2000-02-09 01:28:47 +05:30
|
|
|
/* vi: set sw=4 ts=4: */
|
2000-01-16 03:58:50 +05:30
|
|
|
/*
|
2003-03-19 14:43:01 +05:30
|
|
|
* wc implementation for busybox
|
2000-01-16 03:58:50 +05:30
|
|
|
*
|
2003-03-19 14:43:01 +05:30
|
|
|
* Copyright (C) 2003 Manuel Novoa III <mjn3@codepoet.org>
|
2000-01-16 03:58:50 +05:30
|
|
|
*
|
2010-08-16 23:44:46 +05:30
|
|
|
* Licensed under GPLv2 or later, see file LICENSE in this source tree.
|
2000-01-16 03:58:50 +05:30
|
|
|
*/
|
2003-03-19 14:43:01 +05:30
|
|
|
/* Mar 16, 2003 Manuel Novoa III (mjn3@codepoet.org)
|
|
|
|
*
|
|
|
|
* Rewritten to fix a number of problems and do some size optimizations.
|
2004-03-15 13:59:22 +05:30
|
|
|
* Problems in the previous busybox implementation (besides bloat) included:
|
2003-03-19 14:43:01 +05:30
|
|
|
* 1) broken 'wc -c' optimization (read note below)
|
|
|
|
* 2) broken handling of '-' args
|
|
|
|
* 3) no checking of ferror on EOF returns
|
|
|
|
* 4) isprint() wasn't considered when word counting.
|
|
|
|
*
|
|
|
|
* NOTES:
|
|
|
|
*
|
|
|
|
* The previous busybox wc attempted an optimization using stat for the
|
|
|
|
* case of counting chars only. I omitted that because it was broken.
|
|
|
|
* It didn't take into account the possibility of input coming from a
|
|
|
|
* pipe, or input from a file with file pointer not at the beginning.
|
|
|
|
*
|
|
|
|
* To implement such a speed optimization correctly, not only do you
|
|
|
|
* need the size, but also the file position. Note also that the
|
|
|
|
* file position may be past the end of file. Consider the example
|
|
|
|
* (adapted from example in gnu wc.c)
|
|
|
|
*
|
|
|
|
* echo hello > /tmp/testfile &&
|
2007-07-21 20:38:09 +05:30
|
|
|
* (dd ibs=1k skip=1 count=0 &> /dev/null; wc -c) < /tmp/testfile
|
2003-03-19 14:43:01 +05:30
|
|
|
*
|
|
|
|
* for which 'wc -c' should output '0'.
|
|
|
|
*/
|
2016-11-23 19:16:56 +05:30
|
|
|
//config:config WC
|
2017-07-19 01:31:24 +05:30
|
|
|
//config: bool "wc (4.4 kb)"
|
2016-11-23 19:16:56 +05:30
|
|
|
//config: default y
|
|
|
|
//config: help
|
|
|
|
//config: wc is used to print the number of bytes, words, and lines,
|
|
|
|
//config: in specified files.
|
|
|
|
//config:
|
|
|
|
//config:config FEATURE_WC_LARGE
|
2017-01-10 19:28:54 +05:30
|
|
|
//config: bool "Support very large counts"
|
2016-11-23 19:16:56 +05:30
|
|
|
//config: default y
|
|
|
|
//config: depends on WC
|
|
|
|
//config: help
|
2017-01-10 19:28:54 +05:30
|
|
|
//config: Use "unsigned long long" for counter variables.
|
2016-11-23 19:16:56 +05:30
|
|
|
|
|
|
|
//applet:IF_WC(APPLET(wc, BB_DIR_USR_BIN, BB_SUID_DROP))
|
|
|
|
|
|
|
|
//kbuild:lib-$(CONFIG_WC) += wc.o
|
|
|
|
|
|
|
|
/* BB_AUDIT SUSv3 compliant. */
|
|
|
|
/* http://www.opengroup.org/onlinepubs/007904975/utilities/wc.html */
|
|
|
|
|
2007-05-27 00:30:18 +05:30
|
|
|
#include "libbb.h"
|
2010-10-04 20:38:14 +05:30
|
|
|
#include "unicode.h"
|
2000-01-16 03:58:50 +05:30
|
|
|
|
2009-10-23 01:58:08 +05:30
|
|
|
#if !ENABLE_LOCALE_SUPPORT
|
|
|
|
# undef isprint
|
|
|
|
# undef isspace
|
|
|
|
# define isprint(c) ((unsigned)((c) - 0x20) <= (0x7e - 0x20))
|
|
|
|
# define isspace(c) ((c) == ' ')
|
2003-03-19 14:43:01 +05:30
|
|
|
#endif
|
|
|
|
|
2006-09-30 05:11:59 +05:30
|
|
|
#if ENABLE_FEATURE_WC_LARGE
|
2009-10-23 01:58:08 +05:30
|
|
|
# define COUNT_T unsigned long long
|
|
|
|
# define COUNT_FMT "llu"
|
2006-09-30 05:11:59 +05:30
|
|
|
#else
|
2009-10-23 01:58:08 +05:30
|
|
|
# define COUNT_T unsigned
|
|
|
|
# define COUNT_FMT "u"
|
2006-09-30 05:11:59 +05:30
|
|
|
#endif
|
2006-09-30 05:11:04 +05:30
|
|
|
|
2010-10-04 20:38:14 +05:30
|
|
|
/* We support -m even when UNICODE_SUPPORT is off,
|
|
|
|
* we just don't advertise it in help text,
|
|
|
|
* since it is the same as -c in this case.
|
|
|
|
*/
|
|
|
|
|
|
|
|
//usage:#define wc_trivial_usage
|
|
|
|
//usage: "[-c"IF_UNICODE_SUPPORT("m")"lwL] [FILE]..."
|
|
|
|
//usage:
|
|
|
|
//usage:#define wc_full_usage "\n\n"
|
|
|
|
//usage: "Count lines, words, and bytes for each FILE (or stdin)\n"
|
|
|
|
//usage: "\n -c Count bytes"
|
|
|
|
//usage: IF_UNICODE_SUPPORT(
|
|
|
|
//usage: "\n -m Count characters"
|
|
|
|
//usage: )
|
|
|
|
//usage: "\n -l Count newlines"
|
|
|
|
//usage: "\n -w Count words"
|
|
|
|
//usage: "\n -L Print longest line length"
|
|
|
|
//usage:
|
|
|
|
//usage:#define wc_example_usage
|
|
|
|
//usage: "$ wc /etc/passwd\n"
|
|
|
|
//usage: " 31 46 1365 /etc/passwd\n"
|
|
|
|
|
|
|
|
/* Order is important if we want to be compatible with
|
|
|
|
* column order in "wc -cmlwL" output:
|
|
|
|
*/
|
2003-03-19 14:43:01 +05:30
|
|
|
enum {
|
2011-01-22 22:27:01 +05:30
|
|
|
WC_LINES = 0, /* -l */
|
|
|
|
WC_WORDS = 1, /* -w */
|
|
|
|
WC_UNICHARS = 2, /* -m */
|
|
|
|
WC_BYTES = 3, /* -c */
|
|
|
|
WC_LENGTH = 4, /* -L */
|
2010-10-04 20:38:14 +05:30
|
|
|
NUM_WCS = 5,
|
2001-11-21 14:47:00 +05:30
|
|
|
};
|
2000-01-16 03:58:50 +05:30
|
|
|
|
2007-10-11 15:35:36 +05:30
|
|
|
int wc_main(int argc, char **argv) MAIN_EXTERNALLY_VISIBLE;
|
2008-07-05 14:48:54 +05:30
|
|
|
int wc_main(int argc UNUSED_PARAM, char **argv)
|
2000-01-16 03:58:50 +05:30
|
|
|
{
|
2009-11-18 15:18:09 +05:30
|
|
|
const char *arg;
|
2008-02-19 06:08:10 +05:30
|
|
|
const char *start_fmt = " %9"COUNT_FMT + 1;
|
2006-09-30 05:11:04 +05:30
|
|
|
const char *fname_fmt = " %s\n";
|
|
|
|
COUNT_T *pcounts;
|
2010-10-04 20:34:20 +05:30
|
|
|
COUNT_T counts[NUM_WCS];
|
|
|
|
COUNT_T totals[NUM_WCS];
|
2009-11-18 15:18:09 +05:30
|
|
|
int num_files;
|
2007-01-21 02:58:36 +05:30
|
|
|
smallint status = EXIT_SUCCESS;
|
2006-09-30 05:11:59 +05:30
|
|
|
unsigned print_type;
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2010-10-04 20:38:14 +05:30
|
|
|
init_unicode();
|
|
|
|
|
2011-01-22 22:27:01 +05:30
|
|
|
print_type = getopt32(argv, "lwmcL");
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2003-03-19 14:43:01 +05:30
|
|
|
if (print_type == 0) {
|
2011-01-22 22:27:01 +05:30
|
|
|
print_type = (1 << WC_LINES) | (1 << WC_WORDS) | (1 << WC_BYTES);
|
2001-11-21 18:16:36 +05:30
|
|
|
}
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2003-03-19 14:43:01 +05:30
|
|
|
argv += optind;
|
2006-09-30 05:11:04 +05:30
|
|
|
if (!argv[0]) {
|
2003-03-19 14:43:01 +05:30
|
|
|
*--argv = (char *) bb_msg_standard_input;
|
2006-09-30 05:11:04 +05:30
|
|
|
fname_fmt = "\n";
|
2010-03-09 02:33:24 +05:30
|
|
|
}
|
|
|
|
if (!argv[1]) { /* zero or one filename? */
|
2006-09-30 05:11:04 +05:30
|
|
|
if (!((print_type-1) & print_type)) /* exactly one option? */
|
|
|
|
start_fmt = "%"COUNT_FMT;
|
2001-11-21 18:16:36 +05:30
|
|
|
}
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2003-03-19 14:43:01 +05:30
|
|
|
memset(totals, 0, sizeof(totals));
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2003-03-19 14:43:01 +05:30
|
|
|
pcounts = counts;
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2009-11-18 15:18:09 +05:30
|
|
|
num_files = 0;
|
2010-10-04 20:34:20 +05:30
|
|
|
while ((arg = *argv++) != NULL) {
|
2009-11-18 15:18:09 +05:30
|
|
|
FILE *fp;
|
|
|
|
const char *s;
|
|
|
|
unsigned u;
|
|
|
|
unsigned linepos;
|
|
|
|
smallint in_word;
|
|
|
|
|
2003-03-19 14:43:01 +05:30
|
|
|
++num_files;
|
2006-10-27 04:55:17 +05:30
|
|
|
fp = fopen_or_warn_stdin(arg);
|
2006-09-30 05:11:04 +05:30
|
|
|
if (!fp) {
|
2003-03-19 14:43:01 +05:30
|
|
|
status = EXIT_FAILURE;
|
|
|
|
continue;
|
2001-11-21 18:16:36 +05:30
|
|
|
}
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2003-03-19 14:43:01 +05:30
|
|
|
memset(counts, 0, sizeof(counts));
|
|
|
|
linepos = 0;
|
|
|
|
in_word = 0;
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2010-10-04 20:34:20 +05:30
|
|
|
while (1) {
|
2009-11-18 15:18:09 +05:30
|
|
|
int c;
|
2006-09-30 05:11:59 +05:30
|
|
|
/* Our -w doesn't match GNU wc exactly... oh well */
|
|
|
|
|
2003-03-19 14:43:01 +05:30
|
|
|
c = getc(fp);
|
2009-11-18 15:18:09 +05:30
|
|
|
if (c == EOF) {
|
|
|
|
if (ferror(fp)) {
|
|
|
|
bb_simple_perror_msg(arg);
|
|
|
|
status = EXIT_FAILURE;
|
|
|
|
}
|
2010-10-29 15:16:52 +05:30
|
|
|
goto DO_EOF; /* Treat an EOF as '\r'. */
|
2009-11-18 15:18:09 +05:30
|
|
|
}
|
2010-10-04 20:38:14 +05:30
|
|
|
|
|
|
|
/* Cater for -c and -m */
|
2011-01-22 22:27:01 +05:30
|
|
|
++counts[WC_BYTES];
|
2010-10-04 20:38:14 +05:30
|
|
|
if (unicode_status != UNICODE_ON /* every byte is a new char */
|
|
|
|
|| (c & 0xc0) != 0x80 /* it isn't a 2nd+ byte of a Unicode char */
|
|
|
|
) {
|
|
|
|
++counts[WC_UNICHARS];
|
|
|
|
}
|
2010-10-04 20:34:20 +05:30
|
|
|
|
2010-10-04 20:38:14 +05:30
|
|
|
if (isprint_asciionly(c)) { /* FIXME: not unicode-aware */
|
2003-03-19 14:43:01 +05:30
|
|
|
++linepos;
|
2009-10-23 01:58:08 +05:30
|
|
|
if (!isspace(c)) {
|
2003-03-19 14:43:01 +05:30
|
|
|
in_word = 1;
|
|
|
|
continue;
|
|
|
|
}
|
2009-10-23 01:58:08 +05:30
|
|
|
} else if ((unsigned)(c - 9) <= 4) {
|
2003-03-19 14:43:01 +05:30
|
|
|
/* \t 9
|
|
|
|
* \n 10
|
|
|
|
* \v 11
|
|
|
|
* \f 12
|
|
|
|
* \r 13
|
|
|
|
*/
|
|
|
|
if (c == '\t') {
|
|
|
|
linepos = (linepos | 7) + 1;
|
2010-10-29 15:16:52 +05:30
|
|
|
} else { /* '\n', '\r', '\f', or '\v' */
|
2009-11-18 15:18:09 +05:30
|
|
|
DO_EOF:
|
2003-03-19 14:43:01 +05:30
|
|
|
if (linepos > counts[WC_LENGTH]) {
|
|
|
|
counts[WC_LENGTH] = linepos;
|
|
|
|
}
|
|
|
|
if (c == '\n') {
|
|
|
|
++counts[WC_LINES];
|
|
|
|
}
|
|
|
|
if (c != '\v') {
|
|
|
|
linepos = 0;
|
|
|
|
}
|
|
|
|
}
|
2001-11-21 15:56:28 +05:30
|
|
|
} else {
|
2003-03-19 14:43:01 +05:30
|
|
|
continue;
|
|
|
|
}
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2003-03-19 14:43:01 +05:30
|
|
|
counts[WC_WORDS] += in_word;
|
|
|
|
in_word = 0;
|
|
|
|
if (c == EOF) {
|
|
|
|
break;
|
2001-11-21 15:56:28 +05:30
|
|
|
}
|
2010-10-04 20:34:20 +05:30
|
|
|
}
|
|
|
|
|
|
|
|
fclose_if_not_stdin(fp);
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2003-03-19 14:43:01 +05:30
|
|
|
if (totals[WC_LENGTH] < counts[WC_LENGTH]) {
|
|
|
|
totals[WC_LENGTH] = counts[WC_LENGTH];
|
2001-11-21 18:16:36 +05:30
|
|
|
}
|
2003-03-19 14:43:01 +05:30
|
|
|
totals[WC_LENGTH] -= counts[WC_LENGTH];
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2009-11-18 15:18:09 +05:30
|
|
|
OUTPUT:
|
2006-09-30 05:11:04 +05:30
|
|
|
/* coreutils wc tries hard to print pretty columns
|
2010-10-04 20:34:20 +05:30
|
|
|
* (saves results for all files, finds max col len etc...)
|
2006-09-30 05:11:04 +05:30
|
|
|
* we won't try that hard, it will bloat us too much */
|
|
|
|
s = start_fmt;
|
2003-03-19 14:43:01 +05:30
|
|
|
u = 0;
|
|
|
|
do {
|
|
|
|
if (print_type & (1 << u)) {
|
2006-10-27 04:51:47 +05:30
|
|
|
printf(s, pcounts[u]);
|
2006-09-30 05:11:04 +05:30
|
|
|
s = " %9"COUNT_FMT; /* Ok... restore the leading space. */
|
2003-03-19 14:43:01 +05:30
|
|
|
}
|
|
|
|
totals[u] += pcounts[u];
|
2010-10-04 20:34:20 +05:30
|
|
|
} while (++u < NUM_WCS);
|
2006-10-27 04:51:47 +05:30
|
|
|
printf(fname_fmt, arg);
|
2006-09-30 05:11:04 +05:30
|
|
|
}
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2003-03-19 14:43:01 +05:30
|
|
|
/* If more than one file was processed, we want the totals. To save some
|
|
|
|
* space, we set the pcounts ptr to the totals array. This has the side
|
|
|
|
* effect of trashing the totals array after outputting it, but that's
|
|
|
|
* irrelavent since we no longer need it. */
|
|
|
|
if (num_files > 1) {
|
2010-10-29 15:16:52 +05:30
|
|
|
num_files = 0; /* Make sure we don't get here again. */
|
2006-09-30 05:11:04 +05:30
|
|
|
arg = "total";
|
2003-03-19 14:43:01 +05:30
|
|
|
pcounts = totals;
|
2006-09-30 05:11:04 +05:30
|
|
|
--argv;
|
2003-03-19 14:43:01 +05:30
|
|
|
goto OUTPUT;
|
2000-07-20 05:38:10 +05:30
|
|
|
}
|
2004-03-15 13:59:22 +05:30
|
|
|
|
2006-10-27 04:51:47 +05:30
|
|
|
fflush_stdout_and_exit(status);
|
2000-01-16 03:58:50 +05:30
|
|
|
}
|