Replace str(1) with stris(1) #102
10
Makefile
10
Makefile
@ -42,7 +42,7 @@ BIN = build/bin
|
||||
default: all test
|
||||
|
||||
.PHONY: all
|
||||
all: dj false fop hru intcmp mm npc rpn scrut str strcmp swab true
|
||||
all: dj false fop hru intcmp mm npc rpn scrut strcmp stris swab true
|
||||
|
||||
# keep build/include until bindgen(1) has stdin support
|
||||
# https://github.com/rust-lang/rust-bindgen/issues/2703
|
||||
@ -147,10 +147,10 @@ scrut: build/bin/scrut
|
||||
build/bin/scrut: src/scrut.c build
|
||||
$(CC) $(CFLAGS) -o $@ src/scrut.c
|
||||
|
||||
.PHONY: str
|
||||
str: build/bin/str
|
||||
build/bin/str: src/str.c build
|
||||
$(CC) $(CFLAGS) -o $@ src/str.c
|
||||
.PHONY: stris
|
||||
stris: build/bin/stris
|
||||
build/bin/stris: src/stris.rs build rustlibs
|
||||
$(RUSTC) $(RUSTFLAGS) -o $@ src/stris.rs
|
||||
|
||||
.PHONY: strcmp
|
||||
strcmp: build/bin/strcmp
|
||||
|
||||
59
docs/str.1
59
docs/str.1
@ -1,59 +0,0 @@
|
||||
.\" Copyright (c) 2023–2024 DTB <trinity@trinity.moe>
|
||||
.\" Copyright (c) 2023–2024 Emma Tebibyte <emma@tebibyte.media>
|
||||
.\"
|
||||
.\" This work is licensed under CC BY-SA 4.0. To see a copy of this license,
|
||||
.\" visit <http://creativecommons.org/licenses/by-sa/4.0/>.
|
||||
.\"
|
||||
.TH STR 1 2024-06-17 "Harakit X.X.X"
|
||||
.SH NAME
|
||||
str \(en test string arguments
|
||||
.\"
|
||||
.SH SYNOPSIS
|
||||
|
||||
str
|
||||
.B type string...
|
||||
.\"
|
||||
.SH DESCRIPTION
|
||||
|
||||
Test the character types of string arguments.
|
||||
|
||||
The tests in this program are equivalent to the functions with the same names in
|
||||
.BR ctype.h (0p)
|
||||
and are the methods by which string arguments are tested.
|
||||
.\"
|
||||
.SH DIAGNOSTICS
|
||||
|
||||
If all tests pass, the program will exit successfully. If any of the tests fail,
|
||||
the program will exit unsuccessfully with an error code of 1.
|
||||
|
||||
When invoked incorrectly, a debug message will be printed and the program will
|
||||
exit with the appropriate
|
||||
.BR sysexits.h (3)
|
||||
error code.
|
||||
.\"
|
||||
.SH CAVEATS
|
||||
|
||||
None of an empty string\(cqs contents pass any of the tests, so the program will
|
||||
exit unsuccessfully if one is specified.
|
||||
|
||||
There\(cqs no way of knowing which argument failed the test without re-testing
|
||||
arguments individually.
|
||||
|
||||
If a character in a string isn\(cqt valid ASCII, the program will exit
|
||||
unsuccessfully.
|
||||
.\"
|
||||
.SH AUTHOR
|
||||
|
||||
Written by DTB
|
||||
.MT trinity@trinity.moe
|
||||
.ME .
|
||||
.\"
|
||||
.SH COPYRIGHT
|
||||
|
||||
Copyright \(co 2023 DTB. License AGPLv3+: GNU AGPL version 3 or later
|
||||
<https://gnu.org/licenses/gpl.html>.
|
||||
.\"
|
||||
.SH SEE ALSO
|
||||
.BR ctype (3p),
|
||||
.BR strcmp(1),
|
||||
.BR ascii(7)
|
||||
118
docs/stris.1
Normal file
118
docs/stris.1
Normal file
@ -0,0 +1,118 @@
|
||||
.\" Copyright (c) 2023–2024 DTB <trinity@trinity.moe>
|
||||
.\" Copyright (c) 2023 Emma Tebibyte <emma@tebibyte.media>
|
||||
.\"
|
||||
.\" This work is licensed under CC BY-SA 4.0. To see a copy of this license,
|
||||
.\" visit <http://creativecommons.org/licenses/by-sa/4.0/>.
|
||||
|
||||
.TH STRIS 1
|
||||
|
||||
.SH NAME
|
||||
|
||||
stris \(en test the character types of string arguments
|
||||
|
||||
.SH SYNOPSIS
|
||||
|
||||
stris
|
||||
.RB ( -7bcdlu )
|
||||
.RB ( -i [ inclusions ])
|
||||
.RB [ strings... ]
|
||||
|
||||
.SH DESCRIPTION
|
||||
|
||||
Stris tests each rune in an arbitrary quantity of string arguments, ensuring
|
||||
each meets any of the parameters specified in the program options.
|
||||
|
||||
.SH OPTIONS
|
||||
|
||||
.B -7
|
||||
.RS
|
||||
Tests to see if runes are fit within seven bits; that is, that they are encoded
|
||||
with ASCII.
|
||||
.RE
|
||||
|
||||
.B -b
|
||||
.RS
|
||||
Tests to see if runes are blank or "whitespace"; characters that do not print
|
||||
but fill a predictable amount of space.
|
||||
.RE
|
||||
|
||||
.B -c
|
||||
.RS
|
||||
Tests to see if runes are control characters; characters that are not printing
|
||||
or graphical.
|
||||
.RE
|
||||
|
||||
.B -d
|
||||
.RS
|
||||
Tests to see if runes are numeric. This test does not only allow the ASCII
|
||||
digits but any numeric symbol.
|
||||
.RE
|
||||
|
||||
.B -i
|
||||
.RS
|
||||
Permits, in addition to the given specified parameters, all of the runes
|
||||
supplied in its option argument.
|
||||
.RE
|
||||
|
||||
.B -l
|
||||
.RS
|
||||
Tests to see if runes are in their lower case.
|
||||
.RE
|
||||
|
||||
.B -u
|
||||
.RS
|
||||
Tests to see if runes are in their upper case, or capitalized.
|
||||
.RE
|
||||
|
||||
.SH DIAGNOSTICS
|
||||
|
||||
Stris exits successfully if all runes in all given strings meet any of the
|
||||
specified parameters, or if no parameters were specified but all given strings
|
||||
were legibly encoded. It exits unsuccessfully if the previous is untrue, and if
|
||||
invalid options were given, or if no strings were given, a usage synopsis will
|
||||
be printed to the standard error.
|
||||
|
||||
.SH BUGS
|
||||
|
||||
There's no way of knowing which argument failed the test without re-testing
|
||||
arguments individually.
|
||||
|
||||
Some runes that can losslessly be encoded into ASCII from UTF-8 but in an
|
||||
"overlong encoding", where the rune was encoded with unnecessary leading
|
||||
zeroes causing it to span multiple bytes, won't be detected as ASCII.
|
||||
|
||||
.SH EXAMPLES
|
||||
|
||||
This is an sh(1p) snippet that checks to see if an environment variable is an
|
||||
ASCII digit.
|
||||
|
||||
.RS
|
||||
.R stris -7 "$v" && stris -d "$v" && echo ASCII digit.
|
||||
.RE
|
||||
|
||||
This is an sh(1p) snippet that checks to see if an environment variable is a
|
||||
hexadecimal number.
|
||||
|
||||
.RS
|
||||
.R stris -7 "$v" && stris -di ABCDEFabcdef "$v" && echo Hexadecimal number.
|
||||
.RE
|
||||
|
||||
.SH AUTHOR
|
||||
|
||||
Written by DTB <trinity@trinity.moe>.
|
||||
|
||||
.SH HISTORY
|
||||
|
||||
Stris replaces the former str(1) which took the name of a function from
|
||||
ctype(3) as its first argument and checked the following strings against it;
|
||||
str(1) exited unsuccessfully when it encountered any non-ASCII runes and could
|
||||
only have one parameter specified.
|
||||
|
||||
.SH COPYRIGHT
|
||||
|
||||
Copyright © 2023–2024 DTB. License AGPLv3+: GNU AGPL version 3 or later
|
||||
<https://gnu.org/licenses/gpl.html>.
|
||||
|
||||
.SH SEE ALSO
|
||||
|
||||
ascii(7), ctype(3p), strcmp(1)
|
||||
96
src/str.c
96
src/str.c
@ -1,96 +0,0 @@
|
||||
/*
|
||||
* Copyright (c) 2023 DTB <trinity@trinity.moe>
|
||||
* Copyright (c) 2023 Marceline Cramer <mars@tebibyte.media>
|
||||
* SPDX-License-Identifier: AGPL-3.0-or-later
|
||||
*
|
||||
* This program is free software: you can redistribute it and/or modify it under
|
||||
* the terms of the GNU Affero General Public License as published by the Free
|
||||
* Software Foundation, either version 3 of the License, or (at your option) any
|
||||
* later version.
|
||||
*
|
||||
* This program is distributed in the hope that it will be useful, but WITHOUT
|
||||
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
|
||||
* FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more
|
||||
* details.
|
||||
*
|
||||
* You should have received a copy of the GNU Affero General Public License
|
||||
* along with this program. If not, see https://www.gnu.org/licenses/.
|
||||
*/
|
||||
|
||||
#include <ctype.h>
|
||||
#include <stddef.h> /* NULL */
|
||||
#include <stdio.h> /* fprintf(3), perror(3) */
|
||||
#include <stdlib.h> /* size_t, EXIT_FAILURE */
|
||||
#include <string.h> /* strcmp(3) */
|
||||
#include <sysexits.h> /* EX_OSERR, EX_USAGE */
|
||||
|
||||
#ifdef __OpenBSD__
|
||||
# include <unistd.h> /* pledge(2) */
|
||||
#endif
|
||||
|
||||
char *program_name = "str";
|
||||
|
||||
static struct {
|
||||
char *name;
|
||||
int (*f)(int);
|
||||
} ctypes[] = {
|
||||
{ "isalnum", isalnum },
|
||||
{ "isalpha", isalpha },
|
||||
{ "isblank", isblank },
|
||||
{ "iscntrl", iscntrl },
|
||||
{ "isdigit", isdigit },
|
||||
{ "isxdigit", isxdigit },
|
||||
{ "isgraph", isgraph },
|
||||
{ "islower", islower },
|
||||
{ "isprint", isprint },
|
||||
{ "ispunct", ispunct },
|
||||
{ "isspace", isspace },
|
||||
{ "isupper", isupper },
|
||||
{ NULL, NULL } /* marks end */
|
||||
};
|
||||
|
||||
static int
|
||||
usage(char *argv0) {
|
||||
(void)fprintf(stderr, "Usage: %s type string...\n", argv0);
|
||||
|
||||
return EX_USAGE;
|
||||
}
|
||||
|
||||
int main(int argc, char *argv[]) {
|
||||
size_t ctype; // selected from ctypes.h; index of ctype
|
||||
int retval; // initially fail but becomes success on the first valid char
|
||||
program_name = argv[0] == NULL ? program_name : argv[0];
|
||||
|
||||
#ifdef __OpenBSD__
|
||||
if (pledge("stdio", NULL) == -1) {
|
||||
perror(program_name);
|
||||
return EX_OSERR;
|
||||
}
|
||||
#endif
|
||||
|
||||
if (argc < 3) { return usage(program_name); }
|
||||
|
||||
for ( /* iterate ctypes */
|
||||
ctype = 0;
|
||||
ctypes[ctype].f != NULL /* break at the end of ctypes */
|
||||
&& strcmp(argv[1], ctypes[ctype].name) != 0; /* break at match */
|
||||
++ctype
|
||||
);
|
||||
|
||||
if (ctypes[ctype].f == NULL) { return usage(argv[0]); }
|
||||
|
||||
/* iterate args */
|
||||
for (argv += 2, retval = EXIT_FAILURE; *argv != NULL; ++argv) {
|
||||
for (size_t i = 0; argv[0][i] != '\0'; ++i) { /* iterate arg bytes */
|
||||
/* First checks if argv[0][i] is valid ASCII; ctypes(3) don't
|
||||
* handle non-ASCII. This is bad. */
|
||||
if(
|
||||
(unsigned char)argv[0][i] < 0x80 // argv[0][i] is ASCII,
|
||||
&& !ctypes[ctype].f(argv[0][i]) // so use ctypes(3)
|
||||
) { return EXIT_FAILURE; }
|
||||
else { retval = EXIT_SUCCESS; }
|
||||
}
|
||||
}
|
||||
|
||||
return retval;
|
||||
}
|
||||
87
src/stris.rs
Normal file
87
src/stris.rs
Normal file
@ -0,0 +1,87 @@
|
||||
/*
|
||||
* Copyright (c) 2023–2024 DTB <trinity@trinity.moe>
|
||||
* Copyright (c) 2023 Marceline Cramer <mars@tebibyte.media>
|
||||
* SPDX-License-Identifier: AGPL-3.0-or-later
|
||||
*
|
||||
* This program is free software: you can redistribute it and/or modify it under
|
||||
* the terms of the GNU Affero General Public License as published by the Free
|
||||
* Software Foundation, either version 3 of the License, or (at your option) any
|
||||
* later version.
|
||||
*
|
||||
* This program is distributed in the hope that it will be useful, but WITHOUT
|
||||
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
|
||||
* FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more
|
||||
* details.
|
||||
*
|
||||
* You should have received a copy of the GNU Affero General Public License
|
||||
* along with this program. If not, see https://www.gnu.org/licenses/.
|
||||
*/
|
||||
|
||||
use std::{
|
||||
env::args,
|
||||
process::ExitCode,
|
||||
};
|
||||
|
||||
extern crate getopt;
|
||||
extern crate sysexits;
|
||||
|
||||
use getopt::GetOpt;
|
||||
use sysexits::EX_USAGE;
|
||||
|
||||
struct Reqs {
|
||||
ascii: bool, blank: bool, cntrl: bool, digit: bool, lower: bool,
|
||||
upper: bool, inuse: bool, extra: String
|
||||
}
|
||||
|
||||
fn usage(s: &str) -> ExitCode {
|
||||
eprintln!("Usage: {} [-7bcdlu] [-i inclusions] [strings...]", s);
|
||||
ExitCode::from(EX_USAGE as u8)
|
||||
}
|
||||
|
||||
fn main() -> ExitCode {
|
||||
let argv = args().collect::<Vec<String>>();
|
||||
let mut optind = 1;
|
||||
let mut reqs = Reqs {
|
||||
ascii: false, blank: false, cntrl: false, digit: false, lower: false,
|
||||
upper: false, inuse: false, extra: String::default()
|
||||
};
|
||||
|
||||
while let Some(opt) = argv.getopt("7bcdi:lu") {
|
||||
match opt.opt() {
|
||||
Ok("7") => reqs.ascii = true,
|
||||
Ok("b") => reqs.blank = true,
|
||||
Ok("c") => reqs.cntrl = true,
|
||||
Ok("d") => reqs.digit = true,
|
||||
Ok("i") => reqs.extra = opt.arg().unwrap(),
|
||||
Ok("l") => reqs.lower = true,
|
||||
Ok("u") => reqs.upper = true,
|
||||
_ => { return usage(&argv[0]); }
|
||||
}
|
||||
optind = opt.ind();
|
||||
reqs.inuse = true;
|
||||
}
|
||||
|
||||
if argv.len() == optind { return usage(&argv[0]); }
|
||||
|
||||
drop(argv);
|
||||
|
||||
if reqs.inuse {
|
||||
for arg in args().skip(optind) {
|
||||
for c in arg.chars() {
|
||||
if (reqs.ascii && c.is_ascii())
|
||||
|| (reqs.blank && c.is_whitespace())
|
||||
|| (reqs.cntrl && c.is_control())
|
||||
|| (reqs.digit && c.is_numeric())
|
||||
|| (reqs.lower && c.is_lowercase())
|
||||
|| (reqs.upper && c.is_uppercase())
|
||||
|| reqs.extra.contains(c) {
|
||||
continue;
|
||||
} else {
|
||||
return ExitCode::FAILURE;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
ExitCode::SUCCESS
|
||||
}
|
||||
@ -7,14 +7,19 @@
|
||||
|
||||
.PRAGMA: command_comment
|
||||
|
||||
.PHONY: str_tests
|
||||
str_tests: str_help str_isalpha
|
||||
.PHONY: stris_tests
|
||||
stris_tests: stris_help stris_7 stris_b
|
||||
|
||||
.PHONY: str_help
|
||||
str_help: $(BIN)/str
|
||||
! $(BIN)/str -h
|
||||
.PHONY: stris_help
|
||||
stris_help: $(BIN)/stris
|
||||
! $(BIN)/stris -h
|
||||
|
||||
.PHONY: str_isalpha
|
||||
str_isalpha: $(BIN)/str
|
||||
$(BIN)/str isalpha c
|
||||
! $(BIN)/str isalpha 3
|
||||
.PHONY: stris_7
|
||||
stris_7: $(BIN)/stris
|
||||
$(BIN)/stris -7 !1Aa' '
|
||||
! $(BIN)/stris -7 今日は
|
||||
|
||||
.PHONY: stris_b
|
||||
stris_b: $(BIN)/stris
|
||||
$(BIN)/stris -b "$(printf ' \t\v\r\n')"
|
||||
! $(BIN)/stris -b !1Aa
|
||||
Loading…
Reference in New Issue
Block a user