| 6 |
its pattern matching. On a Unix or Win32 system it can recurse into |
its pattern matching. On a Unix or Win32 system it can recurse into |
| 7 |
directories. |
directories. |
| 8 |
|
|
| 9 |
Copyright (c) 1997-2007 University of Cambridge |
Copyright (c) 1997-2009 University of Cambridge |
| 10 |
|
|
| 11 |
----------------------------------------------------------------------------- |
----------------------------------------------------------------------------- |
| 12 |
Redistribution and use in source and binary forms, with or without |
Redistribution and use in source and binary forms, with or without |
| 38 |
*/ |
*/ |
| 39 |
|
|
| 40 |
#ifdef HAVE_CONFIG_H |
#ifdef HAVE_CONFIG_H |
| 41 |
# include <config.h> |
#include "config.h" |
| 42 |
#endif |
#endif |
| 43 |
|
|
| 44 |
#include <ctype.h> |
#include <ctype.h> |
| 50 |
|
|
| 51 |
#include <sys/types.h> |
#include <sys/types.h> |
| 52 |
#include <sys/stat.h> |
#include <sys/stat.h> |
| 53 |
|
|
| 54 |
#ifdef HAVE_UNISTD_H |
#ifdef HAVE_UNISTD_H |
| 55 |
# include <unistd.h> |
#include <unistd.h> |
| 56 |
|
#endif |
| 57 |
|
|
| 58 |
|
#ifdef SUPPORT_LIBZ |
| 59 |
|
#include <zlib.h> |
| 60 |
|
#endif |
| 61 |
|
|
| 62 |
|
#ifdef SUPPORT_LIBBZ2 |
| 63 |
|
#include <bzlib.h> |
| 64 |
#endif |
#endif |
| 65 |
|
|
| 66 |
#include <pcre.h> |
#include "pcre.h" |
| 67 |
|
|
| 68 |
#define FALSE 0 |
#define FALSE 0 |
| 69 |
#define TRUE 1 |
#define TRUE 1 |
| 71 |
typedef int BOOL; |
typedef int BOOL; |
| 72 |
|
|
| 73 |
#define MAX_PATTERN_COUNT 100 |
#define MAX_PATTERN_COUNT 100 |
| 74 |
|
#define OFFSET_SIZE 99 |
| 75 |
|
|
| 76 |
#if BUFSIZ > 8192 |
#if BUFSIZ > 8192 |
| 77 |
#define MBUFTHIRD BUFSIZ |
#define MBUFTHIRD BUFSIZ |
| 83 |
output. The order is important; it is assumed that a file name is wanted for |
output. The order is important; it is assumed that a file name is wanted for |
| 84 |
all values greater than FN_DEFAULT. */ |
all values greater than FN_DEFAULT. */ |
| 85 |
|
|
| 86 |
enum { FN_NONE, FN_DEFAULT, FN_ONLY, FN_NOMATCH_ONLY, FN_FORCE }; |
enum { FN_NONE, FN_DEFAULT, FN_MATCH_ONLY, FN_NOMATCH_ONLY, FN_FORCE }; |
| 87 |
|
|
| 88 |
|
/* File reading styles */ |
| 89 |
|
|
| 90 |
|
enum { FR_PLAIN, FR_LIBZ, FR_LIBBZ2 }; |
| 91 |
|
|
| 92 |
/* Actions for the -d and -D options */ |
/* Actions for the -d and -D options */ |
| 93 |
|
|
| 102 |
|
|
| 103 |
/* Line ending types */ |
/* Line ending types */ |
| 104 |
|
|
| 105 |
enum { EL_LF, EL_CR, EL_CRLF, EL_ANY }; |
enum { EL_LF, EL_CR, EL_CRLF, EL_ANY, EL_ANYCRLF }; |
| 106 |
|
|
| 107 |
|
|
| 108 |
|
|
| 140 |
|
|
| 141 |
static char *include_pattern = NULL; |
static char *include_pattern = NULL; |
| 142 |
static char *exclude_pattern = NULL; |
static char *exclude_pattern = NULL; |
| 143 |
|
static char *include_dir_pattern = NULL; |
| 144 |
|
static char *exclude_dir_pattern = NULL; |
| 145 |
|
|
| 146 |
static pcre *include_compiled = NULL; |
static pcre *include_compiled = NULL; |
| 147 |
static pcre *exclude_compiled = NULL; |
static pcre *exclude_compiled = NULL; |
| 148 |
|
static pcre *include_dir_compiled = NULL; |
| 149 |
|
static pcre *exclude_dir_compiled = NULL; |
| 150 |
|
|
| 151 |
static int after_context = 0; |
static int after_context = 0; |
| 152 |
static int before_context = 0; |
static int before_context = 0; |
| 159 |
|
|
| 160 |
static BOOL count_only = FALSE; |
static BOOL count_only = FALSE; |
| 161 |
static BOOL do_colour = FALSE; |
static BOOL do_colour = FALSE; |
| 162 |
|
static BOOL file_offsets = FALSE; |
| 163 |
static BOOL hyphenpending = FALSE; |
static BOOL hyphenpending = FALSE; |
| 164 |
static BOOL invert = FALSE; |
static BOOL invert = FALSE; |
| 165 |
|
static BOOL line_offsets = FALSE; |
| 166 |
static BOOL multiline = FALSE; |
static BOOL multiline = FALSE; |
| 167 |
static BOOL number = FALSE; |
static BOOL number = FALSE; |
| 168 |
|
static BOOL omit_zero_count = FALSE; |
| 169 |
static BOOL only_matching = FALSE; |
static BOOL only_matching = FALSE; |
| 170 |
static BOOL quiet = FALSE; |
static BOOL quiet = FALSE; |
| 171 |
static BOOL silent = FALSE; |
static BOOL silent = FALSE; |
| 187 |
/* Options without a single-letter equivalent get a negative value. This can be |
/* Options without a single-letter equivalent get a negative value. This can be |
| 188 |
used to identify them. */ |
used to identify them. */ |
| 189 |
|
|
| 190 |
#define N_COLOUR (-1) |
#define N_COLOUR (-1) |
| 191 |
#define N_EXCLUDE (-2) |
#define N_EXCLUDE (-2) |
| 192 |
#define N_HELP (-3) |
#define N_EXCLUDE_DIR (-3) |
| 193 |
#define N_INCLUDE (-4) |
#define N_HELP (-4) |
| 194 |
#define N_LABEL (-5) |
#define N_INCLUDE (-5) |
| 195 |
#define N_LOCALE (-6) |
#define N_INCLUDE_DIR (-6) |
| 196 |
#define N_NULL (-7) |
#define N_LABEL (-7) |
| 197 |
|
#define N_LOCALE (-8) |
| 198 |
|
#define N_NULL (-9) |
| 199 |
|
#define N_LOFFSETS (-10) |
| 200 |
|
#define N_FOFFSETS (-11) |
| 201 |
|
|
| 202 |
static option_item optionlist[] = { |
static option_item optionlist[] = { |
| 203 |
{ OP_NODATA, N_NULL, NULL, "", " terminate options" }, |
{ OP_NODATA, N_NULL, NULL, "", " terminate options" }, |
| 213 |
{ OP_PATLIST, 'e', NULL, "regex(p)", "specify pattern (may be used more than once)" }, |
{ OP_PATLIST, 'e', NULL, "regex(p)", "specify pattern (may be used more than once)" }, |
| 214 |
{ OP_NODATA, 'F', NULL, "fixed_strings", "patterns are sets of newline-separated strings" }, |
{ OP_NODATA, 'F', NULL, "fixed_strings", "patterns are sets of newline-separated strings" }, |
| 215 |
{ OP_STRING, 'f', &pattern_filename, "file=path", "read patterns from file" }, |
{ OP_STRING, 'f', &pattern_filename, "file=path", "read patterns from file" }, |
| 216 |
|
{ OP_NODATA, N_FOFFSETS, NULL, "file-offsets", "output file offsets, not text" }, |
| 217 |
{ OP_NODATA, 'H', NULL, "with-filename", "force the prefixing filename on output" }, |
{ OP_NODATA, 'H', NULL, "with-filename", "force the prefixing filename on output" }, |
| 218 |
{ OP_NODATA, 'h', NULL, "no-filename", "suppress the prefixing filename on output" }, |
{ OP_NODATA, 'h', NULL, "no-filename", "suppress the prefixing filename on output" }, |
| 219 |
{ OP_NODATA, 'i', NULL, "ignore-case", "ignore case distinctions" }, |
{ OP_NODATA, 'i', NULL, "ignore-case", "ignore case distinctions" }, |
| 220 |
{ OP_NODATA, 'l', NULL, "files-with-matches", "print only FILE names containing matches" }, |
{ OP_NODATA, 'l', NULL, "files-with-matches", "print only FILE names containing matches" }, |
| 221 |
{ OP_NODATA, 'L', NULL, "files-without-match","print only FILE names not containing matches" }, |
{ OP_NODATA, 'L', NULL, "files-without-match","print only FILE names not containing matches" }, |
| 222 |
{ OP_STRING, N_LABEL, &stdin_name, "label=name", "set name for standard input" }, |
{ OP_STRING, N_LABEL, &stdin_name, "label=name", "set name for standard input" }, |
| 223 |
|
{ OP_NODATA, N_LOFFSETS, NULL, "line-offsets", "output line numbers and offsets, not text" }, |
| 224 |
{ OP_STRING, N_LOCALE, &locale, "locale=locale", "use the named locale" }, |
{ OP_STRING, N_LOCALE, &locale, "locale=locale", "use the named locale" }, |
| 225 |
{ OP_NODATA, 'M', NULL, "multiline", "run in multiline mode" }, |
{ OP_NODATA, 'M', NULL, "multiline", "run in multiline mode" }, |
| 226 |
{ OP_STRING, 'N', &newline, "newline=type", "specify newline type (CR, LR, CRLF)" }, |
{ OP_STRING, 'N', &newline, "newline=type", "set newline type (CR, LF, CRLF, ANYCRLF or ANY)" }, |
| 227 |
{ OP_NODATA, 'n', NULL, "line-number", "print line number with output lines" }, |
{ OP_NODATA, 'n', NULL, "line-number", "print line number with output lines" }, |
| 228 |
{ OP_NODATA, 'o', NULL, "only-matching", "show only the part of the line that matched" }, |
{ OP_NODATA, 'o', NULL, "only-matching", "show only the part of the line that matched" }, |
| 229 |
{ OP_NODATA, 'q', NULL, "quiet", "suppress output, just set return code" }, |
{ OP_NODATA, 'q', NULL, "quiet", "suppress output, just set return code" }, |
| 230 |
{ OP_NODATA, 'r', NULL, "recursive", "recursively scan sub-directories" }, |
{ OP_NODATA, 'r', NULL, "recursive", "recursively scan sub-directories" }, |
| 231 |
{ OP_STRING, N_EXCLUDE,&exclude_pattern, "exclude=pattern","exclude matching files when recursing" }, |
{ OP_STRING, N_EXCLUDE,&exclude_pattern, "exclude=pattern","exclude matching files when recursing" }, |
| 232 |
{ OP_STRING, N_INCLUDE,&include_pattern, "include=pattern","include matching files when recursing" }, |
{ OP_STRING, N_INCLUDE,&include_pattern, "include=pattern","include matching files when recursing" }, |
| 233 |
|
{ OP_STRING, N_EXCLUDE_DIR,&exclude_dir_pattern, "exclude_dir=pattern","exclude matching directories when recursing" }, |
| 234 |
|
{ OP_STRING, N_INCLUDE_DIR,&include_dir_pattern, "include_dir=pattern","include matching directories when recursing" }, |
| 235 |
#ifdef JFRIEDL_DEBUG |
#ifdef JFRIEDL_DEBUG |
| 236 |
{ OP_OP_NUMBER, 'S', &S_arg, "jeffS", "replace matched (sub)string with X" }, |
{ OP_OP_NUMBER, 'S', &S_arg, "jeffS", "replace matched (sub)string with X" }, |
| 237 |
#endif |
#endif |
| 255 |
static const char *suffix[] = { |
static const char *suffix[] = { |
| 256 |
"", "\\b", ")$", ")$", "\\E", "\\E\\b", "\\E)$", "\\E)$" }; |
"", "\\b", ")$", ")$", "\\E", "\\E\\b", "\\E)$", "\\E)$" }; |
| 257 |
|
|
| 258 |
/* UTF-8 tables - used only when the newline setting is "all". */ |
/* UTF-8 tables - used only when the newline setting is "any". */ |
| 259 |
|
|
| 260 |
const int utf8_table3[] = { 0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01}; |
const int utf8_table3[] = { 0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01}; |
| 261 |
|
|
| 309 |
if (strcmp(dent->d_name, ".") != 0 && strcmp(dent->d_name, "..") != 0) |
if (strcmp(dent->d_name, ".") != 0 && strcmp(dent->d_name, "..") != 0) |
| 310 |
return dent->d_name; |
return dent->d_name; |
| 311 |
} |
} |
| 312 |
return NULL; /* Keep compiler happy; never executed */ |
/* Control never reaches here */ |
| 313 |
} |
} |
| 314 |
|
|
| 315 |
static void |
static void |
| 344 |
|
|
| 345 |
/* I (Philip Hazel) have no means of testing this code. It was contributed by |
/* I (Philip Hazel) have no means of testing this code. It was contributed by |
| 346 |
Lionel Fourquaux. David Burgess added a patch to define INVALID_FILE_ATTRIBUTES |
Lionel Fourquaux. David Burgess added a patch to define INVALID_FILE_ATTRIBUTES |
| 347 |
when it did not exist. */ |
when it did not exist. David Byron added a patch that moved the #include of |
| 348 |
|
<windows.h> to before the INVALID_FILE_ATTRIBUTES definition rather than after. |
| 349 |
|
*/ |
| 350 |
|
|
| 351 |
#elif HAVE_WINDOWS_H |
#elif HAVE_WINDOWS_H |
| 352 |
|
|
| 356 |
#ifndef WIN32_LEAN_AND_MEAN |
#ifndef WIN32_LEAN_AND_MEAN |
| 357 |
# define WIN32_LEAN_AND_MEAN |
# define WIN32_LEAN_AND_MEAN |
| 358 |
#endif |
#endif |
| 359 |
|
|
| 360 |
|
#include <windows.h> |
| 361 |
|
|
| 362 |
#ifndef INVALID_FILE_ATTRIBUTES |
#ifndef INVALID_FILE_ATTRIBUTES |
| 363 |
#define INVALID_FILE_ATTRIBUTES 0xFFFFFFFF |
#define INVALID_FILE_ATTRIBUTES 0xFFFFFFFF |
| 364 |
#endif |
#endif |
| 365 |
|
|
|
#include <windows.h> |
|
|
|
|
| 366 |
typedef struct directory_type |
typedef struct directory_type |
| 367 |
{ |
{ |
| 368 |
HANDLE handle; |
HANDLE handle; |
| 447 |
|
|
| 448 |
int isregfile(char *filename) |
int isregfile(char *filename) |
| 449 |
{ |
{ |
| 450 |
return !isdirectory(filename) |
return !isdirectory(filename); |
| 451 |
} |
} |
| 452 |
|
|
| 453 |
|
|
| 458 |
static BOOL |
static BOOL |
| 459 |
is_stdout_tty(void) |
is_stdout_tty(void) |
| 460 |
{ |
{ |
| 461 |
FALSE; |
return FALSE; |
| 462 |
} |
} |
| 463 |
|
|
| 464 |
|
|
| 576 |
} |
} |
| 577 |
break; |
break; |
| 578 |
|
|
| 579 |
|
case EL_ANYCRLF: |
| 580 |
|
while (p < endptr) |
| 581 |
|
{ |
| 582 |
|
int extra = 0; |
| 583 |
|
register int c = *((unsigned char *)p); |
| 584 |
|
|
| 585 |
|
if (utf8 && c >= 0xc0) |
| 586 |
|
{ |
| 587 |
|
int gcii, gcss; |
| 588 |
|
extra = utf8_table4[c & 0x3f]; /* Number of additional bytes */ |
| 589 |
|
gcss = 6*extra; |
| 590 |
|
c = (c & utf8_table3[extra]) << gcss; |
| 591 |
|
for (gcii = 1; gcii <= extra; gcii++) |
| 592 |
|
{ |
| 593 |
|
gcss -= 6; |
| 594 |
|
c |= (p[gcii] & 0x3f) << gcss; |
| 595 |
|
} |
| 596 |
|
} |
| 597 |
|
|
| 598 |
|
p += 1 + extra; |
| 599 |
|
|
| 600 |
|
switch (c) |
| 601 |
|
{ |
| 602 |
|
case 0x0a: /* LF */ |
| 603 |
|
*lenptr = 1; |
| 604 |
|
return p; |
| 605 |
|
|
| 606 |
|
case 0x0d: /* CR */ |
| 607 |
|
if (p < endptr && *p == 0x0a) |
| 608 |
|
{ |
| 609 |
|
*lenptr = 2; |
| 610 |
|
p++; |
| 611 |
|
} |
| 612 |
|
else *lenptr = 1; |
| 613 |
|
return p; |
| 614 |
|
|
| 615 |
|
default: |
| 616 |
|
break; |
| 617 |
|
} |
| 618 |
|
} /* End of loop for ANYCRLF case */ |
| 619 |
|
|
| 620 |
|
*lenptr = 0; /* Must have hit the end */ |
| 621 |
|
return endptr; |
| 622 |
|
|
| 623 |
case EL_ANY: |
case EL_ANY: |
| 624 |
while (p < endptr) |
while (p < endptr) |
| 625 |
{ |
{ |
| 718 |
return p; /* But control should never get here */ |
return p; /* But control should never get here */ |
| 719 |
|
|
| 720 |
case EL_ANY: |
case EL_ANY: |
| 721 |
|
case EL_ANYCRLF: |
| 722 |
if (*(--p) == '\n' && p > startptr && p[-1] == '\r') p--; |
if (*(--p) == '\n' && p > startptr && p[-1] == '\r') p--; |
| 723 |
if (utf8) while ((*p & 0xc0) == 0x80) p--; |
if (utf8) while ((*p & 0xc0) == 0x80) p--; |
| 724 |
|
|
| 747 |
} |
} |
| 748 |
else c = *((unsigned char *)pp); |
else c = *((unsigned char *)pp); |
| 749 |
|
|
| 750 |
switch (c) |
if (endlinetype == EL_ANYCRLF) switch (c) |
| 751 |
|
{ |
| 752 |
|
case 0x0a: /* LF */ |
| 753 |
|
case 0x0d: /* CR */ |
| 754 |
|
return p; |
| 755 |
|
|
| 756 |
|
default: |
| 757 |
|
break; |
| 758 |
|
} |
| 759 |
|
|
| 760 |
|
else switch (c) |
| 761 |
{ |
{ |
| 762 |
case 0x0a: /* LF */ |
case 0x0a: /* LF */ |
| 763 |
case 0x0b: /* VT */ |
case 0x0b: /* VT */ |
| 823 |
|
|
| 824 |
|
|
| 825 |
/************************************************* |
/************************************************* |
| 826 |
|
* Apply patterns to subject till one matches * |
| 827 |
|
*************************************************/ |
| 828 |
|
|
| 829 |
|
/* This function is called to run through all patterns, looking for a match. It |
| 830 |
|
is used multiple times for the same subject when colouring is enabled, in order |
| 831 |
|
to find all possible matches. |
| 832 |
|
|
| 833 |
|
Arguments: |
| 834 |
|
matchptr the start of the subject |
| 835 |
|
length the length of the subject to match |
| 836 |
|
offsets the offets vector to fill in |
| 837 |
|
mrc address of where to put the result of pcre_exec() |
| 838 |
|
|
| 839 |
|
Returns: TRUE if there was a match |
| 840 |
|
FALSE if there was no match |
| 841 |
|
invert if there was a non-fatal error |
| 842 |
|
*/ |
| 843 |
|
|
| 844 |
|
static BOOL |
| 845 |
|
match_patterns(char *matchptr, size_t length, int *offsets, int *mrc) |
| 846 |
|
{ |
| 847 |
|
int i; |
| 848 |
|
for (i = 0; i < pattern_count; i++) |
| 849 |
|
{ |
| 850 |
|
*mrc = pcre_exec(pattern_list[i], hints_list[i], matchptr, length, 0, |
| 851 |
|
PCRE_NOTEMPTY, offsets, OFFSET_SIZE); |
| 852 |
|
if (*mrc >= 0) return TRUE; |
| 853 |
|
if (*mrc == PCRE_ERROR_NOMATCH) continue; |
| 854 |
|
fprintf(stderr, "pcregrep: pcre_exec() error %d while matching ", *mrc); |
| 855 |
|
if (pattern_count > 1) fprintf(stderr, "pattern number %d to ", i+1); |
| 856 |
|
fprintf(stderr, "this text:\n"); |
| 857 |
|
fwrite(matchptr, 1, length, stderr); /* In case binary zero included */ |
| 858 |
|
fprintf(stderr, "\n"); |
| 859 |
|
if (error_count == 0 && |
| 860 |
|
(*mrc == PCRE_ERROR_MATCHLIMIT || *mrc == PCRE_ERROR_RECURSIONLIMIT)) |
| 861 |
|
{ |
| 862 |
|
fprintf(stderr, "pcregrep: error %d means that a resource limit " |
| 863 |
|
"was exceeded\n", *mrc); |
| 864 |
|
fprintf(stderr, "pcregrep: check your regex for nested unlimited loops\n"); |
| 865 |
|
} |
| 866 |
|
if (error_count++ > 20) |
| 867 |
|
{ |
| 868 |
|
fprintf(stderr, "pcregrep: too many errors - abandoned\n"); |
| 869 |
|
exit(2); |
| 870 |
|
} |
| 871 |
|
return invert; /* No more matching; don't show the line again */ |
| 872 |
|
} |
| 873 |
|
|
| 874 |
|
return FALSE; /* No match, no errors */ |
| 875 |
|
} |
| 876 |
|
|
| 877 |
|
|
| 878 |
|
|
| 879 |
|
/************************************************* |
| 880 |
* Grep an individual file * |
* Grep an individual file * |
| 881 |
*************************************************/ |
*************************************************/ |
| 882 |
|
|
| 888 |
"before" context printing. |
"before" context printing. |
| 889 |
|
|
| 890 |
Arguments: |
Arguments: |
| 891 |
in the fopened FILE stream |
handle the fopened FILE stream for a normal file |
| 892 |
|
the gzFile pointer when reading is via libz |
| 893 |
|
the BZFILE pointer when reading is via libbz2 |
| 894 |
|
frtype FR_PLAIN, FR_LIBZ, or FR_LIBBZ2 |
| 895 |
printname the file name if it is to be printed for each match |
printname the file name if it is to be printed for each match |
| 896 |
or NULL if the file name is not to be printed |
or NULL if the file name is not to be printed |
| 897 |
it cannot be NULL if filenames[_nomatch]_only is set |
it cannot be NULL if filenames[_nomatch]_only is set |
| 898 |
|
|
| 899 |
Returns: 0 if there was at least one match |
Returns: 0 if there was at least one match |
| 900 |
1 otherwise (no matches) |
1 otherwise (no matches) |
| 901 |
|
2 if there is a read error on a .bz2 file |
| 902 |
*/ |
*/ |
| 903 |
|
|
| 904 |
static int |
static int |
| 905 |
pcregrep(FILE *in, char *printname) |
pcregrep(void *handle, int frtype, char *printname) |
| 906 |
{ |
{ |
| 907 |
int rc = 1; |
int rc = 1; |
| 908 |
int linenumber = 1; |
int linenumber = 1; |
| 909 |
int lastmatchnumber = 0; |
int lastmatchnumber = 0; |
| 910 |
int count = 0; |
int count = 0; |
| 911 |
int offsets[99]; |
int filepos = 0; |
| 912 |
|
int offsets[OFFSET_SIZE]; |
| 913 |
char *lastmatchrestart = NULL; |
char *lastmatchrestart = NULL; |
| 914 |
char buffer[3*MBUFTHIRD]; |
char buffer[3*MBUFTHIRD]; |
| 915 |
char *ptr = buffer; |
char *ptr = buffer; |
| 916 |
char *endptr; |
char *endptr; |
| 917 |
size_t bufflength; |
size_t bufflength; |
| 918 |
BOOL endhyphenpending = FALSE; |
BOOL endhyphenpending = FALSE; |
| 919 |
|
FILE *in = NULL; /* Ensure initialized */ |
| 920 |
|
|
| 921 |
|
#ifdef SUPPORT_LIBZ |
| 922 |
|
gzFile ingz = NULL; |
| 923 |
|
#endif |
| 924 |
|
|
| 925 |
|
#ifdef SUPPORT_LIBBZ2 |
| 926 |
|
BZFILE *inbz2 = NULL; |
| 927 |
|
#endif |
| 928 |
|
|
|
/* Do the first read into the start of the buffer and set up the pointer to |
|
|
end of what we have. */ |
|
| 929 |
|
|
| 930 |
bufflength = fread(buffer, 1, 3*MBUFTHIRD, in); |
/* Do the first read into the start of the buffer and set up the pointer to end |
| 931 |
|
of what we have. In the case of libz, a non-zipped .gz file will be read as a |
| 932 |
|
plain file. However, if a .bz2 file isn't actually bzipped, the first read will |
| 933 |
|
fail. */ |
| 934 |
|
|
| 935 |
|
#ifdef SUPPORT_LIBZ |
| 936 |
|
if (frtype == FR_LIBZ) |
| 937 |
|
{ |
| 938 |
|
ingz = (gzFile)handle; |
| 939 |
|
bufflength = gzread (ingz, buffer, 3*MBUFTHIRD); |
| 940 |
|
} |
| 941 |
|
else |
| 942 |
|
#endif |
| 943 |
|
|
| 944 |
|
#ifdef SUPPORT_LIBBZ2 |
| 945 |
|
if (frtype == FR_LIBBZ2) |
| 946 |
|
{ |
| 947 |
|
inbz2 = (BZFILE *)handle; |
| 948 |
|
bufflength = BZ2_bzread(inbz2, buffer, 3*MBUFTHIRD); |
| 949 |
|
if ((int)bufflength < 0) return 2; /* Gotcha: bufflength is size_t; */ |
| 950 |
|
} /* without the cast it is unsigned. */ |
| 951 |
|
else |
| 952 |
|
#endif |
| 953 |
|
|
| 954 |
|
{ |
| 955 |
|
in = (FILE *)handle; |
| 956 |
|
bufflength = fread(buffer, 1, 3*MBUFTHIRD, in); |
| 957 |
|
} |
| 958 |
|
|
| 959 |
endptr = buffer + bufflength; |
endptr = buffer + bufflength; |
| 960 |
|
|
| 961 |
/* Loop while the current pointer is not at the end of the file. For large |
/* Loop while the current pointer is not at the end of the file. For large |
| 965 |
|
|
| 966 |
while (ptr < endptr) |
while (ptr < endptr) |
| 967 |
{ |
{ |
| 968 |
int i, endlinelength; |
int endlinelength; |
| 969 |
int mrc = 0; |
int mrc = 0; |
| 970 |
BOOL match = FALSE; |
BOOL match; |
| 971 |
|
char *matchptr = ptr; |
| 972 |
char *t = ptr; |
char *t = ptr; |
| 973 |
size_t length, linelength; |
size_t length, linelength; |
| 974 |
|
|
| 975 |
/* At this point, ptr is at the start of a line. We need to find the length |
/* At this point, ptr is at the start of a line. We need to find the length |
| 976 |
of the subject string to pass to pcre_exec(). In multiline mode, it is the |
of the subject string to pass to pcre_exec(). In multiline mode, it is the |
| 977 |
length remainder of the data in the buffer. Otherwise, it is the length of |
length remainder of the data in the buffer. Otherwise, it is the length of |
| 978 |
the next line. After matching, we always advance by the length of the next |
the next line, excluding the terminating newline. After matching, we always |
| 979 |
line. In multiline mode the PCRE_FIRSTLINE option is used for compiling, so |
advance by the length of the next line. In multiline mode the PCRE_FIRSTLINE |
| 980 |
that any match is constrained to be in the first line. */ |
option is used for compiling, so that any match is constrained to be in the |
| 981 |
|
first line. */ |
| 982 |
|
|
| 983 |
t = end_of_line(t, endptr, &endlinelength); |
t = end_of_line(t, endptr, &endlinelength); |
| 984 |
linelength = t - ptr - endlinelength; |
linelength = t - ptr - endlinelength; |
| 985 |
length = multiline? endptr - ptr : linelength; |
length = multiline? (size_t)(endptr - ptr) : linelength; |
| 986 |
|
|
| 987 |
/* Extra processing for Jeffrey Friedl's debugging. */ |
/* Extra processing for Jeffrey Friedl's debugging. */ |
| 988 |
|
|
| 993 |
#include <time.h> |
#include <time.h> |
| 994 |
struct timeval start_time, end_time; |
struct timeval start_time, end_time; |
| 995 |
struct timezone dummy; |
struct timezone dummy; |
| 996 |
|
int i; |
| 997 |
|
|
| 998 |
if (jfriedl_XT) |
if (jfriedl_XT) |
| 999 |
{ |
{ |
| 1019 |
|
|
| 1020 |
|
|
| 1021 |
for (i = 0; i < jfriedl_XR; i++) |
for (i = 0; i < jfriedl_XR; i++) |
| 1022 |
match = (pcre_exec(pattern_list[0], hints_list[0], ptr, length, 0, 0, offsets, 99) >= 0); |
match = (pcre_exec(pattern_list[0], hints_list[0], ptr, length, 0, |
| 1023 |
|
PCRE_NOTEMPTY, offsets, OFFSET_SIZE) >= 0); |
| 1024 |
|
|
| 1025 |
if (gettimeofday(&end_time, &dummy) != 0) |
if (gettimeofday(&end_time, &dummy) != 0) |
| 1026 |
perror("bad gettimeofday"); |
perror("bad gettimeofday"); |
| 1034 |
} |
} |
| 1035 |
#endif |
#endif |
| 1036 |
|
|
| 1037 |
|
/* We come back here after a match when the -o option (only_matching) is set, |
| 1038 |
|
in order to find any further matches in the same line. */ |
| 1039 |
|
|
| 1040 |
/* Run through all the patterns until one matches. Note that we don't include |
ONLY_MATCHING_RESTART: |
|
the final newline in the subject string. */ |
|
| 1041 |
|
|
| 1042 |
for (i = 0; i < pattern_count; i++) |
/* Run through all the patterns until one matches or there is an error other |
| 1043 |
{ |
than NOMATCH. This code is in a subroutine so that it can be re-used for |
| 1044 |
mrc = pcre_exec(pattern_list[i], hints_list[i], ptr, length, 0, 0, |
finding subsequent matches when colouring matched lines. */ |
| 1045 |
offsets, 99); |
|
| 1046 |
if (mrc >= 0) { match = TRUE; break; } |
match = match_patterns(matchptr, length, offsets, &mrc); |
|
if (mrc != PCRE_ERROR_NOMATCH) |
|
|
{ |
|
|
fprintf(stderr, "pcregrep: pcre_exec() error %d while matching ", mrc); |
|
|
if (pattern_count > 1) fprintf(stderr, "pattern number %d to ", i+1); |
|
|
fprintf(stderr, "this line:\n"); |
|
|
fwrite(ptr, 1, linelength, stderr); /* In case binary zero included */ |
|
|
fprintf(stderr, "\n"); |
|
|
if (error_count == 0 && |
|
|
(mrc == PCRE_ERROR_MATCHLIMIT || mrc == PCRE_ERROR_RECURSIONLIMIT)) |
|
|
{ |
|
|
fprintf(stderr, "pcregrep: error %d means that a resource limit " |
|
|
"was exceeded\n", mrc); |
|
|
fprintf(stderr, "pcregrep: check your regex for nested unlimited loops\n"); |
|
|
} |
|
|
if (error_count++ > 20) |
|
|
{ |
|
|
fprintf(stderr, "pcregrep: too many errors - abandoned\n"); |
|
|
exit(2); |
|
|
} |
|
|
match = invert; /* No more matching; don't show the line again */ |
|
|
break; |
|
|
} |
|
|
} |
|
| 1047 |
|
|
| 1048 |
/* If it's a match or a not-match (as required), do what's wanted. */ |
/* If it's a match or a not-match (as required), do what's wanted. */ |
| 1049 |
|
|
| 1062 |
/* If all we want is a file name, there is no need to scan any more lines |
/* If all we want is a file name, there is no need to scan any more lines |
| 1063 |
in the file. */ |
in the file. */ |
| 1064 |
|
|
| 1065 |
else if (filenames == FN_ONLY) |
else if (filenames == FN_MATCH_ONLY) |
| 1066 |
{ |
{ |
| 1067 |
fprintf(stdout, "%s\n", printname); |
fprintf(stdout, "%s\n", printname); |
| 1068 |
return 0; |
return 0; |
| 1073 |
else if (quiet) return 0; |
else if (quiet) return 0; |
| 1074 |
|
|
| 1075 |
/* The --only-matching option prints just the substring that matched, and |
/* The --only-matching option prints just the substring that matched, and |
| 1076 |
does not pring any context. */ |
the --file-offsets and --line-offsets options output offsets for the |
| 1077 |
|
matching substring (they both force --only-matching). None of these options |
| 1078 |
|
prints any context. Afterwards, adjust the start and length, and then jump |
| 1079 |
|
back to look for further matches in the same line. If we are in invert |
| 1080 |
|
mode, however, nothing is printed - this could be still useful because the |
| 1081 |
|
return code is set. */ |
| 1082 |
|
|
| 1083 |
else if (only_matching) |
else if (only_matching) |
| 1084 |
{ |
{ |
| 1085 |
if (printname != NULL) fprintf(stdout, "%s:", printname); |
if (!invert) |
| 1086 |
if (number) fprintf(stdout, "%d:", linenumber); |
{ |
| 1087 |
fwrite(ptr + offsets[0], 1, offsets[1] - offsets[0], stdout); |
if (printname != NULL) fprintf(stdout, "%s:", printname); |
| 1088 |
fprintf(stdout, "\n"); |
if (number) fprintf(stdout, "%d:", linenumber); |
| 1089 |
|
if (line_offsets) |
| 1090 |
|
fprintf(stdout, "%d,%d", (int)(matchptr + offsets[0] - ptr), |
| 1091 |
|
offsets[1] - offsets[0]); |
| 1092 |
|
else if (file_offsets) |
| 1093 |
|
fprintf(stdout, "%d,%d", (int)(filepos + matchptr + offsets[0] - ptr), |
| 1094 |
|
offsets[1] - offsets[0]); |
| 1095 |
|
else |
| 1096 |
|
{ |
| 1097 |
|
if (do_colour) fprintf(stdout, "%c[%sm", 0x1b, colour_string); |
| 1098 |
|
fwrite(matchptr + offsets[0], 1, offsets[1] - offsets[0], stdout); |
| 1099 |
|
if (do_colour) fprintf(stdout, "%c[00m", 0x1b); |
| 1100 |
|
} |
| 1101 |
|
fprintf(stdout, "\n"); |
| 1102 |
|
matchptr += offsets[1]; |
| 1103 |
|
length -= offsets[1]; |
| 1104 |
|
match = FALSE; |
| 1105 |
|
goto ONLY_MATCHING_RESTART; |
| 1106 |
|
} |
| 1107 |
} |
} |
| 1108 |
|
|
| 1109 |
/* This is the default case when none of the above options is set. We print |
/* This is the default case when none of the above options is set. We print |
| 1193 |
|
|
| 1194 |
/* In multiline mode, we want to print to the end of the line in which |
/* In multiline mode, we want to print to the end of the line in which |
| 1195 |
the end of the matched string is found, so we adjust linelength and the |
the end of the matched string is found, so we adjust linelength and the |
| 1196 |
line number appropriately. Because the PCRE_FIRSTLINE option is set, the |
line number appropriately, but only when there actually was a match |
| 1197 |
start of the match will always be before the first newline sequence. */ |
(invert not set). Because the PCRE_FIRSTLINE option is set, the start of |
| 1198 |
|
the match will always be before the first newline sequence. */ |
| 1199 |
|
|
| 1200 |
if (multiline) |
if (multiline) |
| 1201 |
{ |
{ |
| 1202 |
int ellength; |
int ellength; |
| 1203 |
char *endmatch = ptr + offsets[1]; |
char *endmatch = ptr; |
| 1204 |
t = ptr; |
if (!invert) |
|
while (t < endmatch) |
|
| 1205 |
{ |
{ |
| 1206 |
t = end_of_line(t, endptr, &ellength); |
endmatch += offsets[1]; |
| 1207 |
if (t <= endmatch) linenumber++; else break; |
t = ptr; |
| 1208 |
|
while (t < endmatch) |
| 1209 |
|
{ |
| 1210 |
|
t = end_of_line(t, endptr, &ellength); |
| 1211 |
|
if (t <= endmatch) linenumber++; else break; |
| 1212 |
|
} |
| 1213 |
} |
} |
| 1214 |
endmatch = end_of_line(endmatch, endptr, &ellength); |
endmatch = end_of_line(endmatch, endptr, &ellength); |
| 1215 |
linelength = endmatch - ptr - ellength; |
linelength = endmatch - ptr - ellength; |
| 1234 |
else |
else |
| 1235 |
#endif |
#endif |
| 1236 |
|
|
| 1237 |
/* We have to split the line(s) up if colouring. */ |
/* We have to split the line(s) up if colouring, and search for further |
| 1238 |
|
matches. */ |
| 1239 |
|
|
| 1240 |
if (do_colour) |
if (do_colour) |
| 1241 |
{ |
{ |
| 1242 |
|
int last_offset = 0; |
| 1243 |
fwrite(ptr, 1, offsets[0], stdout); |
fwrite(ptr, 1, offsets[0], stdout); |
| 1244 |
fprintf(stdout, "%c[%sm", 0x1b, colour_string); |
fprintf(stdout, "%c[%sm", 0x1b, colour_string); |
| 1245 |
fwrite(ptr + offsets[0], 1, offsets[1] - offsets[0], stdout); |
fwrite(ptr + offsets[0], 1, offsets[1] - offsets[0], stdout); |
| 1246 |
fprintf(stdout, "%c[00m", 0x1b); |
fprintf(stdout, "%c[00m", 0x1b); |
| 1247 |
fwrite(ptr + offsets[1], 1, linelength - offsets[1], stdout); |
for (;;) |
| 1248 |
|
{ |
| 1249 |
|
last_offset += offsets[1]; |
| 1250 |
|
matchptr += offsets[1]; |
| 1251 |
|
length -= offsets[1]; |
| 1252 |
|
if (!match_patterns(matchptr, length, offsets, &mrc)) break; |
| 1253 |
|
fwrite(matchptr, 1, offsets[0], stdout); |
| 1254 |
|
fprintf(stdout, "%c[%sm", 0x1b, colour_string); |
| 1255 |
|
fwrite(matchptr + offsets[0], 1, offsets[1] - offsets[0], stdout); |
| 1256 |
|
fprintf(stdout, "%c[00m", 0x1b); |
| 1257 |
|
} |
| 1258 |
|
fwrite(ptr + last_offset, 1, (linelength + endlinelength) - last_offset, |
| 1259 |
|
stdout); |
| 1260 |
} |
} |
| 1261 |
|
|
| 1262 |
|
/* Not colouring; no need to search for further matches */ |
| 1263 |
|
|
| 1264 |
else fwrite(ptr, 1, linelength + endlinelength, stdout); |
else fwrite(ptr, 1, linelength + endlinelength, stdout); |
| 1265 |
} |
} |
| 1266 |
|
|
| 1275 |
lastmatchnumber = linenumber + 1; |
lastmatchnumber = linenumber + 1; |
| 1276 |
} |
} |
| 1277 |
|
|
| 1278 |
/* Advance to after the newline and increment the line number. */ |
/* For a match in multiline inverted mode (which of course did not cause |
| 1279 |
|
anything to be printed), we have to move on to the end of the match before |
| 1280 |
|
proceeding. */ |
| 1281 |
|
|
| 1282 |
|
if (multiline && invert && match) |
| 1283 |
|
{ |
| 1284 |
|
int ellength; |
| 1285 |
|
char *endmatch = ptr + offsets[1]; |
| 1286 |
|
t = ptr; |
| 1287 |
|
while (t < endmatch) |
| 1288 |
|
{ |
| 1289 |
|
t = end_of_line(t, endptr, &ellength); |
| 1290 |
|
if (t <= endmatch) linenumber++; else break; |
| 1291 |
|
} |
| 1292 |
|
endmatch = end_of_line(endmatch, endptr, &ellength); |
| 1293 |
|
linelength = endmatch - ptr - ellength; |
| 1294 |
|
} |
| 1295 |
|
|
| 1296 |
|
/* Advance to after the newline and increment the line number. The file |
| 1297 |
|
offset to the current line is maintained in filepos. */ |
| 1298 |
|
|
| 1299 |
ptr += linelength + endlinelength; |
ptr += linelength + endlinelength; |
| 1300 |
|
filepos += linelength + endlinelength; |
| 1301 |
linenumber++; |
linenumber++; |
| 1302 |
|
|
| 1303 |
/* If we haven't yet reached the end of the file (the buffer is full), and |
/* If we haven't yet reached the end of the file (the buffer is full), and |
| 1319 |
|
|
| 1320 |
memmove(buffer, buffer + MBUFTHIRD, 2*MBUFTHIRD); |
memmove(buffer, buffer + MBUFTHIRD, 2*MBUFTHIRD); |
| 1321 |
ptr -= MBUFTHIRD; |
ptr -= MBUFTHIRD; |
| 1322 |
|
|
| 1323 |
|
#ifdef SUPPORT_LIBZ |
| 1324 |
|
if (frtype == FR_LIBZ) |
| 1325 |
|
bufflength = 2*MBUFTHIRD + |
| 1326 |
|
gzread (ingz, buffer + 2*MBUFTHIRD, MBUFTHIRD); |
| 1327 |
|
else |
| 1328 |
|
#endif |
| 1329 |
|
|
| 1330 |
|
#ifdef SUPPORT_LIBBZ2 |
| 1331 |
|
if (frtype == FR_LIBBZ2) |
| 1332 |
|
bufflength = 2*MBUFTHIRD + |
| 1333 |
|
BZ2_bzread(inbz2, buffer + 2*MBUFTHIRD, MBUFTHIRD); |
| 1334 |
|
else |
| 1335 |
|
#endif |
| 1336 |
|
|
| 1337 |
bufflength = 2*MBUFTHIRD + fread(buffer + 2*MBUFTHIRD, 1, MBUFTHIRD, in); |
bufflength = 2*MBUFTHIRD + fread(buffer + 2*MBUFTHIRD, 1, MBUFTHIRD, in); |
| 1338 |
|
|
| 1339 |
endptr = buffer + bufflength; |
endptr = buffer + bufflength; |
| 1340 |
|
|
| 1341 |
/* Adjust any last match point */ |
/* Adjust any last match point */ |
| 1366 |
|
|
| 1367 |
if (count_only) |
if (count_only) |
| 1368 |
{ |
{ |
| 1369 |
if (printname != NULL) fprintf(stdout, "%s:", printname); |
if (count > 0 || !omit_zero_count) |
| 1370 |
fprintf(stdout, "%d\n", count); |
{ |
| 1371 |
|
if (printname != NULL && filenames != FN_NONE) |
| 1372 |
|
fprintf(stdout, "%s:", printname); |
| 1373 |
|
fprintf(stdout, "%d\n", count); |
| 1374 |
|
} |
| 1375 |
} |
} |
| 1376 |
|
|
| 1377 |
return rc; |
return rc; |
| 1403 |
{ |
{ |
| 1404 |
int rc = 1; |
int rc = 1; |
| 1405 |
int sep; |
int sep; |
| 1406 |
FILE *in; |
int frtype; |
| 1407 |
|
int pathlen; |
| 1408 |
|
void *handle; |
| 1409 |
|
FILE *in = NULL; /* Ensure initialized */ |
| 1410 |
|
|
| 1411 |
|
#ifdef SUPPORT_LIBZ |
| 1412 |
|
gzFile ingz = NULL; |
| 1413 |
|
#endif |
| 1414 |
|
|
| 1415 |
|
#ifdef SUPPORT_LIBBZ2 |
| 1416 |
|
BZFILE *inbz2 = NULL; |
| 1417 |
|
#endif |
| 1418 |
|
|
| 1419 |
/* If the file name is "-" we scan stdin */ |
/* If the file name is "-" we scan stdin */ |
| 1420 |
|
|
| 1421 |
if (strcmp(pathname, "-") == 0) |
if (strcmp(pathname, "-") == 0) |
| 1422 |
{ |
{ |
| 1423 |
return pcregrep(stdin, |
return pcregrep(stdin, FR_PLAIN, |
| 1424 |
(filenames > FN_DEFAULT || (filenames == FN_DEFAULT && !only_one_at_top))? |
(filenames > FN_DEFAULT || (filenames == FN_DEFAULT && !only_one_at_top))? |
| 1425 |
stdin_name : NULL); |
stdin_name : NULL); |
| 1426 |
} |
} |
| 1427 |
|
|
|
|
|
| 1428 |
/* If the file is a directory, skip if skipping or if we are recursing, scan |
/* If the file is a directory, skip if skipping or if we are recursing, scan |
| 1429 |
each file within it, subject to any include or exclude patterns that were set. |
each file and directory within it, subject to any include or exclude patterns |
| 1430 |
The scanning code is localized so it can be made system-specific. */ |
that were set. The scanning code is localized so it can be made |
| 1431 |
|
system-specific. */ |
| 1432 |
|
|
| 1433 |
if ((sep = isdirectory(pathname)) != 0) |
if ((sep = isdirectory(pathname)) != 0) |
| 1434 |
{ |
{ |
| 1449 |
|
|
| 1450 |
while ((nextfile = readdirectory(dir)) != NULL) |
while ((nextfile = readdirectory(dir)) != NULL) |
| 1451 |
{ |
{ |
| 1452 |
int frc, blen; |
int frc, nflen; |
| 1453 |
sprintf(buffer, "%.512s%c%.128s", pathname, sep, nextfile); |
sprintf(buffer, "%.512s%c%.128s", pathname, sep, nextfile); |
| 1454 |
blen = strlen(buffer); |
nflen = strlen(nextfile); |
| 1455 |
|
|
| 1456 |
if (exclude_compiled != NULL && |
if (isdirectory(buffer)) |
| 1457 |
pcre_exec(exclude_compiled, NULL, buffer, blen, 0, 0, NULL, 0) >= 0) |
{ |
| 1458 |
continue; |
if (exclude_dir_compiled != NULL && |
| 1459 |
|
pcre_exec(exclude_dir_compiled, NULL, nextfile, nflen, 0, 0, NULL, 0) >= 0) |
| 1460 |
if (include_compiled != NULL && |
continue; |
| 1461 |
pcre_exec(include_compiled, NULL, buffer, blen, 0, 0, NULL, 0) < 0) |
|
| 1462 |
continue; |
if (include_dir_compiled != NULL && |
| 1463 |
|
pcre_exec(include_dir_compiled, NULL, nextfile, nflen, 0, 0, NULL, 0) < 0) |
| 1464 |
|
continue; |
| 1465 |
|
} |
| 1466 |
|
else |
| 1467 |
|
{ |
| 1468 |
|
if (exclude_compiled != NULL && |
| 1469 |
|
pcre_exec(exclude_compiled, NULL, nextfile, nflen, 0, 0, NULL, 0) >= 0) |
| 1470 |
|
continue; |
| 1471 |
|
|
| 1472 |
|
if (include_compiled != NULL && |
| 1473 |
|
pcre_exec(include_compiled, NULL, nextfile, nflen, 0, 0, NULL, 0) < 0) |
| 1474 |
|
continue; |
| 1475 |
|
} |
| 1476 |
|
|
| 1477 |
frc = grep_or_recurse(buffer, dir_recurse, FALSE); |
frc = grep_or_recurse(buffer, dir_recurse, FALSE); |
| 1478 |
if (frc > 1) rc = frc; |
if (frc > 1) rc = frc; |
| 1495 |
argument at top level, we don't show the file name, unless we are only showing |
argument at top level, we don't show the file name, unless we are only showing |
| 1496 |
the file name, or the filename was forced (-H). */ |
the file name, or the filename was forced (-H). */ |
| 1497 |
|
|
| 1498 |
in = fopen(pathname, "r"); |
pathlen = strlen(pathname); |
| 1499 |
if (in == NULL) |
|
| 1500 |
|
/* Open using zlib if it is supported and the file name ends with .gz. */ |
| 1501 |
|
|
| 1502 |
|
#ifdef SUPPORT_LIBZ |
| 1503 |
|
if (pathlen > 3 && strcmp(pathname + pathlen - 3, ".gz") == 0) |
| 1504 |
|
{ |
| 1505 |
|
ingz = gzopen(pathname, "rb"); |
| 1506 |
|
if (ingz == NULL) |
| 1507 |
|
{ |
| 1508 |
|
if (!silent) |
| 1509 |
|
fprintf(stderr, "pcregrep: Failed to open %s: %s\n", pathname, |
| 1510 |
|
strerror(errno)); |
| 1511 |
|
return 2; |
| 1512 |
|
} |
| 1513 |
|
handle = (void *)ingz; |
| 1514 |
|
frtype = FR_LIBZ; |
| 1515 |
|
} |
| 1516 |
|
else |
| 1517 |
|
#endif |
| 1518 |
|
|
| 1519 |
|
/* Otherwise open with bz2lib if it is supported and the name ends with .bz2. */ |
| 1520 |
|
|
| 1521 |
|
#ifdef SUPPORT_LIBBZ2 |
| 1522 |
|
if (pathlen > 4 && strcmp(pathname + pathlen - 4, ".bz2") == 0) |
| 1523 |
|
{ |
| 1524 |
|
inbz2 = BZ2_bzopen(pathname, "rb"); |
| 1525 |
|
handle = (void *)inbz2; |
| 1526 |
|
frtype = FR_LIBBZ2; |
| 1527 |
|
} |
| 1528 |
|
else |
| 1529 |
|
#endif |
| 1530 |
|
|
| 1531 |
|
/* Otherwise use plain fopen(). The label is so that we can come back here if |
| 1532 |
|
an attempt to read a .bz2 file indicates that it really is a plain file. */ |
| 1533 |
|
|
| 1534 |
|
#ifdef SUPPORT_LIBBZ2 |
| 1535 |
|
PLAIN_FILE: |
| 1536 |
|
#endif |
| 1537 |
|
{ |
| 1538 |
|
in = fopen(pathname, "rb"); |
| 1539 |
|
handle = (void *)in; |
| 1540 |
|
frtype = FR_PLAIN; |
| 1541 |
|
} |
| 1542 |
|
|
| 1543 |
|
/* All the opening methods return errno when they fail. */ |
| 1544 |
|
|
| 1545 |
|
if (handle == NULL) |
| 1546 |
{ |
{ |
| 1547 |
if (!silent) |
if (!silent) |
| 1548 |
fprintf(stderr, "pcregrep: Failed to open %s: %s\n", pathname, |
fprintf(stderr, "pcregrep: Failed to open %s: %s\n", pathname, |
| 1550 |
return 2; |
return 2; |
| 1551 |
} |
} |
| 1552 |
|
|
| 1553 |
rc = pcregrep(in, (filenames > FN_DEFAULT || |
/* Now grep the file */ |
| 1554 |
|
|
| 1555 |
|
rc = pcregrep(handle, frtype, (filenames > FN_DEFAULT || |
| 1556 |
(filenames == FN_DEFAULT && !only_one_at_top))? pathname : NULL); |
(filenames == FN_DEFAULT && !only_one_at_top))? pathname : NULL); |
| 1557 |
|
|
| 1558 |
|
/* Close in an appropriate manner. */ |
| 1559 |
|
|
| 1560 |
|
#ifdef SUPPORT_LIBZ |
| 1561 |
|
if (frtype == FR_LIBZ) |
| 1562 |
|
gzclose(ingz); |
| 1563 |
|
else |
| 1564 |
|
#endif |
| 1565 |
|
|
| 1566 |
|
/* If it is a .bz2 file and the result is 2, it means that the first attempt to |
| 1567 |
|
read failed. If the error indicates that the file isn't in fact bzipped, try |
| 1568 |
|
again as a normal file. */ |
| 1569 |
|
|
| 1570 |
|
#ifdef SUPPORT_LIBBZ2 |
| 1571 |
|
if (frtype == FR_LIBBZ2) |
| 1572 |
|
{ |
| 1573 |
|
if (rc == 2) |
| 1574 |
|
{ |
| 1575 |
|
int errnum; |
| 1576 |
|
const char *err = BZ2_bzerror(inbz2, &errnum); |
| 1577 |
|
if (errnum == BZ_DATA_ERROR_MAGIC) |
| 1578 |
|
{ |
| 1579 |
|
BZ2_bzclose(inbz2); |
| 1580 |
|
goto PLAIN_FILE; |
| 1581 |
|
} |
| 1582 |
|
else if (!silent) |
| 1583 |
|
fprintf(stderr, "pcregrep: Failed to read %s using bzlib: %s\n", |
| 1584 |
|
pathname, err); |
| 1585 |
|
} |
| 1586 |
|
BZ2_bzclose(inbz2); |
| 1587 |
|
} |
| 1588 |
|
else |
| 1589 |
|
#endif |
| 1590 |
|
|
| 1591 |
|
/* Normal file close */ |
| 1592 |
|
|
| 1593 |
fclose(in); |
fclose(in); |
| 1594 |
|
|
| 1595 |
|
/* Pass back the yield from pcregrep(). */ |
| 1596 |
|
|
| 1597 |
return rc; |
return rc; |
| 1598 |
} |
} |
| 1599 |
|
|
| 1614 |
if (op->one_char > 0) fprintf(stderr, "%c", op->one_char); |
if (op->one_char > 0) fprintf(stderr, "%c", op->one_char); |
| 1615 |
} |
} |
| 1616 |
fprintf(stderr, "] [long options] [pattern] [files]\n"); |
fprintf(stderr, "] [long options] [pattern] [files]\n"); |
| 1617 |
fprintf(stderr, "Type `pcregrep --help' for more information.\n"); |
fprintf(stderr, "Type `pcregrep --help' for more information and the long " |
| 1618 |
|
"options.\n"); |
| 1619 |
return rc; |
return rc; |
| 1620 |
} |
} |
| 1621 |
|
|
| 1634 |
printf("Usage: pcregrep [OPTION]... [PATTERN] [FILE1 FILE2 ...]\n"); |
printf("Usage: pcregrep [OPTION]... [PATTERN] [FILE1 FILE2 ...]\n"); |
| 1635 |
printf("Search for PATTERN in each FILE or standard input.\n"); |
printf("Search for PATTERN in each FILE or standard input.\n"); |
| 1636 |
printf("PATTERN must be present if neither -e nor -f is used.\n"); |
printf("PATTERN must be present if neither -e nor -f is used.\n"); |
| 1637 |
printf("\"-\" can be used as a file name to mean STDIN.\n\n"); |
printf("\"-\" can be used as a file name to mean STDIN.\n"); |
| 1638 |
printf("Example: pcregrep -i 'hello.*world' menu.h main.c\n\n"); |
|
| 1639 |
|
#ifdef SUPPORT_LIBZ |
| 1640 |
|
printf("Files whose names end in .gz are read using zlib.\n"); |
| 1641 |
|
#endif |
| 1642 |
|
|
| 1643 |
|
#ifdef SUPPORT_LIBBZ2 |
| 1644 |
|
printf("Files whose names end in .bz2 are read using bzlib2.\n"); |
| 1645 |
|
#endif |
| 1646 |
|
|
| 1647 |
|
#if defined SUPPORT_LIBZ || defined SUPPORT_LIBBZ2 |
| 1648 |
|
printf("Other files and the standard input are read as plain files.\n\n"); |
| 1649 |
|
#else |
| 1650 |
|
printf("All files are read as plain files, without any interpretation.\n\n"); |
| 1651 |
|
#endif |
| 1652 |
|
|
| 1653 |
|
printf("Example: pcregrep -i 'hello.*world' menu.h main.c\n\n"); |
| 1654 |
printf("Options:\n"); |
printf("Options:\n"); |
| 1655 |
|
|
| 1656 |
for (op = optionlist; op->one_char != 0; op++) |
for (op = optionlist; op->one_char != 0; op++) |
| 1658 |
int n; |
int n; |
| 1659 |
char s[4]; |
char s[4]; |
| 1660 |
if (op->one_char > 0) sprintf(s, "-%c,", op->one_char); else strcpy(s, " "); |
if (op->one_char > 0) sprintf(s, "-%c,", op->one_char); else strcpy(s, " "); |
| 1661 |
printf(" %s --%s%n", s, op->long_name, &n); |
n = 30 - printf(" %s --%s", s, op->long_name); |
|
n = 30 - n; |
|
| 1662 |
if (n < 1) n = 1; |
if (n < 1) n = 1; |
| 1663 |
printf("%.*s%s\n", n, " ", op->help_text); |
printf("%.*s%s\n", n, " ", op->help_text); |
| 1664 |
} |
} |
| 1683 |
{ |
{ |
| 1684 |
switch(letter) |
switch(letter) |
| 1685 |
{ |
{ |
| 1686 |
|
case N_FOFFSETS: file_offsets = TRUE; break; |
| 1687 |
case N_HELP: help(); exit(0); |
case N_HELP: help(); exit(0); |
| 1688 |
|
case N_LOFFSETS: line_offsets = number = TRUE; break; |
| 1689 |
case 'c': count_only = TRUE; break; |
case 'c': count_only = TRUE; break; |
| 1690 |
case 'F': process_options |= PO_FIXED_STRINGS; break; |
case 'F': process_options |= PO_FIXED_STRINGS; break; |
| 1691 |
case 'H': filenames = FN_FORCE; break; |
case 'H': filenames = FN_FORCE; break; |
| 1692 |
case 'h': filenames = FN_NONE; break; |
case 'h': filenames = FN_NONE; break; |
| 1693 |
case 'i': options |= PCRE_CASELESS; break; |
case 'i': options |= PCRE_CASELESS; break; |
| 1694 |
case 'l': filenames = FN_ONLY; break; |
case 'l': omit_zero_count = TRUE; filenames = FN_MATCH_ONLY; break; |
| 1695 |
case 'L': filenames = FN_NOMATCH_ONLY; break; |
case 'L': filenames = FN_NOMATCH_ONLY; break; |
| 1696 |
case 'M': multiline = TRUE; options |= PCRE_MULTILINE|PCRE_FIRSTLINE; break; |
case 'M': multiline = TRUE; options |= PCRE_MULTILINE|PCRE_FIRSTLINE; break; |
| 1697 |
case 'n': number = TRUE; break; |
case 'n': number = TRUE; break; |
| 1781 |
suffix[process_options]); |
suffix[process_options]); |
| 1782 |
pattern_list[pattern_count] = |
pattern_list[pattern_count] = |
| 1783 |
pcre_compile(buffer, options, &error, &errptr, pcretables); |
pcre_compile(buffer, options, &error, &errptr, pcretables); |
| 1784 |
if (pattern_list[pattern_count] != NULL) |
if (pattern_list[pattern_count] != NULL) |
| 1785 |
{ |
{ |
| 1786 |
pattern_count++; |
pattern_count++; |
| 1787 |
return TRUE; |
return TRUE; |
| 1788 |
} |
} |
| 1789 |
|
|
| 1790 |
/* Handle compile errors */ |
/* Handle compile errors */ |
| 1791 |
|
|
| 1843 |
char *p = end_of_line(pattern, eop, &ellength); |
char *p = end_of_line(pattern, eop, &ellength); |
| 1844 |
if (ellength == 0) |
if (ellength == 0) |
| 1845 |
return compile_single_pattern(pattern, options, filename, count); |
return compile_single_pattern(pattern, options, filename, count); |
| 1846 |
sprintf(buffer, "%.*s", p - pattern - ellength, pattern); |
sprintf(buffer, "%.*s", (int)(p - pattern - ellength), pattern); |
| 1847 |
pattern = p; |
pattern = p; |
| 1848 |
if (!compile_single_pattern(buffer, options, filename, count)) |
if (!compile_single_pattern(buffer, options, filename, count)) |
| 1849 |
return FALSE; |
return FALSE; |
| 1876 |
|
|
| 1877 |
/* Set the default line ending value from the default in the PCRE library; |
/* Set the default line ending value from the default in the PCRE library; |
| 1878 |
"lf", "cr", "crlf", and "any" are supported. Anything else is treated as "lf". |
"lf", "cr", "crlf", and "any" are supported. Anything else is treated as "lf". |
| 1879 |
*/ |
Note that the return values from pcre_config(), though derived from the ASCII |
| 1880 |
|
codes, are the same in EBCDIC environments, so we must use the actual values |
| 1881 |
|
rather than escapes such as as '\r'. */ |
| 1882 |
|
|
| 1883 |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &i); |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &i); |
| 1884 |
switch(i) |
switch(i) |
| 1885 |
{ |
{ |
| 1886 |
default: newline = (char *)"lf"; break; |
default: newline = (char *)"lf"; break; |
| 1887 |
case '\r': newline = (char *)"cr"; break; |
case 13: newline = (char *)"cr"; break; |
| 1888 |
case ('\r' << 8) | '\n': newline = (char *)"crlf"; break; |
case (13 << 8) | 10: newline = (char *)"crlf"; break; |
| 1889 |
case -1: newline = (char *)"any"; break; |
case -1: newline = (char *)"any"; break; |
| 1890 |
|
case -2: newline = (char *)"anycrlf"; break; |
| 1891 |
} |
} |
| 1892 |
|
|
| 1893 |
/* Process the options */ |
/* Process the options */ |
| 1945 |
else /* Special case xxx=data */ |
else /* Special case xxx=data */ |
| 1946 |
{ |
{ |
| 1947 |
int oplen = equals - op->long_name; |
int oplen = equals - op->long_name; |
| 1948 |
int arglen = (argequals == NULL)? strlen(arg) : argequals - arg; |
int arglen = (argequals == NULL)? (int)strlen(arg) : argequals - arg; |
| 1949 |
if (oplen == arglen && strncmp(arg, op->long_name, oplen) == 0) |
if (oplen == arglen && strncmp(arg, op->long_name, oplen) == 0) |
| 1950 |
{ |
{ |
| 1951 |
option_data = arg + arglen; |
option_data = arg + arglen; |
| 1964 |
char buff2[24]; |
char buff2[24]; |
| 1965 |
int baselen = opbra - op->long_name; |
int baselen = opbra - op->long_name; |
| 1966 |
sprintf(buff1, "%.*s", baselen, op->long_name); |
sprintf(buff1, "%.*s", baselen, op->long_name); |
| 1967 |
sprintf(buff2, "%s%.*s", buff1, strlen(op->long_name) - baselen - 2, |
sprintf(buff2, "%s%.*s", buff1, |
| 1968 |
opbra + 1); |
(int)strlen(op->long_name) - baselen - 2, opbra + 1); |
| 1969 |
if (strcmp(arg, buff1) == 0 || strcmp(arg, buff2) == 0) |
if (strcmp(arg, buff1) == 0 || strcmp(arg, buff2) == 0) |
| 1970 |
break; |
break; |
| 1971 |
} |
} |
| 2124 |
if (before_context == 0) before_context = both_context; |
if (before_context == 0) before_context = both_context; |
| 2125 |
} |
} |
| 2126 |
|
|
| 2127 |
|
/* Only one of --only-matching, --file-offsets, or --line-offsets is permitted. |
| 2128 |
|
However, the latter two set the only_matching flag. */ |
| 2129 |
|
|
| 2130 |
|
if ((only_matching && (file_offsets || line_offsets)) || |
| 2131 |
|
(file_offsets && line_offsets)) |
| 2132 |
|
{ |
| 2133 |
|
fprintf(stderr, "pcregrep: Cannot mix --only-matching, --file-offsets " |
| 2134 |
|
"and/or --line-offsets\n"); |
| 2135 |
|
exit(usage(2)); |
| 2136 |
|
} |
| 2137 |
|
|
| 2138 |
|
if (file_offsets || line_offsets) only_matching = TRUE; |
| 2139 |
|
|
| 2140 |
/* If a locale has not been provided as an option, see if the LC_CTYPE or |
/* If a locale has not been provided as an option, see if the LC_CTYPE or |
| 2141 |
LC_ALL environment variable is set, and if so, use it. */ |
LC_ALL environment variable is set, and if so, use it. */ |
| 2142 |
|
|
| 2208 |
pcre_options |= PCRE_NEWLINE_ANY; |
pcre_options |= PCRE_NEWLINE_ANY; |
| 2209 |
endlinetype = EL_ANY; |
endlinetype = EL_ANY; |
| 2210 |
} |
} |
| 2211 |
|
else if (strcmp(newline, "anycrlf") == 0 || strcmp(newline, "ANYCRLF") == 0) |
| 2212 |
|
{ |
| 2213 |
|
pcre_options |= PCRE_NEWLINE_ANYCRLF; |
| 2214 |
|
endlinetype = EL_ANYCRLF; |
| 2215 |
|
} |
| 2216 |
else |
else |
| 2217 |
{ |
{ |
| 2218 |
fprintf(stderr, "pcregrep: Invalid newline specifier \"%s\"\n", newline); |
fprintf(stderr, "pcregrep: Invalid newline specifier \"%s\"\n", newline); |
| 2341 |
fprintf(stderr, "pcregrep: Error while studying regex%s: %s\n", s, error); |
fprintf(stderr, "pcregrep: Error while studying regex%s: %s\n", s, error); |
| 2342 |
goto EXIT2; |
goto EXIT2; |
| 2343 |
} |
} |
| 2344 |
hint_count++; |
hint_count++; |
| 2345 |
} |
} |
| 2346 |
|
|
| 2347 |
/* If there are include or exclude patterns, compile them. */ |
/* If there are include or exclude patterns, compile them. */ |
| 2370 |
} |
} |
| 2371 |
} |
} |
| 2372 |
|
|
| 2373 |
|
if (exclude_dir_pattern != NULL) |
| 2374 |
|
{ |
| 2375 |
|
exclude_dir_compiled = pcre_compile(exclude_dir_pattern, 0, &error, &errptr, |
| 2376 |
|
pcretables); |
| 2377 |
|
if (exclude_dir_compiled == NULL) |
| 2378 |
|
{ |
| 2379 |
|
fprintf(stderr, "pcregrep: Error in 'exclude_dir' regex at offset %d: %s\n", |
| 2380 |
|
errptr, error); |
| 2381 |
|
goto EXIT2; |
| 2382 |
|
} |
| 2383 |
|
} |
| 2384 |
|
|
| 2385 |
|
if (include_dir_pattern != NULL) |
| 2386 |
|
{ |
| 2387 |
|
include_dir_compiled = pcre_compile(include_dir_pattern, 0, &error, &errptr, |
| 2388 |
|
pcretables); |
| 2389 |
|
if (include_dir_compiled == NULL) |
| 2390 |
|
{ |
| 2391 |
|
fprintf(stderr, "pcregrep: Error in 'include_dir' regex at offset %d: %s\n", |
| 2392 |
|
errptr, error); |
| 2393 |
|
goto EXIT2; |
| 2394 |
|
} |
| 2395 |
|
} |
| 2396 |
|
|
| 2397 |
/* If there are no further arguments, do the business on stdin and exit. */ |
/* If there are no further arguments, do the business on stdin and exit. */ |
| 2398 |
|
|
| 2399 |
if (i >= argc) |
if (i >= argc) |
| 2400 |
{ |
{ |
| 2401 |
rc = pcregrep(stdin, (filenames > FN_DEFAULT)? stdin_name : NULL); |
rc = pcregrep(stdin, FR_PLAIN, (filenames > FN_DEFAULT)? stdin_name : NULL); |
| 2402 |
goto EXIT; |
goto EXIT; |
| 2403 |
} |
} |
| 2404 |
|
|