'\n", first_char); |
|
813 |
else |
if (need_char < 0) |
814 |
fprintf(outfile, "First char = %d\n", first_char); |
{ |
815 |
} |
fprintf(outfile, "No need char\n"); |
816 |
|
} |
817 |
|
else |
818 |
|
{ |
819 |
|
int ch = need_char & 255; |
820 |
|
const char *caseless = ((need_char & REQ_CASELESS) == 0)? |
821 |
|
"" : " (caseless)"; |
822 |
|
if (isprint(ch)) |
823 |
|
fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless); |
824 |
|
else |
825 |
|
fprintf(outfile, "Need char = %d%s\n", ch, caseless); |
826 |
} |
} |
827 |
} |
} |
828 |
|
|
840 |
extra = pcre_study(re, study_options, &error); |
extra = pcre_study(re, study_options, &error); |
841 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
842 |
if (extra != NULL) free(extra); |
if (extra != NULL) free(extra); |
843 |
fprintf(outfile, " Study time %.2f milliseconds\n", |
fprintf(outfile, " Study time %.3f milliseconds\n", |
844 |
((double)time_taken)/(4 * CLOCKS_PER_SEC)); |
(((double)time_taken * 1000.0) / (double)LOOPREPEAT) / |
845 |
|
(double)CLOCKS_PER_SEC); |
846 |
} |
} |
847 |
|
|
848 |
extra = pcre_study(re, study_options, &error); |
extra = pcre_study(re, study_options, &error); |
851 |
else if (extra == NULL) |
else if (extra == NULL) |
852 |
fprintf(outfile, "Study returned NULL\n"); |
fprintf(outfile, "Study returned NULL\n"); |
853 |
|
|
854 |
/* This looks at internal information. A bit kludgy to do it this |
else if (do_showinfo) |
|
way, but it is useful for testing. */ |
|
|
|
|
|
else if (showinfo || do_debug) |
|
855 |
{ |
{ |
856 |
real_pcre_extra *xx = (real_pcre_extra *)extra; |
size_t size; |
857 |
if ((xx->options & PCRE_STUDY_MAPPED) == 0) |
uschar *start_bits = NULL; |
858 |
|
new_info(re, extra, PCRE_INFO_STUDYSIZE, &size); |
859 |
|
new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits); |
860 |
|
fprintf(outfile, "Study size = %d\n", size); |
861 |
|
if (start_bits == NULL) |
862 |
fprintf(outfile, "No starting character set\n"); |
fprintf(outfile, "No starting character set\n"); |
863 |
else |
else |
864 |
{ |
{ |
867 |
fprintf(outfile, "Starting character set: "); |
fprintf(outfile, "Starting character set: "); |
868 |
for (i = 0; i < 256; i++) |
for (i = 0; i < 256; i++) |
869 |
{ |
{ |
870 |
if ((xx->start_bits[i/8] & (1<<(i%8))) != 0) |
if ((start_bits[i/8] & (1<<(i%8))) != 0) |
871 |
{ |
{ |
872 |
if (c > 75) |
if (c > 75) |
873 |
{ |
{ |
897 |
for (;;) |
for (;;) |
898 |
{ |
{ |
899 |
unsigned char *q; |
unsigned char *q; |
900 |
|
unsigned char *bptr = dbuffer; |
901 |
|
int *use_offsets = offsets; |
902 |
|
int use_size_offsets = size_offsets; |
903 |
|
int callout_data = 0; |
904 |
|
int callout_data_set = 0; |
905 |
int count, c; |
int count, c; |
906 |
int offsets[45]; |
int copystrings = 0; |
907 |
int size_offsets = sizeof(offsets)/sizeof(int); |
int find_match_limit = 0; |
908 |
|
int getstrings = 0; |
909 |
|
int getlist = 0; |
910 |
|
int gmatched = 0; |
911 |
|
int start_offset = 0; |
912 |
|
int g_notempty = 0; |
913 |
|
|
914 |
options = 0; |
options = 0; |
915 |
|
|
916 |
if (infile == stdin) printf(" data> "); |
pcre_callout = callout; |
917 |
|
first_callout = 1; |
918 |
|
callout_extra = 0; |
919 |
|
callout_count = 0; |
920 |
|
callout_fail_count = 999999; |
921 |
|
callout_fail_id = -1; |
922 |
|
|
923 |
|
if (infile == stdin) printf("data> "); |
924 |
if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) |
if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) |
925 |
{ |
{ |
926 |
done = 1; |
done = 1; |
941 |
{ |
{ |
942 |
int i = 0; |
int i = 0; |
943 |
int n = 0; |
int n = 0; |
944 |
|
|
945 |
if (c == '\\') switch ((c = *p++)) |
if (c == '\\') switch ((c = *p++)) |
946 |
{ |
{ |
947 |
case 'a': c = 7; break; |
case 'a': c = 7; break; |
961 |
break; |
break; |
962 |
|
|
963 |
case 'x': |
case 'x': |
964 |
|
|
965 |
|
/* Handle \x{..} specially - new Perl thing for utf8 */ |
966 |
|
|
967 |
|
if (*p == '{') |
968 |
|
{ |
969 |
|
unsigned char *pt = p; |
970 |
|
c = 0; |
971 |
|
while (isxdigit(*(++pt))) |
972 |
|
c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W'); |
973 |
|
if (*pt == '}') |
974 |
|
{ |
975 |
|
unsigned char buff8[8]; |
976 |
|
int ii, utn; |
977 |
|
utn = ord2utf8(c, buff8); |
978 |
|
for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii]; |
979 |
|
c = buff8[ii]; /* Last byte */ |
980 |
|
p = pt + 1; |
981 |
|
break; |
982 |
|
} |
983 |
|
/* Not correct form; fall through */ |
984 |
|
} |
985 |
|
|
986 |
|
/* Ordinary \x */ |
987 |
|
|
988 |
c = 0; |
c = 0; |
989 |
while (i++ < 2 && isxdigit(*p)) |
while (i++ < 2 && isxdigit(*p)) |
990 |
{ |
{ |
1005 |
options |= PCRE_NOTBOL; |
options |= PCRE_NOTBOL; |
1006 |
continue; |
continue; |
1007 |
|
|
1008 |
|
case 'C': |
1009 |
|
if (isdigit(*p)) /* Set copy string */ |
1010 |
|
{ |
1011 |
|
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
1012 |
|
copystrings |= 1 << n; |
1013 |
|
} |
1014 |
|
else if (isalnum(*p)) |
1015 |
|
{ |
1016 |
|
uschar name[256]; |
1017 |
|
uschar *npp = name; |
1018 |
|
while (isalnum(*p)) *npp++ = *p++; |
1019 |
|
*npp = 0; |
1020 |
|
n = pcre_get_stringnumber(re, (char *)name); |
1021 |
|
if (n < 0) |
1022 |
|
fprintf(outfile, "no parentheses with name \"%s\"\n", name); |
1023 |
|
else copystrings |= 1 << n; |
1024 |
|
} |
1025 |
|
else if (*p == '+') |
1026 |
|
{ |
1027 |
|
callout_extra = 1; |
1028 |
|
p++; |
1029 |
|
} |
1030 |
|
else if (*p == '-') |
1031 |
|
{ |
1032 |
|
pcre_callout = NULL; |
1033 |
|
p++; |
1034 |
|
} |
1035 |
|
else if (*p == '!') |
1036 |
|
{ |
1037 |
|
callout_fail_id = 0; |
1038 |
|
p++; |
1039 |
|
while(isdigit(*p)) |
1040 |
|
callout_fail_id = callout_fail_id * 10 + *p++ - '0'; |
1041 |
|
callout_fail_count = 0; |
1042 |
|
if (*p == '!') |
1043 |
|
{ |
1044 |
|
p++; |
1045 |
|
while(isdigit(*p)) |
1046 |
|
callout_fail_count = callout_fail_count * 10 + *p++ - '0'; |
1047 |
|
} |
1048 |
|
} |
1049 |
|
else if (*p == '*') |
1050 |
|
{ |
1051 |
|
int sign = 1; |
1052 |
|
callout_data = 0; |
1053 |
|
if (*(++p) == '-') { sign = -1; p++; } |
1054 |
|
while(isdigit(*p)) |
1055 |
|
callout_data = callout_data * 10 + *p++ - '0'; |
1056 |
|
callout_data *= sign; |
1057 |
|
callout_data_set = 1; |
1058 |
|
} |
1059 |
|
continue; |
1060 |
|
|
1061 |
|
case 'G': |
1062 |
|
if (isdigit(*p)) |
1063 |
|
{ |
1064 |
|
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
1065 |
|
getstrings |= 1 << n; |
1066 |
|
} |
1067 |
|
else if (isalnum(*p)) |
1068 |
|
{ |
1069 |
|
uschar name[256]; |
1070 |
|
uschar *npp = name; |
1071 |
|
while (isalnum(*p)) *npp++ = *p++; |
1072 |
|
*npp = 0; |
1073 |
|
n = pcre_get_stringnumber(re, (char *)name); |
1074 |
|
if (n < 0) |
1075 |
|
fprintf(outfile, "no parentheses with name \"%s\"\n", name); |
1076 |
|
else getstrings |= 1 << n; |
1077 |
|
} |
1078 |
|
continue; |
1079 |
|
|
1080 |
|
case 'L': |
1081 |
|
getlist = 1; |
1082 |
|
continue; |
1083 |
|
|
1084 |
|
case 'M': |
1085 |
|
find_match_limit = 1; |
1086 |
|
continue; |
1087 |
|
|
1088 |
|
case 'N': |
1089 |
|
options |= PCRE_NOTEMPTY; |
1090 |
|
continue; |
1091 |
|
|
1092 |
case 'O': |
case 'O': |
1093 |
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
1094 |
if (n <= (int)(sizeof(offsets)/sizeof(int))) size_offsets = n; |
if (n > size_offsets_max) |
1095 |
|
{ |
1096 |
|
size_offsets_max = n; |
1097 |
|
free(offsets); |
1098 |
|
use_offsets = offsets = malloc(size_offsets_max * sizeof(int)); |
1099 |
|
if (offsets == NULL) |
1100 |
|
{ |
1101 |
|
printf("** Failed to get %d bytes of memory for offsets vector\n", |
1102 |
|
size_offsets_max * sizeof(int)); |
1103 |
|
return 1; |
1104 |
|
} |
1105 |
|
} |
1106 |
|
use_size_offsets = n; |
1107 |
|
if (n == 0) use_offsets = NULL; /* Ensures it can't write to it */ |
1108 |
continue; |
continue; |
1109 |
|
|
1110 |
case 'Z': |
case 'Z': |
1117 |
len = q - dbuffer; |
len = q - dbuffer; |
1118 |
|
|
1119 |
/* Handle matching via the POSIX interface, which does not |
/* Handle matching via the POSIX interface, which does not |
1120 |
support timing. */ |
support timing or playing with the match limit or callout data. */ |
1121 |
|
|
1122 |
|
#if !defined NOPOSIX |
1123 |
if (posix || do_posix) |
if (posix || do_posix) |
1124 |
{ |
{ |
1125 |
int rc; |
int rc; |
1126 |
int eflags = 0; |
int eflags = 0; |
1127 |
regmatch_t pmatch[30]; |
regmatch_t *pmatch = NULL; |
1128 |
|
if (use_size_offsets > 0) |
1129 |
|
pmatch = malloc(sizeof(regmatch_t) * use_size_offsets); |
1130 |
if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL; |
if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL; |
1131 |
if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL; |
if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL; |
1132 |
|
|
1133 |
rc = regexec(&preg, (char *)dbuffer, sizeof(pmatch)/sizeof(regmatch_t), |
rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags); |
|
pmatch, eflags); |
|
1134 |
|
|
1135 |
if (rc != 0) |
if (rc != 0) |
1136 |
{ |
{ |
1140 |
else |
else |
1141 |
{ |
{ |
1142 |
size_t i; |
size_t i; |
1143 |
for (i = 0; i < sizeof(pmatch)/sizeof(regmatch_t); i++) |
for (i = 0; i < (size_t)use_size_offsets; i++) |
1144 |
{ |
{ |
1145 |
if (pmatch[i].rm_so >= 0) |
if (pmatch[i].rm_so >= 0) |
1146 |
{ |
{ |
1147 |
fprintf(outfile, "%2d: ", (int)i); |
fprintf(outfile, "%2d: ", (int)i); |
1148 |
pchars(dbuffer + pmatch[i].rm_so, |
(void)pchars(dbuffer + pmatch[i].rm_so, |
1149 |
pmatch[i].rm_eo - pmatch[i].rm_so); |
pmatch[i].rm_eo - pmatch[i].rm_so, outfile); |
1150 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
1151 |
|
if (i == 0 && do_showrest) |
1152 |
|
{ |
1153 |
|
fprintf(outfile, " 0+ "); |
1154 |
|
(void)pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo, |
1155 |
|
outfile); |
1156 |
|
fprintf(outfile, "\n"); |
1157 |
|
} |
1158 |
} |
} |
1159 |
} |
} |
1160 |
} |
} |
1161 |
|
free(pmatch); |
1162 |
} |
} |
1163 |
|
|
1164 |
/* Handle matching via the native interface */ |
/* Handle matching via the native interface - repeats for /g and /G */ |
1165 |
|
|
1166 |
else |
else |
1167 |
|
#endif /* !defined NOPOSIX */ |
1168 |
|
|
1169 |
|
for (;; gmatched++) /* Loop for /g or /G */ |
1170 |
{ |
{ |
1171 |
if (timeit) |
if (timeit) |
1172 |
{ |
{ |
1173 |
register int i; |
register int i; |
1174 |
clock_t time_taken; |
clock_t time_taken; |
1175 |
clock_t start_time = clock(); |
clock_t start_time = clock(); |
1176 |
for (i = 0; i < 4000; i++) |
for (i = 0; i < LOOPREPEAT; i++) |
1177 |
count = pcre_exec(re, extra, (char *)dbuffer, len, options, offsets, |
count = pcre_exec(re, extra, (char *)bptr, len, |
1178 |
size_offsets); |
start_offset, options | g_notempty, use_offsets, use_size_offsets); |
1179 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
1180 |
fprintf(outfile, "Execute time %.2f milliseconds\n", |
fprintf(outfile, "Execute time %.3f milliseconds\n", |
1181 |
((double)time_taken)/(4 * CLOCKS_PER_SEC)); |
(((double)time_taken * 1000.0) / (double)LOOPREPEAT) / |
1182 |
|
(double)CLOCKS_PER_SEC); |
1183 |
} |
} |
1184 |
|
|
1185 |
count = pcre_exec(re, extra, (char *)dbuffer, len, options, offsets, |
/* If find_match_limit is set, we want to do repeated matches with |
1186 |
size_offsets); |
varying limits in order to find the minimum value. */ |
1187 |
|
|
1188 |
|
if (find_match_limit) |
1189 |
|
{ |
1190 |
|
int min = 0; |
1191 |
|
int mid = 64; |
1192 |
|
int max = -1; |
1193 |
|
|
1194 |
|
if (extra == NULL) |
1195 |
|
{ |
1196 |
|
extra = malloc(sizeof(pcre_extra)); |
1197 |
|
extra->flags = 0; |
1198 |
|
} |
1199 |
|
extra->flags |= PCRE_EXTRA_MATCH_LIMIT; |
1200 |
|
|
1201 |
|
for (;;) |
1202 |
|
{ |
1203 |
|
extra->match_limit = mid; |
1204 |
|
count = pcre_exec(re, extra, (char *)bptr, len, start_offset, |
1205 |
|
options | g_notempty, use_offsets, use_size_offsets); |
1206 |
|
if (count == PCRE_ERROR_MATCHLIMIT) |
1207 |
|
{ |
1208 |
|
/* fprintf(outfile, "Testing match limit = %d\n", mid); */ |
1209 |
|
min = mid; |
1210 |
|
mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2; |
1211 |
|
} |
1212 |
|
else if (count >= 0 || count == PCRE_ERROR_NOMATCH) |
1213 |
|
{ |
1214 |
|
if (mid == min + 1) |
1215 |
|
{ |
1216 |
|
fprintf(outfile, "Minimum match limit = %d\n", mid); |
1217 |
|
break; |
1218 |
|
} |
1219 |
|
/* fprintf(outfile, "Testing match limit = %d\n", mid); */ |
1220 |
|
max = mid; |
1221 |
|
mid = (min + mid)/2; |
1222 |
|
} |
1223 |
|
else break; /* Some other error */ |
1224 |
|
} |
1225 |
|
|
1226 |
|
extra->flags &= ~PCRE_EXTRA_MATCH_LIMIT; |
1227 |
|
} |
1228 |
|
|
1229 |
|
/* If callout_data is set, use the interface with additional data */ |
1230 |
|
|
1231 |
|
else if (callout_data_set) |
1232 |
|
{ |
1233 |
|
if (extra == NULL) |
1234 |
|
{ |
1235 |
|
extra = malloc(sizeof(pcre_extra)); |
1236 |
|
extra->flags = 0; |
1237 |
|
} |
1238 |
|
extra->flags |= PCRE_EXTRA_CALLOUT_DATA; |
1239 |
|
extra->callout_data = (void *)callout_data; |
1240 |
|
count = pcre_exec(re, extra, (char *)bptr, len, start_offset, |
1241 |
|
options | g_notempty, use_offsets, use_size_offsets); |
1242 |
|
extra->flags &= ~PCRE_EXTRA_CALLOUT_DATA; |
1243 |
|
} |
1244 |
|
|
1245 |
|
/* The normal case is just to do the match once, with the default |
1246 |
|
value of match_limit. */ |
1247 |
|
|
1248 |
|
else count = pcre_exec(re, extra, (char *)bptr, len, |
1249 |
|
start_offset, options | g_notempty, use_offsets, use_size_offsets); |
1250 |
|
|
1251 |
if (count == 0) |
if (count == 0) |
1252 |
{ |
{ |
1253 |
fprintf(outfile, "Matched, but too many substrings\n"); |
fprintf(outfile, "Matched, but too many substrings\n"); |
1254 |
count = size_offsets/3; |
count = use_size_offsets/3; |
1255 |
} |
} |
1256 |
|
|
1257 |
|
/* Matched */ |
1258 |
|
|
1259 |
if (count >= 0) |
if (count >= 0) |
1260 |
{ |
{ |
1261 |
int i; |
int i; |
1262 |
count *= 2; |
for (i = 0; i < count * 2; i += 2) |
|
for (i = 0; i < count; i += 2) |
|
1263 |
{ |
{ |
1264 |
if (offsets[i] < 0) |
if (use_offsets[i] < 0) |
1265 |
fprintf(outfile, "%2d: <unset>\n", i/2); |
fprintf(outfile, "%2d: <unset>\n", i/2); |
1266 |
else |
else |
1267 |
{ |
{ |
1268 |
fprintf(outfile, "%2d: ", i/2); |
fprintf(outfile, "%2d: ", i/2); |
1269 |
pchars(dbuffer + offsets[i], offsets[i+1] - offsets[i]); |
(void)pchars(bptr + use_offsets[i], |
1270 |
|
use_offsets[i+1] - use_offsets[i], outfile); |
1271 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
1272 |
|
if (i == 0) |
1273 |
|
{ |
1274 |
|
if (do_showrest) |
1275 |
|
{ |
1276 |
|
fprintf(outfile, " 0+ "); |
1277 |
|
(void)pchars(bptr + use_offsets[i+1], len - use_offsets[i+1], |
1278 |
|
outfile); |
1279 |
|
fprintf(outfile, "\n"); |
1280 |
|
} |
1281 |
|
} |
1282 |
|
} |
1283 |
|
} |
1284 |
|
|
1285 |
|
for (i = 0; i < 32; i++) |
1286 |
|
{ |
1287 |
|
if ((copystrings & (1 << i)) != 0) |
1288 |
|
{ |
1289 |
|
char copybuffer[16]; |
1290 |
|
int rc = pcre_copy_substring((char *)bptr, use_offsets, count, |
1291 |
|
i, copybuffer, sizeof(copybuffer)); |
1292 |
|
if (rc < 0) |
1293 |
|
fprintf(outfile, "copy substring %d failed %d\n", i, rc); |
1294 |
|
else |
1295 |
|
fprintf(outfile, "%2dC %s (%d)\n", i, copybuffer, rc); |
1296 |
|
} |
1297 |
|
} |
1298 |
|
|
1299 |
|
for (i = 0; i < 32; i++) |
1300 |
|
{ |
1301 |
|
if ((getstrings & (1 << i)) != 0) |
1302 |
|
{ |
1303 |
|
const char *substring; |
1304 |
|
int rc = pcre_get_substring((char *)bptr, use_offsets, count, |
1305 |
|
i, &substring); |
1306 |
|
if (rc < 0) |
1307 |
|
fprintf(outfile, "get substring %d failed %d\n", i, rc); |
1308 |
|
else |
1309 |
|
{ |
1310 |
|
fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc); |
1311 |
|
/* free((void *)substring); */ |
1312 |
|
pcre_free_substring(substring); |
1313 |
|
} |
1314 |
|
} |
1315 |
|
} |
1316 |
|
|
1317 |
|
if (getlist) |
1318 |
|
{ |
1319 |
|
const char **stringlist; |
1320 |
|
int rc = pcre_get_substring_list((char *)bptr, use_offsets, count, |
1321 |
|
&stringlist); |
1322 |
|
if (rc < 0) |
1323 |
|
fprintf(outfile, "get substring list failed %d\n", rc); |
1324 |
|
else |
1325 |
|
{ |
1326 |
|
for (i = 0; i < count; i++) |
1327 |
|
fprintf(outfile, "%2dL %s\n", i, stringlist[i]); |
1328 |
|
if (stringlist[i] != NULL) |
1329 |
|
fprintf(outfile, "string list not terminated by NULL\n"); |
1330 |
|
/* free((void *)stringlist); */ |
1331 |
|
pcre_free_substring_list(stringlist); |
1332 |
} |
} |
1333 |
} |
} |
1334 |
} |
} |
1335 |
|
|
1336 |
|
/* Failed to match. If this is a /g or /G loop and we previously set |
1337 |
|
g_notempty after a null match, this is not necessarily the end. |
1338 |
|
We want to advance the start offset, and continue. Fudge the offset |
1339 |
|
values to achieve this. We won't be at the end of the string - that |
1340 |
|
was checked before setting g_notempty. */ |
1341 |
|
|
1342 |
else |
else |
1343 |
{ |
{ |
1344 |
if (count == -1) fprintf(outfile, "No match\n"); |
if (g_notempty != 0) |
1345 |
else fprintf(outfile, "Error %d\n", count); |
{ |
1346 |
|
use_offsets[0] = start_offset; |
1347 |
|
use_offsets[1] = start_offset + 1; |
1348 |
|
} |
1349 |
|
else |
1350 |
|
{ |
1351 |
|
if (gmatched == 0) /* Error if no previous matches */ |
1352 |
|
{ |
1353 |
|
if (count == -1) fprintf(outfile, "No match\n"); |
1354 |
|
else fprintf(outfile, "Error %d\n", count); |
1355 |
|
} |
1356 |
|
break; /* Out of the /g loop */ |
1357 |
|
} |
1358 |
} |
} |
1359 |
} |
|
1360 |
} |
/* If not /g or /G we are done */ |
1361 |
|
|
1362 |
|
if (!do_g && !do_G) break; |
1363 |
|
|
1364 |
|
/* If we have matched an empty string, first check to see if we are at |
1365 |
|
the end of the subject. If so, the /g loop is over. Otherwise, mimic |
1366 |
|
what Perl's /g options does. This turns out to be rather cunning. First |
1367 |
|
we set PCRE_NOTEMPTY and PCRE_ANCHORED and try the match again at the |
1368 |
|
same point. If this fails (picked up above) we advance to the next |
1369 |
|
character. */ |
1370 |
|
|
1371 |
|
g_notempty = 0; |
1372 |
|
if (use_offsets[0] == use_offsets[1]) |
1373 |
|
{ |
1374 |
|
if (use_offsets[0] == len) break; |
1375 |
|
g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED; |
1376 |
|
} |
1377 |
|
|
1378 |
|
/* For /g, update the start offset, leaving the rest alone */ |
1379 |
|
|
1380 |
|
if (do_g) start_offset = use_offsets[1]; |
1381 |
|
|
1382 |
|
/* For /G, update the pointer and length */ |
1383 |
|
|
1384 |
|
else |
1385 |
|
{ |
1386 |
|
bptr += use_offsets[1]; |
1387 |
|
len -= use_offsets[1]; |
1388 |
|
} |
1389 |
|
} /* End of loop for /g and /G */ |
1390 |
|
} /* End of loop for data lines */ |
1391 |
|
|
1392 |
CONTINUE: |
CONTINUE: |
1393 |
|
|
1394 |
|
#if !defined NOPOSIX |
1395 |
if (posix || do_posix) regfree(&preg); |
if (posix || do_posix) regfree(&preg); |
1396 |
|
#endif |
1397 |
|
|
1398 |
if (re != NULL) free(re); |
if (re != NULL) free(re); |
1399 |
if (extra != NULL) free(extra); |
if (extra != NULL) free(extra); |
1400 |
|
if (tables != NULL) |
1401 |
|
{ |
1402 |
|
free((void *)tables); |
1403 |
|
setlocale(LC_CTYPE, "C"); |
1404 |
|
} |
1405 |
} |
} |
1406 |
|
|
1407 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |