| 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 |
#ifndef _WIN32 |
#ifdef SUPPORT_LIBREADLINE |
| 52 |
#include <sys/resource.h> |
#ifdef HAVE_UNISTD_H |
| 53 |
|
#include <unistd.h> |
| 54 |
|
#endif |
| 55 |
|
#include <readline/readline.h> |
| 56 |
|
#include <readline/history.h> |
| 57 |
#endif |
#endif |
| 58 |
|
|
|
#define PCRE_SPY /* For Win32 build, import data, not export */ |
|
| 59 |
|
|
| 60 |
/* We include pcre_internal.h because we need the internal info for displaying |
/* A number of things vary for Windows builds. Originally, pcretest opened its |
| 61 |
the results of pcre_study() and we also need to know about the internal |
input and output without "b"; then I was told that "b" was needed in some |
| 62 |
macros, structures, and other internal data values; pcretest has "inside |
environments, so it was added for release 5.0 to both the input and output. (It |
| 63 |
information" compared to a program that strictly follows the PCRE API. */ |
makes no difference on Unix-like systems.) Later I was told that it is wrong |
| 64 |
|
for the input on Windows. I've now abstracted the modes into two macros that |
| 65 |
|
are set here, to make it easier to fiddle with them, and removed "b" from the |
| 66 |
|
input mode under Windows. */ |
| 67 |
|
|
| 68 |
|
#if defined(_WIN32) || defined(WIN32) |
| 69 |
|
#include <io.h> /* For _setmode() */ |
| 70 |
|
#include <fcntl.h> /* For _O_BINARY */ |
| 71 |
|
#define INPUT_MODE "r" |
| 72 |
|
#define OUTPUT_MODE "wb" |
| 73 |
|
|
| 74 |
|
#define isatty _isatty /* This is what Windows calls them, I'm told */ |
| 75 |
|
#define fileno _fileno |
| 76 |
|
|
| 77 |
|
#else |
| 78 |
|
#include <sys/time.h> /* These two includes are needed */ |
| 79 |
|
#include <sys/resource.h> /* for setrlimit(). */ |
| 80 |
|
#define INPUT_MODE "rb" |
| 81 |
|
#define OUTPUT_MODE "wb" |
| 82 |
|
#endif |
| 83 |
|
|
| 84 |
|
|
| 85 |
|
/* We have to include pcre_internal.h because we need the internal info for |
| 86 |
|
displaying the results of pcre_study() and we also need to know about the |
| 87 |
|
internal macros, structures, and other internal data values; pcretest has |
| 88 |
|
"inside information" compared to a program that strictly follows the PCRE API. |
| 89 |
|
|
| 90 |
|
Although pcre_internal.h does itself include pcre.h, we explicitly include it |
| 91 |
|
here before pcre_internal.h so that the PCRE_EXP_xxx macros get set |
| 92 |
|
appropriately for an application, not for building PCRE. */ |
| 93 |
|
|
| 94 |
|
#include "pcre.h" |
| 95 |
#include "pcre_internal.h" |
#include "pcre_internal.h" |
| 96 |
|
|
| 97 |
/* 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 |
| 105 |
#define _pcre_utf8_table4 utf8_table4 |
#define _pcre_utf8_table4 utf8_table4 |
| 106 |
#define _pcre_utt utt |
#define _pcre_utt utt |
| 107 |
#define _pcre_utt_size utt_size |
#define _pcre_utt_size utt_size |
| 108 |
|
#define _pcre_utt_names utt_names |
| 109 |
#define _pcre_OP_lengths OP_lengths |
#define _pcre_OP_lengths OP_lengths |
| 110 |
|
|
| 111 |
#include "pcre_tables.c" |
#include "pcre_tables.c" |
| 112 |
|
|
| 113 |
/* We also need the pcre_printint() function for printing out compiled |
/* We also need the pcre_printint() function for printing out compiled |
| 114 |
patterns. This function is in a separate file so that it can be included in |
patterns. This function is in a separate file so that it can be included in |
| 115 |
pcre_compile.c when that module is compiled with debugging enabled. */ |
pcre_compile.c when that module is compiled with debugging enabled. |
| 116 |
|
|
| 117 |
|
The definition of the macro PRINTABLE, which determines whether to print an |
| 118 |
|
output character as-is or as a hex value when showing compiled patterns, is |
| 119 |
|
contained in this file. We uses it here also, in cases when the locale has not |
| 120 |
|
been explicitly changed, so as to get consistent output from systems that |
| 121 |
|
differ in their output from isprint() even in the "C" locale. */ |
| 122 |
|
|
| 123 |
#include "pcre_printint.src" |
#include "pcre_printint.src" |
| 124 |
|
|
| 125 |
|
#define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c)) |
| 126 |
|
|
| 127 |
|
|
| 128 |
/* It is possible to compile this test program without including support for |
/* It is possible to compile this test program without including support for |
| 129 |
testing the POSIX interface, though this is not available via the standard |
testing the POSIX interface, though this is not available via the standard |
| 133 |
#include "pcreposix.h" |
#include "pcreposix.h" |
| 134 |
#endif |
#endif |
| 135 |
|
|
| 136 |
/* 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 |
| 137 |
build pcretest without support for UTF8 (define NOUTF8), without the interface |
Exim, to build pcretest without support for UTF8 (define NOUTF8), without the |
| 138 |
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 |
| 139 |
function (define NOINFOCHECK). */ |
"info" function (define NOINFOCHECK). In fact, we automatically cut out the |
| 140 |
|
UTF8 support if PCRE is built without it. */ |
| 141 |
|
|
| 142 |
|
#ifndef SUPPORT_UTF8 |
| 143 |
|
#ifndef NOUTF8 |
| 144 |
|
#define NOUTF8 |
| 145 |
|
#endif |
| 146 |
|
#endif |
| 147 |
|
|
| 148 |
|
|
| 149 |
/* Other parameters */ |
/* Other parameters */ |
| 156 |
#endif |
#endif |
| 157 |
#endif |
#endif |
| 158 |
|
|
| 159 |
|
/* This is the default loop count for timing. */ |
| 160 |
|
|
| 161 |
#define LOOPREPEAT 500000 |
#define LOOPREPEAT 500000 |
| 162 |
|
|
| 163 |
/* Static variables */ |
/* Static variables */ |
| 168 |
static int callout_extra; |
static int callout_extra; |
| 169 |
static int callout_fail_count; |
static int callout_fail_count; |
| 170 |
static int callout_fail_id; |
static int callout_fail_id; |
| 171 |
|
static int debug_lengths; |
| 172 |
static int first_callout; |
static int first_callout; |
| 173 |
|
static int locale_set = 0; |
| 174 |
static int show_malloc; |
static int show_malloc; |
| 175 |
static int use_utf8; |
static int use_utf8; |
| 176 |
static size_t gotten_store; |
static size_t gotten_store; |
| 200 |
Arguments: |
Arguments: |
| 201 |
f the file to read |
f the file to read |
| 202 |
start where in buffer to start (this *must* be within buffer) |
start where in buffer to start (this *must* be within buffer) |
| 203 |
|
prompt for stdin or readline() |
| 204 |
|
|
| 205 |
Returns: pointer to the start of new data |
Returns: pointer to the start of new data |
| 206 |
could be a copy of start, or could be moved |
could be a copy of start, or could be moved |
| 208 |
*/ |
*/ |
| 209 |
|
|
| 210 |
static uschar * |
static uschar * |
| 211 |
extend_inputline(FILE *f, uschar *start) |
extend_inputline(FILE *f, uschar *start, const char *prompt) |
| 212 |
{ |
{ |
| 213 |
uschar *here = start; |
uschar *here = start; |
| 214 |
|
|
| 215 |
for (;;) |
for (;;) |
| 216 |
{ |
{ |
| 217 |
int rlen = buffer_size - (here - buffer); |
int rlen = buffer_size - (here - buffer); |
| 218 |
|
|
| 219 |
if (rlen > 1000) |
if (rlen > 1000) |
| 220 |
{ |
{ |
| 221 |
int dlen; |
int dlen; |
| 222 |
if (fgets((char *)here, rlen, f) == NULL) |
|
| 223 |
return (here == start)? NULL : start; |
/* If libreadline support is required, use readline() to read a line if the |
| 224 |
|
input is a terminal. Note that readline() removes the trailing newline, so |
| 225 |
|
we must put it back again, to be compatible with fgets(). */ |
| 226 |
|
|
| 227 |
|
#ifdef SUPPORT_LIBREADLINE |
| 228 |
|
if (isatty(fileno(f))) |
| 229 |
|
{ |
| 230 |
|
size_t len; |
| 231 |
|
char *s = readline(prompt); |
| 232 |
|
if (s == NULL) return (here == start)? NULL : start; |
| 233 |
|
len = strlen(s); |
| 234 |
|
if (len > 0) add_history(s); |
| 235 |
|
if (len > rlen - 1) len = rlen - 1; |
| 236 |
|
memcpy(here, s, len); |
| 237 |
|
here[len] = '\n'; |
| 238 |
|
here[len+1] = 0; |
| 239 |
|
free(s); |
| 240 |
|
} |
| 241 |
|
else |
| 242 |
|
#endif |
| 243 |
|
|
| 244 |
|
/* Read the next line by normal means, prompting if the file is stdin. */ |
| 245 |
|
|
| 246 |
|
{ |
| 247 |
|
if (f == stdin) printf(prompt); |
| 248 |
|
if (fgets((char *)here, rlen, f) == NULL) |
| 249 |
|
return (here == start)? NULL : start; |
| 250 |
|
} |
| 251 |
|
|
| 252 |
dlen = (int)strlen((char *)here); |
dlen = (int)strlen((char *)here); |
| 253 |
if (dlen > 0 && here[dlen - 1] == '\n') return start; |
if (dlen > 0 && here[dlen - 1] == '\n') return start; |
| 254 |
here += dlen; |
here += dlen; |
| 300 |
|
|
| 301 |
/* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess |
/* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess |
| 302 |
around with conditional compilation, just do the job by hand. It is only used |
around with conditional compilation, just do the job by hand. It is only used |
| 303 |
for unpicking the -o argument, so just keep it simple. |
for unpicking arguments, so just keep it simple. |
| 304 |
|
|
| 305 |
Arguments: |
Arguments: |
| 306 |
str string to be converted |
str string to be converted |
| 398 |
Returns: number of characters placed in the buffer |
Returns: number of characters placed in the buffer |
| 399 |
*/ |
*/ |
| 400 |
|
|
| 401 |
|
#if !defined NOUTF8 |
| 402 |
|
|
| 403 |
static int |
static int |
| 404 |
ord2utf8(int cvalue, uschar *utf8bytes) |
ord2utf8(int cvalue, uschar *utf8bytes) |
| 405 |
{ |
{ |
| 416 |
return i + 1; |
return i + 1; |
| 417 |
} |
} |
| 418 |
|
|
| 419 |
|
#endif |
| 420 |
|
|
| 421 |
|
|
| 422 |
|
|
| 423 |
/************************************************* |
/************************************************* |
| 444 |
{ |
{ |
| 445 |
length -= rc - 1; |
length -= rc - 1; |
| 446 |
p += rc; |
p += rc; |
| 447 |
if (c < 256 && isprint(c)) |
if (PRINTHEX(c)) |
| 448 |
{ |
{ |
| 449 |
if (f != NULL) fprintf(f, "%c", c); |
if (f != NULL) fprintf(f, "%c", c); |
| 450 |
yield++; |
yield++; |
| 451 |
} |
} |
| 452 |
else |
else |
| 453 |
{ |
{ |
| 454 |
int n; |
int n = 4; |
| 455 |
if (f != NULL) fprintf(f, "\\x{%02x}%n", c, &n); |
if (f != NULL) fprintf(f, "\\x{%02x}", c); |
| 456 |
yield += n; |
yield += (n <= 0x000000ff)? 2 : |
| 457 |
|
(n <= 0x00000fff)? 3 : |
| 458 |
|
(n <= 0x0000ffff)? 4 : |
| 459 |
|
(n <= 0x000fffff)? 5 : 6; |
| 460 |
} |
} |
| 461 |
continue; |
continue; |
| 462 |
} |
} |
| 465 |
|
|
| 466 |
/* Not UTF-8, or malformed UTF-8 */ |
/* Not UTF-8, or malformed UTF-8 */ |
| 467 |
|
|
| 468 |
if (isprint(c = *(p++))) |
c = *p++; |
| 469 |
|
if (PRINTHEX(c)) |
| 470 |
{ |
{ |
| 471 |
if (f != NULL) fprintf(f, "%c", c); |
if (f != NULL) fprintf(f, "%c", c); |
| 472 |
yield++; |
yield++; |
| 705 |
|
|
| 706 |
|
|
| 707 |
/************************************************* |
/************************************************* |
| 708 |
|
* Case-independent strncmp() function * |
| 709 |
|
*************************************************/ |
| 710 |
|
|
| 711 |
|
/* |
| 712 |
|
Arguments: |
| 713 |
|
s first string |
| 714 |
|
t second string |
| 715 |
|
n number of characters to compare |
| 716 |
|
|
| 717 |
|
Returns: < 0, = 0, or > 0, according to the comparison |
| 718 |
|
*/ |
| 719 |
|
|
| 720 |
|
static int |
| 721 |
|
strncmpic(uschar *s, uschar *t, int n) |
| 722 |
|
{ |
| 723 |
|
while (n--) |
| 724 |
|
{ |
| 725 |
|
int c = tolower(*s++) - tolower(*t++); |
| 726 |
|
if (c) return c; |
| 727 |
|
} |
| 728 |
|
return 0; |
| 729 |
|
} |
| 730 |
|
|
| 731 |
|
|
| 732 |
|
|
| 733 |
|
/************************************************* |
| 734 |
* Check newline indicator * |
* Check newline indicator * |
| 735 |
*************************************************/ |
*************************************************/ |
| 736 |
|
|
| 737 |
/* 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 |
| 738 |
xxx is LF, CR, or CRLF. 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 |
| 739 |
|
no match. |
| 740 |
|
|
| 741 |
Arguments: |
Arguments: |
| 742 |
p points after the leading '<' |
p points after the leading '<' |
| 748 |
static int |
static int |
| 749 |
check_newline(uschar *p, FILE *f) |
check_newline(uschar *p, FILE *f) |
| 750 |
{ |
{ |
| 751 |
if (strncmp((char *)p, "cr>", 3) == 0) return PCRE_NEWLINE_CR; |
if (strncmpic(p, (uschar *)"cr>", 3) == 0) return PCRE_NEWLINE_CR; |
| 752 |
if (strncmp((char *)p, "lf>", 3) == 0) return PCRE_NEWLINE_LF; |
if (strncmpic(p, (uschar *)"lf>", 3) == 0) return PCRE_NEWLINE_LF; |
| 753 |
if (strncmp((char *)p, "crlf>", 5) == 0) return PCRE_NEWLINE_CRLF; |
if (strncmpic(p, (uschar *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF; |
| 754 |
|
if (strncmpic(p, (uschar *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF; |
| 755 |
|
if (strncmpic(p, (uschar *)"any>", 4) == 0) return PCRE_NEWLINE_ANY; |
| 756 |
|
if (strncmpic(p, (uschar *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF; |
| 757 |
|
if (strncmpic(p, (uschar *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE; |
| 758 |
fprintf(f, "Unknown newline type at: <%s\n", p); |
fprintf(f, "Unknown newline type at: <%s\n", p); |
| 759 |
return 0; |
return 0; |
| 760 |
} |
} |
| 762 |
|
|
| 763 |
|
|
| 764 |
/************************************************* |
/************************************************* |
| 765 |
|
* Usage function * |
| 766 |
|
*************************************************/ |
| 767 |
|
|
| 768 |
|
static void |
| 769 |
|
usage(void) |
| 770 |
|
{ |
| 771 |
|
printf("Usage: pcretest [options] [<input file> [<output file>]]\n\n"); |
| 772 |
|
printf("Input and output default to stdin and stdout.\n"); |
| 773 |
|
#ifdef SUPPORT_LIBREADLINE |
| 774 |
|
printf("If input is a terminal, readline() is used to read from it.\n"); |
| 775 |
|
#else |
| 776 |
|
printf("This version of pcretest is not linked with readline().\n"); |
| 777 |
|
#endif |
| 778 |
|
printf("\nOptions:\n"); |
| 779 |
|
printf(" -b show compiled code (bytecode)\n"); |
| 780 |
|
printf(" -C show PCRE compile-time options and exit\n"); |
| 781 |
|
printf(" -d debug: show compiled code and information (-b and -i)\n"); |
| 782 |
|
#if !defined NODFA |
| 783 |
|
printf(" -dfa force DFA matching for all subjects\n"); |
| 784 |
|
#endif |
| 785 |
|
printf(" -help show usage information\n"); |
| 786 |
|
printf(" -i show information about compiled patterns\n" |
| 787 |
|
" -m output memory used information\n" |
| 788 |
|
" -o <n> set size of offsets vector to <n>\n"); |
| 789 |
|
#if !defined NOPOSIX |
| 790 |
|
printf(" -p use POSIX interface\n"); |
| 791 |
|
#endif |
| 792 |
|
printf(" -q quiet: do not output PCRE version number at start\n"); |
| 793 |
|
printf(" -S <n> set stack size to <n> megabytes\n"); |
| 794 |
|
printf(" -s output store (memory) used information\n" |
| 795 |
|
" -t time compilation and execution\n"); |
| 796 |
|
printf(" -t <n> time compilation and execution, repeating <n> times\n"); |
| 797 |
|
printf(" -tm time execution (matching) only\n"); |
| 798 |
|
printf(" -tm <n> time execution (matching) only, repeating <n> times\n"); |
| 799 |
|
} |
| 800 |
|
|
| 801 |
|
|
| 802 |
|
|
| 803 |
|
/************************************************* |
| 804 |
* Main Program * |
* Main Program * |
| 805 |
*************************************************/ |
*************************************************/ |
| 806 |
|
|
| 815 |
int study_options = 0; |
int study_options = 0; |
| 816 |
int op = 1; |
int op = 1; |
| 817 |
int timeit = 0; |
int timeit = 0; |
| 818 |
|
int timeitm = 0; |
| 819 |
int showinfo = 0; |
int showinfo = 0; |
| 820 |
int showstore = 0; |
int showstore = 0; |
| 821 |
int quiet = 0; |
int quiet = 0; |
| 847 |
dbuffer = (unsigned char *)malloc(buffer_size); |
dbuffer = (unsigned char *)malloc(buffer_size); |
| 848 |
pbuffer = (unsigned char *)malloc(buffer_size); |
pbuffer = (unsigned char *)malloc(buffer_size); |
| 849 |
|
|
| 850 |
/* The outfile variable is static so that new_malloc can use it. The _setmode() |
/* The outfile variable is static so that new_malloc can use it. */ |
|
stuff is some magic that I don't understand, but which apparently does good |
|
|
things in Windows. It's related to line terminations. */ |
|
|
|
|
|
#if defined(_WIN32) || defined(WIN32) |
|
|
_setmode( _fileno( stdout ), 0x8000 ); |
|
|
#endif /* defined(_WIN32) || defined(WIN32) */ |
|
| 851 |
|
|
| 852 |
outfile = stdout; |
outfile = stdout; |
| 853 |
|
|
| 854 |
|
/* The following _setmode() stuff is some Windows magic that tells its runtime |
| 855 |
|
library to translate CRLF into a single LF character. At least, that's what |
| 856 |
|
I've been told: never having used Windows I take this all on trust. Originally |
| 857 |
|
it set 0x8000, but then I was advised that _O_BINARY was better. */ |
| 858 |
|
|
| 859 |
|
#if defined(_WIN32) || defined(WIN32) |
| 860 |
|
_setmode( _fileno( stdout ), _O_BINARY ); |
| 861 |
|
#endif |
| 862 |
|
|
| 863 |
/* Scan options */ |
/* Scan options */ |
| 864 |
|
|
| 865 |
while (argc > 1 && argv[op][0] == '-') |
while (argc > 1 && argv[op][0] == '-') |
| 868 |
|
|
| 869 |
if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0) |
if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0) |
| 870 |
showstore = 1; |
showstore = 1; |
|
else if (strcmp(argv[op], "-t") == 0) timeit = 1; |
|
| 871 |
else if (strcmp(argv[op], "-q") == 0) quiet = 1; |
else if (strcmp(argv[op], "-q") == 0) quiet = 1; |
| 872 |
|
else if (strcmp(argv[op], "-b") == 0) debug = 1; |
| 873 |
else if (strcmp(argv[op], "-i") == 0) showinfo = 1; |
else if (strcmp(argv[op], "-i") == 0) showinfo = 1; |
| 874 |
else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1; |
else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1; |
| 875 |
#if !defined NODFA |
#if !defined NODFA |
| 882 |
op++; |
op++; |
| 883 |
argc--; |
argc--; |
| 884 |
} |
} |
| 885 |
|
else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0) |
| 886 |
|
{ |
| 887 |
|
int both = argv[op][2] == 0; |
| 888 |
|
int temp; |
| 889 |
|
if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr), |
| 890 |
|
*endptr == 0)) |
| 891 |
|
{ |
| 892 |
|
timeitm = temp; |
| 893 |
|
op++; |
| 894 |
|
argc--; |
| 895 |
|
} |
| 896 |
|
else timeitm = LOOPREPEAT; |
| 897 |
|
if (both) timeit = timeitm; |
| 898 |
|
} |
| 899 |
else if (strcmp(argv[op], "-S") == 0 && argc > 2 && |
else if (strcmp(argv[op], "-S") == 0 && argc > 2 && |
| 900 |
((stack_size = get_value((unsigned char *)argv[op+1], &endptr)), |
((stack_size = get_value((unsigned char *)argv[op+1], &endptr)), |
| 901 |
*endptr == 0)) |
*endptr == 0)) |
| 902 |
{ |
{ |
| 903 |
#ifdef _WIN32 |
#if defined(_WIN32) || defined(WIN32) |
| 904 |
printf("PCRE: -S not supported on this OS\n"); |
printf("PCRE: -S not supported on this OS\n"); |
| 905 |
exit(1); |
exit(1); |
| 906 |
#else |
#else |
| 932 |
printf(" %sUnicode properties support\n", rc? "" : "No "); |
printf(" %sUnicode properties support\n", rc? "" : "No "); |
| 933 |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
| 934 |
printf(" Newline sequence is %s\n", (rc == '\r')? "CR" : |
printf(" Newline sequence is %s\n", (rc == '\r')? "CR" : |
| 935 |
(rc == '\n')? "LF" : "CRLF"); |
(rc == '\n')? "LF" : (rc == ('\r'<<8 | '\n'))? "CRLF" : |
| 936 |
|
(rc == -2)? "ANYCRLF" : |
| 937 |
|
(rc == -1)? "ANY" : "???"); |
| 938 |
|
(void)pcre_config(PCRE_CONFIG_BSR, &rc); |
| 939 |
|
printf(" \\R matches %s\n", rc? "CR, LF, or CRLF only" : |
| 940 |
|
"all Unicode newlines"); |
| 941 |
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
| 942 |
printf(" Internal link size = %d\n", rc); |
printf(" Internal link size = %d\n", rc); |
| 943 |
(void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc); |
(void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc); |
| 948 |
printf(" Default recursion depth limit = %d\n", rc); |
printf(" Default recursion depth limit = %d\n", rc); |
| 949 |
(void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc); |
(void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc); |
| 950 |
printf(" Match recursion uses %s\n", rc? "stack" : "heap"); |
printf(" Match recursion uses %s\n", rc? "stack" : "heap"); |
| 951 |
exit(0); |
goto EXIT; |
| 952 |
|
} |
| 953 |
|
else if (strcmp(argv[op], "-help") == 0 || |
| 954 |
|
strcmp(argv[op], "--help") == 0) |
| 955 |
|
{ |
| 956 |
|
usage(); |
| 957 |
|
goto EXIT; |
| 958 |
} |
} |
| 959 |
else |
else |
| 960 |
{ |
{ |
| 961 |
printf("** Unknown or malformed option %s\n", argv[op]); |
printf("** Unknown or malformed option %s\n", argv[op]); |
| 962 |
printf("Usage: pcretest [options] [<input> [<output>]]\n"); |
usage(); |
|
printf(" -C show PCRE compile-time options and exit\n"); |
|
|
printf(" -d debug: show compiled code; implies -i\n"); |
|
|
#if !defined NODFA |
|
|
printf(" -dfa force DFA matching for all subjects\n"); |
|
|
#endif |
|
|
printf(" -i show information about compiled pattern\n" |
|
|
" -m output memory used information\n" |
|
|
" -o <n> set size of offsets vector to <n>\n"); |
|
|
#if !defined NOPOSIX |
|
|
printf(" -p use POSIX interface\n"); |
|
|
#endif |
|
|
printf(" -S <n> set stack size to <n> megabytes\n"); |
|
|
printf(" -s output store (memory) used information\n" |
|
|
" -t time compilation and execution\n"); |
|
| 963 |
yield = 1; |
yield = 1; |
| 964 |
goto EXIT; |
goto EXIT; |
| 965 |
} |
} |
| 974 |
if (offsets == NULL) |
if (offsets == NULL) |
| 975 |
{ |
{ |
| 976 |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
| 977 |
size_offsets_max * sizeof(int)); |
(int)(size_offsets_max * sizeof(int))); |
| 978 |
yield = 1; |
yield = 1; |
| 979 |
goto EXIT; |
goto EXIT; |
| 980 |
} |
} |
| 983 |
|
|
| 984 |
if (argc > 1) |
if (argc > 1) |
| 985 |
{ |
{ |
| 986 |
infile = fopen(argv[op], "rb"); |
infile = fopen(argv[op], INPUT_MODE); |
| 987 |
if (infile == NULL) |
if (infile == NULL) |
| 988 |
{ |
{ |
| 989 |
printf("** Failed to open %s\n", argv[op]); |
printf("** Failed to open %s\n", argv[op]); |
| 994 |
|
|
| 995 |
if (argc > 2) |
if (argc > 2) |
| 996 |
{ |
{ |
| 997 |
outfile = fopen(argv[op+1], "wb"); |
outfile = fopen(argv[op+1], OUTPUT_MODE); |
| 998 |
if (outfile == NULL) |
if (outfile == NULL) |
| 999 |
{ |
{ |
| 1000 |
printf("** Failed to open %s\n", argv[op+1]); |
printf("** Failed to open %s\n", argv[op+1]); |
| 1039 |
int do_showinfo = showinfo; |
int do_showinfo = showinfo; |
| 1040 |
int do_showrest = 0; |
int do_showrest = 0; |
| 1041 |
int do_flip = 0; |
int do_flip = 0; |
| 1042 |
int erroroffset, len, delimiter; |
int erroroffset, len, delimiter, poffset; |
| 1043 |
|
|
| 1044 |
use_utf8 = 0; |
use_utf8 = 0; |
| 1045 |
|
debug_lengths = 1; |
| 1046 |
|
|
| 1047 |
if (infile == stdin) printf(" re> "); |
if (extend_inputline(infile, buffer, " re> ") == NULL) break; |
|
if (extend_inputline(infile, buffer) == NULL) break; |
|
| 1048 |
if (infile != stdin) fprintf(outfile, "%s", (char *)buffer); |
if (infile != stdin) fprintf(outfile, "%s", (char *)buffer); |
| 1049 |
fflush(outfile); |
fflush(outfile); |
| 1050 |
|
|
| 1144 |
|
|
| 1145 |
if (isalnum(delimiter) || delimiter == '\\') |
if (isalnum(delimiter) || delimiter == '\\') |
| 1146 |
{ |
{ |
| 1147 |
fprintf(outfile, "** Delimiter must not be alphameric or \\\n"); |
fprintf(outfile, "** Delimiter must not be alphanumeric or \\\n"); |
| 1148 |
goto SKIP_DATA; |
goto SKIP_DATA; |
| 1149 |
} |
} |
| 1150 |
|
|
| 1151 |
pp = p; |
pp = p; |
| 1152 |
|
poffset = p - buffer; |
| 1153 |
|
|
| 1154 |
for(;;) |
for(;;) |
| 1155 |
{ |
{ |
| 1160 |
pp++; |
pp++; |
| 1161 |
} |
} |
| 1162 |
if (*pp != 0) break; |
if (*pp != 0) break; |
| 1163 |
if (infile == stdin) printf(" > "); |
if ((pp = extend_inputline(infile, pp, " > ")) == NULL) |
|
if ((pp = extend_inputline(infile, pp)) == NULL) |
|
| 1164 |
{ |
{ |
| 1165 |
fprintf(outfile, "** Unexpected EOF\n"); |
fprintf(outfile, "** Unexpected EOF\n"); |
| 1166 |
done = 1; |
done = 1; |
| 1169 |
if (infile != stdin) fprintf(outfile, "%s", (char *)pp); |
if (infile != stdin) fprintf(outfile, "%s", (char *)pp); |
| 1170 |
} |
} |
| 1171 |
|
|
| 1172 |
|
/* The buffer may have moved while being extended; reset the start of data |
| 1173 |
|
pointer to the correct relative point in the buffer. */ |
| 1174 |
|
|
| 1175 |
|
p = buffer + poffset; |
| 1176 |
|
|
| 1177 |
/* If the first character after the delimiter is backslash, make |
/* If the first character after the delimiter is backslash, make |
| 1178 |
the pattern end with backslash. This is purely to provide a way |
the pattern end with backslash. This is purely to provide a way |
| 1179 |
of testing for the error message when a pattern ends with backslash. */ |
of testing for the error message when a pattern ends with backslash. */ |
| 1205 |
|
|
| 1206 |
case '+': do_showrest = 1; break; |
case '+': do_showrest = 1; break; |
| 1207 |
case 'A': options |= PCRE_ANCHORED; break; |
case 'A': options |= PCRE_ANCHORED; break; |
| 1208 |
|
case 'B': do_debug = 1; break; |
| 1209 |
case 'C': options |= PCRE_AUTO_CALLOUT; break; |
case 'C': options |= PCRE_AUTO_CALLOUT; break; |
| 1210 |
case 'D': do_debug = do_showinfo = 1; break; |
case 'D': do_debug = do_showinfo = 1; break; |
| 1211 |
case 'E': options |= PCRE_DOLLAR_ENDONLY; break; |
case 'E': options |= PCRE_DOLLAR_ENDONLY; break; |
| 1223 |
case 'S': do_study = 1; break; |
case 'S': do_study = 1; break; |
| 1224 |
case 'U': options |= PCRE_UNGREEDY; break; |
case 'U': options |= PCRE_UNGREEDY; break; |
| 1225 |
case 'X': options |= PCRE_EXTRA; break; |
case 'X': options |= PCRE_EXTRA; break; |
| 1226 |
|
case 'Z': debug_lengths = 0; break; |
| 1227 |
case '8': options |= PCRE_UTF8; use_utf8 = 1; break; |
case '8': options |= PCRE_UTF8; use_utf8 = 1; break; |
| 1228 |
case '?': options |= PCRE_NO_UTF8_CHECK; break; |
case '?': options |= PCRE_NO_UTF8_CHECK; break; |
| 1229 |
|
|
| 1230 |
case 'L': |
case 'L': |
| 1231 |
ppp = pp; |
ppp = pp; |
| 1232 |
/* The '\r' test here is so that it works on Windows */ |
/* The '\r' test here is so that it works on Windows. */ |
| 1233 |
while (*ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++; |
/* The '0' test is just in case this is an unterminated line. */ |
| 1234 |
|
while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++; |
| 1235 |
*ppp = 0; |
*ppp = 0; |
| 1236 |
if (setlocale(LC_CTYPE, (const char *)pp) == NULL) |
if (setlocale(LC_CTYPE, (const char *)pp) == NULL) |
| 1237 |
{ |
{ |
| 1238 |
fprintf(outfile, "** Failed to set locale \"%s\"\n", pp); |
fprintf(outfile, "** Failed to set locale \"%s\"\n", pp); |
| 1239 |
goto SKIP_DATA; |
goto SKIP_DATA; |
| 1240 |
} |
} |
| 1241 |
|
locale_set = 1; |
| 1242 |
tables = pcre_maketables(); |
tables = pcre_maketables(); |
| 1243 |
pp = ppp; |
pp = ppp; |
| 1244 |
break; |
break; |
| 1252 |
|
|
| 1253 |
case '<': |
case '<': |
| 1254 |
{ |
{ |
| 1255 |
int x = check_newline(pp, outfile); |
if (strncmp((char *)pp, "JS>", 3) == 0) |
| 1256 |
if (x == 0) goto SKIP_DATA; |
{ |
| 1257 |
options |= x; |
options |= PCRE_JAVASCRIPT_COMPAT; |
| 1258 |
while (*pp++ != '>'); |
pp += 3; |
| 1259 |
|
} |
| 1260 |
|
else |
| 1261 |
|
{ |
| 1262 |
|
int x = check_newline(pp, outfile); |
| 1263 |
|
if (x == 0) goto SKIP_DATA; |
| 1264 |
|
options |= x; |
| 1265 |
|
while (*pp++ != '>'); |
| 1266 |
|
} |
| 1267 |
} |
} |
| 1268 |
break; |
break; |
| 1269 |
|
|
| 1313 |
#endif /* !defined NOPOSIX */ |
#endif /* !defined NOPOSIX */ |
| 1314 |
|
|
| 1315 |
{ |
{ |
| 1316 |
if (timeit) |
if (timeit > 0) |
| 1317 |
{ |
{ |
| 1318 |
register int i; |
register int i; |
| 1319 |
clock_t time_taken; |
clock_t time_taken; |
| 1320 |
clock_t start_time = clock(); |
clock_t start_time = clock(); |
| 1321 |
for (i = 0; i < LOOPREPEAT; i++) |
for (i = 0; i < timeit; i++) |
| 1322 |
{ |
{ |
| 1323 |
re = pcre_compile((char *)p, options, &error, &erroroffset, tables); |
re = pcre_compile((char *)p, options, &error, &erroroffset, tables); |
| 1324 |
if (re != NULL) free(re); |
if (re != NULL) free(re); |
| 1325 |
} |
} |
| 1326 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
| 1327 |
fprintf(outfile, "Compile time %.3f milliseconds\n", |
fprintf(outfile, "Compile time %.4f milliseconds\n", |
| 1328 |
(((double)time_taken * 1000.0) / (double)LOOPREPEAT) / |
(((double)time_taken * 1000.0) / (double)timeit) / |
| 1329 |
(double)CLOCKS_PER_SEC); |
(double)CLOCKS_PER_SEC); |
| 1330 |
} |
} |
| 1331 |
|
|
| 1342 |
{ |
{ |
| 1343 |
for (;;) |
for (;;) |
| 1344 |
{ |
{ |
| 1345 |
if (extend_inputline(infile, buffer) == NULL) |
if (extend_inputline(infile, buffer, NULL) == NULL) |
| 1346 |
{ |
{ |
| 1347 |
done = 1; |
done = 1; |
| 1348 |
goto CONTINUE; |
goto CONTINUE; |
| 1377 |
|
|
| 1378 |
if (do_study) |
if (do_study) |
| 1379 |
{ |
{ |
| 1380 |
if (timeit) |
if (timeit > 0) |
| 1381 |
{ |
{ |
| 1382 |
register int i; |
register int i; |
| 1383 |
clock_t time_taken; |
clock_t time_taken; |
| 1384 |
clock_t start_time = clock(); |
clock_t start_time = clock(); |
| 1385 |
for (i = 0; i < LOOPREPEAT; i++) |
for (i = 0; i < timeit; i++) |
| 1386 |
extra = pcre_study(re, study_options, &error); |
extra = pcre_study(re, study_options, &error); |
| 1387 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
| 1388 |
if (extra != NULL) free(extra); |
if (extra != NULL) free(extra); |
| 1389 |
fprintf(outfile, " Study time %.3f milliseconds\n", |
fprintf(outfile, " Study time %.4f milliseconds\n", |
| 1390 |
(((double)time_taken * 1000.0) / (double)LOOPREPEAT) / |
(((double)time_taken * 1000.0) / (double)timeit) / |
| 1391 |
(double)CLOCKS_PER_SEC); |
(double)CLOCKS_PER_SEC); |
| 1392 |
} |
} |
| 1393 |
extra = pcre_study(re, study_options, &error); |
extra = pcre_study(re, study_options, &error); |
| 1405 |
if (do_flip) |
if (do_flip) |
| 1406 |
{ |
{ |
| 1407 |
real_pcre *rre = (real_pcre *)re; |
real_pcre *rre = (real_pcre *)re; |
| 1408 |
rre->magic_number = byteflip(rre->magic_number, sizeof(rre->magic_number)); |
rre->magic_number = |
| 1409 |
|
byteflip(rre->magic_number, sizeof(rre->magic_number)); |
| 1410 |
rre->size = byteflip(rre->size, sizeof(rre->size)); |
rre->size = byteflip(rre->size, sizeof(rre->size)); |
| 1411 |
rre->options = byteflip(rre->options, sizeof(rre->options)); |
rre->options = byteflip(rre->options, sizeof(rre->options)); |
| 1412 |
rre->top_bracket = byteflip(rre->top_bracket, sizeof(rre->top_bracket)); |
rre->flags = (pcre_uint16)byteflip(rre->flags, sizeof(rre->flags)); |
| 1413 |
rre->top_backref = byteflip(rre->top_backref, sizeof(rre->top_backref)); |
rre->top_bracket = |
| 1414 |
rre->first_byte = byteflip(rre->first_byte, sizeof(rre->first_byte)); |
(pcre_uint16)byteflip(rre->top_bracket, sizeof(rre->top_bracket)); |
| 1415 |
rre->req_byte = byteflip(rre->req_byte, sizeof(rre->req_byte)); |
rre->top_backref = |
| 1416 |
rre->name_table_offset = byteflip(rre->name_table_offset, |
(pcre_uint16)byteflip(rre->top_backref, sizeof(rre->top_backref)); |
| 1417 |
|
rre->first_byte = |
| 1418 |
|
(pcre_uint16)byteflip(rre->first_byte, sizeof(rre->first_byte)); |
| 1419 |
|
rre->req_byte = |
| 1420 |
|
(pcre_uint16)byteflip(rre->req_byte, sizeof(rre->req_byte)); |
| 1421 |
|
rre->name_table_offset = (pcre_uint16)byteflip(rre->name_table_offset, |
| 1422 |
sizeof(rre->name_table_offset)); |
sizeof(rre->name_table_offset)); |
| 1423 |
rre->name_entry_size = byteflip(rre->name_entry_size, |
rre->name_entry_size = (pcre_uint16)byteflip(rre->name_entry_size, |
| 1424 |
sizeof(rre->name_entry_size)); |
sizeof(rre->name_entry_size)); |
| 1425 |
rre->name_count = byteflip(rre->name_count, sizeof(rre->name_count)); |
rre->name_count = (pcre_uint16)byteflip(rre->name_count, |
| 1426 |
|
sizeof(rre->name_count)); |
| 1427 |
|
|
| 1428 |
if (extra != NULL) |
if (extra != NULL) |
| 1429 |
{ |
{ |
| 1437 |
|
|
| 1438 |
SHOW_INFO: |
SHOW_INFO: |
| 1439 |
|
|
| 1440 |
|
if (do_debug) |
| 1441 |
|
{ |
| 1442 |
|
fprintf(outfile, "------------------------------------------------------------------\n"); |
| 1443 |
|
pcre_printint(re, outfile, debug_lengths); |
| 1444 |
|
} |
| 1445 |
|
|
| 1446 |
if (do_showinfo) |
if (do_showinfo) |
| 1447 |
{ |
{ |
| 1448 |
unsigned long int get_options, all_options; |
unsigned long int get_options, all_options; |
| 1449 |
#if !defined NOINFOCHECK |
#if !defined NOINFOCHECK |
| 1450 |
int old_first_char, old_options, old_count; |
int old_first_char, old_options, old_count; |
| 1451 |
#endif |
#endif |
| 1452 |
int count, backrefmax, first_char, need_char; |
int count, backrefmax, first_char, need_char, okpartial, jchanged, |
| 1453 |
|
hascrorlf; |
| 1454 |
int nameentrysize, namecount; |
int nameentrysize, namecount; |
| 1455 |
const uschar *nametable; |
const uschar *nametable; |
| 1456 |
|
|
|
if (do_debug) |
|
|
{ |
|
|
fprintf(outfile, "------------------------------------------------------------------\n"); |
|
|
pcre_printint(re, outfile); |
|
|
} |
|
|
|
|
| 1457 |
new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options); |
new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options); |
| 1458 |
new_info(re, NULL, PCRE_INFO_SIZE, &size); |
new_info(re, NULL, PCRE_INFO_SIZE, &size); |
| 1459 |
new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count); |
new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count); |
| 1463 |
new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize); |
new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize); |
| 1464 |
new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount); |
new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount); |
| 1465 |
new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable); |
new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable); |
| 1466 |
|
new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial); |
| 1467 |
|
new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged); |
| 1468 |
|
new_info(re, NULL, PCRE_INFO_HASCRORLF, &hascrorlf); |
| 1469 |
|
|
| 1470 |
#if !defined NOINFOCHECK |
#if !defined NOINFOCHECK |
| 1471 |
old_count = pcre_info(re, &old_options, &old_first_char); |
old_count = pcre_info(re, &old_options, &old_first_char); |
| 1507 |
} |
} |
| 1508 |
} |
} |
| 1509 |
|
|
| 1510 |
/* The NOPARTIAL bit is a private bit in the options, so we have |
if (!okpartial) fprintf(outfile, "Partial matching not supported\n"); |
| 1511 |
to fish it out via out back door */ |
if (hascrorlf) fprintf(outfile, "Contains explicit CR or LF match\n"); |
| 1512 |
|
|
| 1513 |
all_options = ((real_pcre *)re)->options; |
all_options = ((real_pcre *)re)->options; |
| 1514 |
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"); |
|
| 1515 |
|
|
| 1516 |
if (get_options == 0) fprintf(outfile, "No options\n"); |
if (get_options == 0) fprintf(outfile, "No options\n"); |
| 1517 |
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", |
| 1518 |
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
| 1519 |
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
| 1520 |
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
| 1521 |
((get_options & PCRE_MULTILINE) != 0)? " multiline" : "", |
((get_options & PCRE_MULTILINE) != 0)? " multiline" : "", |
| 1522 |
((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "", |
((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "", |
| 1523 |
((get_options & PCRE_DOTALL) != 0)? " dotall" : "", |
((get_options & PCRE_DOTALL) != 0)? " dotall" : "", |
| 1524 |
|
((get_options & PCRE_BSR_ANYCRLF) != 0)? " bsr_anycrlf" : "", |
| 1525 |
|
((get_options & PCRE_BSR_UNICODE) != 0)? " bsr_unicode" : "", |
| 1526 |
((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
| 1527 |
((get_options & PCRE_EXTRA) != 0)? " extra" : "", |
((get_options & PCRE_EXTRA) != 0)? " extra" : "", |
| 1528 |
((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "", |
((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "", |
| 1531 |
((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "", |
((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "", |
| 1532 |
((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : ""); |
((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : ""); |
| 1533 |
|
|
| 1534 |
switch (get_options & PCRE_NEWLINE_CRLF) |
if (jchanged) fprintf(outfile, "Duplicate name status changes\n"); |
| 1535 |
|
|
| 1536 |
|
switch (get_options & PCRE_NEWLINE_BITS) |
| 1537 |
{ |
{ |
| 1538 |
case PCRE_NEWLINE_CR: |
case PCRE_NEWLINE_CR: |
| 1539 |
fprintf(outfile, "Forced newline sequence: CR\n"); |
fprintf(outfile, "Forced newline sequence: CR\n"); |
| 1547 |
fprintf(outfile, "Forced newline sequence: CRLF\n"); |
fprintf(outfile, "Forced newline sequence: CRLF\n"); |
| 1548 |
break; |
break; |
| 1549 |
|
|
| 1550 |
|
case PCRE_NEWLINE_ANYCRLF: |
| 1551 |
|
fprintf(outfile, "Forced newline sequence: ANYCRLF\n"); |
| 1552 |
|
break; |
| 1553 |
|
|
| 1554 |
|
case PCRE_NEWLINE_ANY: |
| 1555 |
|
fprintf(outfile, "Forced newline sequence: ANY\n"); |
| 1556 |
|
break; |
| 1557 |
|
|
| 1558 |
default: |
default: |
| 1559 |
break; |
break; |
| 1560 |
} |
} |
| 1572 |
int ch = first_char & 255; |
int ch = first_char & 255; |
| 1573 |
const char *caseless = ((first_char & REQ_CASELESS) == 0)? |
const char *caseless = ((first_char & REQ_CASELESS) == 0)? |
| 1574 |
"" : " (caseless)"; |
"" : " (caseless)"; |
| 1575 |
if (isprint(ch)) |
if (PRINTHEX(ch)) |
| 1576 |
fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless); |
fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless); |
| 1577 |
else |
else |
| 1578 |
fprintf(outfile, "First char = %d%s\n", ch, caseless); |
fprintf(outfile, "First char = %d%s\n", ch, caseless); |
| 1587 |
int ch = need_char & 255; |
int ch = need_char & 255; |
| 1588 |
const char *caseless = ((need_char & REQ_CASELESS) == 0)? |
const char *caseless = ((need_char & REQ_CASELESS) == 0)? |
| 1589 |
"" : " (caseless)"; |
"" : " (caseless)"; |
| 1590 |
if (isprint(ch)) |
if (PRINTHEX(ch)) |
| 1591 |
fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless); |
fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless); |
| 1592 |
else |
else |
| 1593 |
fprintf(outfile, "Need char = %d%s\n", ch, caseless); |
fprintf(outfile, "Need char = %d%s\n", ch, caseless); |
| 1623 |
fprintf(outfile, "\n "); |
fprintf(outfile, "\n "); |
| 1624 |
c = 2; |
c = 2; |
| 1625 |
} |
} |
| 1626 |
if (isprint(i) && i != ' ') |
if (PRINTHEX(i) && i != ' ') |
| 1627 |
{ |
{ |
| 1628 |
fprintf(outfile, "%c ", i); |
fprintf(outfile, "%c ", i); |
| 1629 |
c += 2; |
c += 2; |
| 1655 |
else |
else |
| 1656 |
{ |
{ |
| 1657 |
uschar sbuf[8]; |
uschar sbuf[8]; |
| 1658 |
sbuf[0] = (true_size >> 24) & 255; |
sbuf[0] = (uschar)((true_size >> 24) & 255); |
| 1659 |
sbuf[1] = (true_size >> 16) & 255; |
sbuf[1] = (uschar)((true_size >> 16) & 255); |
| 1660 |
sbuf[2] = (true_size >> 8) & 255; |
sbuf[2] = (uschar)((true_size >> 8) & 255); |
| 1661 |
sbuf[3] = (true_size) & 255; |
sbuf[3] = (uschar)((true_size) & 255); |
| 1662 |
|
|
| 1663 |
sbuf[4] = (true_study_size >> 24) & 255; |
sbuf[4] = (uschar)((true_study_size >> 24) & 255); |
| 1664 |
sbuf[5] = (true_study_size >> 16) & 255; |
sbuf[5] = (uschar)((true_study_size >> 16) & 255); |
| 1665 |
sbuf[6] = (true_study_size >> 8) & 255; |
sbuf[6] = (uschar)((true_study_size >> 8) & 255); |
| 1666 |
sbuf[7] = (true_study_size) & 255; |
sbuf[7] = (uschar)((true_study_size) & 255); |
| 1667 |
|
|
| 1668 |
if (fwrite(sbuf, 1, 8, f) < 8 || |
if (fwrite(sbuf, 1, 8, f) < 8 || |
| 1669 |
fwrite(re, 1, true_size, f) < true_size) |
fwrite(re, 1, true_size, f) < true_size) |
| 1682 |
strerror(errno)); |
strerror(errno)); |
| 1683 |
} |
} |
| 1684 |
else fprintf(outfile, "Study data written to %s\n", to_file); |
else fprintf(outfile, "Study data written to %s\n", to_file); |
| 1685 |
|
|
| 1686 |
} |
} |
| 1687 |
} |
} |
| 1688 |
fclose(f); |
fclose(f); |
| 1700 |
for (;;) |
for (;;) |
| 1701 |
{ |
{ |
| 1702 |
uschar *q; |
uschar *q; |
| 1703 |
uschar *bptr = dbuffer; |
uschar *bptr; |
| 1704 |
int *use_offsets = offsets; |
int *use_offsets = offsets; |
| 1705 |
int use_size_offsets = size_offsets; |
int use_size_offsets = size_offsets; |
| 1706 |
int callout_data = 0; |
int callout_data = 0; |
| 1737 |
len = 0; |
len = 0; |
| 1738 |
for (;;) |
for (;;) |
| 1739 |
{ |
{ |
| 1740 |
if (infile == stdin) printf("data> "); |
if (extend_inputline(infile, buffer + len, "data> ") == NULL) |
|
if (extend_inputline(infile, buffer + len) == NULL) |
|
| 1741 |
{ |
{ |
| 1742 |
if (len > 0) break; |
if (len > 0) break; |
| 1743 |
done = 1; |
done = 1; |
| 1755 |
p = buffer; |
p = buffer; |
| 1756 |
while (isspace(*p)) p++; |
while (isspace(*p)) p++; |
| 1757 |
|
|
| 1758 |
q = dbuffer; |
bptr = q = dbuffer; |
| 1759 |
while ((c = *p++) != 0) |
while ((c = *p++) != 0) |
| 1760 |
{ |
{ |
| 1761 |
int i = 0; |
int i = 0; |
| 1950 |
if (offsets == NULL) |
if (offsets == NULL) |
| 1951 |
{ |
{ |
| 1952 |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
| 1953 |
size_offsets_max * sizeof(int)); |
(int)(size_offsets_max * sizeof(int))); |
| 1954 |
yield = 1; |
yield = 1; |
| 1955 |
goto EXIT; |
goto EXIT; |
| 1956 |
} |
} |
| 2080 |
|
|
| 2081 |
for (;; gmatched++) /* Loop for /g or /G */ |
for (;; gmatched++) /* Loop for /g or /G */ |
| 2082 |
{ |
{ |
| 2083 |
if (timeit) |
if (timeitm > 0) |
| 2084 |
{ |
{ |
| 2085 |
register int i; |
register int i; |
| 2086 |
clock_t time_taken; |
clock_t time_taken; |
| 2090 |
if (all_use_dfa || use_dfa) |
if (all_use_dfa || use_dfa) |
| 2091 |
{ |
{ |
| 2092 |
int workspace[1000]; |
int workspace[1000]; |
| 2093 |
for (i = 0; i < LOOPREPEAT; i++) |
for (i = 0; i < timeitm; i++) |
| 2094 |
count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset, |
count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset, |
| 2095 |
options | g_notempty, use_offsets, use_size_offsets, workspace, |
options | g_notempty, use_offsets, use_size_offsets, workspace, |
| 2096 |
sizeof(workspace)/sizeof(int)); |
sizeof(workspace)/sizeof(int)); |
| 2098 |
else |
else |
| 2099 |
#endif |
#endif |
| 2100 |
|
|
| 2101 |
for (i = 0; i < LOOPREPEAT; i++) |
for (i = 0; i < timeitm; i++) |
| 2102 |
count = pcre_exec(re, extra, (char *)bptr, len, |
count = pcre_exec(re, extra, (char *)bptr, len, |
| 2103 |
start_offset, options | g_notempty, use_offsets, use_size_offsets); |
start_offset, options | g_notempty, use_offsets, use_size_offsets); |
| 2104 |
|
|
| 2105 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
| 2106 |
fprintf(outfile, "Execute time %.3f milliseconds\n", |
fprintf(outfile, "Execute time %.4f milliseconds\n", |
| 2107 |
(((double)time_taken * 1000.0) / (double)LOOPREPEAT) / |
(((double)time_taken * 1000.0) / (double)timeitm) / |
| 2108 |
(double)CLOCKS_PER_SEC); |
(double)CLOCKS_PER_SEC); |
| 2109 |
} |
} |
| 2110 |
|
|
| 2180 |
|
|
| 2181 |
if (count >= 0) |
if (count >= 0) |
| 2182 |
{ |
{ |
| 2183 |
int i; |
int i, maxcount; |
| 2184 |
|
|
| 2185 |
|
#if !defined NODFA |
| 2186 |
|
if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else |
| 2187 |
|
#endif |
| 2188 |
|
maxcount = use_size_offsets/3; |
| 2189 |
|
|
| 2190 |
|
/* This is a check against a lunatic return value. */ |
| 2191 |
|
|
| 2192 |
|
if (count > maxcount) |
| 2193 |
|
{ |
| 2194 |
|
fprintf(outfile, |
| 2195 |
|
"** PCRE error: returned count %d is too big for offset size %d\n", |
| 2196 |
|
count, use_size_offsets); |
| 2197 |
|
count = use_size_offsets/3; |
| 2198 |
|
if (do_g || do_G) |
| 2199 |
|
{ |
| 2200 |
|
fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G'); |
| 2201 |
|
do_g = do_G = FALSE; /* Break g/G loop */ |
| 2202 |
|
} |
| 2203 |
|
} |
| 2204 |
|
|
| 2205 |
for (i = 0; i < count * 2; i += 2) |
for (i = 0; i < count * 2; i += 2) |
| 2206 |
{ |
{ |
| 2207 |
if (use_offsets[i] < 0) |
if (use_offsets[i] < 0) |
| 2319 |
} |
} |
| 2320 |
|
|
| 2321 |
/* 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 |
| 2322 |
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 |
| 2323 |
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 |
| 2324 |
matching, the advance must be one character, not one byte. Fudge the |
string - that was checked before setting g_notempty. |
| 2325 |
offset values to achieve this. We won't be at the end of the string - |
|
| 2326 |
that was checked before setting g_notempty. */ |
Complication arises in the case when the newline option is "any" or |
| 2327 |
|
"anycrlf". If the previous match was at the end of a line terminated by |
| 2328 |
|
CRLF, an advance of one character just passes the \r, whereas we should |
| 2329 |
|
prefer the longer newline sequence, as does the code in pcre_exec(). |
| 2330 |
|
Fudge the offset value to achieve this. |
| 2331 |
|
|
| 2332 |
|
Otherwise, in the case of UTF-8 matching, the advance must be one |
| 2333 |
|
character, not one byte. */ |
| 2334 |
|
|
| 2335 |
else |
else |
| 2336 |
{ |
{ |
| 2337 |
if (g_notempty != 0) |
if (g_notempty != 0) |
| 2338 |
{ |
{ |
| 2339 |
int onechar = 1; |
int onechar = 1; |
| 2340 |
|
unsigned int obits = ((real_pcre *)re)->options; |
| 2341 |
use_offsets[0] = start_offset; |
use_offsets[0] = start_offset; |
| 2342 |
if (use_utf8) |
if ((obits & PCRE_NEWLINE_BITS) == 0) |
| 2343 |
|
{ |
| 2344 |
|
int d; |
| 2345 |
|
(void)pcre_config(PCRE_CONFIG_NEWLINE, &d); |
| 2346 |
|
obits = (d == '\r')? PCRE_NEWLINE_CR : |
| 2347 |
|
(d == '\n')? PCRE_NEWLINE_LF : |
| 2348 |
|
(d == ('\r'<<8 | '\n'))? PCRE_NEWLINE_CRLF : |
| 2349 |
|
(d == -2)? PCRE_NEWLINE_ANYCRLF : |
| 2350 |
|
(d == -1)? PCRE_NEWLINE_ANY : 0; |
| 2351 |
|
} |
| 2352 |
|
if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY || |
| 2353 |
|
(obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF) |
| 2354 |
|
&& |
| 2355 |
|
start_offset < len - 1 && |
| 2356 |
|
bptr[start_offset] == '\r' && |
| 2357 |
|
bptr[start_offset+1] == '\n') |
| 2358 |
|
onechar++; |
| 2359 |
|
else if (use_utf8) |
| 2360 |
{ |
{ |
| 2361 |
while (start_offset + onechar < len) |
while (start_offset + onechar < len) |
| 2362 |
{ |
{ |
| 2391 |
character. */ |
character. */ |
| 2392 |
|
|
| 2393 |
g_notempty = 0; |
g_notempty = 0; |
| 2394 |
|
|
| 2395 |
if (use_offsets[0] == use_offsets[1]) |
if (use_offsets[0] == use_offsets[1]) |
| 2396 |
{ |
{ |
| 2397 |
if (use_offsets[0] == len) break; |
if (use_offsets[0] == len) break; |
| 2426 |
{ |
{ |
| 2427 |
new_free((void *)tables); |
new_free((void *)tables); |
| 2428 |
setlocale(LC_CTYPE, "C"); |
setlocale(LC_CTYPE, "C"); |
| 2429 |
|
locale_set = 0; |
| 2430 |
} |
} |
| 2431 |
} |
} |
| 2432 |
|
|