Import csplit(1) from OpenBSD

Modified slightly to conform to the sbase style.
This commit is contained in:
sin 2014-10-16 11:58:04 +01:00
parent d1a1c42eb3
commit c0b9a8533f
4 changed files with 619 additions and 1 deletions

View File

@ -42,6 +42,7 @@ SRC = \
cols.c \
comm.c \
cp.c \
csplit.c \
cut.c \
date.c \
dirname.c \

1
TODO
View File

@ -6,7 +6,6 @@ such as vi and sh are also not listed here.
at
awk
bc
csplit
diff
ed
file

166
csplit.1 Normal file
View File

@ -0,0 +1,166 @@
.\" $OpenBSD: csplit.1,v 1.11 2014/02/04 13:26:36 jmc Exp $
.\"
.\" Copyright (c) 2002 Tim J. Robbins.
.\" 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.
.\"
.\" $FreeBSD: src/usr.bin/csplit/csplit.1,v 1.11 2005/01/25 22:29:51 tjr Exp $
.\"
.Dd $Mdocdate: February 4 2014 $
.Dt CSPLIT 1
.Os
.Sh NAME
.Nm csplit
.Nd split files based on context
.Sh SYNOPSIS
.Nm
.Op Fl ks
.Op Fl f Ar prefix
.Op Fl n Ar number
.Ar file args ...
.Sh DESCRIPTION
The
.Nm
utility splits
.Ar file
into pieces using the patterns
.Ar args .
If
.Ar file
is
a dash
.Pq Sq - ,
.Nm
reads from standard input.
.Pp
Files are created with a prefix of
.Dq xx
and two decimal digits.
The size of each file is written to standard output
as it is created.
If an error occurs whilst files are being created,
or a
.Dv HUP ,
.Dv INT ,
or
.Dv TERM
signal is received,
all files previously written are removed.
.Pp
The options are as follows:
.Bl -tag -width indent
.It Fl f Ar prefix
Create file names beginning with
.Ar prefix ,
instead of
.Dq xx .
.It Fl k
Do not remove previously created files if an error occurs or a
.Dv HUP ,
.Dv INT ,
or
.Dv TERM
signal is received.
.It Fl n Ar number
Create file names beginning with
.Ar number
of decimal digits after the prefix,
instead of 2.
.It Fl s
Do not write the size of each output file to standard output as it is
created.
.El
.Pp
The
.Ar args
operands may be a combination of the following patterns:
.Bl -tag -width indent
.It Xo
.Sm off
.No / Ar regexp No /
.Op Oo Cm + | - Oc Ar offset
.Sm on
.Xc
Create a file containing the input from the current line to (but not including)
the next line matching the given basic regular expression.
An optional
.Ar offset
from the line that matched may be specified.
.It Xo
.Sm off
.No % Ar regexp No %
.Op Oo Cm + | - Oc Ar offset
.Sm on
.Xc
Same as above but a file is not created for the output.
.It Ar line_no
Create containing the input from the current line to (but not including)
the specified line number.
.It { Ns Ar num Ns }
Repeat the previous pattern the specified number of times.
If it follows a line number pattern, a new file will be created for each
.Ar line_no
lines,
.Ar num
times.
The first line of the file is line number 1 for historic reasons.
.El
.Pp
After all the patterns have been processed, the remaining input data
(if there is any) will be written to a new file.
.Pp
Requesting to split at a line before the current line number or past the
end of the file will result in an error.
.Sh EXIT STATUS
.Ex -std csplit
.Sh EXAMPLES
Split the
.Xr mdoc 7
file
.Pa foo.1
into one file for each section (up to 21):
.Pp
.Dl "$ csplit -k foo.1 '%^\e.Sh%' '/^\e.Sh/' '{20}'"
.Pp
Split standard input after the first 99 lines and every 100 lines thereafter:
.Pp
.Dl "$ csplit -k - 100 '{19}'"
.Sh SEE ALSO
.Xr sed 1 ,
.Xr split 1 ,
.Xr re_format 7
.Sh STANDARDS
The
.Nm
utility is compliant with the
.St -p1003.1-2008
specification.
.Sh HISTORY
A
.Nm
command appeared in PWB
.Ux .
.Sh BUGS
Input lines are limited to
.Dv LINE_MAX
(2048) bytes in length.

452
csplit.c Normal file
View File

@ -0,0 +1,452 @@
/* $OpenBSD: csplit.c,v 1.5 2014/05/20 01:25:23 guenther Exp $ */
/* $FreeBSD: src/usr.bin/csplit/csplit.c,v 1.9 2004/03/22 11:15:03 tjr Exp $ */
/*-
* Copyright (c) 2002 Tim J. Robbins.
* 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.
*/
/*
* csplit -- split files based on context
*
* This utility splits its input into numbered output files by line number
* or by a regular expression. Regular expression matches have an optional
* offset with them, allowing the split to occur a specified number of
* lines before or after the match.
*
* To handle negative offsets, we stop reading when the match occurs and
* store the offset that the file should have been split at, then use
* this output file as input until all the "overflowed" lines have been read.
* The file is then closed and truncated to the correct length.
*
* We assume that the output files can be seeked upon (ie. they cannot be
* symlinks to named pipes or character devices), but make no such
* assumption about the input.
*/
#include <sys/types.h>
#include <ctype.h>
#include <errno.h>
#include <limits.h>
#include <locale.h>
#include <regex.h>
#include <signal.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include "util.h"
void cleanup(void);
void do_lineno(const char *);
void do_rexp(const char *);
char *get_line(void);
void handlesig(int);
FILE *newfile(void);
void toomuch(FILE *, long);
void usage(void);
/*
* Command line options
*/
const char *prefix; /* File name prefix */
long sufflen; /* Number of decimal digits for suffix */
int sflag; /* Suppress output of file names */
int kflag; /* Keep output if error occurs */
/*
* Other miscellaneous globals (XXX too many)
*/
long lineno; /* Current line number in input file */
long reps; /* Number of repetitions for this pattern */
long nfiles; /* Number of files output so far */
long maxfiles; /* Maximum number of files we can create */
char currfile[PATH_MAX]; /* Current output file */
const char *infn; /* Name of the input file */
FILE *infile; /* Input file handle */
FILE *overfile; /* Overflow file for toomuch() */
off_t truncofs; /* Offset this file should be truncated at */
int doclean; /* Should cleanup() remove output? */
int
main(int argc, char *argv[])
{
struct sigaction sa;
long i;
const char *expr;
char *ep, *p;
FILE *ofp;
kflag = sflag = 0;
prefix = "xx";
sufflen = 2;
ARGBEGIN {
case 'f':
prefix = EARGF(usage());
break;
case 'k':
kflag = 1;
break;
case 'n':
sufflen = estrtol(EARGF(usage()), 10);
break;
case 's':
sflag = 1;
break;
default:
usage();
} ARGEND;
if (sufflen + strlen(prefix) >= PATH_MAX)
eprintf("name too long\n");
if ((infn = *argv++) == NULL)
usage();
if (strcmp(infn, "-") == 0) {
infile = stdin;
infn = "stdin";
} else if ((infile = fopen(infn, "r")) == NULL)
eprintf("fopen %s:", infn);
if (!kflag) {
doclean = 1;
atexit(cleanup);
sa.sa_flags = 0;
sa.sa_handler = handlesig;
sigemptyset(&sa.sa_mask);
sigaddset(&sa.sa_mask, SIGHUP);
sigaddset(&sa.sa_mask, SIGINT);
sigaddset(&sa.sa_mask, SIGTERM);
sigaction(SIGHUP, &sa, NULL);
sigaction(SIGINT, &sa, NULL);
sigaction(SIGTERM, &sa, NULL);
}
lineno = 0;
nfiles = 0;
truncofs = 0;
overfile = NULL;
/* Ensure 10^sufflen < LONG_MAX. */
for (maxfiles = 1, i = 0; i < sufflen; i++) {
if (maxfiles > LONG_MAX / 10)
eprintf("%ld: suffix too long (limit %ld)\n",
sufflen, i);
maxfiles *= 10;
}
/* Create files based on supplied patterns. */
while (nfiles < maxfiles - 1 && (expr = *argv++) != NULL) {
/* Look ahead & see if this pattern has any repetitions. */
if (*argv != NULL && **argv == '{') {
errno = 0;
reps = strtol(*argv + 1, &ep, 10);
if (reps < 0 || *ep != '}' || errno != 0)
eprintf("%s: bad repetition count\n", *argv + 1);
argv++;
} else
reps = 0;
if (*expr == '/' || *expr == '%') {
do {
do_rexp(expr);
} while (reps-- != 0 && nfiles < maxfiles - 1);
} else if (isdigit((unsigned char)*expr))
do_lineno(expr);
else
eprintf("%s: unrecognised pattern\n", expr);
}
/* Copy the rest into a new file. */
if (!feof(infile)) {
ofp = newfile();
while ((p = get_line()) != NULL && fputs(p, ofp) == 0)
;
if (!sflag)
printf("%jd\n", (intmax_t)ftello(ofp));
if (fclose(ofp) != 0)
eprintf("fclose %s:", currfile);
}
toomuch(NULL, 0);
doclean = 0;
return (0);
}
void
usage(void)
{
eprintf("usage: %s [-ks] [-f prefix] [-n number] file args ...\n", argv0);
}
/* ARGSUSED */
void
handlesig(int sig)
{
const char msg[] = "csplit: caught signal, cleaning up\n";
write(STDERR_FILENO, msg, sizeof(msg) - 1);
cleanup();
_exit(2);
}
/* Create a new output file. */
FILE *
newfile(void)
{
FILE *fp;
if ((size_t)snprintf(currfile, sizeof(currfile), "%s%0*ld", prefix,
(int)sufflen, nfiles) >= sizeof(currfile))
eprintf("%s: name too long\n", currfile);
if ((fp = fopen(currfile, "w+")) == NULL)
eprintf("fopen %s:", currfile);
nfiles++;
return (fp);
}
/* Remove partial output, called before exiting. */
void
cleanup(void)
{
char fnbuf[PATH_MAX];
long i;
if (!doclean)
return;
/*
* NOTE: One cannot portably assume to be able to call snprintf() from
* inside a signal handler. It is, however, safe to do on OpenBSD.
*/
for (i = 0; i < nfiles; i++) {
snprintf(fnbuf, sizeof(fnbuf), "%s%0*ld", prefix,
(int)sufflen, i);
unlink(fnbuf);
}
}
/* Read a line from the input into a static buffer. */
char *
get_line(void)
{
static char lbuf[LINE_MAX];
FILE *src;
src = overfile != NULL ? overfile : infile;
again: if (fgets(lbuf, sizeof(lbuf), src) == NULL) {
if (src == overfile) {
src = infile;
goto again;
}
return (NULL);
}
if (ferror(src))
eprintf("%s: read error:", infn);
lineno++;
return (lbuf);
}
/* Conceptually rewind the input (as obtained by get_line()) back `n' lines. */
void
toomuch(FILE *ofp, long n)
{
char buf[BUFSIZ];
size_t i, nread;
if (overfile != NULL) {
/*
* Truncate the previous file we overflowed into back to
* the correct length, close it.
*/
if (fflush(overfile) != 0)
eprintf("overflow\n");
if (ftruncate(fileno(overfile), truncofs) != 0)
eprintf("overflow\n");
if (fclose(overfile) != 0)
eprintf("overflow\n");
overfile = NULL;
}
if (n == 0)
/* Just tidying up */
return;
lineno -= n;
/*
* Wind the overflow file backwards to `n' lines before the
* current one.
*/
do {
if (ftello(ofp) < (off_t)sizeof(buf))
rewind(ofp);
else
fseeko(ofp, -(off_t)sizeof(buf), SEEK_CUR);
if (ferror(ofp))
eprintf("%s: can't seek:", currfile);
if ((nread = fread(buf, 1, sizeof(buf), ofp)) == 0)
eprintf("can't read overflowed output\n");
if (fseeko(ofp, -(off_t)nread, SEEK_CUR) != 0)
eprintf("fseeko %s:", currfile);
for (i = 1; i <= nread; i++)
if (buf[nread - i] == '\n' && n-- == 0)
break;
if (ftello(ofp) == 0)
break;
} while (n > 0);
if (fseeko(ofp, (off_t)(nread - i + 1), SEEK_CUR) != 0)
eprintf("fseeko %s:", currfile);
/*
* get_line() will read from here. Next call will truncate to
* truncofs in this file.
*/
overfile = ofp;
truncofs = ftello(overfile);
}
/* Handle splits for /regexp/ and %regexp% patterns. */
void
do_rexp(const char *expr)
{
regex_t cre;
intmax_t nwritten;
long ofs;
int first;
char *ecopy, *ep, *p, *pofs, *re;
FILE *ofp;
if ((ecopy = strdup(expr)) == NULL)
eprintf("strdup:");
re = ecopy + 1;
if ((pofs = strrchr(ecopy, *expr)) == NULL || pofs[-1] == '\\')
eprintf("%s: missing trailing %c\n", expr, *expr);
*pofs++ = '\0';
if (*pofs != '\0') {
errno = 0;
ofs = strtol(pofs, &ep, 10);
if (*ep != '\0' || errno != 0)
eprintf("%s: bad offset\n", pofs);
} else
ofs = 0;
if (regcomp(&cre, re, REG_NOSUB) != 0)
eprintf("%s: bad regular expression'n", re);
if (*expr == '/')
/* /regexp/: Save results to a file. */
ofp = newfile();
else {
/* %regexp%: Make a temporary file for overflow. */
if ((ofp = tmpfile()) == NULL)
eprintf("tmpfile:");
}
/* Read and output lines until we get a match. */
first = 1;
while ((p = get_line()) != NULL) {
if (fputs(p, ofp) != 0)
break;
if (!first && regexec(&cre, p, 0, NULL, 0) == 0)
break;
first = 0;
}
if (p == NULL)
eprintf("%s: no match\n", re);
if (ofs <= 0) {
/*
* Negative (or zero) offset: throw back any lines we should
* not have read yet.
*/
if (p != NULL) {
toomuch(ofp, -ofs + 1);
nwritten = (intmax_t)truncofs;
} else
nwritten = (intmax_t)ftello(ofp);
} else {
/*
* Positive offset: copy the requested number of lines
* after the match.
*/
while (--ofs > 0 && (p = get_line()) != NULL)
fputs(p, ofp);
toomuch(NULL, 0);
nwritten = (intmax_t)ftello(ofp);
if (fclose(ofp) != 0)
eprintf("fclose %s:", currfile);
}
if (!sflag && *expr == '/')
printf("%jd\n", nwritten);
regfree(&cre);
free(ecopy);
}
/* Handle splits based on line number. */
void
do_lineno(const char *expr)
{
long lastline, tgtline;
char *ep, *p;
FILE *ofp;
errno = 0;
tgtline = strtol(expr, &ep, 10);
if (tgtline <= 0 || errno != 0 || *ep != '\0')
eprintf("%s: bad line number\n", expr);
lastline = tgtline;
if (lastline <= lineno)
eprintf("%s: can't go backwards\n", expr);
while (nfiles < maxfiles - 1) {
ofp = newfile();
while (lineno + 1 != lastline) {
if ((p = get_line()) == NULL)
eprintf("%ld: out of range\n", lastline);
if (fputs(p, ofp) != 0)
break;
}
if (!sflag)
printf("%jd\n", (intmax_t)ftello(ofp));
if (fclose(ofp) != 0)
eprintf("fclose %s:", currfile);
if (reps-- == 0)
break;
lastline += tgtline;
}
}