| 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> |
| 48 |
#include <locale.h> |
#include <locale.h> |
| 49 |
#include <errno.h> |
#include <errno.h> |
| 50 |
|
|
| 51 |
|
#ifdef SUPPORT_LIBREADLINE |
| 52 |
|
#include <unistd.h> |
| 53 |
|
#include <readline/readline.h> |
| 54 |
|
#include <readline/history.h> |
| 55 |
|
#endif |
| 56 |
|
|
| 57 |
|
|
| 58 |
/* A number of things vary for Windows builds. Originally, pcretest opened its |
/* A number of things vary for Windows builds. Originally, pcretest opened its |
| 59 |
input and output without "b"; then I was told that "b" was needed in some |
input and output without "b"; then I was told that "b" was needed in some |
| 77 |
#endif |
#endif |
| 78 |
|
|
| 79 |
|
|
| 80 |
#define PCRE_SPY /* For Win32 build, import data, not export */ |
/* We have to include pcre_internal.h because we need the internal info for |
| 81 |
|
displaying the results of pcre_study() and we also need to know about the |
| 82 |
/* We include pcre_internal.h because we need the internal info for displaying |
internal macros, structures, and other internal data values; pcretest has |
| 83 |
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. |
| 84 |
macros, structures, and other internal data values; pcretest has "inside |
|
| 85 |
information" compared to a program that strictly follows the PCRE API. */ |
Although pcre_internal.h does itself include pcre.h, we explicitly include it |
| 86 |
|
here before pcre_internal.h so that the PCRE_EXP_xxx macros get set |
| 87 |
|
appropriately for an application, not for building PCRE. */ |
| 88 |
|
|
| 89 |
|
#include "pcre.h" |
| 90 |
#include "pcre_internal.h" |
#include "pcre_internal.h" |
| 91 |
|
|
| 92 |
/* 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 |
| 100 |
#define _pcre_utf8_table4 utf8_table4 |
#define _pcre_utf8_table4 utf8_table4 |
| 101 |
#define _pcre_utt utt |
#define _pcre_utt utt |
| 102 |
#define _pcre_utt_size utt_size |
#define _pcre_utt_size utt_size |
| 103 |
|
#define _pcre_utt_names utt_names |
| 104 |
#define _pcre_OP_lengths OP_lengths |
#define _pcre_OP_lengths OP_lengths |
| 105 |
|
|
| 106 |
#include "pcre_tables.c" |
#include "pcre_tables.c" |
| 163 |
static int callout_extra; |
static int callout_extra; |
| 164 |
static int callout_fail_count; |
static int callout_fail_count; |
| 165 |
static int callout_fail_id; |
static int callout_fail_id; |
| 166 |
|
static int debug_lengths; |
| 167 |
static int first_callout; |
static int first_callout; |
| 168 |
static int locale_set = 0; |
static int locale_set = 0; |
| 169 |
static int show_malloc; |
static int show_malloc; |
| 195 |
Arguments: |
Arguments: |
| 196 |
f the file to read |
f the file to read |
| 197 |
start where in buffer to start (this *must* be within buffer) |
start where in buffer to start (this *must* be within buffer) |
| 198 |
|
prompt for stdin or readline() |
| 199 |
|
|
| 200 |
Returns: pointer to the start of new data |
Returns: pointer to the start of new data |
| 201 |
could be a copy of start, or could be moved |
could be a copy of start, or could be moved |
| 203 |
*/ |
*/ |
| 204 |
|
|
| 205 |
static uschar * |
static uschar * |
| 206 |
extend_inputline(FILE *f, uschar *start) |
extend_inputline(FILE *f, uschar *start, const char *prompt) |
| 207 |
{ |
{ |
| 208 |
uschar *here = start; |
uschar *here = start; |
| 209 |
|
|
| 214 |
if (rlen > 1000) |
if (rlen > 1000) |
| 215 |
{ |
{ |
| 216 |
int dlen; |
int dlen; |
| 217 |
if (fgets((char *)here, rlen, f) == NULL) |
|
| 218 |
return (here == start)? NULL : start; |
/* If libreadline support is required, use readline() to read a line if the |
| 219 |
|
input is a terminal. Note that readline() removes the trailing newline, so |
| 220 |
|
we must put it back again, to be compatible with fgets(). */ |
| 221 |
|
|
| 222 |
|
#ifdef SUPPORT_LIBREADLINE |
| 223 |
|
if (isatty(fileno(f))) |
| 224 |
|
{ |
| 225 |
|
size_t len; |
| 226 |
|
char *s = readline(prompt); |
| 227 |
|
if (s == NULL) return (here == start)? NULL : start; |
| 228 |
|
len = strlen(s); |
| 229 |
|
if (len > 0) add_history(s); |
| 230 |
|
if (len > rlen - 1) len = rlen - 1; |
| 231 |
|
memcpy(here, s, len); |
| 232 |
|
here[len] = '\n'; |
| 233 |
|
here[len+1] = 0; |
| 234 |
|
free(s); |
| 235 |
|
} |
| 236 |
|
else |
| 237 |
|
#endif |
| 238 |
|
|
| 239 |
|
/* Read the next line by normal means, prompting if the file is stdin. */ |
| 240 |
|
|
| 241 |
|
{ |
| 242 |
|
if (f == stdin) printf(prompt); |
| 243 |
|
if (fgets((char *)here, rlen, f) == NULL) |
| 244 |
|
return (here == start)? NULL : start; |
| 245 |
|
} |
| 246 |
|
|
| 247 |
dlen = (int)strlen((char *)here); |
dlen = (int)strlen((char *)here); |
| 248 |
if (dlen > 0 && here[dlen - 1] == '\n') return start; |
if (dlen > 0 && here[dlen - 1] == '\n') return start; |
| 249 |
here += dlen; |
here += dlen; |
| 700 |
|
|
| 701 |
|
|
| 702 |
/************************************************* |
/************************************************* |
| 703 |
|
* Case-independent strncmp() function * |
| 704 |
|
*************************************************/ |
| 705 |
|
|
| 706 |
|
/* |
| 707 |
|
Arguments: |
| 708 |
|
s first string |
| 709 |
|
t second string |
| 710 |
|
n number of characters to compare |
| 711 |
|
|
| 712 |
|
Returns: < 0, = 0, or > 0, according to the comparison |
| 713 |
|
*/ |
| 714 |
|
|
| 715 |
|
static int |
| 716 |
|
strncmpic(uschar *s, uschar *t, int n) |
| 717 |
|
{ |
| 718 |
|
while (n--) |
| 719 |
|
{ |
| 720 |
|
int c = tolower(*s++) - tolower(*t++); |
| 721 |
|
if (c) return c; |
| 722 |
|
} |
| 723 |
|
return 0; |
| 724 |
|
} |
| 725 |
|
|
| 726 |
|
|
| 727 |
|
|
| 728 |
|
/************************************************* |
| 729 |
* Check newline indicator * |
* Check newline indicator * |
| 730 |
*************************************************/ |
*************************************************/ |
| 731 |
|
|
| 732 |
/* 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 |
| 733 |
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 |
| 734 |
|
no match. |
| 735 |
|
|
| 736 |
Arguments: |
Arguments: |
| 737 |
p points after the leading '<' |
p points after the leading '<' |
| 743 |
static int |
static int |
| 744 |
check_newline(uschar *p, FILE *f) |
check_newline(uschar *p, FILE *f) |
| 745 |
{ |
{ |
| 746 |
if (strncmp((char *)p, "cr>", 3) == 0) return PCRE_NEWLINE_CR; |
if (strncmpic(p, (uschar *)"cr>", 3) == 0) return PCRE_NEWLINE_CR; |
| 747 |
if (strncmp((char *)p, "lf>", 3) == 0) return PCRE_NEWLINE_LF; |
if (strncmpic(p, (uschar *)"lf>", 3) == 0) return PCRE_NEWLINE_LF; |
| 748 |
if (strncmp((char *)p, "crlf>", 5) == 0) return PCRE_NEWLINE_CRLF; |
if (strncmpic(p, (uschar *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF; |
| 749 |
if (strncmp((char *)p, "any>", 4) == 0) return PCRE_NEWLINE_ANY; |
if (strncmpic(p, (uschar *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF; |
| 750 |
|
if (strncmpic(p, (uschar *)"any>", 4) == 0) return PCRE_NEWLINE_ANY; |
| 751 |
|
if (strncmpic(p, (uschar *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF; |
| 752 |
|
if (strncmpic(p, (uschar *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE; |
| 753 |
fprintf(f, "Unknown newline type at: <%s\n", p); |
fprintf(f, "Unknown newline type at: <%s\n", p); |
| 754 |
return 0; |
return 0; |
| 755 |
} |
} |
| 763 |
static void |
static void |
| 764 |
usage(void) |
usage(void) |
| 765 |
{ |
{ |
| 766 |
printf("Usage: pcretest [options] [<input> [<output>]]\n"); |
printf("Usage: pcretest [options] [<input file> [<output file>]]\n\n"); |
| 767 |
|
printf("Input and output default to stdin and stdout.\n"); |
| 768 |
|
#ifdef SUPPORT_LIBREADLINE |
| 769 |
|
printf("If input is a terminal, readline() is used to read from it.\n"); |
| 770 |
|
#else |
| 771 |
|
printf("This version of pcretest is not linked with readline().\n"); |
| 772 |
|
#endif |
| 773 |
|
printf("\nOptions:\n"); |
| 774 |
printf(" -b show compiled code (bytecode)\n"); |
printf(" -b show compiled code (bytecode)\n"); |
| 775 |
printf(" -C show PCRE compile-time options and exit\n"); |
printf(" -C show PCRE compile-time options and exit\n"); |
| 776 |
printf(" -d debug: show compiled code and information (-b and -i)\n"); |
printf(" -d debug: show compiled code and information (-b and -i)\n"); |
| 928 |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
| 929 |
printf(" Newline sequence is %s\n", (rc == '\r')? "CR" : |
printf(" Newline sequence is %s\n", (rc == '\r')? "CR" : |
| 930 |
(rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" : |
(rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" : |
| 931 |
|
(rc == -2)? "ANYCRLF" : |
| 932 |
(rc == -1)? "ANY" : "???"); |
(rc == -1)? "ANY" : "???"); |
| 933 |
|
(void)pcre_config(PCRE_CONFIG_BSR, &rc); |
| 934 |
|
printf(" \\R matches %s\n", rc? "CR, LF, or CRLF only" : |
| 935 |
|
"all Unicode newlines"); |
| 936 |
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
| 937 |
printf(" Internal link size = %d\n", rc); |
printf(" Internal link size = %d\n", rc); |
| 938 |
(void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc); |
(void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc); |
| 969 |
if (offsets == NULL) |
if (offsets == NULL) |
| 970 |
{ |
{ |
| 971 |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
| 972 |
size_offsets_max * sizeof(int)); |
(int)(size_offsets_max * sizeof(int))); |
| 973 |
yield = 1; |
yield = 1; |
| 974 |
goto EXIT; |
goto EXIT; |
| 975 |
} |
} |
| 1029 |
size_t size, regex_gotten_store; |
size_t size, regex_gotten_store; |
| 1030 |
int do_study = 0; |
int do_study = 0; |
| 1031 |
int do_debug = debug; |
int do_debug = debug; |
|
int debug_lengths = 1; |
|
| 1032 |
int do_G = 0; |
int do_G = 0; |
| 1033 |
int do_g = 0; |
int do_g = 0; |
| 1034 |
int do_showinfo = showinfo; |
int do_showinfo = showinfo; |
| 1037 |
int erroroffset, len, delimiter, poffset; |
int erroroffset, len, delimiter, poffset; |
| 1038 |
|
|
| 1039 |
use_utf8 = 0; |
use_utf8 = 0; |
| 1040 |
|
debug_lengths = 1; |
| 1041 |
|
|
| 1042 |
if (infile == stdin) printf(" re> "); |
if (extend_inputline(infile, buffer, " re> ") == NULL) break; |
|
if (extend_inputline(infile, buffer) == NULL) break; |
|
| 1043 |
if (infile != stdin) fprintf(outfile, "%s", (char *)buffer); |
if (infile != stdin) fprintf(outfile, "%s", (char *)buffer); |
| 1044 |
fflush(outfile); |
fflush(outfile); |
| 1045 |
|
|
| 1139 |
|
|
| 1140 |
if (isalnum(delimiter) || delimiter == '\\') |
if (isalnum(delimiter) || delimiter == '\\') |
| 1141 |
{ |
{ |
| 1142 |
fprintf(outfile, "** Delimiter must not be alphameric or \\\n"); |
fprintf(outfile, "** Delimiter must not be alphanumeric or \\\n"); |
| 1143 |
goto SKIP_DATA; |
goto SKIP_DATA; |
| 1144 |
} |
} |
| 1145 |
|
|
| 1155 |
pp++; |
pp++; |
| 1156 |
} |
} |
| 1157 |
if (*pp != 0) break; |
if (*pp != 0) break; |
| 1158 |
if (infile == stdin) printf(" > "); |
if ((pp = extend_inputline(infile, pp, " > ")) == NULL) |
|
if ((pp = extend_inputline(infile, pp)) == NULL) |
|
| 1159 |
{ |
{ |
| 1160 |
fprintf(outfile, "** Unexpected EOF\n"); |
fprintf(outfile, "** Unexpected EOF\n"); |
| 1161 |
done = 1; |
done = 1; |
| 1247 |
|
|
| 1248 |
case '<': |
case '<': |
| 1249 |
{ |
{ |
| 1250 |
int x = check_newline(pp, outfile); |
if (strncmp((char *)pp, "JS>", 3) == 0) |
| 1251 |
if (x == 0) goto SKIP_DATA; |
{ |
| 1252 |
options |= x; |
options |= PCRE_JAVASCRIPT_COMPAT; |
| 1253 |
while (*pp++ != '>'); |
pp += 3; |
| 1254 |
|
} |
| 1255 |
|
else |
| 1256 |
|
{ |
| 1257 |
|
int x = check_newline(pp, outfile); |
| 1258 |
|
if (x == 0) goto SKIP_DATA; |
| 1259 |
|
options |= x; |
| 1260 |
|
while (*pp++ != '>'); |
| 1261 |
|
} |
| 1262 |
} |
} |
| 1263 |
break; |
break; |
| 1264 |
|
|
| 1337 |
{ |
{ |
| 1338 |
for (;;) |
for (;;) |
| 1339 |
{ |
{ |
| 1340 |
if (extend_inputline(infile, buffer) == NULL) |
if (extend_inputline(infile, buffer, NULL) == NULL) |
| 1341 |
{ |
{ |
| 1342 |
done = 1; |
done = 1; |
| 1343 |
goto CONTINUE; |
goto CONTINUE; |
| 1400 |
if (do_flip) |
if (do_flip) |
| 1401 |
{ |
{ |
| 1402 |
real_pcre *rre = (real_pcre *)re; |
real_pcre *rre = (real_pcre *)re; |
| 1403 |
rre->magic_number = byteflip(rre->magic_number, sizeof(rre->magic_number)); |
rre->magic_number = |
| 1404 |
|
byteflip(rre->magic_number, sizeof(rre->magic_number)); |
| 1405 |
rre->size = byteflip(rre->size, sizeof(rre->size)); |
rre->size = byteflip(rre->size, sizeof(rre->size)); |
| 1406 |
rre->options = byteflip(rre->options, sizeof(rre->options)); |
rre->options = byteflip(rre->options, sizeof(rre->options)); |
| 1407 |
rre->top_bracket = byteflip(rre->top_bracket, sizeof(rre->top_bracket)); |
rre->flags = (pcre_uint16)byteflip(rre->flags, sizeof(rre->flags)); |
| 1408 |
rre->top_backref = byteflip(rre->top_backref, sizeof(rre->top_backref)); |
rre->top_bracket = |
| 1409 |
rre->first_byte = byteflip(rre->first_byte, sizeof(rre->first_byte)); |
(pcre_uint16)byteflip(rre->top_bracket, sizeof(rre->top_bracket)); |
| 1410 |
rre->req_byte = byteflip(rre->req_byte, sizeof(rre->req_byte)); |
rre->top_backref = |
| 1411 |
rre->name_table_offset = byteflip(rre->name_table_offset, |
(pcre_uint16)byteflip(rre->top_backref, sizeof(rre->top_backref)); |
| 1412 |
|
rre->first_byte = |
| 1413 |
|
(pcre_uint16)byteflip(rre->first_byte, sizeof(rre->first_byte)); |
| 1414 |
|
rre->req_byte = |
| 1415 |
|
(pcre_uint16)byteflip(rre->req_byte, sizeof(rre->req_byte)); |
| 1416 |
|
rre->name_table_offset = (pcre_uint16)byteflip(rre->name_table_offset, |
| 1417 |
sizeof(rre->name_table_offset)); |
sizeof(rre->name_table_offset)); |
| 1418 |
rre->name_entry_size = byteflip(rre->name_entry_size, |
rre->name_entry_size = (pcre_uint16)byteflip(rre->name_entry_size, |
| 1419 |
sizeof(rre->name_entry_size)); |
sizeof(rre->name_entry_size)); |
| 1420 |
rre->name_count = byteflip(rre->name_count, sizeof(rre->name_count)); |
rre->name_count = (pcre_uint16)byteflip(rre->name_count, |
| 1421 |
|
sizeof(rre->name_count)); |
| 1422 |
|
|
| 1423 |
if (extra != NULL) |
if (extra != NULL) |
| 1424 |
{ |
{ |
| 1444 |
#if !defined NOINFOCHECK |
#if !defined NOINFOCHECK |
| 1445 |
int old_first_char, old_options, old_count; |
int old_first_char, old_options, old_count; |
| 1446 |
#endif |
#endif |
| 1447 |
int count, backrefmax, first_char, need_char; |
int count, backrefmax, first_char, need_char, okpartial, jchanged, |
| 1448 |
|
hascrorlf; |
| 1449 |
int nameentrysize, namecount; |
int nameentrysize, namecount; |
| 1450 |
const uschar *nametable; |
const uschar *nametable; |
| 1451 |
|
|
| 1458 |
new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize); |
new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize); |
| 1459 |
new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount); |
new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount); |
| 1460 |
new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable); |
new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable); |
| 1461 |
|
new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial); |
| 1462 |
|
new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged); |
| 1463 |
|
new_info(re, NULL, PCRE_INFO_HASCRORLF, &hascrorlf); |
| 1464 |
|
|
| 1465 |
#if !defined NOINFOCHECK |
#if !defined NOINFOCHECK |
| 1466 |
old_count = pcre_info(re, &old_options, &old_first_char); |
old_count = pcre_info(re, &old_options, &old_first_char); |
| 1502 |
} |
} |
| 1503 |
} |
} |
| 1504 |
|
|
| 1505 |
/* The NOPARTIAL bit is a private bit in the options, so we have |
if (!okpartial) fprintf(outfile, "Partial matching not supported\n"); |
| 1506 |
to fish it out via out back door */ |
if (hascrorlf) fprintf(outfile, "Contains explicit CR or LF match\n"); |
| 1507 |
|
|
| 1508 |
all_options = ((real_pcre *)re)->options; |
all_options = ((real_pcre *)re)->options; |
| 1509 |
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"); |
|
| 1510 |
|
|
| 1511 |
if (get_options == 0) fprintf(outfile, "No options\n"); |
if (get_options == 0) fprintf(outfile, "No options\n"); |
| 1512 |
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%s%s\n", |
| 1513 |
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
| 1514 |
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
| 1515 |
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
| 1516 |
((get_options & PCRE_MULTILINE) != 0)? " multiline" : "", |
((get_options & PCRE_MULTILINE) != 0)? " multiline" : "", |
| 1517 |
((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "", |
((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "", |
| 1518 |
((get_options & PCRE_DOTALL) != 0)? " dotall" : "", |
((get_options & PCRE_DOTALL) != 0)? " dotall" : "", |
| 1519 |
|
((get_options & PCRE_BSR_ANYCRLF) != 0)? " bsr_anycrlf" : "", |
| 1520 |
|
((get_options & PCRE_BSR_UNICODE) != 0)? " bsr_unicode" : "", |
| 1521 |
((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
| 1522 |
((get_options & PCRE_EXTRA) != 0)? " extra" : "", |
((get_options & PCRE_EXTRA) != 0)? " extra" : "", |
| 1523 |
((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "", |
((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "", |
| 1526 |
((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "", |
((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "", |
| 1527 |
((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : ""); |
((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : ""); |
| 1528 |
|
|
| 1529 |
|
if (jchanged) fprintf(outfile, "Duplicate name status changes\n"); |
| 1530 |
|
|
| 1531 |
switch (get_options & PCRE_NEWLINE_BITS) |
switch (get_options & PCRE_NEWLINE_BITS) |
| 1532 |
{ |
{ |
| 1533 |
case PCRE_NEWLINE_CR: |
case PCRE_NEWLINE_CR: |
| 1542 |
fprintf(outfile, "Forced newline sequence: CRLF\n"); |
fprintf(outfile, "Forced newline sequence: CRLF\n"); |
| 1543 |
break; |
break; |
| 1544 |
|
|
| 1545 |
|
case PCRE_NEWLINE_ANYCRLF: |
| 1546 |
|
fprintf(outfile, "Forced newline sequence: ANYCRLF\n"); |
| 1547 |
|
break; |
| 1548 |
|
|
| 1549 |
case PCRE_NEWLINE_ANY: |
case PCRE_NEWLINE_ANY: |
| 1550 |
fprintf(outfile, "Forced newline sequence: ANY\n"); |
fprintf(outfile, "Forced newline sequence: ANY\n"); |
| 1551 |
break; |
break; |
| 1650 |
else |
else |
| 1651 |
{ |
{ |
| 1652 |
uschar sbuf[8]; |
uschar sbuf[8]; |
| 1653 |
sbuf[0] = (true_size >> 24) & 255; |
sbuf[0] = (uschar)((true_size >> 24) & 255); |
| 1654 |
sbuf[1] = (true_size >> 16) & 255; |
sbuf[1] = (uschar)((true_size >> 16) & 255); |
| 1655 |
sbuf[2] = (true_size >> 8) & 255; |
sbuf[2] = (uschar)((true_size >> 8) & 255); |
| 1656 |
sbuf[3] = (true_size) & 255; |
sbuf[3] = (uschar)((true_size) & 255); |
| 1657 |
|
|
| 1658 |
sbuf[4] = (true_study_size >> 24) & 255; |
sbuf[4] = (uschar)((true_study_size >> 24) & 255); |
| 1659 |
sbuf[5] = (true_study_size >> 16) & 255; |
sbuf[5] = (uschar)((true_study_size >> 16) & 255); |
| 1660 |
sbuf[6] = (true_study_size >> 8) & 255; |
sbuf[6] = (uschar)((true_study_size >> 8) & 255); |
| 1661 |
sbuf[7] = (true_study_size) & 255; |
sbuf[7] = (uschar)((true_study_size) & 255); |
| 1662 |
|
|
| 1663 |
if (fwrite(sbuf, 1, 8, f) < 8 || |
if (fwrite(sbuf, 1, 8, f) < 8 || |
| 1664 |
fwrite(re, 1, true_size, f) < true_size) |
fwrite(re, 1, true_size, f) < true_size) |
| 1695 |
for (;;) |
for (;;) |
| 1696 |
{ |
{ |
| 1697 |
uschar *q; |
uschar *q; |
| 1698 |
uschar *bptr = dbuffer; |
uschar *bptr; |
| 1699 |
int *use_offsets = offsets; |
int *use_offsets = offsets; |
| 1700 |
int use_size_offsets = size_offsets; |
int use_size_offsets = size_offsets; |
| 1701 |
int callout_data = 0; |
int callout_data = 0; |
| 1732 |
len = 0; |
len = 0; |
| 1733 |
for (;;) |
for (;;) |
| 1734 |
{ |
{ |
| 1735 |
if (infile == stdin) printf("data> "); |
if (extend_inputline(infile, buffer + len, "data> ") == NULL) |
|
if (extend_inputline(infile, buffer + len) == NULL) |
|
| 1736 |
{ |
{ |
| 1737 |
if (len > 0) break; |
if (len > 0) break; |
| 1738 |
done = 1; |
done = 1; |
| 1750 |
p = buffer; |
p = buffer; |
| 1751 |
while (isspace(*p)) p++; |
while (isspace(*p)) p++; |
| 1752 |
|
|
| 1753 |
q = dbuffer; |
bptr = q = dbuffer; |
| 1754 |
while ((c = *p++) != 0) |
while ((c = *p++) != 0) |
| 1755 |
{ |
{ |
| 1756 |
int i = 0; |
int i = 0; |
| 1945 |
if (offsets == NULL) |
if (offsets == NULL) |
| 1946 |
{ |
{ |
| 1947 |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
| 1948 |
size_offsets_max * sizeof(int)); |
(int)(size_offsets_max * sizeof(int))); |
| 1949 |
yield = 1; |
yield = 1; |
| 1950 |
goto EXIT; |
goto EXIT; |
| 1951 |
} |
} |
| 2314 |
} |
} |
| 2315 |
|
|
| 2316 |
/* 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 |
| 2317 |
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 |
| 2318 |
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 |
| 2319 |
matching, the advance must be one character, not one byte. Fudge the |
string - that was checked before setting g_notempty. |
| 2320 |
offset values to achieve this. We won't be at the end of the string - |
|
| 2321 |
that was checked before setting g_notempty. */ |
Complication arises in the case when the newline option is "any" or |
| 2322 |
|
"anycrlf". If the previous match was at the end of a line terminated by |
| 2323 |
|
CRLF, an advance of one character just passes the \r, whereas we should |
| 2324 |
|
prefer the longer newline sequence, as does the code in pcre_exec(). |
| 2325 |
|
Fudge the offset value to achieve this. |
| 2326 |
|
|
| 2327 |
|
Otherwise, in the case of UTF-8 matching, the advance must be one |
| 2328 |
|
character, not one byte. */ |
| 2329 |
|
|
| 2330 |
else |
else |
| 2331 |
{ |
{ |
| 2332 |
if (g_notempty != 0) |
if (g_notempty != 0) |
| 2333 |
{ |
{ |
| 2334 |
int onechar = 1; |
int onechar = 1; |
| 2335 |
|
unsigned int obits = ((real_pcre *)re)->options; |
| 2336 |
use_offsets[0] = start_offset; |
use_offsets[0] = start_offset; |
| 2337 |
if (use_utf8) |
if ((obits & PCRE_NEWLINE_BITS) == 0) |
| 2338 |
|
{ |
| 2339 |
|
int d; |
| 2340 |
|
(void)pcre_config(PCRE_CONFIG_NEWLINE, &d); |
| 2341 |
|
obits = (d == '\r')? PCRE_NEWLINE_CR : |
| 2342 |
|
(d == '\n')? PCRE_NEWLINE_LF : |
| 2343 |
|
(d == ('\r'<<8 | '\n'))? PCRE_NEWLINE_CRLF : |
| 2344 |
|
(d == -2)? PCRE_NEWLINE_ANYCRLF : |
| 2345 |
|
(d == -1)? PCRE_NEWLINE_ANY : 0; |
| 2346 |
|
} |
| 2347 |
|
if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY || |
| 2348 |
|
(obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF) |
| 2349 |
|
&& |
| 2350 |
|
start_offset < len - 1 && |
| 2351 |
|
bptr[start_offset] == '\r' && |
| 2352 |
|
bptr[start_offset+1] == '\n') |
| 2353 |
|
onechar++; |
| 2354 |
|
else if (use_utf8) |
| 2355 |
{ |
{ |
| 2356 |
while (start_offset + onechar < len) |
while (start_offset + onechar < len) |
| 2357 |
{ |
{ |
| 2386 |
character. */ |
character. */ |
| 2387 |
|
|
| 2388 |
g_notempty = 0; |
g_notempty = 0; |
| 2389 |
|
|
| 2390 |
if (use_offsets[0] == use_offsets[1]) |
if (use_offsets[0] == use_offsets[1]) |
| 2391 |
{ |
{ |
| 2392 |
if (use_offsets[0] == len) break; |
if (use_offsets[0] == len) break; |