| 79 |
#define fileno _fileno |
#define fileno _fileno |
| 80 |
#endif |
#endif |
| 81 |
|
|
| 82 |
|
/* A user sent this fix for Borland Builder 5 under Windows. */ |
| 83 |
|
|
| 84 |
|
#ifdef __BORLANDC__ |
| 85 |
|
#define _setmode(handle, mode) setmode(handle, mode) |
| 86 |
|
#endif |
| 87 |
|
|
| 88 |
|
/* Not Windows */ |
| 89 |
|
|
| 90 |
#else |
#else |
| 91 |
#include <sys/time.h> /* These two includes are needed */ |
#include <sys/time.h> /* These two includes are needed */ |
| 92 |
#include <sys/resource.h> /* for setrlimit(). */ |
#include <sys/resource.h> /* for setrlimit(). */ |
| 112 |
external symbols to prevent clashes. */ |
external symbols to prevent clashes. */ |
| 113 |
|
|
| 114 |
#define _pcre_ucp_gentype ucp_gentype |
#define _pcre_ucp_gentype ucp_gentype |
| 115 |
|
#define _pcre_ucp_typerange ucp_typerange |
| 116 |
#define _pcre_utf8_table1 utf8_table1 |
#define _pcre_utf8_table1 utf8_table1 |
| 117 |
#define _pcre_utf8_table1_size utf8_table1_size |
#define _pcre_utf8_table1_size utf8_table1_size |
| 118 |
#define _pcre_utf8_table2 utf8_table2 |
#define _pcre_utf8_table2 utf8_table2 |
| 190 |
static int show_malloc; |
static int show_malloc; |
| 191 |
static int use_utf8; |
static int use_utf8; |
| 192 |
static size_t gotten_store; |
static size_t gotten_store; |
| 193 |
|
static size_t first_gotten_store = 0; |
| 194 |
|
static const unsigned char *last_callout_mark = NULL; |
| 195 |
|
|
| 196 |
/* The buffers grow automatically if very long input lines are encountered. */ |
/* The buffers grow automatically if very long input lines are encountered. */ |
| 197 |
|
|
| 198 |
static int buffer_size = 50000; |
static int buffer_size = 50000; |
| 199 |
static uschar *buffer = NULL; |
static pcre_uint8 *buffer = NULL; |
| 200 |
static uschar *dbuffer = NULL; |
static pcre_uint8 *dbuffer = NULL; |
| 201 |
static uschar *pbuffer = NULL; |
static pcre_uint8 *pbuffer = NULL; |
| 202 |
|
|
| 203 |
|
/* Textual explanations for runtime error codes */ |
| 204 |
|
|
| 205 |
|
static const char *errtexts[] = { |
| 206 |
|
NULL, /* 0 is no error */ |
| 207 |
|
NULL, /* NOMATCH is handled specially */ |
| 208 |
|
"NULL argument passed", |
| 209 |
|
"bad option value", |
| 210 |
|
"magic number missing", |
| 211 |
|
"unknown opcode - pattern overwritten?", |
| 212 |
|
"no more memory", |
| 213 |
|
NULL, /* never returned by pcre_exec() or pcre_dfa_exec() */ |
| 214 |
|
"match limit exceeded", |
| 215 |
|
"callout error code", |
| 216 |
|
NULL, /* BADUTF8 is handled specially */ |
| 217 |
|
"bad UTF-8 offset", |
| 218 |
|
NULL, /* PARTIAL is handled specially */ |
| 219 |
|
"not used - internal error", |
| 220 |
|
"internal error - pattern overwritten?", |
| 221 |
|
"bad count value", |
| 222 |
|
"item unsupported for DFA matching", |
| 223 |
|
"backreference condition or recursion test not supported for DFA matching", |
| 224 |
|
"match limit not supported for DFA matching", |
| 225 |
|
"workspace size exceeded in DFA matching", |
| 226 |
|
"too much recursion for DFA matching", |
| 227 |
|
"recursion limit exceeded", |
| 228 |
|
"not used - internal error", |
| 229 |
|
"invalid combination of newline options", |
| 230 |
|
"bad offset value", |
| 231 |
|
NULL, /* SHORTUTF8 is handled specially */ |
| 232 |
|
"nested recursion at the same subject position", |
| 233 |
|
"JIT stack limit reached" |
| 234 |
|
}; |
| 235 |
|
|
| 236 |
|
|
| 237 |
/************************************************* |
/************************************************* |
| 582 |
#endif /* HAVE_STRERROR */ |
#endif /* HAVE_STRERROR */ |
| 583 |
|
|
| 584 |
|
|
| 585 |
|
/************************************************* |
| 586 |
|
* JIT memory callback * |
| 587 |
|
*************************************************/ |
| 588 |
|
|
| 589 |
|
static pcre_jit_stack* jit_callback(void *arg) |
| 590 |
|
{ |
| 591 |
|
return (pcre_jit_stack *)arg; |
| 592 |
|
} |
| 593 |
|
|
| 594 |
|
|
| 595 |
/************************************************* |
/************************************************* |
| 615 |
NULL if no data read and EOF reached |
NULL if no data read and EOF reached |
| 616 |
*/ |
*/ |
| 617 |
|
|
| 618 |
static uschar * |
static pcre_uint8 * |
| 619 |
extend_inputline(FILE *f, uschar *start, const char *prompt) |
extend_inputline(FILE *f, pcre_uint8 *start, const char *prompt) |
| 620 |
{ |
{ |
| 621 |
uschar *here = start; |
pcre_uint8 *here = start; |
| 622 |
|
|
| 623 |
for (;;) |
for (;;) |
| 624 |
{ |
{ |
| 665 |
else |
else |
| 666 |
{ |
{ |
| 667 |
int new_buffer_size = 2*buffer_size; |
int new_buffer_size = 2*buffer_size; |
| 668 |
uschar *new_buffer = (unsigned char *)malloc(new_buffer_size); |
pcre_uint8 *new_buffer = (unsigned char *)malloc(new_buffer_size); |
| 669 |
uschar *new_dbuffer = (unsigned char *)malloc(new_buffer_size); |
pcre_uint8 *new_dbuffer = (unsigned char *)malloc(new_buffer_size); |
| 670 |
uschar *new_pbuffer = (unsigned char *)malloc(new_buffer_size); |
pcre_uint8 *new_pbuffer = (unsigned char *)malloc(new_buffer_size); |
| 671 |
|
|
| 672 |
if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL) |
if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL) |
| 673 |
{ |
{ |
| 809 |
#if !defined NOUTF8 |
#if !defined NOUTF8 |
| 810 |
|
|
| 811 |
static int |
static int |
| 812 |
ord2utf8(int cvalue, uschar *utf8bytes) |
ord2utf8(int cvalue, pcre_uint8 *utf8bytes) |
| 813 |
{ |
{ |
| 814 |
register int i, j; |
register int i, j; |
| 815 |
for (i = 0; i < utf8_table1_size; i++) |
for (i = 0; i < utf8_table1_size; i++) |
| 972 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
| 973 |
first_callout = 0; |
first_callout = 0; |
| 974 |
|
|
| 975 |
|
if (cb->mark != last_callout_mark) |
| 976 |
|
{ |
| 977 |
|
fprintf(outfile, "Latest Mark: %s\n", |
| 978 |
|
(cb->mark == NULL)? "<unset>" : (char *)(cb->mark)); |
| 979 |
|
last_callout_mark = cb->mark; |
| 980 |
|
} |
| 981 |
|
|
| 982 |
if (cb->callout_data != NULL) |
if (cb->callout_data != NULL) |
| 983 |
{ |
{ |
| 984 |
int callout_data = *((int *)(cb->callout_data)); |
int callout_data = *((int *)(cb->callout_data)); |
| 998 |
* Local malloc functions * |
* Local malloc functions * |
| 999 |
*************************************************/ |
*************************************************/ |
| 1000 |
|
|
| 1001 |
/* Alternative malloc function, to test functionality and show the size of the |
/* Alternative malloc function, to test functionality and save the size of a |
| 1002 |
compiled re. */ |
compiled re, which is the first store request that pcre_compile() makes. The |
| 1003 |
|
show_malloc variable is set only during matching. */ |
| 1004 |
|
|
| 1005 |
static void *new_malloc(size_t size) |
static void *new_malloc(size_t size) |
| 1006 |
{ |
{ |
| 1007 |
void *block = malloc(size); |
void *block = malloc(size); |
| 1008 |
gotten_store = size; |
gotten_store = size; |
| 1009 |
|
if (first_gotten_store == 0) first_gotten_store = size; |
| 1010 |
if (show_malloc) |
if (show_malloc) |
| 1011 |
fprintf(outfile, "malloc %3d %p\n", (int)size, block); |
fprintf(outfile, "malloc %3d %p\n", (int)size, block); |
| 1012 |
return block; |
return block; |
| 1019 |
free(block); |
free(block); |
| 1020 |
} |
} |
| 1021 |
|
|
|
|
|
| 1022 |
/* For recursion malloc/free, to test stacking calls */ |
/* For recursion malloc/free, to test stacking calls */ |
| 1023 |
|
|
| 1024 |
static void *stack_malloc(size_t size) |
static void *stack_malloc(size_t size) |
| 1074 |
*************************************************/ |
*************************************************/ |
| 1075 |
|
|
| 1076 |
static int |
static int |
| 1077 |
check_match_limit(pcre *re, pcre_extra *extra, uschar *bptr, int len, |
check_match_limit(pcre *re, pcre_extra *extra, pcre_uint8 *bptr, int len, |
| 1078 |
int start_offset, int options, int *use_offsets, int use_size_offsets, |
int start_offset, int options, int *use_offsets, int use_size_offsets, |
| 1079 |
int flag, unsigned long int *limit, int errnumber, const char *msg) |
int flag, unsigned long int *limit, int errnumber, const char *msg) |
| 1080 |
{ |
{ |
| 1134 |
*/ |
*/ |
| 1135 |
|
|
| 1136 |
static int |
static int |
| 1137 |
strncmpic(uschar *s, uschar *t, int n) |
strncmpic(pcre_uint8 *s, pcre_uint8 *t, int n) |
| 1138 |
{ |
{ |
| 1139 |
while (n--) |
while (n--) |
| 1140 |
{ |
{ |
| 1161 |
*/ |
*/ |
| 1162 |
|
|
| 1163 |
static int |
static int |
| 1164 |
check_newline(uschar *p, FILE *f) |
check_newline(pcre_uint8 *p, FILE *f) |
| 1165 |
{ |
{ |
| 1166 |
if (strncmpic(p, (uschar *)"cr>", 3) == 0) return PCRE_NEWLINE_CR; |
if (strncmpic(p, (pcre_uint8 *)"cr>", 3) == 0) return PCRE_NEWLINE_CR; |
| 1167 |
if (strncmpic(p, (uschar *)"lf>", 3) == 0) return PCRE_NEWLINE_LF; |
if (strncmpic(p, (pcre_uint8 *)"lf>", 3) == 0) return PCRE_NEWLINE_LF; |
| 1168 |
if (strncmpic(p, (uschar *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF; |
if (strncmpic(p, (pcre_uint8 *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF; |
| 1169 |
if (strncmpic(p, (uschar *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF; |
if (strncmpic(p, (pcre_uint8 *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF; |
| 1170 |
if (strncmpic(p, (uschar *)"any>", 4) == 0) return PCRE_NEWLINE_ANY; |
if (strncmpic(p, (pcre_uint8 *)"any>", 4) == 0) return PCRE_NEWLINE_ANY; |
| 1171 |
if (strncmpic(p, (uschar *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF; |
if (strncmpic(p, (pcre_uint8 *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF; |
| 1172 |
if (strncmpic(p, (uschar *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE; |
if (strncmpic(p, (pcre_uint8 *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE; |
| 1173 |
fprintf(f, "Unknown newline type at: <%s\n", p); |
fprintf(f, "Unknown newline type at: <%s\n", p); |
| 1174 |
return 0; |
return 0; |
| 1175 |
} |
} |
| 1207 |
#endif |
#endif |
| 1208 |
printf(" -q quiet: do not output PCRE version number at start\n"); |
printf(" -q quiet: do not output PCRE version number at start\n"); |
| 1209 |
printf(" -S <n> set stack size to <n> megabytes\n"); |
printf(" -S <n> set stack size to <n> megabytes\n"); |
| 1210 |
printf(" -s output store (memory) used information\n" |
printf(" -s force each pattern to be studied at basic level\n" |
| 1211 |
|
" -s+ force each pattern to be studied, using JIT if available\n" |
| 1212 |
" -t time compilation and execution\n"); |
" -t time compilation and execution\n"); |
| 1213 |
printf(" -t <n> time compilation and execution, repeating <n> times\n"); |
printf(" -t <n> time compilation and execution, repeating <n> times\n"); |
| 1214 |
printf(" -tm time execution (matching) only\n"); |
printf(" -tm time execution (matching) only\n"); |
| 1236 |
int timeitm = 0; |
int timeitm = 0; |
| 1237 |
int showinfo = 0; |
int showinfo = 0; |
| 1238 |
int showstore = 0; |
int showstore = 0; |
| 1239 |
|
int force_study = -1; |
| 1240 |
|
int force_study_options = 0; |
| 1241 |
int quiet = 0; |
int quiet = 0; |
| 1242 |
int size_offsets = 45; |
int size_offsets = 45; |
| 1243 |
int size_offsets_max; |
int size_offsets_max; |
| 1251 |
int yield = 0; |
int yield = 0; |
| 1252 |
int stack_size; |
int stack_size; |
| 1253 |
|
|
| 1254 |
|
pcre_jit_stack *jit_stack = NULL; |
| 1255 |
|
|
| 1256 |
|
|
| 1257 |
/* These vectors store, end-to-end, a list of captured substring names. Assume |
/* These vectors store, end-to-end, a list of captured substring names. Assume |
| 1258 |
that 1024 is plenty long enough for the few names we'll be testing. */ |
that 1024 is plenty long enough for the few names we'll be testing. */ |
| 1259 |
|
|
| 1260 |
uschar copynames[1024]; |
pcre_uchar copynames[1024]; |
| 1261 |
uschar getnames[1024]; |
pcre_uchar getnames[1024]; |
| 1262 |
|
|
| 1263 |
uschar *copynamesptr; |
pcre_uchar *copynamesptr; |
| 1264 |
uschar *getnamesptr; |
pcre_uchar *getnamesptr; |
| 1265 |
|
|
| 1266 |
/* Get buffers from malloc() so that Electric Fence will check their misuse |
/* Get buffers from malloc() so that Electric Fence will check their misuse |
| 1267 |
when I am debugging. They grow automatically when very long lines are read. */ |
when I am debugging. They grow automatically when very long lines are read. */ |
| 1268 |
|
|
| 1269 |
buffer = (unsigned char *)malloc(buffer_size); |
buffer = (pcre_uint8 *)malloc(buffer_size); |
| 1270 |
dbuffer = (unsigned char *)malloc(buffer_size); |
dbuffer = (pcre_uint8 *)malloc(buffer_size); |
| 1271 |
pbuffer = (unsigned char *)malloc(buffer_size); |
pbuffer = (pcre_uint8 *)malloc(buffer_size); |
| 1272 |
|
|
| 1273 |
/* The outfile variable is static so that new_malloc can use it. */ |
/* The outfile variable is static so that new_malloc can use it. */ |
| 1274 |
|
|
| 1289 |
{ |
{ |
| 1290 |
unsigned char *endptr; |
unsigned char *endptr; |
| 1291 |
|
|
| 1292 |
if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0) |
if (strcmp(argv[op], "-m") == 0) showstore = 1; |
| 1293 |
showstore = 1; |
else if (strcmp(argv[op], "-s") == 0) force_study = 0; |
| 1294 |
|
else if (strcmp(argv[op], "-s+") == 0) |
| 1295 |
|
{ |
| 1296 |
|
force_study = 1; |
| 1297 |
|
force_study_options = PCRE_STUDY_JIT_COMPILE; |
| 1298 |
|
} |
| 1299 |
else if (strcmp(argv[op], "-q") == 0) quiet = 1; |
else if (strcmp(argv[op], "-q") == 0) quiet = 1; |
| 1300 |
else if (strcmp(argv[op], "-b") == 0) debug = 1; |
else if (strcmp(argv[op], "-b") == 0) debug = 1; |
| 1301 |
else if (strcmp(argv[op], "-i") == 0) showinfo = 1; |
else if (strcmp(argv[op], "-i") == 0) showinfo = 1; |
| 1329 |
((stack_size = get_value((unsigned char *)argv[op+1], &endptr)), |
((stack_size = get_value((unsigned char *)argv[op+1], &endptr)), |
| 1330 |
*endptr == 0)) |
*endptr == 0)) |
| 1331 |
{ |
{ |
| 1332 |
#if defined(_WIN32) || defined(WIN32) |
#if defined(_WIN32) || defined(WIN32) || defined(__minix) |
| 1333 |
printf("PCRE: -S not supported on this OS\n"); |
printf("PCRE: -S not supported on this OS\n"); |
| 1334 |
exit(1); |
exit(1); |
| 1335 |
#else |
#else |
| 1360 |
printf(" %sUTF-8 support\n", rc? "" : "No "); |
printf(" %sUTF-8 support\n", rc? "" : "No "); |
| 1361 |
(void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc); |
(void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc); |
| 1362 |
printf(" %sUnicode properties support\n", rc? "" : "No "); |
printf(" %sUnicode properties support\n", rc? "" : "No "); |
| 1363 |
|
(void)pcre_config(PCRE_CONFIG_JIT, &rc); |
| 1364 |
|
if (rc) |
| 1365 |
|
printf(" Just-in-time compiler support\n"); |
| 1366 |
|
else |
| 1367 |
|
printf(" No just-in-time compiler support\n"); |
| 1368 |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
| 1369 |
/* Note that these values are always the ASCII values, even |
/* Note that these values are always the ASCII values, even |
| 1370 |
in EBCDIC environments. CR is 13 and NL is 10. */ |
in EBCDIC environments. CR is 13 and NL is 10. */ |
| 1470 |
const unsigned char *tables = NULL; |
const unsigned char *tables = NULL; |
| 1471 |
unsigned long int true_size, true_study_size = 0; |
unsigned long int true_size, true_study_size = 0; |
| 1472 |
size_t size, regex_gotten_store; |
size_t size, regex_gotten_store; |
| 1473 |
|
int do_allcaps = 0; |
| 1474 |
int do_mark = 0; |
int do_mark = 0; |
| 1475 |
int do_study = 0; |
int do_study = 0; |
| 1476 |
|
int no_force_study = 0; |
| 1477 |
int do_debug = debug; |
int do_debug = debug; |
| 1478 |
int do_G = 0; |
int do_G = 0; |
| 1479 |
int do_g = 0; |
int do_g = 0; |
| 1480 |
int do_showinfo = showinfo; |
int do_showinfo = showinfo; |
| 1481 |
int do_showrest = 0; |
int do_showrest = 0; |
| 1482 |
|
int do_showcaprest = 0; |
| 1483 |
int do_flip = 0; |
int do_flip = 0; |
| 1484 |
int erroroffset, len, delimiter, poffset; |
int erroroffset, len, delimiter, poffset; |
| 1485 |
|
|
| 1499 |
if (*p == '<' && strchr((char *)(p+1), '<') == NULL) |
if (*p == '<' && strchr((char *)(p+1), '<') == NULL) |
| 1500 |
{ |
{ |
| 1501 |
unsigned long int magic, get_options; |
unsigned long int magic, get_options; |
| 1502 |
uschar sbuf[8]; |
pcre_uint8 sbuf[8]; |
| 1503 |
FILE *f; |
FILE *f; |
| 1504 |
|
|
| 1505 |
p++; |
p++; |
| 1522 |
(sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7]; |
(sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7]; |
| 1523 |
|
|
| 1524 |
re = (real_pcre *)new_malloc(true_size); |
re = (real_pcre *)new_malloc(true_size); |
| 1525 |
regex_gotten_store = gotten_store; |
regex_gotten_store = first_gotten_store; |
| 1526 |
|
|
| 1527 |
if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ; |
if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ; |
| 1528 |
|
|
| 1541 |
} |
} |
| 1542 |
} |
} |
| 1543 |
|
|
| 1544 |
fprintf(outfile, "Compiled regex%s loaded from %s\n", |
fprintf(outfile, "Compiled pattern%s loaded from %s\n", |
| 1545 |
do_flip? " (byte-inverted)" : "", p); |
do_flip? " (byte-inverted)" : "", p); |
| 1546 |
|
|
| 1547 |
/* Need to know if UTF-8 for printing data strings */ |
/* Need to know if UTF-8 for printing data strings */ |
| 1549 |
new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options); |
new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options); |
| 1550 |
use_utf8 = (get_options & PCRE_UTF8) != 0; |
use_utf8 = (get_options & PCRE_UTF8) != 0; |
| 1551 |
|
|
| 1552 |
/* Now see if there is any following study data */ |
/* Now see if there is any following study data. */ |
| 1553 |
|
|
| 1554 |
if (true_study_size != 0) |
if (true_study_size != 0) |
| 1555 |
{ |
{ |
| 1565 |
{ |
{ |
| 1566 |
FAIL_READ: |
FAIL_READ: |
| 1567 |
fprintf(outfile, "Failed to read data from %s\n", p); |
fprintf(outfile, "Failed to read data from %s\n", p); |
| 1568 |
if (extra != NULL) new_free(extra); |
if (extra != NULL) pcre_free_study(extra); |
| 1569 |
if (re != NULL) new_free(re); |
if (re != NULL) new_free(re); |
| 1570 |
fclose(f); |
fclose(f); |
| 1571 |
continue; |
continue; |
| 1645 |
case 's': options |= PCRE_DOTALL; break; |
case 's': options |= PCRE_DOTALL; break; |
| 1646 |
case 'x': options |= PCRE_EXTENDED; break; |
case 'x': options |= PCRE_EXTENDED; break; |
| 1647 |
|
|
| 1648 |
case '+': do_showrest = 1; break; |
case '+': |
| 1649 |
|
if (do_showrest) do_showcaprest = 1; else do_showrest = 1; |
| 1650 |
|
break; |
| 1651 |
|
|
| 1652 |
|
case '=': do_allcaps = 1; break; |
| 1653 |
case 'A': options |= PCRE_ANCHORED; break; |
case 'A': options |= PCRE_ANCHORED; break; |
| 1654 |
case 'B': do_debug = 1; break; |
case 'B': do_debug = 1; break; |
| 1655 |
case 'C': options |= PCRE_AUTO_CALLOUT; break; |
case 'C': options |= PCRE_AUTO_CALLOUT; break; |
| 1667 |
case 'P': do_posix = 1; break; |
case 'P': do_posix = 1; break; |
| 1668 |
#endif |
#endif |
| 1669 |
|
|
| 1670 |
case 'S': do_study = 1; break; |
case 'S': |
| 1671 |
|
if (do_study == 0) |
| 1672 |
|
{ |
| 1673 |
|
do_study = 1; |
| 1674 |
|
if (*pp == '+') |
| 1675 |
|
{ |
| 1676 |
|
study_options |= PCRE_STUDY_JIT_COMPILE; |
| 1677 |
|
pp++; |
| 1678 |
|
} |
| 1679 |
|
} |
| 1680 |
|
else |
| 1681 |
|
{ |
| 1682 |
|
do_study = 0; |
| 1683 |
|
no_force_study = 1; |
| 1684 |
|
} |
| 1685 |
|
break; |
| 1686 |
|
|
| 1687 |
case 'U': options |= PCRE_UNGREEDY; break; |
case 'U': options |= PCRE_UNGREEDY; break; |
| 1688 |
case 'W': options |= PCRE_UCP; break; |
case 'W': options |= PCRE_UCP; break; |
| 1689 |
case 'X': options |= PCRE_EXTRA; break; |
case 'X': options |= PCRE_EXTRA; break; |
| 1690 |
|
case 'Y': options |= PCRE_NO_START_OPTIMISE; break; |
| 1691 |
case 'Z': debug_lengths = 0; break; |
case 'Z': debug_lengths = 0; break; |
| 1692 |
case '8': options |= PCRE_UTF8; use_utf8 = 1; break; |
case '8': options |= PCRE_UTF8; use_utf8 = 1; break; |
| 1693 |
case '?': options |= PCRE_NO_UTF8_CHECK; break; |
case '?': options |= PCRE_NO_UTF8_CHECK; break; |
| 1736 |
|
|
| 1737 |
case '<': |
case '<': |
| 1738 |
{ |
{ |
| 1739 |
if (strncmpic(pp, (uschar *)"JS>", 3) == 0) |
if (strncmpic(pp, (pcre_uint8 *)"JS>", 3) == 0) |
| 1740 |
{ |
{ |
| 1741 |
options |= PCRE_JAVASCRIPT_COMPAT; |
options |= PCRE_JAVASCRIPT_COMPAT; |
| 1742 |
pp += 3; |
pp += 3; |
| 1780 |
if ((options & PCRE_UCP) != 0) cflags |= REG_UCP; |
if ((options & PCRE_UCP) != 0) cflags |= REG_UCP; |
| 1781 |
if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY; |
if ((options & PCRE_UNGREEDY) != 0) cflags |= REG_UNGREEDY; |
| 1782 |
|
|
| 1783 |
|
first_gotten_store = 0; |
| 1784 |
rc = regcomp(&preg, (char *)p, cflags); |
rc = regcomp(&preg, (char *)p, cflags); |
| 1785 |
|
|
| 1786 |
/* Compilation failed; go back for another re, skipping to blank line |
/* Compilation failed; go back for another re, skipping to blank line |
| 1818 |
(double)CLOCKS_PER_SEC); |
(double)CLOCKS_PER_SEC); |
| 1819 |
} |
} |
| 1820 |
|
|
| 1821 |
|
first_gotten_store = 0; |
| 1822 |
re = pcre_compile((char *)p, options, &error, &erroroffset, tables); |
re = pcre_compile((char *)p, options, &error, &erroroffset, tables); |
| 1823 |
|
|
| 1824 |
/* Compilation failed; go back for another re, skipping to blank line |
/* Compilation failed; go back for another re, skipping to blank line |
| 1853 |
new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options); |
new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options); |
| 1854 |
if ((get_options & PCRE_UTF8) != 0) use_utf8 = 1; |
if ((get_options & PCRE_UTF8) != 0) use_utf8 = 1; |
| 1855 |
|
|
| 1856 |
/* Print information if required. There are now two info-returning |
/* Extract the size for possible writing before possibly flipping it, |
| 1857 |
functions. The old one has a limited interface and returns only limited |
and remember the store that was got. */ |
| 1858 |
data. Check that it agrees with the newer one. */ |
|
| 1859 |
|
true_size = ((real_pcre *)re)->size; |
| 1860 |
|
regex_gotten_store = first_gotten_store; |
| 1861 |
|
|
| 1862 |
|
/* Output code size information if requested */ |
| 1863 |
|
|
| 1864 |
if (log_store) |
if (log_store) |
| 1865 |
fprintf(outfile, "Memory allocation (code space): %d\n", |
fprintf(outfile, "Memory allocation (code space): %d\n", |
| 1866 |
(int)(gotten_store - |
(int)(first_gotten_store - |
| 1867 |
sizeof(real_pcre) - |
sizeof(real_pcre) - |
| 1868 |
((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size)); |
((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size)); |
| 1869 |
|
|
| 1870 |
/* Extract the size for possible writing before possibly flipping it, |
/* If -s or /S was present, study the regex to generate additional info to |
| 1871 |
and remember the store that was got. */ |
help with the matching, unless the pattern has the SS option, which |
| 1872 |
|
suppresses the effect of /S (used for a few test patterns where studying is |
| 1873 |
|
never sensible). */ |
| 1874 |
|
|
| 1875 |
true_size = ((real_pcre *)re)->size; |
if (do_study || (force_study >= 0 && !no_force_study)) |
|
regex_gotten_store = gotten_store; |
|
|
|
|
|
/* If /S was present, study the regexp to generate additional info to |
|
|
help with the matching. */ |
|
|
|
|
|
if (do_study) |
|
| 1876 |
{ |
{ |
| 1877 |
if (timeit > 0) |
if (timeit > 0) |
| 1878 |
{ |
{ |
| 1880 |
clock_t time_taken; |
clock_t time_taken; |
| 1881 |
clock_t start_time = clock(); |
clock_t start_time = clock(); |
| 1882 |
for (i = 0; i < timeit; i++) |
for (i = 0; i < timeit; i++) |
| 1883 |
extra = pcre_study(re, study_options, &error); |
extra = pcre_study(re, study_options | force_study_options, &error); |
| 1884 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
| 1885 |
if (extra != NULL) free(extra); |
if (extra != NULL) pcre_free_study(extra); |
| 1886 |
fprintf(outfile, " Study time %.4f milliseconds\n", |
fprintf(outfile, " Study time %.4f milliseconds\n", |
| 1887 |
(((double)time_taken * 1000.0) / (double)timeit) / |
(((double)time_taken * 1000.0) / (double)timeit) / |
| 1888 |
(double)CLOCKS_PER_SEC); |
(double)CLOCKS_PER_SEC); |
| 1889 |
} |
} |
| 1890 |
extra = pcre_study(re, study_options, &error); |
extra = pcre_study(re, study_options | force_study_options, &error); |
| 1891 |
if (error != NULL) |
if (error != NULL) |
| 1892 |
fprintf(outfile, "Failed to study: %s\n", error); |
fprintf(outfile, "Failed to study: %s\n", error); |
| 1893 |
else if (extra != NULL) |
else if (extra != NULL) |
| 1894 |
|
{ |
| 1895 |
true_study_size = ((pcre_study_data *)(extra->study_data))->size; |
true_study_size = ((pcre_study_data *)(extra->study_data))->size; |
| 1896 |
|
if (log_store) |
| 1897 |
|
{ |
| 1898 |
|
size_t jitsize; |
| 1899 |
|
new_info(re, extra, PCRE_INFO_JITSIZE, &jitsize); |
| 1900 |
|
if (jitsize != 0) |
| 1901 |
|
fprintf(outfile, "Memory allocation (JIT code): %d\n", jitsize); |
| 1902 |
|
} |
| 1903 |
|
} |
| 1904 |
} |
} |
| 1905 |
|
|
| 1906 |
/* If /K was present, we set up for handling MARK data. */ |
/* If /K was present, we set up for handling MARK data. */ |
| 1933 |
(pcre_uint16)byteflip(rre->top_bracket, sizeof(rre->top_bracket)); |
(pcre_uint16)byteflip(rre->top_bracket, sizeof(rre->top_bracket)); |
| 1934 |
rre->top_backref = |
rre->top_backref = |
| 1935 |
(pcre_uint16)byteflip(rre->top_backref, sizeof(rre->top_backref)); |
(pcre_uint16)byteflip(rre->top_backref, sizeof(rre->top_backref)); |
| 1936 |
rre->first_byte = |
rre->first_char = |
| 1937 |
(pcre_uint16)byteflip(rre->first_byte, sizeof(rre->first_byte)); |
(pcre_uint16)byteflip(rre->first_char, sizeof(rre->first_char)); |
| 1938 |
rre->req_byte = |
rre->req_char = |
| 1939 |
(pcre_uint16)byteflip(rre->req_byte, sizeof(rre->req_byte)); |
(pcre_uint16)byteflip(rre->req_char, sizeof(rre->req_char)); |
| 1940 |
rre->name_table_offset = (pcre_uint16)byteflip(rre->name_table_offset, |
rre->name_table_offset = (pcre_uint16)byteflip(rre->name_table_offset, |
| 1941 |
sizeof(rre->name_table_offset)); |
sizeof(rre->name_table_offset)); |
| 1942 |
rre->name_entry_size = (pcre_uint16)byteflip(rre->name_entry_size, |
rre->name_entry_size = (pcre_uint16)byteflip(rre->name_entry_size, |
| 1953 |
} |
} |
| 1954 |
} |
} |
| 1955 |
|
|
| 1956 |
/* Extract information from the compiled data if required */ |
/* Extract information from the compiled data if required. There are now |
| 1957 |
|
two info-returning functions. The old one has a limited interface and |
| 1958 |
|
returns only limited data. Check that it agrees with the newer one. */ |
| 1959 |
|
|
| 1960 |
SHOW_INFO: |
SHOW_INFO: |
| 1961 |
|
|
| 1976 |
int count, backrefmax, first_char, need_char, okpartial, jchanged, |
int count, backrefmax, first_char, need_char, okpartial, jchanged, |
| 1977 |
hascrorlf; |
hascrorlf; |
| 1978 |
int nameentrysize, namecount; |
int nameentrysize, namecount; |
| 1979 |
const uschar *nametable; |
const pcre_uchar *nametable; |
| 1980 |
|
|
| 1981 |
new_info(re, NULL, PCRE_INFO_SIZE, &size); |
new_info(re, NULL, PCRE_INFO_SIZE, &size); |
| 1982 |
new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count); |
new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count); |
| 2037 |
if (do_flip) all_options = byteflip(all_options, sizeof(all_options)); |
if (do_flip) all_options = byteflip(all_options, sizeof(all_options)); |
| 2038 |
|
|
| 2039 |
if (get_options == 0) fprintf(outfile, "No options\n"); |
if (get_options == 0) fprintf(outfile, "No options\n"); |
| 2040 |
else fprintf(outfile, "Options:%s%s%s%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%s%s%s%s\n", |
| 2041 |
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
| 2042 |
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
| 2043 |
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
| 2053 |
((get_options & PCRE_UTF8) != 0)? " utf8" : "", |
((get_options & PCRE_UTF8) != 0)? " utf8" : "", |
| 2054 |
((get_options & PCRE_UCP) != 0)? " ucp" : "", |
((get_options & PCRE_UCP) != 0)? " ucp" : "", |
| 2055 |
((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "", |
((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "", |
| 2056 |
|
((get_options & PCRE_NO_START_OPTIMIZE) != 0)? " no_start_optimize" : "", |
| 2057 |
((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : ""); |
((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : ""); |
| 2058 |
|
|
| 2059 |
if (jchanged) fprintf(outfile, "Duplicate name status changes\n"); |
if (jchanged) fprintf(outfile, "Duplicate name status changes\n"); |
| 2094 |
} |
} |
| 2095 |
else |
else |
| 2096 |
{ |
{ |
| 2097 |
int ch = first_char & 255; |
const char *caseless = |
| 2098 |
const char *caseless = ((first_char & REQ_CASELESS) == 0)? |
((((real_pcre *)re)->flags & PCRE_FCH_CASELESS) == 0)? |
| 2099 |
"" : " (caseless)"; |
"" : " (caseless)"; |
| 2100 |
if (PRINTHEX(ch)) |
|
| 2101 |
fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless); |
if (PRINTHEX(first_char)) |
| 2102 |
|
fprintf(outfile, "First char = \'%c\'%s\n", first_char, caseless); |
| 2103 |
else |
else |
| 2104 |
fprintf(outfile, "First char = %d%s\n", ch, caseless); |
fprintf(outfile, "First char = %d%s\n", first_char, caseless); |
| 2105 |
} |
} |
| 2106 |
|
|
| 2107 |
if (need_char < 0) |
if (need_char < 0) |
| 2110 |
} |
} |
| 2111 |
else |
else |
| 2112 |
{ |
{ |
| 2113 |
int ch = need_char & 255; |
const char *caseless = |
| 2114 |
const char *caseless = ((need_char & REQ_CASELESS) == 0)? |
((((real_pcre *)re)->flags & PCRE_RCH_CASELESS) == 0)? |
| 2115 |
"" : " (caseless)"; |
"" : " (caseless)"; |
| 2116 |
if (PRINTHEX(ch)) |
|
| 2117 |
fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless); |
if (PRINTHEX(need_char)) |
| 2118 |
|
fprintf(outfile, "Need char = \'%c\'%s\n", need_char, caseless); |
| 2119 |
else |
else |
| 2120 |
fprintf(outfile, "Need char = %d%s\n", ch, caseless); |
fprintf(outfile, "Need char = %d%s\n", need_char, caseless); |
| 2121 |
} |
} |
| 2122 |
|
|
| 2123 |
/* Don't output study size; at present it is in any case a fixed |
/* Don't output study size; at present it is in any case a fixed |
| 2124 |
value, but it varies, depending on the computer architecture, and |
value, but it varies, depending on the computer architecture, and |
| 2125 |
so messes up the test suite. (And with the /F option, it might be |
so messes up the test suite. (And with the /F option, it might be |
| 2126 |
flipped.) */ |
flipped.) If study was forced by an external -s, don't show this |
| 2127 |
|
information unless -i or -d was also present. This means that, except |
| 2128 |
|
when auto-callouts are involved, the output from runs with and without |
| 2129 |
|
-s should be identical. */ |
| 2130 |
|
|
| 2131 |
if (do_study) |
if (do_study || (force_study >= 0 && showinfo && !no_force_study)) |
| 2132 |
{ |
{ |
| 2133 |
if (extra == NULL) |
if (extra == NULL) |
| 2134 |
fprintf(outfile, "Study returned NULL\n"); |
fprintf(outfile, "Study returned NULL\n"); |
| 2135 |
else |
else |
| 2136 |
{ |
{ |
| 2137 |
uschar *start_bits = NULL; |
pcre_uint8 *start_bits = NULL; |
| 2138 |
int minlength; |
int minlength; |
| 2139 |
|
|
| 2140 |
new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength); |
new_info(re, extra, PCRE_INFO_MINLENGTH, &minlength); |
| 2172 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
| 2173 |
} |
} |
| 2174 |
} |
} |
| 2175 |
|
|
| 2176 |
|
/* Show this only if the JIT was set by /S, not by -s. */ |
| 2177 |
|
|
| 2178 |
|
if ((study_options & PCRE_STUDY_JIT_COMPILE) != 0) |
| 2179 |
|
{ |
| 2180 |
|
int jit; |
| 2181 |
|
new_info(re, extra, PCRE_INFO_JIT, &jit); |
| 2182 |
|
if (jit) |
| 2183 |
|
fprintf(outfile, "JIT study was successful\n"); |
| 2184 |
|
else |
| 2185 |
|
#ifdef SUPPORT_JIT |
| 2186 |
|
fprintf(outfile, "JIT study was not successful\n"); |
| 2187 |
|
#else |
| 2188 |
|
fprintf(outfile, "JIT support is not available in this version of PCRE\n"); |
| 2189 |
|
#endif |
| 2190 |
|
} |
| 2191 |
} |
} |
| 2192 |
} |
} |
| 2193 |
|
|
| 2204 |
} |
} |
| 2205 |
else |
else |
| 2206 |
{ |
{ |
| 2207 |
uschar sbuf[8]; |
pcre_uint8 sbuf[8]; |
| 2208 |
sbuf[0] = (uschar)((true_size >> 24) & 255); |
sbuf[0] = (pcre_uint8)((true_size >> 24) & 255); |
| 2209 |
sbuf[1] = (uschar)((true_size >> 16) & 255); |
sbuf[1] = (pcre_uint8)((true_size >> 16) & 255); |
| 2210 |
sbuf[2] = (uschar)((true_size >> 8) & 255); |
sbuf[2] = (pcre_uint8)((true_size >> 8) & 255); |
| 2211 |
sbuf[3] = (uschar)((true_size) & 255); |
sbuf[3] = (pcre_uint8)((true_size) & 255); |
| 2212 |
|
|
| 2213 |
sbuf[4] = (uschar)((true_study_size >> 24) & 255); |
sbuf[4] = (pcre_uint8)((true_study_size >> 24) & 255); |
| 2214 |
sbuf[5] = (uschar)((true_study_size >> 16) & 255); |
sbuf[5] = (pcre_uint8)((true_study_size >> 16) & 255); |
| 2215 |
sbuf[6] = (uschar)((true_study_size >> 8) & 255); |
sbuf[6] = (pcre_uint8)((true_study_size >> 8) & 255); |
| 2216 |
sbuf[7] = (uschar)((true_study_size) & 255); |
sbuf[7] = (pcre_uint8)((true_study_size) & 255); |
| 2217 |
|
|
| 2218 |
if (fwrite(sbuf, 1, 8, f) < 8 || |
if (fwrite(sbuf, 1, 8, f) < 8 || |
| 2219 |
fwrite(re, 1, true_size, f) < true_size) |
fwrite(re, 1, true_size, f) < true_size) |
| 2222 |
} |
} |
| 2223 |
else |
else |
| 2224 |
{ |
{ |
| 2225 |
fprintf(outfile, "Compiled regex written to %s\n", to_file); |
fprintf(outfile, "Compiled pattern written to %s\n", to_file); |
| 2226 |
|
|
| 2227 |
|
/* If there is study data, write it. */ |
| 2228 |
|
|
| 2229 |
if (extra != NULL) |
if (extra != NULL) |
| 2230 |
{ |
{ |
| 2231 |
if (fwrite(extra->study_data, 1, true_study_size, f) < |
if (fwrite(extra->study_data, 1, true_study_size, f) < |
| 2235 |
strerror(errno)); |
strerror(errno)); |
| 2236 |
} |
} |
| 2237 |
else fprintf(outfile, "Study data written to %s\n", to_file); |
else fprintf(outfile, "Study data written to %s\n", to_file); |
|
|
|
| 2238 |
} |
} |
| 2239 |
} |
} |
| 2240 |
fclose(f); |
fclose(f); |
| 2241 |
} |
} |
| 2242 |
|
|
| 2243 |
new_free(re); |
new_free(re); |
| 2244 |
if (extra != NULL) new_free(extra); |
if (extra != NULL) pcre_free_study(extra); |
| 2245 |
if (locale_set) |
if (locale_set) |
| 2246 |
{ |
{ |
| 2247 |
new_free((void *)tables); |
new_free((void *)tables); |
| 2256 |
|
|
| 2257 |
for (;;) |
for (;;) |
| 2258 |
{ |
{ |
| 2259 |
uschar *q; |
pcre_uint8 *q; |
| 2260 |
uschar *bptr; |
pcre_uint8 *bptr; |
| 2261 |
int *use_offsets = offsets; |
int *use_offsets = offsets; |
| 2262 |
int use_size_offsets = size_offsets; |
int use_size_offsets = size_offsets; |
| 2263 |
int callout_data = 0; |
int callout_data = 0; |
| 2269 |
int getlist = 0; |
int getlist = 0; |
| 2270 |
int gmatched = 0; |
int gmatched = 0; |
| 2271 |
int start_offset = 0; |
int start_offset = 0; |
| 2272 |
|
int start_offset_sign = 1; |
| 2273 |
int g_notempty = 0; |
int g_notempty = 0; |
| 2274 |
int use_dfa = 0; |
int use_dfa = 0; |
| 2275 |
|
|
| 2283 |
|
|
| 2284 |
pcre_callout = callout; |
pcre_callout = callout; |
| 2285 |
first_callout = 1; |
first_callout = 1; |
| 2286 |
|
last_callout_mark = NULL; |
| 2287 |
callout_extra = 0; |
callout_extra = 0; |
| 2288 |
callout_count = 0; |
callout_count = 0; |
| 2289 |
callout_fail_count = 999999; |
callout_fail_count = 999999; |
| 2362 |
{ |
{ |
| 2363 |
unsigned char *pt = p; |
unsigned char *pt = p; |
| 2364 |
c = 0; |
c = 0; |
| 2365 |
while (isxdigit(*(++pt))) |
|
| 2366 |
c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W'); |
/* We used to have "while (isxdigit(*(++pt)))" here, but it fails |
| 2367 |
|
when isxdigit() is a macro that refers to its argument more than |
| 2368 |
|
once. This is banned by the C Standard, but apparently happens in at |
| 2369 |
|
least one MacOS environment. */ |
| 2370 |
|
|
| 2371 |
|
for (pt++; isxdigit(*pt); pt++) |
| 2372 |
|
c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'a' - 10); |
| 2373 |
if (*pt == '}') |
if (*pt == '}') |
| 2374 |
{ |
{ |
| 2375 |
unsigned char buff8[8]; |
unsigned char buff8[8]; |
| 2399 |
c = 0; |
c = 0; |
| 2400 |
while (i++ < 2 && isxdigit(*p)) |
while (i++ < 2 && isxdigit(*p)) |
| 2401 |
{ |
{ |
| 2402 |
c = c * 16 + tolower(*p) - ((isdigit(*p))? '0' : 'W'); |
c = c * 16 + tolower(*p) - ((isdigit(*p))? '0' : 'a' - 10); |
| 2403 |
p++; |
p++; |
| 2404 |
} |
} |
| 2405 |
break; |
break; |
| 2409 |
continue; |
continue; |
| 2410 |
|
|
| 2411 |
case '>': |
case '>': |
| 2412 |
|
if (*p == '-') |
| 2413 |
|
{ |
| 2414 |
|
start_offset_sign = -1; |
| 2415 |
|
p++; |
| 2416 |
|
} |
| 2417 |
while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0'; |
while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0'; |
| 2418 |
|
start_offset *= start_offset_sign; |
| 2419 |
continue; |
continue; |
| 2420 |
|
|
| 2421 |
case 'A': /* Option setting */ |
case 'A': /* Option setting */ |
| 2434 |
} |
} |
| 2435 |
else if (isalnum(*p)) |
else if (isalnum(*p)) |
| 2436 |
{ |
{ |
| 2437 |
uschar *npp = copynamesptr; |
pcre_uchar *npp = copynamesptr; |
| 2438 |
while (isalnum(*p)) *npp++ = *p++; |
while (isalnum(*p)) *npp++ = *p++; |
| 2439 |
*npp++ = 0; |
*npp++ = 0; |
| 2440 |
*npp = 0; |
*npp = 0; |
| 2504 |
} |
} |
| 2505 |
else if (isalnum(*p)) |
else if (isalnum(*p)) |
| 2506 |
{ |
{ |
| 2507 |
uschar *npp = getnamesptr; |
pcre_uchar *npp = getnamesptr; |
| 2508 |
while (isalnum(*p)) *npp++ = *p++; |
while (isalnum(*p)) *npp++ = *p++; |
| 2509 |
*npp++ = 0; |
*npp++ = 0; |
| 2510 |
*npp = 0; |
*npp = 0; |
| 2515 |
} |
} |
| 2516 |
continue; |
continue; |
| 2517 |
|
|
| 2518 |
|
case 'J': |
| 2519 |
|
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
| 2520 |
|
if (extra != NULL |
| 2521 |
|
&& (extra->flags & PCRE_EXTRA_EXECUTABLE_JIT) != 0 |
| 2522 |
|
&& extra->executable_jit != NULL) |
| 2523 |
|
{ |
| 2524 |
|
if (jit_stack != NULL) pcre_jit_stack_free(jit_stack); |
| 2525 |
|
jit_stack = pcre_jit_stack_alloc(1, n * 1024); |
| 2526 |
|
pcre_assign_jit_stack(extra, jit_callback, jit_stack); |
| 2527 |
|
} |
| 2528 |
|
continue; |
| 2529 |
|
|
| 2530 |
case 'L': |
case 'L': |
| 2531 |
getlist = 1; |
getlist = 1; |
| 2532 |
continue; |
continue; |
| 2685 |
(void)pchars(dbuffer + pmatch[i].rm_so, |
(void)pchars(dbuffer + pmatch[i].rm_so, |
| 2686 |
pmatch[i].rm_eo - pmatch[i].rm_so, outfile); |
pmatch[i].rm_eo - pmatch[i].rm_so, outfile); |
| 2687 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
| 2688 |
if (i == 0 && do_showrest) |
if (do_showcaprest || (i == 0 && do_showrest)) |
| 2689 |
{ |
{ |
| 2690 |
fprintf(outfile, " 0+ "); |
fprintf(outfile, "%2d+ ", (int)i); |
| 2691 |
(void)pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo, |
(void)pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo, |
| 2692 |
outfile); |
outfile); |
| 2693 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
| 2737 |
|
|
| 2738 |
/* If find_match_limit is set, we want to do repeated matches with |
/* If find_match_limit is set, we want to do repeated matches with |
| 2739 |
varying limits in order to find the minimum value for the match limit and |
varying limits in order to find the minimum value for the match limit and |
| 2740 |
for the recursion limit. */ |
for the recursion limit. The match limits are relevant only to the normal |
| 2741 |
|
running of pcre_exec(), so disable the JIT optimization. This makes it |
| 2742 |
|
possible to run the same set of tests with and without JIT externally |
| 2743 |
|
requested. */ |
| 2744 |
|
|
| 2745 |
if (find_match_limit) |
if (find_match_limit) |
| 2746 |
{ |
{ |
| 2749 |
extra = (pcre_extra *)malloc(sizeof(pcre_extra)); |
extra = (pcre_extra *)malloc(sizeof(pcre_extra)); |
| 2750 |
extra->flags = 0; |
extra->flags = 0; |
| 2751 |
} |
} |
| 2752 |
|
else extra->flags &= ~PCRE_EXTRA_EXECUTABLE_JIT; |
| 2753 |
|
|
| 2754 |
(void)check_match_limit(re, extra, bptr, len, start_offset, |
(void)check_match_limit(re, extra, bptr, len, start_offset, |
| 2755 |
options|g_notempty, use_offsets, use_size_offsets, |
options|g_notempty, use_offsets, use_size_offsets, |
| 2833 |
} |
} |
| 2834 |
} |
} |
| 2835 |
|
|
| 2836 |
|
/* do_allcaps requests showing of all captures in the pattern, to check |
| 2837 |
|
unset ones at the end. */ |
| 2838 |
|
|
| 2839 |
|
if (do_allcaps) |
| 2840 |
|
{ |
| 2841 |
|
new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count); |
| 2842 |
|
count++; /* Allow for full match */ |
| 2843 |
|
if (count * 2 > use_size_offsets) count = use_size_offsets/2; |
| 2844 |
|
} |
| 2845 |
|
|
| 2846 |
|
/* Output the captured substrings */ |
| 2847 |
|
|
| 2848 |
for (i = 0; i < count * 2; i += 2) |
for (i = 0; i < count * 2; i += 2) |
| 2849 |
{ |
{ |
| 2850 |
if (use_offsets[i] < 0) |
if (use_offsets[i] < 0) |
| 2851 |
|
{ |
| 2852 |
|
if (use_offsets[i] != -1) |
| 2853 |
|
fprintf(outfile, "ERROR: bad negative value %d for offset %d\n", |
| 2854 |
|
use_offsets[i], i); |
| 2855 |
|
if (use_offsets[i+1] != -1) |
| 2856 |
|
fprintf(outfile, "ERROR: bad negative value %d for offset %d\n", |
| 2857 |
|
use_offsets[i+1], i+1); |
| 2858 |
fprintf(outfile, "%2d: <unset>\n", i/2); |
fprintf(outfile, "%2d: <unset>\n", i/2); |
| 2859 |
|
} |
| 2860 |
else |
else |
| 2861 |
{ |
{ |
| 2862 |
fprintf(outfile, "%2d: ", i/2); |
fprintf(outfile, "%2d: ", i/2); |
| 2863 |
(void)pchars(bptr + use_offsets[i], |
(void)pchars(bptr + use_offsets[i], |
| 2864 |
use_offsets[i+1] - use_offsets[i], outfile); |
use_offsets[i+1] - use_offsets[i], outfile); |
| 2865 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
| 2866 |
if (i == 0) |
if (do_showcaprest || (i == 0 && do_showrest)) |
| 2867 |
{ |
{ |
| 2868 |
if (do_showrest) |
fprintf(outfile, "%2d+ ", i/2); |
| 2869 |
{ |
(void)pchars(bptr + use_offsets[i+1], len - use_offsets[i+1], |
| 2870 |
fprintf(outfile, " 0+ "); |
outfile); |
| 2871 |
(void)pchars(bptr + use_offsets[i+1], len - use_offsets[i+1], |
fprintf(outfile, "\n"); |
|
outfile); |
|
|
fprintf(outfile, "\n"); |
|
|
} |
|
| 2872 |
} |
} |
| 2873 |
} |
} |
| 2874 |
} |
} |
| 2948 |
fprintf(outfile, "%2dL %s\n", i, stringlist[i]); |
fprintf(outfile, "%2dL %s\n", i, stringlist[i]); |
| 2949 |
if (stringlist[i] != NULL) |
if (stringlist[i] != NULL) |
| 2950 |
fprintf(outfile, "string list not terminated by NULL\n"); |
fprintf(outfile, "string list not terminated by NULL\n"); |
|
/* free((void *)stringlist); */ |
|
| 2951 |
pcre_free_substring_list(stringlist); |
pcre_free_substring_list(stringlist); |
| 2952 |
} |
} |
| 2953 |
} |
} |
| 2974 |
to advance the start offset, and continue. We won't be at the end of the |
to advance the start offset, and continue. We won't be at the end of the |
| 2975 |
string - that was checked before setting g_notempty. |
string - that was checked before setting g_notempty. |
| 2976 |
|
|
| 2977 |
Complication arises in the case when the newline option is "any" or |
Complication arises in the case when the newline convention is "any", |
| 2978 |
"anycrlf". If the previous match was at the end of a line terminated by |
"crlf", or "anycrlf". If the previous match was at the end of a line |
| 2979 |
CRLF, an advance of one character just passes the \r, whereas we should |
terminated by CRLF, an advance of one character just passes the \r, |
| 2980 |
prefer the longer newline sequence, as does the code in pcre_exec(). |
whereas we should prefer the longer newline sequence, as does the code in |
| 2981 |
Fudge the offset value to achieve this. |
pcre_exec(). Fudge the offset value to achieve this. We check for a |
| 2982 |
|
newline setting in the pattern; if none was set, use pcre_config() to |
| 2983 |
|
find the default. |
| 2984 |
|
|
| 2985 |
Otherwise, in the case of UTF-8 matching, the advance must be one |
Otherwise, in the case of UTF-8 matching, the advance must be one |
| 2986 |
character, not one byte. */ |
character, not one byte. */ |
| 3005 |
(d == -1)? PCRE_NEWLINE_ANY : 0; |
(d == -1)? PCRE_NEWLINE_ANY : 0; |
| 3006 |
} |
} |
| 3007 |
if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY || |
if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY || |
| 3008 |
|
(obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_CRLF || |
| 3009 |
(obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF) |
(obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF) |
| 3010 |
&& |
&& |
| 3011 |
start_offset < len - 1 && |
start_offset < len - 1 && |
| 3016 |
{ |
{ |
| 3017 |
while (start_offset + onechar < len) |
while (start_offset + onechar < len) |
| 3018 |
{ |
{ |
| 3019 |
int tb = bptr[start_offset+onechar]; |
if ((bptr[start_offset+onechar] & 0xc0) != 0x80) break; |
| 3020 |
if (tb <= 127) break; |
onechar++; |
|
tb &= 0xc0; |
|
|
if (tb != 0 && tb != 0xc0) onechar++; |
|
| 3021 |
} |
} |
| 3022 |
} |
} |
| 3023 |
use_offsets[1] = start_offset + onechar; |
use_offsets[1] = start_offset + onechar; |
| 3024 |
} |
} |
| 3025 |
else |
else |
| 3026 |
{ |
{ |
| 3027 |
if (count == PCRE_ERROR_NOMATCH) |
switch(count) |
| 3028 |
{ |
{ |
| 3029 |
|
case PCRE_ERROR_NOMATCH: |
| 3030 |
if (gmatched == 0) |
if (gmatched == 0) |
| 3031 |
{ |
{ |
| 3032 |
if (markptr == NULL) fprintf(outfile, "No match\n"); |
if (markptr == NULL) fprintf(outfile, "No match\n"); |
| 3033 |
else fprintf(outfile, "No match, mark = %s\n", markptr); |
else fprintf(outfile, "No match, mark = %s\n", markptr); |
| 3034 |
} |
} |
| 3035 |
|
break; |
| 3036 |
|
|
| 3037 |
|
case PCRE_ERROR_BADUTF8: |
| 3038 |
|
case PCRE_ERROR_SHORTUTF8: |
| 3039 |
|
fprintf(outfile, "Error %d (%s UTF-8 string)", count, |
| 3040 |
|
(count == PCRE_ERROR_BADUTF8)? "bad" : "short"); |
| 3041 |
|
if (use_size_offsets >= 2) |
| 3042 |
|
fprintf(outfile, " offset=%d reason=%d", use_offsets[0], |
| 3043 |
|
use_offsets[1]); |
| 3044 |
|
fprintf(outfile, "\n"); |
| 3045 |
|
break; |
| 3046 |
|
|
| 3047 |
|
default: |
| 3048 |
|
if (count < 0 && (-count) < sizeof(errtexts)/sizeof(const char *)) |
| 3049 |
|
fprintf(outfile, "Error %d (%s)\n", count, errtexts[-count]); |
| 3050 |
|
else |
| 3051 |
|
fprintf(outfile, "Error %d (Unexpected value)\n", count); |
| 3052 |
|
break; |
| 3053 |
} |
} |
| 3054 |
else fprintf(outfile, "Error %d\n", count); |
|
| 3055 |
break; /* Out of the /g loop */ |
break; /* Out of the /g loop */ |
| 3056 |
} |
} |
| 3057 |
} |
} |
| 3098 |
#endif |
#endif |
| 3099 |
|
|
| 3100 |
if (re != NULL) new_free(re); |
if (re != NULL) new_free(re); |
| 3101 |
if (extra != NULL) new_free(extra); |
if (extra != NULL) pcre_free_study(extra); |
| 3102 |
if (locale_set) |
if (locale_set) |
| 3103 |
{ |
{ |
| 3104 |
new_free((void *)tables); |
new_free((void *)tables); |
| 3105 |
setlocale(LC_CTYPE, "C"); |
setlocale(LC_CTYPE, "C"); |
| 3106 |
locale_set = 0; |
locale_set = 0; |
| 3107 |
} |
} |
| 3108 |
|
if (jit_stack != NULL) |
| 3109 |
|
{ |
| 3110 |
|
pcre_jit_stack_free(jit_stack); |
| 3111 |
|
jit_stack = NULL; |
| 3112 |
|
} |
| 3113 |
} |
} |
| 3114 |
|
|
| 3115 |
if (infile == stdin) fprintf(outfile, "\n"); |
if (infile == stdin) fprintf(outfile, "\n"); |