semios/grep
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 3 Wiki Activity

initial commit

Browse Source 
Signed-off-by: sisungo <[email protected]>
This commit is contained in:
sisungo
2026-07-06 18:37:51 +08:00
commit b9c952d01d
10 changed files with 3051 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.gitignore
+1
View File
@@ -0,0 +1 @@
/grep
build.sh
Executable
+5
View File
@@ -0,0 +1,5 @@
#!/bin/sh
cd $(dirname $0)
${CC} -O3 ${CFLAGS} -o ./grep *.c
file.c
+248
View File
@@ -0,0 +1,248 @@
/* $NetBSD: file.c,v 1.5 2011/02/16 18:35:39 joerg Exp $ */
/* $OpenBSD: file.c,v 1.11 2010/07/02 20:48:48 nicm Exp $ */
/*-
* SPDX-License-Identifier: BSD-2-Clause
*
* Copyright (c) 1999 James Howard and Dag-Erling Smørgrav
* Copyright (C) 2008-2010 Gabor Kovesdan <[email protected]>
* Copyright (C) 2010 Dimitry Andric <[email protected]>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/param.h>
#include <sys/mman.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <err.h>
#include <errno.h>
#include <fcntl.h>
#include <stddef.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <wchar.h>
#include <wctype.h>
#include "grep.h"
#define MAXBUFSIZ (32 * 1024)
#define LNBUFBUMP 80
static char *buffer;
static char *bufpos;
static size_t bufrem;
static size_t fsiz;
static char *lnbuf;
static size_t lnbuflen;
static inline int
grep_refill(struct file *f)
{
ssize_t nr;
if (filebehave == FILE_MMAP)
return (0);
bufpos = buffer;
bufrem = 0;
nr = read(f->fd, buffer, MAXBUFSIZ);
if (nr < 0 && errno == EISDIR)
nr = 0;
if (nr < 0)
return (-1);
bufrem = nr;
return (0);
}
static inline int
grep_lnbufgrow(size_t newlen)
{
if (lnbuflen < newlen) {
lnbuf = grep_realloc(lnbuf, newlen);
lnbuflen = newlen;
}
return (0);
}
char *
grep_fgetln(struct file *f, struct parsec *pc)
{
char *p;
size_t len;
size_t off;
ptrdiff_t diff;
/* Fill the buffer, if necessary */
if (bufrem == 0 && grep_refill(f) != 0)
goto error;
if (bufrem == 0) {
/* Return zero length to indicate EOF */
pc->ln.len= 0;
return (bufpos);
}
/* Look for a newline in the remaining part of the buffer */
if ((p = memchr(bufpos, fileeol, bufrem)) != NULL) {
++p; /* advance over newline */
len = p - bufpos;
if (grep_lnbufgrow(len + 1))
goto error;
memcpy(lnbuf, bufpos, len);
bufrem -= len;
bufpos = p;
pc->ln.len = len;
lnbuf[len] = '\0';
return (lnbuf);
}
/* We have to copy the current buffered data to the line buffer */
for (len = bufrem, off = 0; ; len += bufrem) {
/* Make sure there is room for more data */
if (grep_lnbufgrow(len + LNBUFBUMP))
goto error;
memcpy(lnbuf + off, bufpos, len - off);
/* With FILE_MMAP, this is EOF; there's no more to refill */
if (filebehave == FILE_MMAP) {
bufrem -= len;
break;
}
off = len;
/* Fetch more to try and find EOL/EOF */
if (grep_refill(f) != 0)
goto error;
if (bufrem == 0)
/* EOF: return partial line */
break;
if ((p = memchr(bufpos, fileeol, bufrem)) == NULL)
continue;
/* got it: finish up the line (like code above) */
++p;
diff = p - bufpos;
len += diff;
if (grep_lnbufgrow(len + 1))
goto error;
memcpy(lnbuf + off, bufpos, diff);
bufrem -= diff;
bufpos = p;
break;
}
pc->ln.len = len;
lnbuf[len] = '\0';
return (lnbuf);
error:
pc->ln.len = 0;
return (NULL);
}
/*
* Opens a file for processing.
*/
struct file *
grep_open(const char *path)
{
struct file *f;
f = grep_malloc(sizeof *f);
memset(f, 0, sizeof *f);
if (path == NULL) {
/* Processing stdin implies --line-buffered. */
lbflag = true;
f->fd = STDIN_FILENO;
} else if ((f->fd = open(path, O_RDONLY)) == -1)
goto error1;
if (filebehave == FILE_MMAP) {
struct stat st;
if (fstat(f->fd, &st) == -1 || !S_ISREG(st.st_mode))
filebehave = FILE_STDIO;
else {
int flags = MAP_PRIVATE;
#ifdef MAP_PREFAULT_READ
flags |= MAP_PREFAULT_READ;
#endif
fsiz = st.st_size;
buffer = mmap(NULL, fsiz, PROT_READ, flags,
f->fd, (off_t)0);
if (buffer == MAP_FAILED)
filebehave = FILE_STDIO;
else {
bufrem = st.st_size;
bufpos = buffer;
madvise(buffer, st.st_size, MADV_SEQUENTIAL);
}
}
}
if ((buffer == NULL) || (buffer == MAP_FAILED))
buffer = grep_malloc(MAXBUFSIZ);
/* Fill read buffer, also catches errors early */
if (bufrem == 0 && grep_refill(f) != 0)
goto error2;
/* Check for binary stuff, if necessary */
if (binbehave != BINFILE_TEXT && fileeol != '\0' &&
memchr(bufpos, '\0', bufrem) != NULL)
f->binary = true;
return (f);
error2:
close(f->fd);
error1:
free(f);
return (NULL);
}
/*
* Closes a file.
*/
void
grep_close(struct file *f)
{
close(f->fd);
/* Reset read buffer and line buffer */
if (filebehave == FILE_MMAP) {
munmap(buffer, fsiz);
buffer = NULL;
}
bufpos = buffer;
bufrem = 0;
free(lnbuf);
lnbuf = NULL;
lnbuflen = 0;
}
grep.1
+592
View File
@@ -0,0 +1,592 @@
.\" $NetBSD: grep.1,v 1.2 2011/02/16 01:31:33 joerg Exp $
.\" $OpenBSD: grep.1,v 1.38 2010/04/05 06:30:59 jmc Exp $
.\" Copyright (c) 1980, 1990, 1993
.\" The Regents of the University of California. All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\" 3. Neither the name of the University nor the names of its contributors
.\" may be used to endorse or promote products derived from this software
.\" without specific prior written permission.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.Dd December 18, 2022
.Dt GREP 1
.Os
.Sh NAME
.Nm grep ,
.Nm egrep ,
.Nm fgrep ,
.Nm rgrep
.Nd file pattern searcher
.Sh SYNOPSIS
.Nm grep
.Bk -words
.Op Fl abcdDEFGHhIiLlmnOopqRSsUVvwxz
.Op Fl A Ar num
.Op Fl B Ar num
.Op Fl C Ar num
.Op Fl e Ar pattern
.Op Fl f Ar file
.Op Fl Fl binary-files= Ns Ar value
.Op Fl Fl color Ns Op Cm = Ns Ar when
.Op Fl Fl colour Ns Op Cm = Ns Ar when
.Op Fl Fl context= Ns Ar num
.Op Fl Fl label
.Op Fl Fl line-buffered
.Op Fl Fl null
.Op Ar pattern
.Op Ar
.Ek
.Sh DESCRIPTION
The
.Nm grep
utility searches any given input files,
selecting lines that match one or more patterns.
By default, a pattern matches an input line if the regular expression
(RE) in the pattern matches the input line
without its trailing newline.
An empty expression matches every line.
Each input line that matches at least one of the patterns is written
to the standard output.
.Pp
.Nm grep
is used for simple patterns and
basic regular expressions
.Pq BREs ;
.Nm egrep
can handle extended regular expressions
.Pq EREs .
See
.Xr re_format 7
for more information on regular expressions.
.Nm fgrep
is quicker than both
.Nm grep
and
.Nm egrep ,
but can only handle fixed patterns
(i.e., it does not interpret regular expressions).
Patterns may consist of one or more lines,
allowing any of the pattern lines to match a portion of the input.
.Pp
The following options are available:
.Bl -tag -width indent
.It Fl A Ar num , Fl Fl after-context= Ns Ar num
Print
.Ar num
lines of trailing context after each match.
See also the
.Fl B
and
.Fl C
options.
.It Fl a , Fl Fl text
Treat all files as ASCII text.
Normally
.Nm
will simply print
.Dq Binary file ... matches
if files contain binary characters.
Use of this option forces
.Nm
to output lines matching the specified pattern.
.It Fl B Ar num , Fl Fl before-context= Ns Ar num
Print
.Ar num
lines of leading context before each match.
See also the
.Fl A
and
.Fl C
options.
.It Fl b , Fl Fl byte-offset
The offset in bytes of a matched pattern is
displayed in front of the respective matched line.
.It Fl C Ar num , Fl Fl context= Ns Ar num
Print
.Ar num
lines of leading and trailing context surrounding each match.
See also the
.Fl A
and
.Fl B
options.
.It Fl c , Fl Fl count
Only a count of selected lines is written to standard output.
.It Fl Fl colour= Ns Oo Ar when Oc , Fl Fl color= Ns Oo Ar when Oc
Mark up the matching text with the expression stored in the
.Ev GREP_COLOR
environment variable.
The possible values of
.Ar when
are
.Dq Cm never ,
.Dq Cm always
and
.Dq Cm auto .
.It Fl D Ar action , Fl Fl devices= Ns Ar action
Specify the demanded
.Ar action
for devices, FIFOs and sockets.
The default
.Ar action
is
.Dq Cm read ,
which means, that they are read as if they were normal files.
If the
.Ar action
is set to
.Dq Cm skip ,
devices are silently skipped.
.It Fl d Ar action , Fl Fl directories= Ns Ar action
Specify the demanded
.Ar action
for directories.
It is
.Dq Cm read
by default, which means that the directories
are read in the same manner as normal files.
Other possible values are
.Dq Cm skip
to silently ignore the directories, and
.Dq Cm recurse
to read them recursively, which has the same effect as the
.Fl R
and
.Fl r
option.
.It Fl E , Fl Fl extended-regexp
Interpret
.Ar pattern
as an extended regular expression
(i.e., force
.Nm grep
to behave as
.Nm egrep ) .
.It Fl e Ar pattern , Fl Fl regexp= Ns Ar pattern
Specify a
.Ar pattern
used during the search of the input:
an input line is selected if it matches any of the specified patterns.
This option is most useful when multiple
.Fl e
options are used to specify multiple patterns,
or when a
.Ar pattern
begins with a dash
.Pq Sq - .
.It Fl Fl exclude Ar pattern
If specified, it excludes files matching the given
filename
.Ar pattern
from the search.
Note that
.Fl Fl exclude
and
.Fl Fl include
patterns are processed in the order given.
If a name matches multiple patterns, the latest matching rule wins.
If no
.Fl Fl include
pattern is specified, all files are searched that are
not excluded.
Patterns are matched to the full path specified,
not only to the filename component.
.It Fl Fl exclude-dir Ar pattern
If
.Fl R
is specified, it excludes directories matching the
given filename
.Ar pattern
from the search.
Note that
.Fl Fl exclude-dir
and
.Fl Fl include-dir
patterns are processed in the order given.
If a name matches multiple patterns, the latest matching rule wins.
If no
.Fl Fl include-dir
pattern is specified, all directories are searched that are
not excluded.
.It Fl F , Fl Fl fixed-strings
Interpret
.Ar pattern
as a set of fixed strings
(i.e., force
.Nm grep
to behave as
.Nm fgrep ) .
.It Fl f Ar file , Fl Fl file= Ns Ar file
Read one or more newline separated patterns from
.Ar file .
Empty pattern lines match every input line.
Newlines are not considered part of a pattern.
If
.Ar file
is empty, nothing is matched.
.It Fl G , Fl Fl basic-regexp
Interpret
.Ar pattern
as a basic regular expression
(i.e., force
.Nm grep
to behave as traditional
.Nm grep ) .
.It Fl H
Always print filename headers with output lines.
.It Fl h , Fl Fl no-filename
Never print filename headers
.Pq i.e., filenames
with output lines.
.It Fl Fl help
Print a brief help message.
.It Fl I
Ignore binary files.
This option is equivalent to the
.Dq Fl Fl binary-files= Ns Cm without-match
option.
.It Fl i , Fl Fl ignore-case
Perform case insensitive matching.
By default,
.Nm grep
is case sensitive.
.It Fl Fl include Ar pattern
If specified, only files matching the given filename
.Ar pattern
are searched.
Note that
.Fl Fl include
and
.Fl Fl exclude
patterns are processed in the order given.
If a name matches multiple patterns, the latest matching rule wins.
Patterns are matched to the full path specified,
not only to the filename component.
.It Fl Fl include-dir Ar pattern
If
.Fl R
is specified, only directories matching the given filename
.Ar pattern
are searched.
Note that
.Fl Fl include-dir
and
.Fl Fl exclude-dir
patterns are processed in the order given.
If a name matches multiple patterns, the latest matching rule wins.
.It Fl L , Fl Fl files-without-match
Only the names of files not containing selected lines are written to
standard output.
Pathnames are listed once per file searched.
If the standard input is searched, the string
.Dq (standard input)
is written unless a
.Fl Fl label
is specified.
.It Fl l , Fl Fl files-with-matches
Only the names of files containing selected lines are written to
standard output.
.Nm grep
will only search a file until a match has been found,
making searches potentially less expensive.
Pathnames are listed once per file searched.
If the standard input is searched, the string
.Dq (standard input)
is written unless a
.Fl Fl label
is specified.
.It Fl Fl label
Label to use in place of
.Dq (standard input)
for a file name where a file name would normally be printed.
This option applies to
.Fl H ,
.Fl L ,
and
.Fl l .
.It Fl Fl mmap
Use
.Xr mmap 2
instead of
.Xr read 2
to read input, which can result in better performance under some
circumstances but can cause undefined behaviour.
.It Fl m Ar num , Fl Fl max-count= Ns Ar num
Stop reading the file after
.Ar num
matches.
.It Fl n , Fl Fl line-number
Each output line is preceded by its relative line number in the file,
starting at line 1.
The line number counter is reset for each file processed.
This option is ignored if
.Fl c ,
.Fl L ,
.Fl l ,
or
.Fl q
is
specified.
.It Fl Fl null
Prints a zero-byte after the file name.
.It Fl O
If
.Fl R
is specified, follow symbolic links only if they were explicitly listed
on the command line.
The default is not to follow symbolic links.
.It Fl o , Fl Fl only-matching
Prints only the matching part of the lines.
.It Fl p
If
.Fl R
is specified, no symbolic links are followed.
This is the default.
.It Fl q , Fl Fl quiet , Fl Fl silent
Quiet mode:
suppress normal output.
.Nm grep
will only search a file until a match has been found,
making searches potentially less expensive.
.It Fl R , Fl r , Fl Fl recursive
Recursively search subdirectories listed.
(i.e., force
.Nm grep
to behave as
.Nm rgrep ) .
.It Fl S
If
.Fl R
is specified, all symbolic links are followed.
The default is not to follow symbolic links.
.It Fl s , Fl Fl no-messages
Silent mode.
Nonexistent and unreadable files are ignored
(i.e., their error messages are suppressed).
.It Fl U , Fl Fl binary
Search binary files, but do not attempt to print them.
.It Fl u
This option has no effect and is provided only for compatibility with GNU grep.
.It Fl V , Fl Fl version
Display version information and exit.
.It Fl v , Fl Fl invert-match
Selected lines are those
.Em not
matching any of the specified patterns.
.It Fl w , Fl Fl word-regexp
The expression is searched for as a word (as if surrounded by
.Sq [[:<:]]
and
.Sq [[:>:]] ;
see
.Xr re_format 7 ) .
This option has no effect if
.Fl x
is also specified.
.It Fl x , Fl Fl line-regexp
Only input lines selected against an entire fixed string or regular
expression are considered to be matching lines.
.It Fl y
Equivalent to
.Fl i .
Obsoleted.
.It Fl z , Fl Fl null-data
Treat input and output data as sequences of lines terminated by a
zero-byte instead of a newline.
.It Fl Fl binary-files= Ns Ar value
Controls searching and printing of binary files.
Options are:
.Bl -tag -compact -width "binary (default)"
.It Cm binary No (default)
Search binary files but do not print them.
.It Cm without-match
Do not search binary files.
.It Cm text
Treat all files as text.
.El
.It Fl Fl line-buffered
Force output to be line buffered.
By default, output is line buffered when standard output is a terminal
and block buffered otherwise.
.El
.Pp
If no file arguments are specified, the standard input is used.
Additionally,
.Dq Cm -
may be used in place of a file name, anywhere that a file name is accepted, to
read from standard input.
This includes both
.Fl f
and file arguments.
.Sh ENVIRONMENT
The following environment variables affect the execution of
.Nm :
.Bl -tag -width "GREP_OPTIONS"
.It Ev GREP_COLOR
This variable specifies the color used to highlight matched (non-empty) text.
.It Ev GREP_OPTIONS
This variable specifies default options
to be placed in front of any explicit options.
It may cause problems with portable scripts.
.It Ev TERM
This variable specifies the type name of the terminal, console or display-device
type to be used.
See
.Xr term 7 .
.El
.Sh EXIT STATUS
The
.Nm grep
utility exits with one of the following values:
.Pp
.Bl -tag -width flag -compact
.It Li 0
One or more lines were selected.
.It Li 1
No lines were selected.
.It Li \*(Gt1
An error occurred.
.El
.Sh EXAMPLES
.Bl -dash
.It
Find all occurrences of the pattern
.Sq patricia
in a file:
.Pp
.Dl $ grep 'patricia' myfile
.It
Same as above but looking only for complete words:
.Pp
.Dl $ grep -w 'patricia' myfile
.It
Count occurrences of the exact pattern
.Sq FOO
:
.Pp
.Dl $ grep -c FOO myfile
.It
Same as above but ignoring case:
.Pp
.Dl $ grep -c -i FOO myfile
.It
Find all occurrences of the pattern
.Ql .Pp
at the beginning of a line:
.Pp
.Dl $ grep '^\e.Pp' myfile
.Pp
The apostrophes ensure the entire expression is evaluated by
.Nm grep
instead of by the user's shell.
The caret
.Ql ^
matches the null string at the beginning of a line,
and the
.Ql \e
escapes the
.Ql \&. ,
which would otherwise match any character.
.It
Find all lines in a file which do not contain the words
.Sq foo
or
.Sq bar :
.Pp
.Dl $ grep -v -e 'foo' -e 'bar' myfile
.It
Peruse the file
.Sq calendar
looking for either 19, 20, or 25 using extended regular expressions:
.Pp
.Dl $ egrep '19|20|25' calendar
.It
Show matching lines and the name of the
.Sq *.h
files which contain the pattern
.Sq FIXME .
Do the search recursively from the
.Pa /usr/src/sys/arm
directory
.Pp
.Dl $ grep -H -R FIXME --include="*.h" /usr/src/sys/arm/
.It
Same as above but show only the name of the matching file:
.Pp
.Dl $ grep -l -R FIXME --include="*.h" /usr/src/sys/arm/
.It
Show lines containing the text
.Sq foo .
The matching part of the output is colored and every line is prefixed with
the line number and the offset in the file for those lines that matched.
.Pp
.Dl $ grep -b --colour -n foo myfile
.It
Show lines that match the extended regular expression patterns read from the
standard input:
.Pp
.Dl $ echo -e 'Free\enBSD\enAll.*reserved' | grep -E -f - myfile
.It
Show lines from the output of the
.Xr pciconf 8
command matching the specified extended regular expression along with
three lines of leading context and one line of trailing context:
.Pp
.Dl $ pciconf -lv | grep -B3 -A1 -E 'class.*=.*storage'
.It
Suppress any output and use the exit status to show an appropriate message:
.Pp
.Dl $ grep -q foo myfile && echo File matches
.El
.Sh SEE ALSO
.Xr ed 1 ,
.Xr ex 1 ,
.Xr sed 1 ,
.Xr zgrep 1 ,
.Xr re_format 7
.Sh STANDARDS
The
.Nm
utility is compliant with the
.St -p1003.1-2008
specification.
.Pp
The flags
.Op Fl AaBbCDdGHhILmopRSUVw
are extensions to that specification, and the behaviour of the
.Fl f
flag when used with an empty pattern file is left undefined.
.Pp
All long options are provided for compatibility with
GNU versions of this utility.
.Pp
Historic versions of the
.Nm grep
utility also supported the flags
.Op Fl ruy .
This implementation supports those options;
however, their use is strongly discouraged.
.Sh HISTORY
The
.Nm grep
command first appeared in
.At v6 .
grep.c
+724
View File
@@ -0,0 +1,724 @@
/* $NetBSD: grep.c,v 1.6 2011/04/18 03:48:23 joerg Exp $ */
/* $OpenBSD: grep.c,v 1.42 2010/07/02 22:18:03 tedu Exp $ */
/*-
* SPDX-License-Identifier: BSD-2-Clause
*
* Copyright (c) 1999 James Howard and Dag-Erling Smørgrav
* Copyright (C) 2008-2009 Gabor Kovesdan <[email protected]>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <ctype.h>
#include <err.h>
#include <errno.h>
#include <fcntl.h>
#include <getopt.h>
#include <limits.h>
#include <libgen.h>
#include <locale.h>
#include <stdbool.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include "grep.h"
const char *errstr[] = {
"",
/* 1*/ "(standard input)",
/* 2*/ "unknown %s option",
/* 3*/ "usage: %s [-abcDEFGHhIiLlmnOopqRSsUVvwxz] [-A num] [-B num] [-C num]\n",
/* 4*/ "\t[-e pattern] [-f file] [--binary-files=value] [--color=when]\n",
/* 5*/ "\t[--context=num] [--directories=action] [--label] [--line-buffered]\n",
/* 6*/ "\t[--null] [pattern] [file ...]\n",
/* 7*/ "Binary file %s matches\n",
/* 8*/ "%s (BSD grep, GNU compatible) %s\n",
};
/* Flags passed to regcomp() and regexec() */
int cflags = REG_NOSUB | REG_NEWLINE;
int eflags = REG_STARTEND;
bool matchall;
/* Searching patterns */
unsigned int patterns;
static unsigned int pattern_sz;
struct pat *pattern;
regex_t *r_pattern;
/* Filename exclusion/inclusion patterns */
unsigned int fpatterns, dpatterns;
static unsigned int fpattern_sz, dpattern_sz;
struct epat *dpattern, *fpattern;
/* For regex errors */
char re_error[RE_ERROR_BUF + 1];
/* Command-line flags */
long long Aflag; /* -A x: print x lines trailing each match */
long long Bflag; /* -B x: print x lines leading each match */
bool Hflag; /* -H: always print file name */
bool Lflag; /* -L: only show names of files with no matches */
bool bflag; /* -b: show block numbers for each match */
bool cflag; /* -c: only show a count of matching lines */
bool hflag; /* -h: don't print filename headers */
bool iflag; /* -i: ignore case */
bool lflag; /* -l: only show names of files with matches */
bool mflag; /* -m x: stop reading the files after x matches */
long long mcount; /* count for -m */
long long mlimit; /* requested value for -m */
char fileeol; /* indicator for eol */
bool nflag; /* -n: show line numbers in front of matching lines */
bool oflag; /* -o: print only matching part */
bool qflag; /* -q: quiet mode (don't output anything) */
bool sflag; /* -s: silent mode (ignore errors) */
bool vflag; /* -v: only show non-matching lines */
bool wflag; /* -w: pattern must start and end on word boundaries */
bool xflag; /* -x: pattern must match entire line */
bool lbflag; /* --line-buffered */
bool nullflag; /* --null */
char *label; /* --label */
const char *color; /* --color */
int grepbehave = GREP_BASIC; /* -EFG: type of the regex */
int binbehave = BINFILE_BIN; /* -aIU: handling of binary files */
int filebehave = FILE_STDIO;
int devbehave = DEV_READ; /* -D: handling of devices */
int dirbehave = DIR_READ; /* -dRr: handling of directories */
int linkbehave = LINK_SKIP; /* -OpS: handling of symlinks */
bool dexclude, dinclude; /* --exclude-dir and --include-dir */
bool fexclude, finclude; /* --exclude and --include */
enum {
BIN_OPT = CHAR_MAX + 1,
COLOR_OPT,
HELP_OPT,
MMAP_OPT,
LINEBUF_OPT,
LABEL_OPT,
NULL_OPT,
R_EXCLUDE_OPT,
R_INCLUDE_OPT,
R_DEXCLUDE_OPT,
R_DINCLUDE_OPT
};
static inline const char *init_color(const char *);
/* Housekeeping */
bool file_err; /* file reading error */
/*
* Prints usage information and returns 2.
*/
static void
usage(void)
{
fprintf(stderr, errstr[3], getprogname());
fprintf(stderr, "%s", errstr[4]);
fprintf(stderr, "%s", errstr[5]);
fprintf(stderr, "%s", errstr[6]);
exit(2);
}
static const char *optstr = "0123456789A:B:C:D:EFGHILOSRUVabcd:e:f:hilm:nopqrsuvwxyz";
static const struct option long_options[] =
{
{"binary-files", required_argument, NULL, BIN_OPT},
{"help", no_argument, NULL, HELP_OPT},
{"mmap", no_argument, NULL, MMAP_OPT},
{"line-buffered", no_argument, NULL, LINEBUF_OPT},
{"label", required_argument, NULL, LABEL_OPT},
{"null", no_argument, NULL, NULL_OPT},
{"color", optional_argument, NULL, COLOR_OPT},
{"colour", optional_argument, NULL, COLOR_OPT},
{"exclude", required_argument, NULL, R_EXCLUDE_OPT},
{"include", required_argument, NULL, R_INCLUDE_OPT},
{"exclude-dir", required_argument, NULL, R_DEXCLUDE_OPT},
{"include-dir", required_argument, NULL, R_DINCLUDE_OPT},
{"after-context", required_argument, NULL, 'A'},
{"text", no_argument, NULL, 'a'},
{"before-context", required_argument, NULL, 'B'},
{"byte-offset", no_argument, NULL, 'b'},
{"context", optional_argument, NULL, 'C'},
{"count", no_argument, NULL, 'c'},
{"devices", required_argument, NULL, 'D'},
{"directories", required_argument, NULL, 'd'},
{"extended-regexp", no_argument, NULL, 'E'},
{"regexp", required_argument, NULL, 'e'},
{"fixed-strings", no_argument, NULL, 'F'},
{"file", required_argument, NULL, 'f'},
{"basic-regexp", no_argument, NULL, 'G'},
{"no-filename", no_argument, NULL, 'h'},
{"with-filename", no_argument, NULL, 'H'},
{"ignore-case", no_argument, NULL, 'i'},
{"files-with-matches", no_argument, NULL, 'l'},
{"files-without-match", no_argument, NULL, 'L'},
{"max-count", required_argument, NULL, 'm'},
{"line-number", no_argument, NULL, 'n'},
{"only-matching", no_argument, NULL, 'o'},
{"quiet", no_argument, NULL, 'q'},
{"silent", no_argument, NULL, 'q'},
{"recursive", no_argument, NULL, 'r'},
{"no-messages", no_argument, NULL, 's'},
{"binary", no_argument, NULL, 'U'},
{"unix-byte-offsets", no_argument, NULL, 'u'},
{"invert-match", no_argument, NULL, 'v'},
{"version", no_argument, NULL, 'V'},
{"word-regexp", no_argument, NULL, 'w'},
{"line-regexp", no_argument, NULL, 'x'},
{"null-data", no_argument, NULL, 'z'},
{NULL, no_argument, NULL, 0}
};
/*
* Adds a searching pattern to the internal array.
*/
static void
add_pattern(char *pat, size_t len)
{
/* Check if we can do a shortcut */
if (len == 0) {
matchall = true;
return;
}
/* Increase size if necessary */
if (patterns == pattern_sz) {
pattern_sz *= 2;
pattern = grep_realloc(pattern, ++pattern_sz *
sizeof(struct pat));
}
if (len > 0 && pat[len - 1] == '\n')
--len;
/* pat may not be NUL-terminated */
pattern[patterns].pat = grep_malloc(len + 1);
memcpy(pattern[patterns].pat, pat, len);
pattern[patterns].len = len;
pattern[patterns].pat[len] = '\0';
++patterns;
}
/*
* Adds a file include/exclude pattern to the internal array.
*/
static void
add_fpattern(const char *pat, int mode)
{
/* Increase size if necessary */
if (fpatterns == fpattern_sz) {
fpattern_sz *= 2;
fpattern = grep_realloc(fpattern, ++fpattern_sz *
sizeof(struct epat));
}
fpattern[fpatterns].pat = grep_strdup(pat);
fpattern[fpatterns].mode = mode;
++fpatterns;
}
/*
* Adds a directory include/exclude pattern to the internal array.
*/
static void
add_dpattern(const char *pat, int mode)
{
/* Increase size if necessary */
if (dpatterns == dpattern_sz) {
dpattern_sz *= 2;
dpattern = grep_realloc(dpattern, ++dpattern_sz *
sizeof(struct epat));
}
dpattern[dpatterns].pat = grep_strdup(pat);
dpattern[dpatterns].mode = mode;
++dpatterns;
}
/*
* Reads searching patterns from a file and adds them with add_pattern().
*/
static void
read_patterns(const char *fn)
{
struct stat st;
FILE *f;
char *line;
size_t len;
ssize_t rlen;
if (strcmp(fn, "-") == 0)
f = stdin;
else if ((f = fopen(fn, "r")) == NULL)
err(2, "%s", fn);
if ((fstat(fileno(f), &st) == -1) || (S_ISDIR(st.st_mode))) {
fclose(f);
return;
}
len = 0;
line = NULL;
while ((rlen = getline(&line, &len, f)) != -1) {
if (line[0] == '\0')
continue;
add_pattern(line, line[0] == '\n' ? 0 : (size_t)rlen);
}
free(line);
if (ferror(f))
err(2, "%s", fn);
if (strcmp(fn, "-") != 0)
fclose(f);
}
static inline const char *
init_color(const char *d)
{
char *c;
c = getenv("GREP_COLOR");
return (c != NULL && c[0] != '\0' ? c : d);
}
int
main(int argc, char *argv[])
{
char **aargv, **eargv, *eopts;
char *ep;
const char *pn;
long long l;
unsigned int aargc, eargc, i;
int c, lastc, needpattern, newarg, prevoptind;
bool matched;
setlocale(LC_ALL, "");
/*
* Check how we've bene invoked to determine the behavior we should
* exhibit. In this way we can have all the functionalities in one
* binary without the need of scripting and using ugly hacks.
*/
pn = getprogname();
switch (pn[0]) {
case 'e':
grepbehave = GREP_EXTENDED;
break;
case 'f':
grepbehave = GREP_FIXED;
break;
case 'r':
dirbehave = DIR_RECURSE;
Hflag = true;
break;
}
lastc = '\0';
newarg = 1;
prevoptind = 1;
needpattern = 1;
fileeol = '\n';
eopts = getenv("GREP_OPTIONS");
/* support for extra arguments in GREP_OPTIONS */
eargc = 0;
if (eopts != NULL && eopts[0] != '\0') {
char *str;
/* make an estimation of how many extra arguments we have */
for (unsigned int j = 0; j < strlen(eopts); j++)
if (eopts[j] == ' ')
eargc++;
eargv = (char **)grep_malloc(sizeof(char *) * (eargc + 1));
eargc = 0;
/* parse extra arguments */
while ((str = strsep(&eopts, " ")) != NULL)
if (str[0] != '\0')
eargv[eargc++] = grep_strdup(str);
aargv = (char **)grep_calloc(eargc + argc + 1,
sizeof(char *));
aargv[0] = argv[0];
for (i = 0; i < eargc; i++)
aargv[i + 1] = eargv[i];
for (int j = 1; j < argc; j++, i++)
aargv[i + 1] = argv[j];
aargc = eargc + argc;
} else {
aargv = argv;
aargc = argc;
}
while (((c = getopt_long(aargc, aargv, optstr, long_options, NULL)) !=
-1)) {
switch (c) {
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
if (newarg || !isdigit(lastc))
Aflag = 0;
else if (Aflag > LLONG_MAX / 10 - 1) {
errno = ERANGE;
err(2, NULL);
}
Aflag = Bflag = (Aflag * 10) + (c - '0');
break;
case 'C':
if (optarg == NULL) {
Aflag = Bflag = 2;
break;
}
/* FALLTHROUGH */
case 'A':
/* FALLTHROUGH */
case 'B':
errno = 0;
l = strtoll(optarg, &ep, 10);
if (errno == ERANGE || errno == EINVAL)
err(2, NULL);
else if (ep[0] != '\0') {
errno = EINVAL;
err(2, NULL);
} else if (l < 0) {
errno = EINVAL;
err(2, "context argument must be non-negative");
}
if (c == 'A')
Aflag = l;
else if (c == 'B')
Bflag = l;
else
Aflag = Bflag = l;
break;
case 'a':
binbehave = BINFILE_TEXT;
break;
case 'b':
bflag = true;
break;
case 'c':
cflag = true;
break;
case 'D':
if (strcasecmp(optarg, "skip") == 0)
devbehave = DEV_SKIP;
else if (strcasecmp(optarg, "read") == 0)
devbehave = DEV_READ;
else
errx(2, errstr[2], "--devices");
break;
case 'd':
if (strcasecmp("recurse", optarg) == 0) {
Hflag = true;
dirbehave = DIR_RECURSE;
} else if (strcasecmp("skip", optarg) == 0)
dirbehave = DIR_SKIP;
else if (strcasecmp("read", optarg) == 0)
dirbehave = DIR_READ;
else
errx(2, errstr[2], "--directories");
break;
case 'E':
grepbehave = GREP_EXTENDED;
break;
case 'e':
{
char *token;
char *string = optarg;
while ((token = strsep(&string, "\n")) != NULL)
add_pattern(token, strlen(token));
}
needpattern = 0;
break;
case 'F':
grepbehave = GREP_FIXED;
break;
case 'f':
read_patterns(optarg);
needpattern = 0;
break;
case 'G':
grepbehave = GREP_BASIC;
break;
case 'H':
Hflag = true;
break;
case 'h':
Hflag = false;
hflag = true;
break;
case 'I':
binbehave = BINFILE_SKIP;
break;
case 'i':
case 'y':
iflag = true;
cflags |= REG_ICASE;
break;
case 'L':
lflag = false;
Lflag = true;
break;
case 'l':
Lflag = false;
lflag = true;
break;
case 'm':
mflag = true;
errno = 0;
mlimit = mcount = strtoll(optarg, &ep, 10);
if (((errno == ERANGE) && (mcount == LLONG_MAX)) ||
((errno == EINVAL) && (mcount == 0)))
err(2, NULL);
else if (ep[0] != '\0') {
errno = EINVAL;
err(2, NULL);
}
break;
case 'n':
nflag = true;
break;
case 'O':
linkbehave = LINK_EXPLICIT;
break;
case 'o':
oflag = true;
cflags &= ~REG_NOSUB;
break;
case 'p':
linkbehave = LINK_SKIP;
break;
case 'q':
qflag = true;
break;
case 'S':
linkbehave = LINK_READ;
break;
case 'R':
case 'r':
dirbehave = DIR_RECURSE;
Hflag = true;
break;
case 's':
sflag = true;
break;
case 'U':
binbehave = BINFILE_BIN;
break;
case 'u':
case MMAP_OPT:
filebehave = FILE_MMAP;
break;
case 'V':
printf(errstr[8], getprogname(), VERSION);
exit(0);
case 'v':
vflag = true;
break;
case 'w':
wflag = true;
cflags &= ~REG_NOSUB;
break;
case 'x':
xflag = true;
cflags &= ~REG_NOSUB;
break;
case 'z':
fileeol = '\0';
cflags &= ~REG_NEWLINE;
break;
case BIN_OPT:
if (strcasecmp("binary", optarg) == 0)
binbehave = BINFILE_BIN;
else if (strcasecmp("without-match", optarg) == 0)
binbehave = BINFILE_SKIP;
else if (strcasecmp("text", optarg) == 0)
binbehave = BINFILE_TEXT;
else
errx(2, errstr[2], "--binary-files");
break;
case COLOR_OPT:
color = NULL;
if (optarg == NULL || strcasecmp("auto", optarg) == 0 ||
strcasecmp("tty", optarg) == 0 ||
strcasecmp("if-tty", optarg) == 0) {
char *term;
term = getenv("TERM");
if (isatty(STDOUT_FILENO) && term != NULL &&
strcasecmp(term, "dumb") != 0)
color = init_color("01;31");
} else if (strcasecmp("always", optarg) == 0 ||
strcasecmp("yes", optarg) == 0 ||
strcasecmp("force", optarg) == 0) {
color = init_color("01;31");
} else if (strcasecmp("never", optarg) != 0 &&
strcasecmp("none", optarg) != 0 &&
strcasecmp("no", optarg) != 0)
errx(2, errstr[2], "--color");
cflags &= ~REG_NOSUB;
break;
case LABEL_OPT:
label = optarg;
break;
case LINEBUF_OPT:
lbflag = true;
break;
case NULL_OPT:
nullflag = true;
break;
case R_INCLUDE_OPT:
finclude = true;
add_fpattern(optarg, INCL_PAT);
break;
case R_EXCLUDE_OPT:
fexclude = true;
add_fpattern(optarg, EXCL_PAT);
break;
case R_DINCLUDE_OPT:
dinclude = true;
add_dpattern(optarg, INCL_PAT);
break;
case R_DEXCLUDE_OPT:
dexclude = true;
add_dpattern(optarg, EXCL_PAT);
break;
case HELP_OPT:
default:
usage();
}
lastc = c;
newarg = optind != prevoptind;
prevoptind = optind;
}
aargc -= optind;
aargv += optind;
/* xflag takes precedence, don't confuse the matching bits. */
if (wflag && xflag)
wflag = false;
/* Fail if we don't have any pattern */
if (aargc == 0 && needpattern)
usage();
/* Process patterns from command line */
if (aargc != 0 && needpattern) {
char *token;
char *string = *aargv;
while ((token = strsep(&string, "\n")) != NULL)
add_pattern(token, strlen(token));
--aargc;
++aargv;
}
switch (grepbehave) {
case GREP_BASIC:
break;
case GREP_FIXED:
/*
* regex(3) implementations that support fixed-string searches generally
* define either REG_NOSPEC or REG_LITERAL. Set the appropriate flag
* here. If neither are defined, GREP_FIXED later implies that the
* internal literal matcher should be used. Other cflags that have
* the same interpretation as REG_NOSPEC and REG_LITERAL should be
* similarly added here, and grep.h should be amended to take this into
* consideration when defining WITH_INTERNAL_NOSPEC.
*/
#if defined(REG_NOSPEC)
cflags |= REG_NOSPEC;
#elif defined(REG_LITERAL)
cflags |= REG_LITERAL;
#endif
break;
case GREP_EXTENDED:
cflags |= REG_EXTENDED;
break;
default:
/* NOTREACHED */
usage();
}
r_pattern = grep_calloc(patterns, sizeof(*r_pattern));
#ifdef WITH_INTERNAL_NOSPEC
if (grepbehave != GREP_FIXED) {
#else
{
#endif
/* Check if cheating is allowed (always is for fgrep). */
for (i = 0; i < patterns; ++i) {
c = regcomp(&r_pattern[i], pattern[i].pat, cflags);
if (c != 0) {
regerror(c, &r_pattern[i], re_error,
RE_ERROR_BUF);
errx(2, "%s", re_error);
}
}
}
if (lbflag)
setlinebuf(stdout);
if ((aargc == 0 || aargc == 1) && !Hflag)
hflag = true;
initqueue();
if (aargc == 0 && dirbehave != DIR_RECURSE)
exit(!procfile("-"));
if (dirbehave == DIR_RECURSE)
matched = grep_tree(aargv);
else
for (matched = false; aargc--; ++aargv) {
if ((finclude || fexclude) && !file_matching(*aargv))
continue;
if (procfile(*aargv))
matched = true;
}
if (Lflag)
matched = !matched;
/*
* Calculate the correct return value according to the
* results and the command line option.
*/
exit(matched ? (file_err ? (qflag ? 0 : 2) : 0) : (file_err ? 2 : 1));
}
grep.h
+159
View File
@@ -0,0 +1,159 @@
/* $NetBSD: grep.h,v 1.5 2011/02/27 17:33:37 joerg Exp $ */
/* $OpenBSD: grep.h,v 1.15 2010/04/05 03:03:55 tedu Exp $ */
/*-
* SPDX-License-Identifier: BSD-2-Clause
*
* Copyright (c) 1999 James Howard and Dag-Erling Smørgrav
* Copyright (c) 2008-2009 Gabor Kovesdan <[email protected]>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <bzlib.h>
#include <limits.h>
#include <regex.h>
#include <stdbool.h>
#include <stdio.h>
#include <zlib.h>
extern const char *errstr[];
#define VERSION "2.6.0-FreeBSD"
#define GREP_FIXED 0
#define GREP_BASIC 1
#define GREP_EXTENDED 2
#if !defined(REG_NOSPEC) && !defined(REG_LITERAL)
#define WITH_INTERNAL_NOSPEC
#endif
#define BINFILE_BIN 0
#define BINFILE_SKIP 1
#define BINFILE_TEXT 2
#define FILE_STDIO 0
#define FILE_MMAP 1
#define DIR_READ 0
#define DIR_SKIP 1
#define DIR_RECURSE 2
#define DEV_READ 0
#define DEV_SKIP 1
#define LINK_READ 0
#define LINK_EXPLICIT 1
#define LINK_SKIP 2
#define EXCL_PAT 0
#define INCL_PAT 1
#define MAX_MATCHES 32
struct file {
int fd;
bool binary;
};
struct str {
off_t boff;
off_t off;
size_t len;
char *dat;
char *file;
int line_no;
};
struct pat {
char *pat;
int len;
};
struct epat {
char *pat;
int mode;
};
/*
* Parsing context; used to hold things like matches made and
* other useful bits
*/
struct parsec {
regmatch_t matches[MAX_MATCHES]; /* Matches made */
/* XXX TODO: This should be a chunk, not a line */
struct str ln; /* Current line */
size_t lnstart; /* Position in line */
size_t matchidx; /* Latest match index */
int printed; /* Metadata printed? */
bool binary; /* Binary file? */
bool cntlines; /* Count lines? */
};
/* Flags passed to regcomp() and regexec() */
extern int cflags, eflags;
/* Command line flags */
extern bool Eflag, Fflag, Gflag, Hflag, Lflag,
bflag, cflag, hflag, iflag, lflag, mflag, nflag, oflag,
qflag, sflag, vflag, wflag, xflag;
extern bool dexclude, dinclude, fexclude, finclude, lbflag, nullflag;
extern long long Aflag, Bflag;
extern long long mcount;
extern long long mlimit;
extern char fileeol;
extern char *label;
extern const char *color;
extern int binbehave, devbehave, dirbehave, filebehave, grepbehave, linkbehave;
extern bool file_err, matchall;
extern unsigned int dpatterns, fpatterns, patterns;
extern struct pat *pattern;
extern struct epat *dpattern, *fpattern;
extern regex_t *er_pattern, *r_pattern;
/* For regex errors */
#define RE_ERROR_BUF 512
extern char re_error[RE_ERROR_BUF + 1]; /* Seems big enough */
/* util.c */
bool file_matching(const char *fname);
bool procfile(const char *fn);
bool grep_tree(char **argv);
void *grep_malloc(size_t size);
void *grep_calloc(size_t nmemb, size_t size);
void *grep_realloc(void *ptr, size_t size);
char *grep_strdup(const char *str);
void grep_printline(struct str *line, int sep);
/* queue.c */
void initqueue(void);
bool enqueue(struct str *x);
void printqueue(void);
void clearqueue(void);
/* file.c */
void grep_close(struct file *f);
struct file *grep_open(const char *path);
char *grep_fgetln(struct file *f, struct parsec *pc);
queue.c
+142
View File
@@ -0,0 +1,142 @@
/* $NetBSD: queue.c,v 1.5 2011/08/31 16:24:57 plunky Exp $ */
/*-
* SPDX-License-Identifier: BSD-2-Clause
*
* Copyright (c) 1999 James Howard and Dag-Erling Smørgrav
* All rights reserved.
* Copyright (c) 2020 Kyle Evans <[email protected]>
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
/*
* A really poor man's queue. It does only what it has to and gets out of
* Dodge. It is used in place of <sys/queue.h> to get a better performance.
*/
#include <sys/param.h>
#include <sys/queue.h>
#include <stdlib.h>
#include <string.h>
#include "grep.h"
typedef struct str qentry_t;
static long long filled;
static qentry_t *qend, *qpool;
/*
* qnext is the next entry to populate. qlist is where the list actually
* starts, for the purposes of printing.
*/
static qentry_t *qlist, *qnext;
void
initqueue(void)
{
qlist = qnext = qpool = grep_calloc(Bflag, sizeof(*qpool));
qend = qpool + (Bflag - 1);
}
static qentry_t *
advqueue(qentry_t *itemp)
{
if (itemp == qend)
return (qpool);
return (itemp + 1);
}
/*
* Enqueue another line; return true if we've dequeued a line as a result
*/
bool
enqueue(struct str *x)
{
qentry_t *item;
bool rotated;
item = qnext;
qnext = advqueue(qnext);
rotated = false;
if (filled < Bflag) {
filled++;
} else if (filled == Bflag) {
/* We had already filled up coming in; just rotate. */
qlist = advqueue(qlist);
rotated = true;
free(item->dat);
}
/* len + 1 for NUL-terminator */
item->dat = grep_malloc(sizeof(char) * x->len + 1);
item->len = x->len;
item->line_no = x->line_no;
item->boff = x->boff;
item->off = x->off;
memcpy(item->dat, x->dat, x->len);
item->dat[x->len] = '\0';
item->file = x->file;
return (rotated);
}
void
printqueue(void)
{
qentry_t *item;
item = qlist;
do {
/* Buffer must have ended early. */
if (item->dat == NULL)
break;
grep_printline(item, '-');
free(item->dat);
item->dat = NULL;
item = advqueue(item);
} while (item != qlist);
qlist = qnext = qpool;
filled = 0;
}
void
clearqueue(void)
{
qentry_t *item;
item = qlist;
do {
free(item->dat);
item->dat = NULL;
item = advqueue(item);
} while (item != qlist);
qlist = qnext = qpool;
filled = 0;
}
util.c
+856
View File
@@ -0,0 +1,856 @@
/* $NetBSD: util.c,v 1.9 2011/02/27 17:33:37 joerg Exp $ */
/* $OpenBSD: util.c,v 1.39 2010/07/02 22:18:03 tedu Exp $ */
/*-
* SPDX-License-Identifier: BSD-2-Clause
*
* Copyright (c) 1999 James Howard and Dag-Erling Smørgrav
* Copyright (C) 2008-2010 Gabor Kovesdan <[email protected]>
* Copyright (C) 2017 Kyle Evans <[email protected]>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <ctype.h>
#include <err.h>
#include <errno.h>
#include <fnmatch.h>
#include <fts.h>
#include <libgen.h>
#include <stdbool.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <wchar.h>
#include <wctype.h>
#include "grep.h"
static bool first_match = true;
/*
* Match printing context
*/
struct mprintc {
long long tail; /* Number of trailing lines to record */
int last_outed; /* Number of lines since last output */
bool doctx; /* Printing context? */
bool printmatch; /* Printing matches? */
bool same_file; /* Same file as previously printed? */
};
static void procmatch_match(struct mprintc *mc, struct parsec *pc);
static void procmatch_nomatch(struct mprintc *mc, struct parsec *pc);
static bool procmatches(struct mprintc *mc, struct parsec *pc, bool matched);
#ifdef WITH_INTERNAL_NOSPEC
static int litexec(const struct pat *pat, const char *string,
size_t nmatch, regmatch_t pmatch[]);
#endif
static bool procline(struct parsec *pc);
static bool printline(struct parsec *pc, int sep, size_t *last_out);
static void printline_metadata(struct str *line, int sep);
bool
file_matching(const char *fname)
{
char *fname_base, *fname_buf;
bool ret;
ret = finclude ? false : true;
fname_buf = strdup(fname);
if (fname_buf == NULL)
err(2, "strdup");
fname_base = basename(fname_buf);
for (unsigned int i = 0; i < fpatterns; ++i) {
if (fnmatch(fpattern[i].pat, fname, 0) == 0 ||
fnmatch(fpattern[i].pat, fname_base, 0) == 0)
/*
* The last pattern matched wins exclusion/inclusion
* rights, so we can't reasonably bail out early here.
*/
ret = (fpattern[i].mode != EXCL_PAT);
}
free(fname_buf);
return (ret);
}
static inline bool
dir_matching(const char *dname)
{
bool ret;
ret = dinclude ? false : true;
for (unsigned int i = 0; i < dpatterns; ++i) {
if (dname != NULL && fnmatch(dpattern[i].pat, dname, 0) == 0)
/*
* The last pattern matched wins exclusion/inclusion
* rights, so we can't reasonably bail out early here.
*/
ret = (dpattern[i].mode != EXCL_PAT);
}
return (ret);
}
/*
* Processes a directory when a recursive search is performed with
* the -R option. Each appropriate file is passed to procfile().
*/
bool
grep_tree(char **argv)
{
FTS *fts;
FTSENT *p;
int fts_flags;
bool matched, ok;
const char *wd[] = { ".", NULL };
matched = false;
/* This switch effectively initializes 'fts_flags' */
switch(linkbehave) {
case LINK_EXPLICIT:
fts_flags = FTS_COMFOLLOW | FTS_PHYSICAL;
break;
case LINK_SKIP:
fts_flags = FTS_PHYSICAL;
break;
default:
fts_flags = FTS_LOGICAL | FTS_NOSTAT;
}
fts_flags |= FTS_NOCHDIR;
fts = fts_open((argv[0] == NULL) ?
__DECONST(char * const *, wd) : argv, fts_flags, NULL);
if (fts == NULL)
err(2, "fts_open");
while (errno = 0, (p = fts_read(fts)) != NULL) {
switch (p->fts_info) {
case FTS_DNR:
case FTS_ERR:
case FTS_NS:
file_err = true;
if(!sflag)
warnc(p->fts_errno, "%s", p->fts_path);
break;
case FTS_D:
if (dexclude || dinclude)
if (!dir_matching(p->fts_name) ||
!dir_matching(p->fts_path))
fts_set(fts, p, FTS_SKIP);
break;
case FTS_DC:
/* Print a warning for recursive directory loop */
warnx("warning: %s: recursive directory loop",
p->fts_path);
break;
case FTS_DP:
break;
case FTS_SL:
/*
* Skip symlinks for LINK_EXPLICIT and
* LINK_SKIP. Note that due to FTS_COMFOLLOW,
* symlinks on the command line are followed
* for LINK_EXPLICIT and not reported as
* symlinks.
*/
break;
default:
/* Check for file exclusion/inclusion */
ok = true;
if (fexclude || finclude)
ok &= file_matching(p->fts_path);
if (ok && procfile(p->fts_path))
matched = true;
break;
}
}
if (errno != 0)
err(2, "fts_read");
fts_close(fts);
return (matched);
}
static void
procmatch_match(struct mprintc *mc, struct parsec *pc)
{
if (mc->doctx) {
if (!first_match && (!mc->same_file || mc->last_outed > 0))
printf("--\n");
if (Bflag > 0)
printqueue();
mc->tail = Aflag;
}
/* Print the matching line, but only if not quiet/binary */
if (mc->printmatch) {
size_t last_out;
bool terminated;
last_out = 0;
terminated = printline(pc, ':', &last_out);
while (pc->matchidx >= MAX_MATCHES) {
/* Reset matchidx and try again */
pc->matchidx = 0;
if (procline(pc) == !vflag)
terminated = printline(pc, ':', &last_out);
else
break;
}
/*
* The above loop processes the entire line as long as we keep
* hitting the maximum match count. At this point, we know
* that there's nothing left to be printed and can terminate the
* line.
*/
if (!terminated)
printline(pc, ':', &last_out);
first_match = false;
mc->same_file = true;
mc->last_outed = 0;
}
}
static void
procmatch_nomatch(struct mprintc *mc, struct parsec *pc)
{
/* Deal with any -A context as needed */
if (mc->tail > 0) {
grep_printline(&pc->ln, '-');
mc->tail--;
if (Bflag > 0)
clearqueue();
} else if (Bflag == 0 || (Bflag > 0 && enqueue(&pc->ln)))
/*
* Enqueue non-matching lines for -B context. If we're not
* actually doing -B context or if the enqueue resulted in a
* line being rotated out, then go ahead and increment
* last_outed to signify a gap between context/match.
*/
++mc->last_outed;
}
/*
* Process any matches in the current parsing context, return a boolean
* indicating whether we should halt any further processing or not. 'true' to
* continue processing, 'false' to halt.
*/
static bool
procmatches(struct mprintc *mc, struct parsec *pc, bool matched)
{
if (mflag && mcount <= 0) {
/*
* We already hit our match count, but we need to keep dumping
* lines until we've lost our tail.
*/
grep_printline(&pc->ln, '-');
mc->tail--;
return (mc->tail != 0);
}
/*
* XXX TODO: This should loop over pc->matches and handle things on a
* line-by-line basis, setting up a `struct str` as needed.
*/
/* Deal with any -B context or context separators */
if (matched) {
procmatch_match(mc, pc);
/* Count the matches if we have a match limit */
if (mflag) {
/* XXX TODO: Decrement by number of matched lines */
mcount -= 1;
if (mcount <= 0)
return (mc->tail != 0);
}
} else if (mc->doctx)
procmatch_nomatch(mc, pc);
return (true);
}
/*
* Opens a file and processes it. Each file is processed line-by-line
* passing the lines to procline().
*/
bool
procfile(const char *fn)
{
struct parsec pc;
struct mprintc mc;
struct file *f;
struct stat sb;
mode_t s;
int lines;
bool line_matched;
if (strcmp(fn, "-") == 0) {
fn = label != NULL ? label : errstr[1];
f = grep_open(NULL);
} else {
if (stat(fn, &sb) == 0) {
/* Check if we need to process the file */
s = sb.st_mode & S_IFMT;
if (dirbehave == DIR_SKIP && s == S_IFDIR)
return (false);
if (devbehave == DEV_SKIP && (s == S_IFIFO ||
s == S_IFCHR || s == S_IFBLK || s == S_IFSOCK))
return (false);
}
f = grep_open(fn);
}
if (f == NULL) {
file_err = true;
if (!sflag)
warn("%s", fn);
return (false);
}
pc.ln.file = grep_strdup(fn);
pc.ln.line_no = 0;
pc.ln.len = 0;
pc.ln.boff = 0;
pc.ln.off = -1;
pc.binary = f->binary;
pc.cntlines = false;
memset(&mc, 0, sizeof(mc));
mc.printmatch = true;
if ((pc.binary && binbehave == BINFILE_BIN) || cflag || qflag ||
lflag || Lflag)
mc.printmatch = false;
if (mc.printmatch && (Aflag != 0 || Bflag != 0))
mc.doctx = true;
if (mc.printmatch && (Aflag != 0 || Bflag != 0 || mflag || nflag))
pc.cntlines = true;
mcount = mlimit;
for (lines = 0; lines == 0 || !(lflag || qflag); ) {
/*
* XXX TODO: We need to revisit this in a chunking world. We're
* not going to be doing per-line statistics because of the
* overhead involved. procmatches can figure that stuff out as
* needed. */
/* Reset per-line statistics */
pc.printed = 0;
pc.matchidx = 0;
pc.lnstart = 0;
pc.ln.boff = 0;
pc.ln.off += pc.ln.len + 1;
/* XXX TODO: Grab a chunk */
if ((pc.ln.dat = grep_fgetln(f, &pc)) == NULL ||
pc.ln.len == 0)
break;
if (pc.ln.len > 0 && pc.ln.dat[pc.ln.len - 1] == fileeol)
--pc.ln.len;
pc.ln.line_no++;
/* Return if we need to skip a binary file */
if (pc.binary && binbehave == BINFILE_SKIP) {
grep_close(f);
free(pc.ln.file);
free(f);
return (0);
}
if (mflag && mcount <= 0) {
/*
* Short-circuit, already hit match count and now we're
* just picking up any remaining pieces.
*/
if (!procmatches(&mc, &pc, false))
break;
continue;
}
line_matched = procline(&pc) == !vflag;
if (line_matched)
++lines;
/* Halt processing if we hit our match limit */
if (!procmatches(&mc, &pc, line_matched))
break;
}
if (Bflag > 0)
clearqueue();
grep_close(f);
if (cflag && !qflag) {
if (!hflag)
printf("%s:", pc.ln.file);
printf("%u\n", lines);
}
if (lflag && !qflag && lines != 0)
printf("%s%c", fn, nullflag ? 0 : '\n');
if (Lflag && !qflag && lines == 0)
printf("%s%c", fn, nullflag ? 0 : '\n');
if (lines != 0 && !cflag && !lflag && !Lflag &&
binbehave == BINFILE_BIN && f->binary && !qflag)
printf(errstr[7], fn);
free(pc.ln.file);
free(f);
return (lines != 0);
}
#ifdef WITH_INTERNAL_NOSPEC
/*
* Internal implementation of literal string search within a string, modeled
* after regexec(3), for use when the regex(3) implementation doesn't offer
* either REG_NOSPEC or REG_LITERAL. This does not apply in the default FreeBSD
* config, but in other scenarios such as building against libgnuregex or on
* some non-FreeBSD OSes.
*/
static int
litexec(const struct pat *pat, const char *string, size_t nmatch,
regmatch_t pmatch[])
{
char *(*strstr_fn)(const char *, const char *);
char *sub, *subject;
const char *search;
size_t idx, n, ofs, stringlen;
if (cflags & REG_ICASE)
strstr_fn = strcasestr;
else
strstr_fn = strstr;
idx = 0;
ofs = pmatch[0].rm_so;
stringlen = pmatch[0].rm_eo;
if (ofs >= stringlen)
return (REG_NOMATCH);
subject = strndup(string, stringlen);
if (subject == NULL)
return (REG_ESPACE);
for (n = 0; ofs < stringlen;) {
search = (subject + ofs);
if ((unsigned long)pat->len > strlen(search))
break;
sub = strstr_fn(search, pat->pat);
/*
* Ignoring the empty string possibility due to context: grep optimizes
* for empty patterns and will never reach this point.
*/
if (sub == NULL)
break;
++n;
/* Fill in pmatch if necessary */
if (nmatch > 0) {
pmatch[idx].rm_so = ofs + (sub - search);
pmatch[idx].rm_eo = pmatch[idx].rm_so + pat->len;
if (++idx == nmatch)
break;
ofs = pmatch[idx].rm_so + 1;
} else
/* We only needed to know if we match or not */
break;
}
free(subject);
if (n > 0 && nmatch > 0)
for (n = idx; n < nmatch; ++n)
pmatch[n].rm_so = pmatch[n].rm_eo = -1;
return (n > 0 ? 0 : REG_NOMATCH);
}
#endif /* WITH_INTERNAL_NOSPEC */
#define iswword(x) (iswalnum((x)) || (x) == L'_')
/*
* Processes a line comparing it with the specified patterns. Each pattern
* is looped to be compared along with the full string, saving each and every
* match, which is necessary to colorize the output and to count the
* matches. The matching lines are passed to printline() to display the
* appropriate output.
*/
static bool
procline(struct parsec *pc)
{
regmatch_t pmatch, lastmatch, chkmatch;
wchar_t wbegin, wend;
size_t st, nst;
unsigned int i;
int r = 0, leflags = eflags;
size_t startm = 0, matchidx;
unsigned int retry;
bool lastmatched, matched;
matchidx = pc->matchidx;
/* Null pattern shortcuts. */
if (matchall) {
if (xflag && pc->ln.len == 0) {
/* Matches empty lines (-x). */
return (true);
} else if (!wflag && !xflag) {
/* Matches every line (no -w or -x). */
return (true);
}
/*
* If we only have the NULL pattern, whether we match or not
* depends on if we got here with -w or -x. If either is set,
* the answer is no. If we have other patterns, we'll defer
* to them.
*/
if (patterns == 0) {
return (!(wflag || xflag));
}
} else if (patterns == 0) {
/* Pattern file with no patterns. */
return (false);
}
matched = false;
st = pc->lnstart;
nst = 0;
/* Initialize to avoid a false positive warning from GCC. */
lastmatch.rm_so = lastmatch.rm_eo = 0;
/* Loop to process the whole line */
while (st <= pc->ln.len) {
lastmatched = false;
startm = matchidx;
retry = 0;
if (st > 0 && pc->ln.dat[st - 1] != fileeol)
leflags |= REG_NOTBOL;
/* Loop to compare with all the patterns */
for (i = 0; i < patterns; i++) {
pmatch.rm_so = st;
pmatch.rm_eo = pc->ln.len;
#ifdef WITH_INTERNAL_NOSPEC
if (grepbehave == GREP_FIXED)
r = litexec(&pattern[i], pc->ln.dat, 1, &pmatch);
else
#endif
r = regexec(&r_pattern[i], pc->ln.dat, 1, &pmatch,
leflags);
if (r != 0)
continue;
/* Check for full match */
if (xflag && (pmatch.rm_so != 0 ||
(size_t)pmatch.rm_eo != pc->ln.len))
continue;
/* Check for whole word match */
if (wflag) {
wbegin = wend = L' ';
if (pmatch.rm_so != 0 &&
sscanf(&pc->ln.dat[pmatch.rm_so - 1],
"%lc", &wbegin) != 1)
r = REG_NOMATCH;
else if ((size_t)pmatch.rm_eo !=
pc->ln.len &&
sscanf(&pc->ln.dat[pmatch.rm_eo],
"%lc", &wend) != 1)
r = REG_NOMATCH;
else if (iswword(wbegin) ||
iswword(wend))
r = REG_NOMATCH;
/*
* If we're doing whole word matching and we
* matched once, then we should try the pattern
* again after advancing just past the start of
* the earliest match. This allows the pattern
* to match later on in the line and possibly
* still match a whole word.
*/
if (r == REG_NOMATCH &&
(retry == pc->lnstart ||
(unsigned int)pmatch.rm_so + 1 < retry))
retry = pmatch.rm_so + 1;
if (r == REG_NOMATCH)
continue;
}
lastmatched = true;
lastmatch = pmatch;
if (matchidx == 0)
matched = true;
/*
* Replace previous match if the new one is earlier
* and/or longer. This will lead to some amount of
* extra work if -o/--color are specified, but it's
* worth it from a correctness point of view.
*/
if (matchidx > startm) {
chkmatch = pc->matches[matchidx - 1];
if (pmatch.rm_so < chkmatch.rm_so ||
(pmatch.rm_so == chkmatch.rm_so &&
(pmatch.rm_eo - pmatch.rm_so) >
(chkmatch.rm_eo - chkmatch.rm_so))) {
pc->matches[matchidx - 1] = pmatch;
nst = pmatch.rm_eo;
}
} else {
/* Advance as normal if not */
pc->matches[matchidx++] = pmatch;
nst = pmatch.rm_eo;
}
/* avoid excessive matching - skip further patterns */
if ((color == NULL && !oflag) || qflag || lflag ||
matchidx >= MAX_MATCHES) {
pc->lnstart = nst;
lastmatched = false;
break;
}
}
/*
* Advance to just past the start of the earliest match, try
* again just in case we still have a chance to match later in
* the string.
*/
if (!lastmatched && retry > pc->lnstart) {
st = retry;
continue;
}
/* XXX TODO: We will need to keep going, since we're chunky */
/* One pass if we are not recording matches */
if (!wflag && ((color == NULL && !oflag) || qflag || lflag || Lflag))
break;
/* If we didn't have any matches or REG_NOSUB set */
if (!lastmatched || (cflags & REG_NOSUB))
nst = pc->ln.len;
if (!lastmatched)
/* No matches */
break;
else if (st == nst && lastmatch.rm_so == lastmatch.rm_eo)
/* Zero-length match -- advance one more so we don't get stuck */
nst++;
/* Advance st based on previous matches */
st = nst;
pc->lnstart = st;
}
/* Reflect the new matchidx in the context */
pc->matchidx = matchidx;
return matched;
}
/*
* Safe malloc() for internal use.
*/
void *
grep_malloc(size_t size)
{
void *ptr;
if (size == 0)
return (NULL);
if ((ptr = malloc(size)) == NULL)
err(2, "malloc");
return (ptr);
}
/*
* Safe calloc() for internal use.
*/
void *
grep_calloc(size_t nmemb, size_t size)
{
void *ptr;
if (nmemb == 0 || size == 0)
return (NULL);
if ((ptr = calloc(nmemb, size)) == NULL)
err(2, "calloc");
return (ptr);
}
/*
* Safe realloc() for internal use.
*/
void *
grep_realloc(void *ptr, size_t size)
{
if ((ptr = realloc(ptr, size)) == NULL)
err(2, "realloc");
return (ptr);
}
/*
* Safe strdup() for internal use.
*/
char *
grep_strdup(const char *str)
{
char *ret;
if ((ret = strdup(str)) == NULL)
err(2, "strdup");
return (ret);
}
/*
* Print an entire line as-is, there are no inline matches to consider. This is
* used for printing context.
*/
void grep_printline(struct str *line, int sep) {
printline_metadata(line, sep);
fwrite(line->dat, line->len, 1, stdout);
putchar(fileeol);
fflush(stdout);
}
static void
printline_metadata(struct str *line, int sep)
{
bool printsep;
printsep = false;
if (!hflag) {
if (!nullflag) {
fputs(line->file, stdout);
printsep = true;
} else {
printf("%s", line->file);
putchar(0);
}
}
if (nflag) {
if (printsep)
putchar(sep);
printf("%d", line->line_no);
printsep = true;
}
if (bflag) {
if (printsep)
putchar(sep);
printf("%lld", (long long)(line->off + line->boff));
printsep = true;
}
if (printsep)
putchar(sep);
}
/*
* Prints a matching line according to the command line options. We need
* *last_out to be populated on entry in case this is just a continuation of
* matches within the same line.
*
* Returns true if the line was terminated, false if it was not.
*/
static bool
printline(struct parsec *pc, int sep, size_t *last_out)
{
size_t a = *last_out;
size_t i, matchidx;
regmatch_t match;
bool terminated;
/*
* Nearly all paths below will terminate the line by default, but it is
* avoided in some circumstances in case we don't have the full context
* available here.
*/
terminated = true;
/* If matchall, everything matches but don't actually print for -o */
if (oflag && matchall)
return (terminated);
matchidx = pc->matchidx;
/* --color and -o */
if ((oflag || color) && (pc->printed > 0 || matchidx > 0)) {
/* Only print metadata once per line if --color */
if (!oflag && pc->printed == 0) {
printline_metadata(&pc->ln, sep);
}
for (i = 0; i < matchidx; i++) {
match = pc->matches[i];
/* Don't output zero length matches */
if (match.rm_so == match.rm_eo)
continue;
/*
* Metadata is printed on a per-line basis, so every
* match gets file metadata with the -o flag.
*/
if (oflag) {
pc->ln.boff = match.rm_so;
printline_metadata(&pc->ln, sep);
} else {
fwrite(pc->ln.dat + a, match.rm_so - a, 1,
stdout);
}
if (color)
fprintf(stdout, "\33[%sm\33[K", color);
fwrite(pc->ln.dat + match.rm_so,
match.rm_eo - match.rm_so, 1, stdout);
if (color)
fprintf(stdout, "\33[m\33[K");
a = match.rm_eo;
if (oflag)
putchar('\n');
}
/*
* Don't terminate if we reached the match limit; we may have
* other matches on this line to process.
*/
*last_out = a;
if (!oflag && matchidx != MAX_MATCHES) {
if (pc->ln.len - a > 0) {
fwrite(pc->ln.dat + a, pc->ln.len - a, 1,
stdout);
*last_out = pc->ln.len;
}
putchar('\n');
fflush(stdout);
} else if (!oflag) {
/*
* -o is terminated on every match output, so this
* branch is only designed to capture MAX_MATCHES in a
* line which may be a signal to us for a lack of
* context. The caller will know more and call us again
* to terminate if it needs to.
*/
terminated = false;
} else {
fflush(stdout);
}
} else
grep_printline(&pc->ln, sep);
pc->printed++;
return (terminated);
}
zgrep.1
+94
View File
@@ -0,0 +1,94 @@
.\" Copyright (c) 2018 Baptiste Daroussin <[email protected]>
.\" All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.Dd July 20, 2020
.Dt ZGREP 1
.Os
.Sh NAME
.Nm zgrep ,
.Nm zegrep ,
.Nm zfgrep ,
.Nm bzgrep ,
.Nm bzegrep ,
.Nm bzfgrep ,
.Nm lzgrep ,
.Nm lzegrep ,
.Nm lzfgrep ,
.Nm xzgrep ,
.Nm xzegrep ,
.Nm xzfgrep ,
.Nm zstdgrep ,
.Nm zstdegrep ,
.Nm zstdfgrep
.Nd grep compressed files
.Sh SYNOPSIS
.Nm
.Oo Ar flags Oc Ar files
.Nm zegrep
.Oo Ar flags Oc Ar files
.Nm zfgrep
.Oo Ar flags Oc Ar files
.Pp
.Nm bzgrep
.Oo Ar flags Oc Ar files
.Nm bzegrep
.Oo Ar flags Oc Ar files
.Nm bzfgrep
.Oo Ar flags Oc Ar files
.Pp
.Nm lzgrep
.Oo Ar flags Oc Ar files
.Nm lzegrep
.Oo Ar flags Oc Ar files
.Nm lzfgrep
.Oo Ar flags Oc Ar files
.Pp
.Nm xzgrep
.Oo Ar flags Oc Ar files
.Nm xzegrep
.Oo Ar flags Oc Ar files
.Nm xzfgrep
.Oo Ar flags Oc Ar files
.Pp
.Nm zstdgrep
.Oo Ar flags Oc Ar files
.Nm zstdegrep
.Oo Ar flags Oc Ar files
.Nm zstdfgrep
.Oo Ar flags Oc Ar files
.Sh DESCRIPTION
Allow
.Xr grep 1
to read compressed files.
.Sh SEE ALSO
.Xr bzip2 1 ,
.Xr grep 1 ,
.Xr gzip 1 ,
.Xr xz 1 ,
.Xr zstd 1
.Sh AUTHORS
This version of the
.Nm
utility was written by
.An Thomas Klausner Aq Mt [email protected] .
zgrep.sh
Executable
+230
View File
@@ -0,0 +1,230 @@
#!/bin/sh
#
# Copyright (c) 2003 Thomas Klausner.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
# 1. Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# 2. Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution.
#
# THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
# IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
# OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
# IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
# INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
# NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
# THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#
set -u
grep=grep
zcat=zstdcat
endofopts=0
pattern_file=0
pattern_found=0
grep_args=""
hyphen=0
silent=0
prg=${0##*/}
# handle being called 'zegrep' or 'zfgrep'
case ${prg} in
*egrep)
grep_args="-E";;
*fgrep)
grep_args="-F";;
esac
catargs="-f"
case ${prg} in
zstd*)
cattool="/usr/bin/zstdcat"
catargs="-fq"
;;
bz*)
cattool="/usr/bin/bzcat"
;;
z*)
cattool="/usr/bin/zcat"
;;
xz*)
cattool="/usr/bin/xzcat"
;;
lz*)
cattool="/usr/bin/lzcat"
;;
*)
echo "Invalid command: ${prg}" >&2
exit 1
;;
esac
# skip all options and pass them on to grep taking care of options
# with arguments, and if -e was supplied
while [ $# -gt 0 -a ${endofopts} -eq 0 ]
do
case $1 in
# from GNU grep-2.6.0 -- keep in sync!
--)
shift
endofopts=1
;;
--file=*)
pattern_file=1
grep_args="${grep_args} ${1}"
shift
;;
--regexp=*)
if [ ${pattern_found} -ne 0 ]; then
grep_args="${grep_args} -e ${pattern}"
fi
pattern="${1#--regexp=}"
pattern_found=1
shift
;;
-h|--no-filename)
silent=1
shift
;;
-V|--version)
exec ${grep} -V
;;
--*)
grep_args="${grep_args} $1"
shift
;;
-[EFGHILOSUVabchilnopqsuvwxyz]*)
post="${1#-?}"
pre=${1%${post}}
grep_args="${grep_args} ${pre}"
shift
# Put back partial arg
set -- "-${post}" "$@"
;;
-[ABCDdefm])
if [ $# -lt 2 ]
then
echo "${prg}: missing argument for $1 flag" >&2
exit 1
fi
case $1 in
-e)
if [ ${pattern_found} -ne 0 ]; then
grep_args="${grep_args} -e ${pattern}"
fi
pattern="$2"
pattern_found=1
shift 2
continue
;;
-f)
pattern_file=1
;;
*)
;;
esac
grep_args="${grep_args} $1 $2"
shift 2
;;
-[ABCDdefm]*)
post="${1#-e}"
case ${1} in
-e*)
if [ ${pattern_found} -ne 0 ]; then
grep_args="${grep_args} -e ${pattern}"
fi
pattern="${post}"
pattern_found=1
shift
continue
;;
-f*)
pattern_file=1
;;
*)
;;
esac
grep_args="${grep_args} ${post}"
shift
;;
-)
hyphen=1
shift
;;
-r|-R)
echo "${prg}: the ${1} flag is not currently supported" >&2
exit 1
;;
-?)
grep_args="${grep_args} $1"
shift
;;
*)
# pattern to grep for
endofopts=1
;;
esac
done
# if no -e option was found, take next argument as grep-pattern
if [ ${pattern_file} -eq 0 -a ${pattern_found} -eq 0 ]
then
if [ $# -ge 1 ]; then
pattern="$1"
shift
elif [ ${hyphen} -gt 0 ]; then
pattern="-"
else
echo "${prg}: missing pattern" >&2
exit 1
fi
pattern_found=1
fi
# Clean up possible leading blank
grep_args="${grep_args# }"
# call grep ...
if [ $# -lt 1 ]
then
# ... on stdin
if [ ${pattern_file} -eq 0 ]; then
${cattool} ${catargs} - | ${grep} ${grep_args} -e "${pattern}" -- -
else
${cattool} ${catargs} - | ${grep} ${grep_args} -- -
fi
ret=$?
else
# ... on all files given on the command line
if [ ${silent} -lt 1 -a $# -gt 1 ]; then
grep_args="-H ${grep_args}"
fi
# Succeed if any file matches. First assume no match.
ret=1
for file; do
if [ ${pattern_file} -eq 0 ]; then
${cattool} ${catargs} -- "${file}" |
${grep} --label="${file}" ${grep_args} -e "${pattern}" -- -
else
${cattool} ${catargs} -- "${file}" |
${grep} --label="${file}" ${grep_args} -- -
fi
this_ret=$?
# A match (0) overrides a no-match (1). An error (>=2) overrides all.
if [ ${this_ret} -eq 0 -a ${ret} -eq 1 ] || [ ${this_ret} -ge 2 ]; then
ret=${this_ret}
fi
done
fi
exit ${ret}