busybox/coreutils/sort.c

649 lines
18 KiB
C
Raw Normal View History

/* vi: set sw=4 ts=4: */
1999-12-22 01:30:35 +05:30
/*
* SuS3 compliant sort implementation for busybox
1999-12-22 01:30:35 +05:30
*
* Copyright (C) 2004 by Rob Landley <rob@landley.net>
1999-12-22 01:30:35 +05:30
*
* MAINTAINER: Rob Landley <rob@landley.net>
2006-09-17 21:58:10 +05:30
*
* Licensed under GPLv2 or later, see file LICENSE in this source tree.
1999-12-22 01:30:35 +05:30
*
* See SuS3 sort standard at:
* http://www.opengroup.org/onlinepubs/007904975/utilities/sort.html
1999-12-22 01:30:35 +05:30
*/
//config:config SORT
//config: bool "sort (7.7 kb)"
//config: default y
//config: help
//config: sort is used to sort lines of text in specified files.
//config:
//config:config FEATURE_SORT_BIG
//config: bool "Full SuSv3 compliant sort (support -ktcbdfiogM)"
//config: default y
//config: depends on SORT
//config: help
//config: Without this, sort only supports -rusz, and an integer version
//config: of -n. Selecting this adds sort keys, floating point support, and
//config: more. This adds a little over 3k to a nonstatic build on x86.
//config:
//config: The SuSv3 sort standard is available at:
//config: http://www.opengroup.org/onlinepubs/007904975/utilities/sort.html
//config:
//config:config FEATURE_SORT_OPTIMIZE_MEMORY
//config: bool "Use less memory (but might be slower)"
//config: default n # defaults to N since we are size-paranoid tribe
//config: depends on SORT
//config: help
//config: Attempt to use less memory (by storing only one copy
//config: of duplicated lines, and such). Useful if you work on huge files.
//applet:IF_SORT(APPLET_NOEXEC(sort, sort, BB_DIR_USR_BIN, BB_SUID_DROP, sort))
//kbuild:lib-$(CONFIG_SORT) += sort.o
1999-12-22 01:30:35 +05:30
//usage:#define sort_trivial_usage
//usage: "[-nru"
//usage: IF_FEATURE_SORT_BIG("gMcszbdfiokt] [-o FILE] [-k START[.OFS][OPTS][,END[.OFS][OPTS]] [-t CHAR")
//usage: "] [FILE]..."
//usage:#define sort_full_usage "\n\n"
//usage: "Sort lines of text\n"
//usage: IF_FEATURE_SORT_BIG(
//usage: "\n -o FILE Output to FILE"
//usage: "\n -c Check whether input is sorted"
//usage: "\n -b Ignore leading blanks"
//usage: "\n -f Ignore case"
//usage: "\n -i Ignore unprintable characters"
//usage: "\n -d Dictionary order (blank or alphanumeric only)"
//usage: )
//-h, --human-numeric-sort: compare human readable numbers (e.g., 2K 1G)
//usage: "\n -n Sort numbers"
//usage: IF_FEATURE_SORT_BIG(
//usage: "\n -g General numerical sort"
//usage: "\n -M Sort month"
//usage: "\n -V Sort version"
//usage: "\n -t CHAR Field separator"
//usage: "\n -k N[,M] Sort by Nth field"
//usage: )
//usage: "\n -r Reverse sort order"
//usage: "\n -s Stable (don't sort ties alphabetically)"
//usage: "\n -u Suppress duplicate lines"
//usage: "\n -z NUL terminated input and output"
///////: "\n -m Ignored for GNU compatibility"
///////: "\n -S BUFSZ Ignored for GNU compatibility"
///////: "\n -T TMPDIR Ignored for GNU compatibility"
//usage:
//usage:#define sort_example_usage
//usage: "$ echo -e \"e\\nf\\nb\\nd\\nc\\na\" | sort\n"
//usage: "a\n"
//usage: "b\n"
//usage: "c\n"
//usage: "d\n"
//usage: "e\n"
//usage: "f\n"
//usage: IF_FEATURE_SORT_BIG(
//usage: "$ echo -e \"c 3\\nb 2\\nd 2\" | $SORT -k 2,2n -k 1,1r\n"
//usage: "d 2\n"
//usage: "b 2\n"
//usage: "c 3\n"
//usage: )
//usage: ""
#include "libbb.h"
1999-12-22 01:30:35 +05:30
/* These are sort types */
enum {
FLAG_n = 1 << 0, /* Numeric sort */
FLAG_g = 1 << 1, /* Sort using strtod() */
FLAG_M = 1 << 2, /* Sort date */
FLAG_V = 1 << 3, /* Sort version */
/* ucsz apply to root level only, not keys. b at root level implies bb */
FLAG_u = 1 << 4, /* Unique */
FLAG_c = 1 << 5, /* Check: no output, exit(!ordered) */
FLAG_s = 1 << 6, /* Stable sort, no ascii fallback at end */
FLAG_z = 1 << 7, /* Input and output is NUL terminated, not \n */
/* These can be applied to search keys, the previous four can't */
FLAG_b = 1 << 8, /* Ignore leading blanks */
FLAG_r = 1 << 9, /* Reverse */
FLAG_d = 1 << 10, /* Ignore !(isalnum()|isspace()) */
FLAG_f = 1 << 11, /* Force uppercase */
FLAG_i = 1 << 12, /* Ignore !isprint() */
FLAG_m = 1 << 13, /* ignored: merge already sorted files; do not sort */
FLAG_S = 1 << 14, /* ignored: -S, --buffer-size=SIZE */
FLAG_T = 1 << 15, /* ignored: -T, --temporary-directory=DIR */
FLAG_o = 1 << 16,
FLAG_k = 1 << 17,
FLAG_t = 1 << 18,
FLAG_bb = 0x80000000, /* Ignore trailing blanks */
FLAG_no_tie_break = 0x40000000,
};
static const char sort_opt_str[] ALIGN1 = "^"
"ngMVucszbrdfimS:T:o:k:*t:"
"\0" "o--o:t--t"/*-t, -o: at most one of each*/;
/*
* OPT_STR must not be string literal, needs to have stable address:
* code uses "strchr(OPT_STR,c) - OPT_STR" idiom.
*/
#define OPT_STR (sort_opt_str + 1)
#if ENABLE_FEATURE_SORT_BIG
static char key_separator;
static struct sort_key {
struct sort_key *next_key; /* linked list */
unsigned range[4]; /* start word, start char, end word, end char */
unsigned flags;
} *key_list;
1999-12-22 06:00:29 +05:30
/* This is a NOEXEC applet. Be very careful! */
static char *get_key(char *str, struct sort_key *key, int flags)
1999-12-22 06:00:29 +05:30
{
int start = start; /* for compiler */
int end;
int len, j;
unsigned i;
/* Special case whole string, so we don't have to make a copy */
if (key->range[0] == 1 && !key->range[1] && !key->range[2] && !key->range[3]
&& !(flags & (FLAG_b | FLAG_d | FLAG_f | FLAG_i | FLAG_bb))
) {
return str;
}
/* Find start of key on first pass, end on second pass */
len = strlen(str);
for (j = 0; j < 2; j++) {
if (!key->range[2*j])
end = len;
/* Loop through fields */
else {
unsigned char ch = 0;
end = 0;
for (i = 1; i < key->range[2*j] + j; i++) {
if (key_separator) {
/* Skip body of key and separator */
while ((ch = str[end]) != '\0') {
end++;
if (ch == key_separator)
break;
}
} else {
/* Skip leading blanks */
while (isspace(str[end]))
end++;
/* Skip body of key */
while (str[end] != '\0') {
if (isspace(str[end]))
break;
end++;
}
}
}
/* Remove last delim: "abc:def:" => "abc:def" */
if (j && ch) {
//if (str[end-1] != key_separator)
// bb_error_msg(_and_die("BUG! "
// "str[start:%d,end:%d]:'%.*s'",
// start, end, (int)(end-start), &str[start]);
end--;
}
}
if (!j) start = end;
}
/* Strip leading whitespace if necessary */
if (flags & FLAG_b)
/* not using skip_whitespace() for speed */
while (isspace(str[start])) start++;
/* Strip trailing whitespace if necessary */
if (flags & FLAG_bb)
while (end > start && isspace(str[end-1])) end--;
/* -kSTART,N.ENDCHAR: honor ENDCHAR (1-based) */
if (key->range[3]) {
end = key->range[3];
if (end > len) end = len;
}
/* -kN.STARTCHAR[,...]: honor STARTCHAR (1-based) */
if (key->range[1]) {
start += key->range[1] - 1;
if (start > len) start = len;
}
/* Make the copy */
if (end < start)
end = start;
str = xstrndup(str+start, end-start);
/* Handle -d */
if (flags & FLAG_d) {
for (start = end = 0; str[end]; end++)
if (isspace(str[end]) || isalnum(str[end]))
str[start++] = str[end];
str[start] = '\0';
}
/* Handle -i */
if (flags & FLAG_i) {
for (start = end = 0; str[end]; end++)
if (isprint_asciionly(str[end]))
str[start++] = str[end];
str[start] = '\0';
}
/* Handle -f */
if (flags & FLAG_f)
for (i = 0; str[i]; i++)
str[i] = toupper(str[i]);
return str;
1999-12-22 06:00:29 +05:30
}
1999-12-22 01:30:35 +05:30
static struct sort_key *add_key(void)
1999-12-22 01:30:35 +05:30
{
struct sort_key **pkey = &key_list;
while (*pkey)
pkey = &((*pkey)->next_key);
return *pkey = xzalloc(sizeof(struct sort_key));
}
#define GET_LINE(fp) \
((option_mask32 & FLAG_z) \
? bb_get_chunk_from_file(fp, NULL) \
: xmalloc_fgetline(fp))
#else
#define GET_LINE(fp) xmalloc_fgetline(fp)
#endif
2003-03-19 14:43:01 +05:30
/* Iterate through keys list and perform comparisons */
static int compare_keys(const void *xarg, const void *yarg)
{
int flags = option_mask32, retval = 0;
char *x, *y;
2003-03-19 14:43:01 +05:30
#if ENABLE_FEATURE_SORT_BIG
struct sort_key *key;
2006-01-25 05:38:53 +05:30
for (key = key_list; !retval && key; key = key->next_key) {
flags = key->flags ? key->flags : option_mask32;
/* Chop out and modify key chunks, handling -dfib */
x = get_key(*(char **)xarg, key, flags);
y = get_key(*(char **)yarg, key, flags);
#else
/* This curly bracket serves no purpose but to match the nesting
* level of the for () loop we're not using */
{
x = *(char **)xarg;
y = *(char **)yarg;
#endif
/* Perform actual comparison */
switch (flags & (FLAG_n | FLAG_g | FLAG_M | FLAG_V)) {
default:
libbb: reduce the overhead of single parameter bb_error_msg() calls Back in 2007, commit 0c97c9d43707 ("'simple' error message functions by Loic Grenie") introduced bb_simple_perror_msg() to allow for a lower overhead call to bb_perror_msg() when only a string was being printed with no parameters. This saves space for some CPU architectures because it avoids the overhead of a call to a variadic function. However there has never been a simple version of bb_error_msg(), and since 2007 many new calls to bb_perror_msg() have been added that only take a single parameter and so could have been using bb_simple_perror_message(). This changeset introduces 'simple' versions of bb_info_msg(), bb_error_msg(), bb_error_msg_and_die(), bb_herror_msg() and bb_herror_msg_and_die(), and replaces all calls that only take a single parameter, or use something like ("%s", arg), with calls to the corresponding 'simple' version. Since it is likely that single parameter calls to the variadic functions may be accidentally reintroduced in the future a new debugging config option WARN_SIMPLE_MSG has been introduced. This uses some macro magic which will cause any such calls to generate a warning, but this is turned off by default to avoid use of the unpleasant macros in normal circumstances. This is a large changeset due to the number of calls that have been replaced. The only files that contain changes other than simple substitution of function calls are libbb.h, libbb/herror_msg.c, libbb/verror_msg.c and libbb/xfuncs_printf.c. In miscutils/devfsd.c, networking/udhcp/common.h and util-linux/mdev.c additonal macros have been added for logging so that single parameter and multiple parameter logging variants exist. The amount of space saved varies considerably by architecture, and was found to be as follows (for 'defconfig' using GCC 7.4): Arm: -92 bytes MIPS: -52 bytes PPC: -1836 bytes x86_64: -938 bytes Note that for the MIPS architecture only an exception had to be made disabling the 'simple' calls for 'udhcp' (in networking/udhcp/common.h) because it made these files larger on MIPS. Signed-off-by: James Byrne <james.byrne@origamienergy.com> Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2019-07-02 15:05:03 +05:30
bb_simple_error_msg_and_die("unknown sort type");
break;
#if defined(HAVE_STRVERSCMP) && HAVE_STRVERSCMP == 1
case FLAG_V:
retval = strverscmp(x, y);
break;
#endif
/* Ascii sort */
case 0:
#if ENABLE_LOCALE_SUPPORT
retval = strcoll(x, y);
#else
retval = strcmp(x, y);
#endif
break;
#if ENABLE_FEATURE_SORT_BIG
case FLAG_g: {
char *xx, *yy;
//TODO: needs setlocale(LC_NUMERIC, "C")?
2006-11-26 21:18:54 +05:30
double dx = strtod(x, &xx);
double dy = strtod(y, &yy);
/* not numbers < NaN < -infinity < numbers < +infinity) */
if (x == xx)
retval = (y == yy ? 0 : -1);
else if (y == yy)
retval = 1;
/* Check for isnan */
else if (dx != dx)
retval = (dy != dy) ? 0 : -1;
else if (dy != dy)
retval = 1;
/* Check for infinity. Could underflow, but it avoids libm. */
else if (1.0 / dx == 0.0) {
if (dx < 0)
retval = (1.0 / dy == 0.0 && dy < 0) ? 0 : -1;
else
retval = (1.0 / dy == 0.0 && dy > 0) ? 0 : 1;
} else if (1.0 / dy == 0.0)
retval = (dy < 0) ? 1 : -1;
else
retval = (dx > dy) ? 1 : ((dx < dy) ? -1 : 0);
break;
}
case FLAG_M: {
struct tm thyme;
int dx;
char *xx, *yy;
1999-12-22 01:30:35 +05:30
xx = strptime(x, "%b", &thyme);
dx = thyme.tm_mon;
yy = strptime(y, "%b", &thyme);
if (!xx)
retval = (!yy) ? 0 : -1;
else if (!yy)
retval = 1;
else
retval = dx - thyme.tm_mon;
break;
}
/* Full floating point version of -n */
case FLAG_n: {
2006-11-26 21:18:54 +05:30
double dx = atof(x);
double dy = atof(y);
retval = (dx > dy) ? 1 : ((dx < dy) ? -1 : 0);
break;
}
2006-11-26 21:18:54 +05:30
} /* switch */
/* Free key copies. */
if (x != *(char **)xarg) free(x);
if (y != *(char **)yarg) free(y);
/* if (retval) break; - done by for () anyway */
#else
/* Integer version of -n for tiny systems */
case FLAG_n:
retval = atoi(x) - atoi(y);
break;
2006-11-26 21:18:54 +05:30
} /* switch */
#endif
} /* for */
if (retval == 0) {
/* So far lines are "the same" */
if (option_mask32 & FLAG_s) {
/* "Stable sort": later line is "greater than",
* IOW: do not allow qsort() to swap equal lines.
*/
uint32_t *p32;
uint32_t x32, y32;
char *line;
unsigned len;
line = *(char**)xarg;
len = (strlen(line) + 4) & (~3u);
p32 = (void*)(line + len);
x32 = *p32;
line = *(char**)yarg;
len = (strlen(line) + 4) & (~3u);
p32 = (void*)(line + len);
y32 = *p32;
/* If x > y, 1, else -1 */
retval = (x32 > y32) * 2 - 1;
} else
if (!(option_mask32 & FLAG_no_tie_break)) {
/* fallback sort */
flags = option_mask32;
retval = strcmp(*(char **)xarg, *(char **)yarg);
}
}
if (flags & FLAG_r)
return -retval;
return retval;
}
#if ENABLE_FEATURE_SORT_BIG
static unsigned str2u(char **str)
{
unsigned long lu;
if (!isdigit((*str)[0]))
libbb: reduce the overhead of single parameter bb_error_msg() calls Back in 2007, commit 0c97c9d43707 ("'simple' error message functions by Loic Grenie") introduced bb_simple_perror_msg() to allow for a lower overhead call to bb_perror_msg() when only a string was being printed with no parameters. This saves space for some CPU architectures because it avoids the overhead of a call to a variadic function. However there has never been a simple version of bb_error_msg(), and since 2007 many new calls to bb_perror_msg() have been added that only take a single parameter and so could have been using bb_simple_perror_message(). This changeset introduces 'simple' versions of bb_info_msg(), bb_error_msg(), bb_error_msg_and_die(), bb_herror_msg() and bb_herror_msg_and_die(), and replaces all calls that only take a single parameter, or use something like ("%s", arg), with calls to the corresponding 'simple' version. Since it is likely that single parameter calls to the variadic functions may be accidentally reintroduced in the future a new debugging config option WARN_SIMPLE_MSG has been introduced. This uses some macro magic which will cause any such calls to generate a warning, but this is turned off by default to avoid use of the unpleasant macros in normal circumstances. This is a large changeset due to the number of calls that have been replaced. The only files that contain changes other than simple substitution of function calls are libbb.h, libbb/herror_msg.c, libbb/verror_msg.c and libbb/xfuncs_printf.c. In miscutils/devfsd.c, networking/udhcp/common.h and util-linux/mdev.c additonal macros have been added for logging so that single parameter and multiple parameter logging variants exist. The amount of space saved varies considerably by architecture, and was found to be as follows (for 'defconfig' using GCC 7.4): Arm: -92 bytes MIPS: -52 bytes PPC: -1836 bytes x86_64: -938 bytes Note that for the MIPS architecture only an exception had to be made disabling the 'simple' calls for 'udhcp' (in networking/udhcp/common.h) because it made these files larger on MIPS. Signed-off-by: James Byrne <james.byrne@origamienergy.com> Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2019-07-02 15:05:03 +05:30
bb_simple_error_msg_and_die("bad field specification");
lu = strtoul(*str, str, 10);
if ((sizeof(long) > sizeof(int) && lu > INT_MAX) || !lu)
libbb: reduce the overhead of single parameter bb_error_msg() calls Back in 2007, commit 0c97c9d43707 ("'simple' error message functions by Loic Grenie") introduced bb_simple_perror_msg() to allow for a lower overhead call to bb_perror_msg() when only a string was being printed with no parameters. This saves space for some CPU architectures because it avoids the overhead of a call to a variadic function. However there has never been a simple version of bb_error_msg(), and since 2007 many new calls to bb_perror_msg() have been added that only take a single parameter and so could have been using bb_simple_perror_message(). This changeset introduces 'simple' versions of bb_info_msg(), bb_error_msg(), bb_error_msg_and_die(), bb_herror_msg() and bb_herror_msg_and_die(), and replaces all calls that only take a single parameter, or use something like ("%s", arg), with calls to the corresponding 'simple' version. Since it is likely that single parameter calls to the variadic functions may be accidentally reintroduced in the future a new debugging config option WARN_SIMPLE_MSG has been introduced. This uses some macro magic which will cause any such calls to generate a warning, but this is turned off by default to avoid use of the unpleasant macros in normal circumstances. This is a large changeset due to the number of calls that have been replaced. The only files that contain changes other than simple substitution of function calls are libbb.h, libbb/herror_msg.c, libbb/verror_msg.c and libbb/xfuncs_printf.c. In miscutils/devfsd.c, networking/udhcp/common.h and util-linux/mdev.c additonal macros have been added for logging so that single parameter and multiple parameter logging variants exist. The amount of space saved varies considerably by architecture, and was found to be as follows (for 'defconfig' using GCC 7.4): Arm: -92 bytes MIPS: -52 bytes PPC: -1836 bytes x86_64: -938 bytes Note that for the MIPS architecture only an exception had to be made disabling the 'simple' calls for 'udhcp' (in networking/udhcp/common.h) because it made these files larger on MIPS. Signed-off-by: James Byrne <james.byrne@origamienergy.com> Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2019-07-02 15:05:03 +05:30
bb_simple_error_msg_and_die("bad field specification");
return lu;
}
#endif
int sort_main(int argc, char **argv) MAIN_EXTERNALLY_VISIBLE;
2008-07-05 14:48:54 +05:30
int sort_main(int argc UNUSED_PARAM, char **argv)
{
char **lines;
char *str_ignored, *str_o, *str_t;
llist_t *lst_k = NULL;
int i;
int linecount;
unsigned opts;
#if ENABLE_FEATURE_SORT_OPTIMIZE_MEMORY
bool can_drop_dups;
size_t prev_len = 0;
char *prev_line = (char*) "";
/* Postpone optimizing if the input is small, < 16k lines:
* even just free()ing duplicate lines takes time.
*/
size_t count_to_optimize_dups = 0x3fff;
#endif
xfunc_error_retval = 2;
/* Parse command line options */
opts = getopt32(argv,
sort_opt_str,
getopt32: remove opt_complementary function old new delta vgetopt32 1318 1392 +74 runsvdir_main 703 713 +10 bb_make_directory 423 425 +2 collect_cpu 546 545 -1 opt_chars 3 - -3 opt_complementary 4 - -4 tftpd_main 567 562 -5 ntp_init 476 471 -5 zcip_main 1266 1256 -10 xxd_main 428 418 -10 whois_main 140 130 -10 who_main 463 453 -10 which_main 212 202 -10 wget_main 2535 2525 -10 watchdog_main 291 281 -10 watch_main 222 212 -10 vlock_main 399 389 -10 uuencode_main 332 322 -10 uudecode_main 316 306 -10 unlink_main 45 35 -10 udhcpd_main 1482 1472 -10 udhcpc_main 2762 2752 -10 tune2fs_main 290 280 -10 tunctl_main 366 356 -10 truncate_main 218 208 -10 tr_main 518 508 -10 time_main 1134 1124 -10 tftp_main 286 276 -10 telnetd_main 1873 1863 -10 tcpudpsvd_main 1785 1775 -10 taskset_main 521 511 -10 tar_main 1009 999 -10 tail_main 1644 1634 -10 syslogd_main 1967 1957 -10 switch_root_main 368 358 -10 svlogd_main 1454 1444 -10 sv 1296 1286 -10 stat_main 104 94 -10 start_stop_daemon_main 1028 1018 -10 split_main 542 532 -10 sort_main 796 786 -10 slattach_main 624 614 -10 shuf_main 504 494 -10 setsid_main 96 86 -10 setserial_main 1132 1122 -10 setfont_main 388 378 -10 setconsole_main 78 68 -10 sendmail_main 1209 1199 -10 sed_main 677 667 -10 script_main 1077 1067 -10 run_parts_main 325 315 -10 rtcwake_main 454 444 -10 rm_main 175 165 -10 reformime_main 119 109 -10 readlink_main 123 113 -10 rdate_main 246 236 -10 pwdx_main 189 179 -10 pstree_main 317 307 -10 pscan_main 663 653 -10 popmaildir_main 818 808 -10 pmap_main 80 70 -10 nc_main 1042 1032 -10 mv_main 558 548 -10 mountpoint_main 477 467 -10 mount_main 1264 1254 -10 modprobe_main 768 758 -10 modinfo_main 333 323 -10 mktemp_main 200 190 -10 mkswap_main 324 314 -10 mkfs_vfat_main 1489 1479 -10 microcom_main 715 705 -10 md5_sha1_sum_main 521 511 -10 man_main 867 857 -10 makedevs_main 1052 1042 -10 ls_main 563 553 -10 losetup_main 432 422 -10 loadfont_main 89 79 -10 ln_main 524 514 -10 link_main 75 65 -10 ipcalc_main 544 534 -10 iostat_main 2397 2387 -10 install_main 768 758 -10 id_main 480 470 -10 i2cset_main 1239 1229 -10 i2cget_main 380 370 -10 i2cdump_main 1482 1472 -10 i2cdetect_main 682 672 -10 hwclock_main 406 396 -10 httpd_main 741 731 -10 grep_main 837 827 -10 getty_main 1559 1549 -10 fuser_main 297 287 -10 ftpgetput_main 345 335 -10 ftpd_main 2232 2222 -10 fstrim_main 251 241 -10 fsfreeze_main 77 67 -10 fsck_minix_main 2921 2911 -10 flock_main 314 304 -10 flashcp_main 740 730 -10 flash_eraseall_main 833 823 -10 fdformat_main 532 522 -10 expand_main 680 670 -10 eject_main 335 325 -10 dumpleases_main 630 620 -10 du_main 314 304 -10 dos2unix_main 441 431 -10 diff_main 1350 1340 -10 df_main 1064 1054 -10 date_main 1095 1085 -10 cut_main 961 951 -10 cryptpw_main 228 218 -10 crontab_main 575 565 -10 crond_main 1149 1139 -10 cp_main 370 360 -10 common_traceroute_main 3834 3824 -10 common_ping_main 1767 1757 -10 comm_main 239 229 -10 cmp_main 655 645 -10 chrt_main 379 369 -10 chpst_main 704 694 -10 chpasswd_main 308 298 -10 chown_main 171 161 -10 chmod_main 158 148 -10 cat_main 428 418 -10 bzip2_main 120 110 -10 blkdiscard_main 264 254 -10 base64_main 221 211 -10 arping_main 1665 1655 -10 ar_main 556 546 -10 adjtimex_main 406 396 -10 adduser_main 882 872 -10 addgroup_main 411 401 -10 acpid_main 1198 1188 -10 optstring 11 - -11 opt_string 18 - -18 OPT_STR 25 - -25 ubi_tools_main 1288 1258 -30 ls_options 31 - -31 ------------------------------------------------------------------------------ (add/remove: 0/6 grow/shrink: 3/129 up/down: 86/-1383) Total: -1297 bytes text data bss dec hex filename 915428 485 6876 922789 e14a5 busybox_old 914629 485 6872 921986 e1182 busybox_unstripped Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-09 01:25:02 +05:30
&str_ignored, &str_ignored, &str_o, &lst_k, &str_t
);
#if ENABLE_FEATURE_SORT_OPTIMIZE_MEMORY
/* Can drop dups only if -u but no "complicating" options,
* IOW: if we do a full line compares. Safe options:
* -o FILE Output to FILE
* -z Lines are terminated by NUL, not newline
* -r Reverse sort order
* -s Stable (don't sort ties alphabetically)
* Not sure about these:
* -b Ignore leading blanks
* -f Ignore case
* -i Ignore unprintable characters
* -d Dictionary order (blank or alphanumeric only)
* -n Sort numbers
* -g General numerical sort
* -M Sort month
*/
can_drop_dups = ((opts & ~(FLAG_o|FLAG_z|FLAG_r|FLAG_s)) == FLAG_u);
/* Stable sort needs every line to be uniquely allocated,
* disable optimization to reuse strings:
*/
if (opts & FLAG_s)
count_to_optimize_dups = (size_t)-1L;
#endif
/* global b strips leading and trailing spaces */
if (opts & FLAG_b)
option_mask32 |= FLAG_bb;
#if ENABLE_FEATURE_SORT_BIG
if (opts & FLAG_t) {
if (!str_t[0] || str_t[1])
libbb: reduce the overhead of single parameter bb_error_msg() calls Back in 2007, commit 0c97c9d43707 ("'simple' error message functions by Loic Grenie") introduced bb_simple_perror_msg() to allow for a lower overhead call to bb_perror_msg() when only a string was being printed with no parameters. This saves space for some CPU architectures because it avoids the overhead of a call to a variadic function. However there has never been a simple version of bb_error_msg(), and since 2007 many new calls to bb_perror_msg() have been added that only take a single parameter and so could have been using bb_simple_perror_message(). This changeset introduces 'simple' versions of bb_info_msg(), bb_error_msg(), bb_error_msg_and_die(), bb_herror_msg() and bb_herror_msg_and_die(), and replaces all calls that only take a single parameter, or use something like ("%s", arg), with calls to the corresponding 'simple' version. Since it is likely that single parameter calls to the variadic functions may be accidentally reintroduced in the future a new debugging config option WARN_SIMPLE_MSG has been introduced. This uses some macro magic which will cause any such calls to generate a warning, but this is turned off by default to avoid use of the unpleasant macros in normal circumstances. This is a large changeset due to the number of calls that have been replaced. The only files that contain changes other than simple substitution of function calls are libbb.h, libbb/herror_msg.c, libbb/verror_msg.c and libbb/xfuncs_printf.c. In miscutils/devfsd.c, networking/udhcp/common.h and util-linux/mdev.c additonal macros have been added for logging so that single parameter and multiple parameter logging variants exist. The amount of space saved varies considerably by architecture, and was found to be as follows (for 'defconfig' using GCC 7.4): Arm: -92 bytes MIPS: -52 bytes PPC: -1836 bytes x86_64: -938 bytes Note that for the MIPS architecture only an exception had to be made disabling the 'simple' calls for 'udhcp' (in networking/udhcp/common.h) because it made these files larger on MIPS. Signed-off-by: James Byrne <james.byrne@origamienergy.com> Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2019-07-02 15:05:03 +05:30
bb_simple_error_msg_and_die("bad -t parameter");
key_separator = str_t[0];
}
/* note: below this point we use option_mask32, not opts,
* since that reduces register pressure and makes code smaller */
/* Parse sort key */
while (lst_k) {
enum {
FLAG_allowed_for_k =
FLAG_n | /* Numeric sort */
FLAG_g | /* Sort using strtod() */
FLAG_M | /* Sort date */
FLAG_b | /* Ignore leading blanks */
FLAG_r | /* Reverse */
FLAG_d | /* Ignore !(isalnum()|isspace()) */
FLAG_f | /* Force uppercase */
FLAG_i | /* Ignore !isprint() */
0
};
struct sort_key *key = add_key();
char *str_k = llist_pop(&lst_k);
i = 0; /* i==0 before comma, 1 after (-k3,6) */
while (*str_k) {
/* Start of range */
/* Cannot use bb_strtou - suffix can be a letter */
key->range[2*i] = str2u(&str_k);
if (*str_k == '.') {
str_k++;
key->range[2*i+1] = str2u(&str_k);
}
while (*str_k) {
int flag;
const char *idx;
if (*str_k == ',' && !i++) {
str_k++;
break;
} /* no else needed: fall through to syntax error
because comma isn't in OPT_STR */
idx = strchr(OPT_STR, *str_k);
if (!idx)
libbb: reduce the overhead of single parameter bb_error_msg() calls Back in 2007, commit 0c97c9d43707 ("'simple' error message functions by Loic Grenie") introduced bb_simple_perror_msg() to allow for a lower overhead call to bb_perror_msg() when only a string was being printed with no parameters. This saves space for some CPU architectures because it avoids the overhead of a call to a variadic function. However there has never been a simple version of bb_error_msg(), and since 2007 many new calls to bb_perror_msg() have been added that only take a single parameter and so could have been using bb_simple_perror_message(). This changeset introduces 'simple' versions of bb_info_msg(), bb_error_msg(), bb_error_msg_and_die(), bb_herror_msg() and bb_herror_msg_and_die(), and replaces all calls that only take a single parameter, or use something like ("%s", arg), with calls to the corresponding 'simple' version. Since it is likely that single parameter calls to the variadic functions may be accidentally reintroduced in the future a new debugging config option WARN_SIMPLE_MSG has been introduced. This uses some macro magic which will cause any such calls to generate a warning, but this is turned off by default to avoid use of the unpleasant macros in normal circumstances. This is a large changeset due to the number of calls that have been replaced. The only files that contain changes other than simple substitution of function calls are libbb.h, libbb/herror_msg.c, libbb/verror_msg.c and libbb/xfuncs_printf.c. In miscutils/devfsd.c, networking/udhcp/common.h and util-linux/mdev.c additonal macros have been added for logging so that single parameter and multiple parameter logging variants exist. The amount of space saved varies considerably by architecture, and was found to be as follows (for 'defconfig' using GCC 7.4): Arm: -92 bytes MIPS: -52 bytes PPC: -1836 bytes x86_64: -938 bytes Note that for the MIPS architecture only an exception had to be made disabling the 'simple' calls for 'udhcp' (in networking/udhcp/common.h) because it made these files larger on MIPS. Signed-off-by: James Byrne <james.byrne@origamienergy.com> Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2019-07-02 15:05:03 +05:30
bb_simple_error_msg_and_die("unknown key option");
flag = 1 << (idx - OPT_STR);
if (flag & ~FLAG_allowed_for_k)
libbb: reduce the overhead of single parameter bb_error_msg() calls Back in 2007, commit 0c97c9d43707 ("'simple' error message functions by Loic Grenie") introduced bb_simple_perror_msg() to allow for a lower overhead call to bb_perror_msg() when only a string was being printed with no parameters. This saves space for some CPU architectures because it avoids the overhead of a call to a variadic function. However there has never been a simple version of bb_error_msg(), and since 2007 many new calls to bb_perror_msg() have been added that only take a single parameter and so could have been using bb_simple_perror_message(). This changeset introduces 'simple' versions of bb_info_msg(), bb_error_msg(), bb_error_msg_and_die(), bb_herror_msg() and bb_herror_msg_and_die(), and replaces all calls that only take a single parameter, or use something like ("%s", arg), with calls to the corresponding 'simple' version. Since it is likely that single parameter calls to the variadic functions may be accidentally reintroduced in the future a new debugging config option WARN_SIMPLE_MSG has been introduced. This uses some macro magic which will cause any such calls to generate a warning, but this is turned off by default to avoid use of the unpleasant macros in normal circumstances. This is a large changeset due to the number of calls that have been replaced. The only files that contain changes other than simple substitution of function calls are libbb.h, libbb/herror_msg.c, libbb/verror_msg.c and libbb/xfuncs_printf.c. In miscutils/devfsd.c, networking/udhcp/common.h and util-linux/mdev.c additonal macros have been added for logging so that single parameter and multiple parameter logging variants exist. The amount of space saved varies considerably by architecture, and was found to be as follows (for 'defconfig' using GCC 7.4): Arm: -92 bytes MIPS: -52 bytes PPC: -1836 bytes x86_64: -938 bytes Note that for the MIPS architecture only an exception had to be made disabling the 'simple' calls for 'udhcp' (in networking/udhcp/common.h) because it made these files larger on MIPS. Signed-off-by: James Byrne <james.byrne@origamienergy.com> Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2019-07-02 15:05:03 +05:30
bb_simple_error_msg_and_die("unknown sort type");
/* b after ',' means strip _trailing_ space */
if (i && flag == FLAG_b)
flag = FLAG_bb;
key->flags |= flag;
str_k++;
}
}
2003-03-19 14:43:01 +05:30
}
#endif
/* Open input files and read data */
argv += optind;
if (!*argv)
*--argv = (char*)"-";
linecount = 0;
lines = NULL;
do {
/* coreutils 6.9 compat: abort on first open error,
* do not continue to next file: */
FILE *fp = xfopen_stdin(*argv);
for (;;) {
char *line = GET_LINE(fp);
if (!line)
break;
#if ENABLE_FEATURE_SORT_OPTIMIZE_MEMORY
if (count_to_optimize_dups != 0)
count_to_optimize_dups--;
if (count_to_optimize_dups == 0) {
size_t len;
char *new_line;
/* On kernel/linux/arch/ *.[ch] files,
* this reduces memory usage by 6%.
* yes | head -99999999 | sort
* goes down from 1900Mb to 380 Mb.
*/
len = strlen(line);
if (len <= prev_len) {
new_line = prev_line + (prev_len - len);
if (strcmp(line, new_line) == 0) {
/* it's a tail of the prev line */
if (can_drop_dups && prev_len == len) {
/* it's identical to prev line */
free(line);
continue;
}
free(line);
line = new_line;
/* continue using longer prev_line
* for future tail tests.
*/
goto skip;
}
}
prev_len = len;
prev_line = line;
skip: ;
}
#else
//TODO: lighter version which only drops total dups if can_drop_dups == true
#endif
lines = xrealloc_vector(lines, 6, linecount);
lines[linecount++] = line;
2003-03-19 14:43:01 +05:30
}
fclose_if_not_stdin(fp);
} while (*++argv);
#if ENABLE_FEATURE_SORT_BIG
/* If no key, perform alphabetic sort */
if (!key_list)
add_key()->range[0] = 1;
/* Handle -c */
if (option_mask32 & FLAG_c) {
int j = (option_mask32 & FLAG_u) ? -1 : 0;
for (i = 1; i < linecount; i++) {
if (compare_keys(&lines[i-1], &lines[i]) > j) {
fprintf(stderr, "Check line %u\n", i);
return EXIT_FAILURE;
}
}
return EXIT_SUCCESS;
}
#endif
/* For stable sort, store original line position beyond terminating NUL */
if (option_mask32 & FLAG_s) {
for (i = 0; i < linecount; i++) {
uint32_t *p32;
char *line;
unsigned len;
line = lines[i];
len = (strlen(line) + 4) & (~3u);
lines[i] = line = xrealloc(line, len + 4);
p32 = (void*)(line + len);
*p32 = i;
}
/*option_mask32 |= FLAG_no_tie_break;*/
/* ^^^redundant: if FLAG_s, compare_keys() does no tie break */
}
/* Perform the actual sort */
qsort(lines, linecount, sizeof(lines[0]), compare_keys);
/* Handle -u */
if (option_mask32 & FLAG_u) {
int j = 0;
/* coreutils 6.3 drop lines for which only key is the same
* -- disabling last-resort compare, or else compare_keys()
* will be the same only for completely identical lines.
*/
option_mask32 |= FLAG_no_tie_break;
for (i = 1; i < linecount; i++) {
if (compare_keys(&lines[j], &lines[i]) == 0)
free(lines[i]);
else
lines[++j] = lines[i];
}
if (linecount)
linecount = j+1;
}
/* Print it */
#if ENABLE_FEATURE_SORT_BIG
/* Open output file _after_ we read all input ones */
if (option_mask32 & FLAG_o)
xmove_fd(xopen(str_o, O_WRONLY|O_CREAT|O_TRUNC), STDOUT_FILENO);
#endif
{
int ch = (option_mask32 & FLAG_z) ? '\0' : '\n';
for (i = 0; i < linecount; i++)
printf("%s%c", lines[i], ch);
}
2006-10-27 04:51:47 +05:30
fflush_stdout_and_exit(EXIT_SUCCESS);
2000-12-21 02:19:56 +05:30
}