| 36 |
*/ |
*/ |
| 37 |
|
|
| 38 |
|
|
| 39 |
|
#ifdef HAVE_CONFIG_H |
| 40 |
|
#include <config.h> |
| 41 |
|
#endif |
| 42 |
|
|
| 43 |
#include <ctype.h> |
#include <ctype.h> |
| 44 |
#include <stdio.h> |
#include <stdio.h> |
| 45 |
#include <string.h> |
#include <string.h> |
| 71 |
#endif |
#endif |
| 72 |
|
|
| 73 |
|
|
| 74 |
#define PCRE_SPY /* For Win32 build, import data, not export */ |
/* We have to include pcre_internal.h because we need the internal info for |
| 75 |
|
displaying the results of pcre_study() and we also need to know about the |
| 76 |
/* We include pcre_internal.h because we need the internal info for displaying |
internal macros, structures, and other internal data values; pcretest has |
| 77 |
the results of pcre_study() and we also need to know about the internal |
"inside information" compared to a program that strictly follows the PCRE API. |
| 78 |
macros, structures, and other internal data values; pcretest has "inside |
|
| 79 |
information" compared to a program that strictly follows the PCRE API. */ |
Although pcre_internal.h does itself include pcre.h, we explicitly include it |
| 80 |
|
here before pcre_internal.h so that the PCRE_EXP_xxx macros get set |
| 81 |
|
appropriately for an application, not for building PCRE. */ |
| 82 |
|
|
| 83 |
|
#include "pcre.h" |
| 84 |
#include "pcre_internal.h" |
#include "pcre_internal.h" |
| 85 |
|
|
| 86 |
/* We need access to the data tables that PCRE uses. So as not to have to keep |
/* We need access to the data tables that PCRE uses. So as not to have to keep |
| 121 |
#include "pcreposix.h" |
#include "pcreposix.h" |
| 122 |
#endif |
#endif |
| 123 |
|
|
| 124 |
/* It is also possible, for the benefit of the version imported into Exim, to |
/* It is also possible, for the benefit of the version currently imported into |
| 125 |
build pcretest without support for UTF8 (define NOUTF8), without the interface |
Exim, to build pcretest without support for UTF8 (define NOUTF8), without the |
| 126 |
to the DFA matcher (NODFA), and without the doublecheck of the old "info" |
interface to the DFA matcher (NODFA), and without the doublecheck of the old |
| 127 |
function (define NOINFOCHECK). */ |
"info" function (define NOINFOCHECK). In fact, we automatically cut out the |
| 128 |
|
UTF8 support if PCRE is built without it. */ |
| 129 |
|
|
| 130 |
|
#ifndef SUPPORT_UTF8 |
| 131 |
|
#ifndef NOUTF8 |
| 132 |
|
#define NOUTF8 |
| 133 |
|
#endif |
| 134 |
|
#endif |
| 135 |
|
|
| 136 |
|
|
| 137 |
/* Other parameters */ |
/* Other parameters */ |
| 667 |
*************************************************/ |
*************************************************/ |
| 668 |
|
|
| 669 |
/* This is used both at compile and run-time to check for <xxx> escapes, where |
/* This is used both at compile and run-time to check for <xxx> escapes, where |
| 670 |
xxx is LF, CR, CRLF, or ANY. Print a message and return 0 if there is no match. |
xxx is LF, CR, CRLF, ANYCRLF, or ANY. Print a message and return 0 if there is |
| 671 |
|
no match. |
| 672 |
|
|
| 673 |
Arguments: |
Arguments: |
| 674 |
p points after the leading '<' |
p points after the leading '<' |
| 683 |
if (strncmp((char *)p, "cr>", 3) == 0) return PCRE_NEWLINE_CR; |
if (strncmp((char *)p, "cr>", 3) == 0) return PCRE_NEWLINE_CR; |
| 684 |
if (strncmp((char *)p, "lf>", 3) == 0) return PCRE_NEWLINE_LF; |
if (strncmp((char *)p, "lf>", 3) == 0) return PCRE_NEWLINE_LF; |
| 685 |
if (strncmp((char *)p, "crlf>", 5) == 0) return PCRE_NEWLINE_CRLF; |
if (strncmp((char *)p, "crlf>", 5) == 0) return PCRE_NEWLINE_CRLF; |
| 686 |
|
if (strncmp((char *)p, "anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF; |
| 687 |
if (strncmp((char *)p, "any>", 4) == 0) return PCRE_NEWLINE_ANY; |
if (strncmp((char *)p, "any>", 4) == 0) return PCRE_NEWLINE_ANY; |
| 688 |
fprintf(f, "Unknown newline type at: <%s\n", p); |
fprintf(f, "Unknown newline type at: <%s\n", p); |
| 689 |
return 0; |
return 0; |
| 856 |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
| 857 |
printf(" Newline sequence is %s\n", (rc == '\r')? "CR" : |
printf(" Newline sequence is %s\n", (rc == '\r')? "CR" : |
| 858 |
(rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" : |
(rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" : |
| 859 |
|
(rc == -2)? "ANYCRLF" : |
| 860 |
(rc == -1)? "ANY" : "???"); |
(rc == -1)? "ANY" : "???"); |
| 861 |
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
| 862 |
printf(" Internal link size = %d\n", rc); |
printf(" Internal link size = %d\n", rc); |
| 868 |
printf(" Default recursion depth limit = %d\n", rc); |
printf(" Default recursion depth limit = %d\n", rc); |
| 869 |
(void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc); |
(void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc); |
| 870 |
printf(" Match recursion uses %s\n", rc? "stack" : "heap"); |
printf(" Match recursion uses %s\n", rc? "stack" : "heap"); |
| 871 |
exit(0); |
goto EXIT; |
| 872 |
} |
} |
| 873 |
else if (strcmp(argv[op], "-help") == 0 || |
else if (strcmp(argv[op], "-help") == 0 || |
| 874 |
strcmp(argv[op], "--help") == 0) |
strcmp(argv[op], "--help") == 0) |
| 894 |
if (offsets == NULL) |
if (offsets == NULL) |
| 895 |
{ |
{ |
| 896 |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
| 897 |
size_offsets_max * sizeof(int)); |
(int)(size_offsets_max * sizeof(int))); |
| 898 |
yield = 1; |
yield = 1; |
| 899 |
goto EXIT; |
goto EXIT; |
| 900 |
} |
} |
| 954 |
size_t size, regex_gotten_store; |
size_t size, regex_gotten_store; |
| 955 |
int do_study = 0; |
int do_study = 0; |
| 956 |
int do_debug = debug; |
int do_debug = debug; |
| 957 |
|
int debug_lengths = 1; |
| 958 |
int do_G = 0; |
int do_G = 0; |
| 959 |
int do_g = 0; |
int do_g = 0; |
| 960 |
int do_showinfo = showinfo; |
int do_showinfo = showinfo; |
| 1145 |
case 'S': do_study = 1; break; |
case 'S': do_study = 1; break; |
| 1146 |
case 'U': options |= PCRE_UNGREEDY; break; |
case 'U': options |= PCRE_UNGREEDY; break; |
| 1147 |
case 'X': options |= PCRE_EXTRA; break; |
case 'X': options |= PCRE_EXTRA; break; |
| 1148 |
|
case 'Z': debug_lengths = 0; break; |
| 1149 |
case '8': options |= PCRE_UTF8; use_utf8 = 1; break; |
case '8': options |= PCRE_UTF8; use_utf8 = 1; break; |
| 1150 |
case '?': options |= PCRE_NO_UTF8_CHECK; break; |
case '?': options |= PCRE_NO_UTF8_CHECK; break; |
| 1151 |
|
|
| 1347 |
if (do_debug) |
if (do_debug) |
| 1348 |
{ |
{ |
| 1349 |
fprintf(outfile, "------------------------------------------------------------------\n"); |
fprintf(outfile, "------------------------------------------------------------------\n"); |
| 1350 |
pcre_printint(re, outfile); |
pcre_printint(re, outfile, debug_lengths); |
| 1351 |
} |
} |
| 1352 |
|
|
| 1353 |
if (do_showinfo) |
if (do_showinfo) |
| 1356 |
#if !defined NOINFOCHECK |
#if !defined NOINFOCHECK |
| 1357 |
int old_first_char, old_options, old_count; |
int old_first_char, old_options, old_count; |
| 1358 |
#endif |
#endif |
| 1359 |
int count, backrefmax, first_char, need_char; |
int count, backrefmax, first_char, need_char, okpartial, jchanged; |
| 1360 |
int nameentrysize, namecount; |
int nameentrysize, namecount; |
| 1361 |
const uschar *nametable; |
const uschar *nametable; |
| 1362 |
|
|
| 1369 |
new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize); |
new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize); |
| 1370 |
new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount); |
new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount); |
| 1371 |
new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable); |
new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable); |
| 1372 |
|
new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial); |
| 1373 |
|
new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged); |
| 1374 |
|
|
| 1375 |
#if !defined NOINFOCHECK |
#if !defined NOINFOCHECK |
| 1376 |
old_count = pcre_info(re, &old_options, &old_first_char); |
old_count = pcre_info(re, &old_options, &old_first_char); |
| 1412 |
} |
} |
| 1413 |
} |
} |
| 1414 |
|
|
| 1415 |
/* The NOPARTIAL bit is a private bit in the options, so we have |
if (!okpartial) fprintf(outfile, "Partial matching not supported\n"); |
|
to fish it out via out back door */ |
|
| 1416 |
|
|
| 1417 |
all_options = ((real_pcre *)re)->options; |
all_options = ((real_pcre *)re)->options; |
| 1418 |
if (do_flip) |
if (do_flip) all_options = byteflip(all_options, sizeof(all_options)); |
|
{ |
|
|
all_options = byteflip(all_options, sizeof(all_options)); |
|
|
} |
|
|
|
|
|
if ((all_options & PCRE_NOPARTIAL) != 0) |
|
|
fprintf(outfile, "Partial matching not supported\n"); |
|
| 1419 |
|
|
| 1420 |
if (get_options == 0) fprintf(outfile, "No options\n"); |
if (get_options == 0) fprintf(outfile, "No options\n"); |
| 1421 |
else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s%s%s\n", |
else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s%s%s\n", |
| 1433 |
((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "", |
((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "", |
| 1434 |
((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : ""); |
((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : ""); |
| 1435 |
|
|
| 1436 |
|
if (jchanged) fprintf(outfile, "Duplicate name status changes\n"); |
| 1437 |
|
|
| 1438 |
switch (get_options & PCRE_NEWLINE_BITS) |
switch (get_options & PCRE_NEWLINE_BITS) |
| 1439 |
{ |
{ |
| 1440 |
case PCRE_NEWLINE_CR: |
case PCRE_NEWLINE_CR: |
| 1449 |
fprintf(outfile, "Forced newline sequence: CRLF\n"); |
fprintf(outfile, "Forced newline sequence: CRLF\n"); |
| 1450 |
break; |
break; |
| 1451 |
|
|
| 1452 |
|
case PCRE_NEWLINE_ANYCRLF: |
| 1453 |
|
fprintf(outfile, "Forced newline sequence: ANYCRLF\n"); |
| 1454 |
|
break; |
| 1455 |
|
|
| 1456 |
case PCRE_NEWLINE_ANY: |
case PCRE_NEWLINE_ANY: |
| 1457 |
fprintf(outfile, "Forced newline sequence: ANY\n"); |
fprintf(outfile, "Forced newline sequence: ANY\n"); |
| 1458 |
break; |
break; |
| 1602 |
for (;;) |
for (;;) |
| 1603 |
{ |
{ |
| 1604 |
uschar *q; |
uschar *q; |
| 1605 |
uschar *bptr = dbuffer; |
uschar *bptr; |
| 1606 |
int *use_offsets = offsets; |
int *use_offsets = offsets; |
| 1607 |
int use_size_offsets = size_offsets; |
int use_size_offsets = size_offsets; |
| 1608 |
int callout_data = 0; |
int callout_data = 0; |
| 1658 |
p = buffer; |
p = buffer; |
| 1659 |
while (isspace(*p)) p++; |
while (isspace(*p)) p++; |
| 1660 |
|
|
| 1661 |
q = dbuffer; |
bptr = q = dbuffer; |
| 1662 |
while ((c = *p++) != 0) |
while ((c = *p++) != 0) |
| 1663 |
{ |
{ |
| 1664 |
int i = 0; |
int i = 0; |
| 1853 |
if (offsets == NULL) |
if (offsets == NULL) |
| 1854 |
{ |
{ |
| 1855 |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
| 1856 |
size_offsets_max * sizeof(int)); |
(int)(size_offsets_max * sizeof(int))); |
| 1857 |
yield = 1; |
yield = 1; |
| 1858 |
goto EXIT; |
goto EXIT; |
| 1859 |
} |
} |
| 2222 |
} |
} |
| 2223 |
|
|
| 2224 |
/* Failed to match. If this is a /g or /G loop and we previously set |
/* Failed to match. If this is a /g or /G loop and we previously set |
| 2225 |
g_notempty after a null match, this is not necessarily the end. |
g_notempty after a null match, this is not necessarily the end. We want |
| 2226 |
We want to advance the start offset, and continue. In the case of UTF-8 |
to advance the start offset, and continue. We won't be at the end of the |
| 2227 |
matching, the advance must be one character, not one byte. Fudge the |
string - that was checked before setting g_notempty. |
| 2228 |
offset values to achieve this. We won't be at the end of the string - |
|
| 2229 |
that was checked before setting g_notempty. */ |
Complication arises in the case when the newline option is "any" or |
| 2230 |
|
"anycrlf". If the previous match was at the end of a line terminated by |
| 2231 |
|
CRLF, an advance of one character just passes the \r, whereas we should |
| 2232 |
|
prefer the longer newline sequence, as does the code in pcre_exec(). |
| 2233 |
|
Fudge the offset value to achieve this. |
| 2234 |
|
|
| 2235 |
|
Otherwise, in the case of UTF-8 matching, the advance must be one |
| 2236 |
|
character, not one byte. */ |
| 2237 |
|
|
| 2238 |
else |
else |
| 2239 |
{ |
{ |
| 2240 |
if (g_notempty != 0) |
if (g_notempty != 0) |
| 2241 |
{ |
{ |
| 2242 |
int onechar = 1; |
int onechar = 1; |
| 2243 |
|
unsigned int obits = ((real_pcre *)re)->options; |
| 2244 |
use_offsets[0] = start_offset; |
use_offsets[0] = start_offset; |
| 2245 |
if (use_utf8) |
if ((obits & PCRE_NEWLINE_BITS) == 0) |
| 2246 |
|
{ |
| 2247 |
|
int d; |
| 2248 |
|
(void)pcre_config(PCRE_CONFIG_NEWLINE, &d); |
| 2249 |
|
obits = (d == '\r')? PCRE_NEWLINE_CR : |
| 2250 |
|
(d == '\n')? PCRE_NEWLINE_LF : |
| 2251 |
|
(d == ('\r'<<8 | '\n'))? PCRE_NEWLINE_CRLF : |
| 2252 |
|
(d == -2)? PCRE_NEWLINE_ANYCRLF : |
| 2253 |
|
(d == -1)? PCRE_NEWLINE_ANY : 0; |
| 2254 |
|
} |
| 2255 |
|
if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY || |
| 2256 |
|
(obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF) |
| 2257 |
|
&& |
| 2258 |
|
start_offset < len - 1 && |
| 2259 |
|
bptr[start_offset] == '\r' && |
| 2260 |
|
bptr[start_offset+1] == '\n') |
| 2261 |
|
onechar++; |
| 2262 |
|
else if (use_utf8) |
| 2263 |
{ |
{ |
| 2264 |
while (start_offset + onechar < len) |
while (start_offset + onechar < len) |
| 2265 |
{ |
{ |
| 2294 |
character. */ |
character. */ |
| 2295 |
|
|
| 2296 |
g_notempty = 0; |
g_notempty = 0; |
| 2297 |
|
|
| 2298 |
if (use_offsets[0] == use_offsets[1]) |
if (use_offsets[0] == use_offsets[1]) |
| 2299 |
{ |
{ |
| 2300 |
if (use_offsets[0] == len) break; |
if (use_offsets[0] == len) break; |