git: dad64f0e7cad - main - wc: Clean up and modernize.
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Tue, 14 Feb 2023 22:26:59 UTC
The branch main has been updated by des:
URL: https://cgit.FreeBSD.org/src/commit/?id=dad64f0e7cadb5913d6463fc014f59f8706d316c
commit dad64f0e7cadb5913d6463fc014f59f8706d316c
Author: Dag-Erling Smørgrav <des@FreeBSD.org>
AuthorDate: 2023-02-14 22:26:34 +0000
Commit: Dag-Erling Smørgrav <des@FreeBSD.org>
CommitDate: 2023-02-14 22:26:47 +0000
wc: Clean up and modernize.
* Drop <err.h>, which is unnecessary since we use libxo.
* As per POSIX, report an error if output fails.
* Fix some type mismatches.
* Use bool instead of int where appropriate.
* Avoid repeatedly checking for a null filename.
* Miscellaneous other tidying.
* Add tests (partly derived from work performed by SHENG-YI HONG <i19780219111@kimo.com>).
Sponsored by: Klara, Inc.
Reviewed by: kevans
Differential Revision: https://reviews.freebsd.org/D38496
---
etc/mtree/BSD.tests.dist | 2 +
usr.bin/wc/Makefile | 3 +
usr.bin/wc/tests/Makefile | 8 ++
usr.bin/wc/tests/wc_test.sh | 197 ++++++++++++++++++++++++++++++++++++++++++++
usr.bin/wc/wc.c | 142 ++++++++++++++++---------------
5 files changed, 279 insertions(+), 73 deletions(-)
diff --git a/etc/mtree/BSD.tests.dist b/etc/mtree/BSD.tests.dist
index 910db16f0d15..859b5c5ace7f 100644
--- a/etc/mtree/BSD.tests.dist
+++ b/etc/mtree/BSD.tests.dist
@@ -1120,6 +1120,8 @@
..
vmstat
..
+ wc
+ ..
xargs
..
xinstall
diff --git a/usr.bin/wc/Makefile b/usr.bin/wc/Makefile
index 550b718e1478..060c67193e00 100644
--- a/usr.bin/wc/Makefile
+++ b/usr.bin/wc/Makefile
@@ -12,4 +12,7 @@ LIBADD+= cap_fileargs
CFLAGS+=-DWITH_CASPER
.endif
+HAS_TESTS=
+SUBDIR.${MK_TESTS}= tests
+
.include <bsd.prog.mk>
diff --git a/usr.bin/wc/tests/Makefile b/usr.bin/wc/tests/Makefile
new file mode 100644
index 000000000000..81e6c4246544
--- /dev/null
+++ b/usr.bin/wc/tests/Makefile
@@ -0,0 +1,8 @@
+# $FreeBSD$
+
+PACKAGE= tests
+
+ATF_TESTS_SH= wc_test
+BINDIR= ${TESTSDIR}
+
+.include <bsd.test.mk>
diff --git a/usr.bin/wc/tests/wc_test.sh b/usr.bin/wc/tests/wc_test.sh
new file mode 100755
index 000000000000..574f2daacaca
--- /dev/null
+++ b/usr.bin/wc/tests/wc_test.sh
@@ -0,0 +1,197 @@
+#
+# Copyright (c) 2023 Klara, Inc.
+#
+# SPDX-License-Identifier: BSD-2-Clause
+#
+
+#
+# These tests need to run in a multibyte locale with non-localized
+# error messages.
+#
+export LC_CTYPE=C.UTF-8
+export LC_MESSAGES=C
+
+#
+# Sample text containing multibyte characters
+#
+tv="Der bode en underlig gråsprængt en
+på den yderste nøgne ø; –
+han gjorde visst intet menneske mén
+hverken på land eller sjø;
+dog stundom gnistred hans øjne stygt, –
+helst mod uroligt vejr, –
+og da mente folk, at han var forrykt,
+og da var der få, som uden frykt
+kom Terje Vigen nær.
+"
+tvl=10
+tvw=55
+tvc=300
+tvm=283
+tvcL=42
+tvmL=39
+
+#
+# Run a series of tests using the same input file. The first argument
+# is the name of the file. The next three are the expected line,
+# word, and byte counts. The optional fifth is the expected character
+# count; if not provided, it is expected to be identical to the byte
+# count.
+#
+atf_check_wc() {
+ local file="$1"
+ local l="$2"
+ local w="$3"
+ local c="$4"
+ local m="${5-$4}"
+
+ atf_check -o match:"^ +${l} +${w} +${c}\$" wc <"${file}"
+ atf_check -o match:"^ +${l}\$" wc -l <"${file}"
+ atf_check -o match:"^ +${w}\$" wc -w <"${file}"
+ atf_check -o match:"^ +${c}\$" wc -c <"${file}"
+ atf_check -o match:"^ +${m}\$" wc -m <"${file}"
+ atf_check -o match:"^ +${l} +${w} +${c} ${file}\$" wc "$file"
+ atf_check -o match:"^ +${l} ${file}\$" wc -l "$file"
+ atf_check -o match:"^ +${w} ${file}\$" wc -w "$file"
+ atf_check -o match:"^ +${c} ${file}\$" wc -c "$file"
+ atf_check -o match:"^ +${m} ${file}\$" wc -m "$file"
+}
+
+atf_test_case basic
+basic_head()
+{
+ atf_set "descr" "Basic test case"
+}
+basic_body()
+{
+ printf "a b\n" >foo
+ atf_check_wc foo 1 2 4
+}
+
+atf_test_case blank
+blank_head()
+{
+ atf_set "descr" "Input containing only blank lines"
+}
+blank_body()
+{
+ printf "\n\n\n" >foo
+ atf_check_wc foo 3 0 3
+}
+
+atf_test_case empty
+empty_head()
+{
+ atf_set "descr" "Empty input"
+}
+empty_body()
+{
+ printf "" >foo
+ atf_check_wc foo 0 0 0
+}
+
+atf_test_case invalid
+invalid_head()
+{
+ atf_set "descr" "Invalid multibyte input"
+}
+invalid_body()
+{
+ printf "a\377b\n" >foo
+ atf_check \
+ -e match:"Illegal byte sequence" \
+ -o match:"^ +4 foo$" \
+ wc -m foo
+}
+
+atf_test_case multiline
+multiline_head()
+{
+ atf_set "descr" "Multiline, multibyte input"
+}
+multiline_body()
+{
+ printf "%s\n" "$tv" >foo
+ atf_check_wc foo $tvl $tvw $tvc $tvm
+ # longest line in bytes
+ atf_check -o match:"^ +$tvc +$tvcL foo" wc -cL foo
+ atf_check -o match:"^ +$tvc +$tvcL" wc -cL <foo
+ # longest line in characters
+ atf_check -o match:"^ +$tvm +$tvmL foo" wc -mL foo
+ atf_check -o match:"^ +$tvm +$tvmL" wc -mL <foo
+}
+
+atf_test_case multiline_repeated
+multiline_repeated_head()
+{
+ atf_set "descr" "Multiline input exceeding the input buffer size"
+}
+multiline_repeated_body()
+{
+ local c=0
+ while [ $c -lt 1000 ] ; do
+ printf "%1\$s\n%1\$s\n%1\$s\n%1\$s\n%1\$s\n" "$tv"
+ c=$((c+5))
+ done >foo
+ atf_check_wc foo $((tvl*c)) $((tvw*c)) $((tvc*c)) $((tvm*c))
+}
+
+atf_test_case total
+total_head()
+{
+ atf_set "descr" "Multiple inputs"
+}
+total_body()
+{
+ printf "%s\n" "$tv" >foo
+ printf "%s\n" "$tv" >bar
+ atf_check \
+ -o match:"^ +$((tvl*2)) +$((tvw*2)) +$((tvc*2)) total$" \
+ wc foo bar
+}
+
+atf_test_case unterminated
+unterminated_head()
+{
+ atf_set "descr" "Input not ending in newline"
+}
+unterminated_body()
+{
+ printf "a b" >foo
+ atf_check_wc foo 0 2 3
+}
+
+atf_test_case usage
+usage_head()
+{
+ atf_set "descr" "Trigger usage message"
+}
+usage_body()
+{
+ atf_check -s exit:1 -e match:"usage: wc" wc -\?
+}
+
+atf_test_case whitespace
+whitespace_head()
+{
+ atf_set "descr" "Input containing only whitespace and newlines"
+}
+whitespace_body()
+{
+ printf "\n \n\t\n" >foo
+ atf_check_wc foo 3 0 5
+}
+
+atf_init_test_cases()
+{
+ atf_add_test_case basic
+ atf_add_test_case blank
+ atf_add_test_case empty
+ atf_add_test_case invalid
+ atf_add_test_case multiline
+ atf_add_test_case multiline_repeated
+ atf_add_test_case total
+ atf_add_test_case unterminated
+ atf_add_test_case usage
+ atf_add_test_case whitespace
+}
diff --git a/usr.bin/wc/wc.c b/usr.bin/wc/wc.c
index a12c13a3e36d..9c129917dd04 100644
--- a/usr.bin/wc/wc.c
+++ b/usr.bin/wc/wc.c
@@ -50,10 +50,10 @@ __FBSDID("$FreeBSD$");
#include <capsicum_helpers.h>
#include <ctype.h>
-#include <err.h>
#include <errno.h>
#include <fcntl.h>
#include <locale.h>
+#include <stdbool.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
@@ -66,9 +66,11 @@ __FBSDID("$FreeBSD$");
#include <libcasper.h>
#include <casper/cap_fileargs.h>
+static const char *stdin_filename = "stdin";
+
static fileargs_t *fa;
static uintmax_t tlinect, twordct, tcharct, tlongline;
-static int doline, doword, dochar, domulti, dolongline;
+static bool doline, doword, dochar, domulti, dolongline;
static volatile sig_atomic_t siginfo;
static xo_handle_t *stderr_handle;
@@ -102,26 +104,26 @@ main(int argc, char *argv[])
argc = xo_parse_args(argc, argv);
if (argc < 0)
- return (argc);
+ exit(EXIT_FAILURE);
while ((ch = getopt(argc, argv, "clmwL")) != -1)
switch((char)ch) {
case 'l':
- doline = 1;
+ doline = true;
break;
case 'w':
- doword = 1;
+ doword = true;
break;
case 'c':
- dochar = 1;
- domulti = 0;
+ dochar = true;
+ domulti = false;
break;
case 'L':
- dolongline = 1;
+ dolongline = true;
break;
case 'm':
- domulti = 1;
- dochar = 0;
+ domulti = true;
+ dochar = false;
break;
case '?':
default:
@@ -134,27 +136,17 @@ main(int argc, char *argv[])
fa = fileargs_init(argc, argv, O_RDONLY, 0,
cap_rights_init(&rights, CAP_READ, CAP_FSTAT), FA_OPEN);
- if (fa == NULL) {
- xo_warn("Unable to init casper");
- exit(1);
- }
-
+ if (fa == NULL)
+ xo_err(EXIT_FAILURE, "Unable to initialize casper");
caph_cache_catpages();
- if (caph_limit_stdio() < 0) {
- xo_warn("Unable to limit stdio");
- fileargs_free(fa);
- exit(1);
- }
-
- if (caph_enter_casper() < 0) {
- xo_warn("Unable to enter capability mode");
- fileargs_free(fa);
- exit(1);
- }
+ if (caph_limit_stdio() < 0)
+ xo_err(EXIT_FAILURE, "Unable to limit stdio");
+ if (caph_enter_casper() < 0)
+ xo_err(EXIT_FAILURE, "Unable to enter capability mode");
/* Wc's flags are on by default. */
- if (doline + doword + dochar + domulti + dolongline == 0)
- doline = doword = dochar = 1;
+ if (!(doline || doword || dochar || domulti || dolongline))
+ doline = doword = dochar = true;
stderr_handle = xo_create_to_file(stderr, XO_STYLE_TEXT, 0);
xo_open_container("wc");
@@ -162,19 +154,19 @@ main(int argc, char *argv[])
errors = 0;
total = 0;
- if (!*argv) {
- xo_open_instance("file");
- if (cnt((char *)NULL) != 0)
+ if (argc == 0) {
+ xo_open_instance("file");
+ if (cnt(NULL) != 0)
++errors;
- xo_close_instance("file");
+ xo_close_instance("file");
} else {
- do {
- xo_open_instance("file");
- if (cnt(*argv) != 0)
+ while (argc--) {
+ xo_open_instance("file");
+ if (cnt(*argv++) != 0)
++errors;
- xo_close_instance("file");
+ xo_close_instance("file");
++total;
- } while(*++argv);
+ }
}
xo_close_list("file");
@@ -187,8 +179,9 @@ main(int argc, char *argv[])
fileargs_free(fa);
xo_close_container("wc");
- xo_finish();
- exit(errors == 0 ? 0 : 1);
+ if (xo_finish() < 0)
+ xo_err(EXIT_FAILURE, "stdout");
+ exit(errors == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
}
static void
@@ -212,7 +205,7 @@ show_cnt(const char *file, uintmax_t linect, uintmax_t wordct,
xo_emit_h(xop, " {:characters/%7ju/%ju}", charct);
if (dolongline)
xo_emit_h(xop, " {:long-lines/%7ju/%ju}", llct);
- if (file != NULL)
+ if (file != stdin_filename)
xo_emit_h(xop, " {:filename/%s}\n", file);
else
xo_emit_h(xop, "\n");
@@ -221,20 +214,21 @@ show_cnt(const char *file, uintmax_t linect, uintmax_t wordct,
static int
cnt(const char *file)
{
+ char buf[MAXBSIZE], *p;
struct stat sb;
+ mbstate_t mbs;
uintmax_t linect, wordct, charct, llct, tmpll;
- int fd, len, warned;
+ ssize_t len;
size_t clen;
- short gotsp;
- u_char *p;
- u_char buf[MAXBSIZE];
+ int fd;
wchar_t wch;
- mbstate_t mbs;
+ bool gotsp, warned;
linect = wordct = charct = llct = tmpll = 0;
- if (file == NULL)
+ if (file == NULL) {
fd = STDIN_FILENO;
- else if ((fd = fileargs_open(fa, file)) < 0) {
+ file = stdin_filename;
+ } else if ((fd = fileargs_open(fa, file)) < 0) {
xo_warn("%s: open", file);
return (1);
}
@@ -246,7 +240,7 @@ cnt(const char *file)
*/
if (doline == 0 && dolongline == 0) {
if (fstat(fd, &sb)) {
- xo_warn("%s: fstat", file != NULL ? file : "stdin");
+ xo_warn("%s: fstat", file);
(void)close(fd);
return (1);
}
@@ -265,9 +259,9 @@ cnt(const char *file)
* lines than to get words, since the word count requires locale
* handling.
*/
- while ((len = read(fd, buf, MAXBSIZE))) {
- if (len == -1) {
- xo_warn("%s: read", file != NULL ? file : "stdin");
+ while ((len = read(fd, buf, sizeof(buf)))) {
+ if (len < 0) {
+ xo_warn("%s: read", file);
(void)close(fd);
return (1);
}
@@ -275,14 +269,16 @@ cnt(const char *file)
show_cnt(file, linect, wordct, charct, llct);
charct += len;
if (doline || dolongline) {
- for (p = buf; len--; ++p)
+ for (p = buf; len > 0; --len, ++p) {
if (*p == '\n') {
if (tmpll > llct)
llct = tmpll;
tmpll = 0;
++linect;
- } else
+ } else {
tmpll++;
+ }
+ }
}
}
reset_siginfo();
@@ -297,12 +293,12 @@ cnt(const char *file)
return (0);
/* Do it the hard way... */
-word: gotsp = 1;
- warned = 0;
+word: gotsp = true;
+ warned = false;
memset(&mbs, 0, sizeof(mbs));
- while ((len = read(fd, buf, MAXBSIZE)) != 0) {
- if (len == -1) {
- xo_warn("%s: read", file != NULL ? file : "stdin");
+ while ((len = read(fd, buf, sizeof(buf))) != 0) {
+ if (len < 0) {
+ xo_warn("%s: read", file);
(void)close(fd);
return (1);
}
@@ -313,21 +309,20 @@ word: gotsp = 1;
if (!domulti || MB_CUR_MAX == 1) {
clen = 1;
wch = (unsigned char)*p;
- } else if ((clen = mbrtowc(&wch, p, len, &mbs)) ==
- (size_t)-1) {
+ } else if ((clen = mbrtowc(&wch, p, len, &mbs)) == 0) {
+ clen = 1;
+ } else if (clen == (size_t)-1) {
if (!warned) {
errno = EILSEQ;
- xo_warn("%s",
- file != NULL ? file : "stdin");
- warned = 1;
+ xo_warn("%s", file);
+ warned = true;
}
memset(&mbs, 0, sizeof(mbs));
clen = 1;
wch = (unsigned char)*p;
- } else if (clen == (size_t)-2)
+ } else if (clen == (size_t)-2) {
break;
- else if (clen == 0)
- clen = 1;
+ }
charct++;
if (wch != L'\n')
tmpll++;
@@ -339,18 +334,19 @@ word: gotsp = 1;
tmpll = 0;
++linect;
}
- if (iswspace(wch))
- gotsp = 1;
- else if (gotsp) {
- gotsp = 0;
+ if (iswspace(wch)) {
+ gotsp = true;
+ } else if (gotsp) {
+ gotsp = false;
++wordct;
}
}
}
reset_siginfo();
- if (domulti && MB_CUR_MAX > 1)
+ if (domulti && MB_CUR_MAX > 1) {
if (mbrtowc(NULL, NULL, 0, &mbs) == (size_t)-1 && !warned)
- xo_warn("%s", file != NULL ? file : "stdin");
+ xo_warn("%s", file);
+ }
if (doline)
tlinect += linect;
if (doword)
@@ -368,5 +364,5 @@ static void
usage(void)
{
xo_error("usage: wc [-Lclmw] [file ...]\n");
- exit(1);
+ exit(EXIT_FAILURE);
}