stris(1): fix bug only checking the first rune of strings

Makefile

@@ -28,7 +28,7 @@ RUSTLIBS = --extern getopt=build/o/libgetopt.rlib \
 CFLAGS += -I$(SYSEXITS)
 
 .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
 
 build:
 	# keep build/include until bindgen(1) has stdin support
@@ -102,15 +102,11 @@ mm: build/bin/mm
 build/bin/mm: src/mm.c build
 	$(CC) $(CFLAGS) -o $@ src/mm.c
 
+
 .PHONY: npc
 npc: build/bin/npc
 build/bin/npc: src/npc.c build
-	$(CC) $(CFLAGS) -o $@ src/npc.c
-
-.PHONY: peek
-peek: build/bin/peek
-build/bin/peek: src/peek.c build
-	$(CC) $(CFLAGS) -o $@ src/peek.c
+	$(CC) $(CFLAGAS) -o $@ src/npc.c
 
 .PHONY: rpn
 rpn: build/bin/rpn
@@ -122,10 +118,13 @@ 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 build/o/libgetopt.rlib \
+		build/o/libsysexits.rlib
+	$(RUSTC) $(RUSTFLAGS) --extern getopt=build/o/libgetopt.rlib \
+		--extern sysexits=build/o/libsysexits.rlib \
+		-o $@ src/stris.rs
 
 .PHONY: strcmp
 strcmp: build/bin/strcmp
@@ -134,7 +133,8 @@ build/bin/strcmp: src/strcmp.c build
 
 .PHONY: swab
 swab: build/bin/swab
-build/bin/swab: src/swab.rs build build/o/libsysexits.rlib
+build/bin/swab: src/swab.rs build build/o/libgetopt.rlib \
+		build/o/libsysexits.rlib
 	$(RUSTC) $(RUSTFLAGS) --extern getopt=build/o/libgetopt.rlib \
 		--extern sysexits=build/o/libsysexits.rlib \
 		-o $@ src/swab.rs

docs/peek.1

@@ -1,90 +0,0 @@
-.\" Copyright (c) 2023-2024 DTB <trinity@trinity.moe>
-.\"
-.\" 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 PEEK 1
-
-.SH NAME
-
-peek \(en read from the standard input, furtively
-
-.SH SYNOPSIS
-
-peek
-.RB ( -i )
-
-.SH DESCRIPTION
-
-Peek reads input from the standard input with terminal echo disabled.
-
-.SH OPTIONS
-
-.B -i
-.RS
-Allows input to come from sources other than terminals (i.e. a pipe).
-.RE
-
-.SH DIAGNOSTICS
-
-Peek prints an error message and exits with the appropriate status from
-sysexits.h(3) if it encounters an unrecoverable error.
-
-Peek tries to handle the signal SIGINT (^C) to ensure the user's terminal is
-still usable after premature termination; if the signal can't be handled, it
-prints an error message and continues. If peek is interrupted, it exits
-unsuccessfully, without an error message.
-
-.SH RATIONALE
-
-This tool was originally written to accept passwords in shell scripts, as an
-extremely simple alternative to the GNU Privacy Guard project's pinentry(1).
-Accepting input without showing what is being typed is useful when keying in
-secrets in public settings or places with installed surveillance cameras.
-
-.SH BUGS
-
-This does nothing to prevent others seeing the keyboard being used to input
-secrets or mask the sound of typing. Audio or video recordings of typing can be
-used to determine what was input without needing to see the characters appear
-on the screen.
-
-Accepting secrets in shell scripts is probably not adviseable.
-
-On systems that support it, the ioctl(2) command TIOCSTI can be used to insert
-characters into the standard input going to peek. This doesn't allow snooping
-but can be used for general mischief.
-
-.SH EXAMPLES
-
-This is an sh(1p) command line that hashes a given password. It uses head(1p)
-to only accept one line of input, xargs(1p) and printf(1p) to strip the
-trailing newline, htpasswd(1) from Apache's utilities to hash the input with
-the bcrypt algorithm, and cut(1p) to print only the resulting hash:
-
-.RS
-.R $ peek | head -n 1 | xargs printf '%s' | htpasswd -nBi _ | cut -d : -f 2
-.RE
-
-This is an sh(1p) command line that lets a user blindly write into a text file,
-only able to see written lines. Some writers have the habit of prematurely
-revising their work and use tools like this to prevent it. It uses mm(1) to
-pipe the output of peek to both the standard error and the regular file
-writing.txt.
-
-.RS
-.R $ echo Input ^D to quit. && peek | mm -eo - >writing.txt
-.RE
-
-.SH AUTHOR
-
-Written by DTB <trinity@trinity.moe>.
-
-.SH COPYRIGHT
-
-Copyright (c) 2023-2024 DTB. License AGPLv3+: GNU AGPL version 3 or later
-<https://gnu.org/licenses/gpl.html>.
-
-.SH SEE ALSO
-
-ioctl(2), ioctl_tty(2), read(1), sh(1)

docs/str.1

@@ -1,58 +0,0 @@
-.\" 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 STR 1
-
-.SH NAME
-
-str \(en test the character types of string arguments
-
-.SH SYNOPSIS
-
-str
-.RB [ type ]
-.RB [ string... ]
-
-.SH DESCRIPTION
-
-Str tests each character in an arbitrary quantity of string arguments against
-the function of the same name within ctype(3).
-
-.SH DIAGNOSTICS
-
-Str exits successfully if all tests pass and unsuccessfully if a test failed.
-.PP
-Str will exit unsuccessfully if a string is empty, as none of its contents
-passed the test.
-.PP
-Str will print a message to standard error and exit unsuccessfully if used
-improperly.
-
-.SH DEPRECATED FEATURES
-
-Str used to have an "isvalue" type as an extension to ctype(3). This was
-removed in favor of using strcmp(1) to compare strings against the empty string
-('').
-
-.SH BUGS
-
-There's no way of knowing which argument failed the test without re-testing
-arguments individually.
-.PP
-If a character in a string isn't valid ASCII str will exit unsuccessfully.
-
-.SH AUTHOR
-
-Written by DTB <trinity@trinity.moe>.
-
-.SH COPYRIGHT
-
-Copyright © 2023 DTB. License AGPLv3+: GNU AGPL version 3 or later
-<https://gnu.org/licenses/gpl.html>.
-
-.SH SEE ALSO
-
-ctype(3p), strcmp(1), ascii(7)

docs/stris.1

@@ -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)

src/mm.c

@@ -106,6 +106,15 @@ oserr(char *s, char *r){
 	} \
 	return retval
 
+/* Prints a usage text, in which s is the program being run (i.e. argv[0]), and
+ * returns an exit status appropriate for a usage error. */
+int usage(char *s){
+
+	fprintf(stderr, "Usage: %s (-aenu) (-i [input])... (-o [output])...\n", s);
+
+	return EX_USAGE;
+}
+
 int main(int argc, char *argv[]){
 	int c;
 	struct Files files[2]; /* {read, write} */
@@ -178,12 +187,15 @@ int main(int argc, char *argv[]){
 				k = 1;
 				break;
 			default:
-				fprintf(stderr, "Usage: %s (-aenu) (-i [input])..."
-					" (-o [output])...\n", argv[0]);
-				retval = EX_USAGE;
+				retval = usage(argv[0]);
 				terminate;
 			}
 
+	if(optind != argc){
+		retval = usage(argv[0]);
+		terminate;
+	}
+
 	files[0].s += files[0].s == 0;
 	files[1].s += files[1].s == 0;
 	

src/peek.c

@@ -1,118 +0,0 @@
-/*
- * Copyright (c) 2023–2024 DTB <trinity@trinity.moe>
- * 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 <signal.h> /* sigaction(2), signal(2), struct sigaction, SIGINT */
-#include <stdio.h> /* fprintf(3), fgetc(3), perror(3), fputc(3), stderr, stdin,
-                    * stdout, EOF, NULL */
-#include <stdlib.h> /* exit(3), EXIT_FAILURE */
-#if !defined EX_IOERR || !defined EX_OK || !defined EX_USAGE
-#	include <sysexits.h>
-#endif
-#include <termios.h> /* tcgetattr(3), tcsetattr(3), struct termios, ECHO */
-#include <unistd.h> /* getopt(3), isatty(3), STDIN_FILENO */
-
-static char *program_name = "peek";
-
-/* Restores terminal echo; otherwise when a user ^Cs the terminal would
- * continue to not display typed text. If sig isn't zero, this will terminate
- * the program. */
-static void restore_echo(int sig){
-	static struct termios t;
-
-	tcgetattr(STDIN_FILENO, &t);
-	t.c_lflag |= ECHO;
-	tcsetattr(STDIN_FILENO, TCSAFLUSH, &t);
-
-	/* If, for whatever ungodly reason, exit(3) returns, the user will notice
-	 * their typed characters on the screen. */
-	if(sig != 0)
-		exit(EXIT_FAILURE);
-
-	return;
-}
-
-static int ioerr(char *s){ perror(s); restore_echo(0); return EX_IOERR; }
-static int usage(char *s){
-	fprintf(stderr, "Usage: %s (-1enot) (-p [program [arguments...]])\n", s);
-	return EX_USAGE;
-}
-
-int main(int argc, char *argv[]){
-
-	if(argc < 1)
-		return usage(program_name);
-
-	{ /* option parsing */
-		char allow_nonterminals;
-		int c;
-
-		allow_nonterminals = 0;
-		while((c = getopt(argc, argv, "i")) != -1)
-			switch(c){
-			case 'i':	allow_nonterminals = 1; break;
-			default:	return usage(argv[0]);
-			}
-
-		if(argc > optind)
-			return usage(argv[0]);
-
-		if(!allow_nonterminals && isatty(STDIN_FILENO) != 1){
-			fprintf(stderr, "%s: Must be run in a terminal"
-			    " (option -i skips this check)\n", argv[0]);
-			return EX_USAGE;
-		}
-	}
-
-	{ /* install signal handler */
-		/* There isn't a difference in functionality between the signal(2) and
-		 * sigaction(2) methods. sigaction(2) is vastly preferred for
-		 * portability but some configurations can only use signal(2). */
-		/* Errors aren't terminating because the worst that happens is some
-		 * terminal phooeyness if things go awry. */
-#if defined _POSIX_C_SOURCE
-		struct sigaction act = { 0 };
-
-		act.sa_handler = restore_echo;
-		if(sigaction(SIGINT, &act, NULL) != 0)
-			perror(argv[0]);
-#else
-		if(signal(SIGINT, restore_echo) == SIG_ERR)
-			perror(argv[0]);
-#endif
-	}
-	
-	{ /* banish terminal echo */
-		struct termios t;
-
-		tcgetattr(STDIN_FILENO, &t);
-		t.c_lflag ^= ECHO;
-		tcsetattr(STDIN_FILENO, TCSAFLUSH, &t);
-	}
-
-	{ /* actual input loop */
-		int c;
-
-		while((c = fgetc(stdin)) != EOF)
-			if(fputc(c, stdout) == EOF)
-				return ioerr(argv[0]);
-	}
-
-	restore_echo(0);
-
-	return EX_OK;
-}

src/scrut.c

@@ -1,5 +1,6 @@
 /*
- * Copyright (c) 2023 DTB <trinity@trinity.moe>
+ * Copyright (c) 2023–2024 DTB <trinity@trinity.moe>
+ * Copyright (c) 2024 Emma Tebibyte <emma@tebibyte.media>
  * SPDX-License-Identifier: AGPL-3.0-or-later
  *
  * This program is free software: you can redistribute it and/or modify it under
@@ -17,13 +18,15 @@
  */
 
 #include <stdio.h> /* fprintf(3), stderr, NULL */
-#include <stdlib.h> /* EXIT_FAILURE */
+#include <stdlib.h> /* EXIT_FAILURE, EXIT_SUCCESS */
 #include <string.h> /* memset(3), strchr(3) */
+#ifndef EX_USAGE
+#	include <sysexits.h>
+#endif
 #include <unistd.h> /* access(3), getopt(3), F_OK, R_OK, W_OK, X_OK */
 #include <sys/stat.h> /* lstat(3), stat struct, S_ISBLK, S_ISCHR, S_ISDIR,
                        * S_ISFIFO, S_ISGID, S_ISREG, S_ISLNK, S_ISSOCK,
                        * S_ISUID, S_ISVTX */
-#include <sysexits.h>
 
 static char args[] = "bcdefghkprsuwxLS";
 static char ops[(sizeof args) / (sizeof *args)];
@@ -57,7 +60,7 @@ int main(int argc, char *argv[]){
 
 	argv += optind;
 	do{	if(access(*argv, F_OK) != 0 || lstat(*argv, &buf) == -1)
-			return 1; /* doesn't exist or isn't stattable */
+			return EXIT_FAILURE; /* doesn't exist or isn't stattable */
 
 		for(i = 0; ops[i] != '\0'; ++i)
 			if(ops[i] == 'e')
@@ -97,8 +100,8 @@ usage:				fprintf(stderr, "Usage: %s (-%s) [file...]\n",
 						&& !S_ISLNK(buf.st_mode))
 					|| (ops[i] == 'S'
 						&& !S_ISSOCK(buf.st_mode)))
-				return 1;
+				return EXIT_FAILURE;
 	}while(*++argv != NULL);
 
-	return 0;
+	return EXIT_SUCCESS;
 }

src/str.c

@@ -1,75 +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) */
-#include <stdlib.h> /* EXIT_FAILURE */
-#include <string.h> /* strcmp(3) */
-#include <sysexits.h>
-
-static 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 }
-};
-
-int main(int argc, char *argv[]){
-	int ctype;
-	int i;
-	int r;
-
-	if(argc >= 3){
-		for(ctype = 0; ctype < (sizeof ctypes) / (sizeof *ctypes);
-				++ctype)
-			if(strcmp(argv[1], ctypes[ctype].name) == 0)
-				goto pass;
-	}
-
-	fprintf(stderr, "Usage: %s [type] [string...]\n",
-		argv[0] == NULL ? program_name : argv[0]);
-
-	return EX_USAGE;
-
-pass:	for(argv += 2, r = 1; *argv != NULL; ++argv)
-		for(i = 0; argv[0][i] != '\0'; ++i)
-			/* 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 && !ctypes[ctype].f(argv[0][i]))
-				return 1;
-			else
-				r = 0;
-
-	return r;
-}

src/stris.rs

@@ -0,0 +1,93 @@
+/*
+ * 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;
+use getopt::{ Opt, Parser };
+
+extern crate sysexits;
+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 opts = Parser::new(&argv, "7bcdi:lu");
+	let mut reqs = Reqs {
+		ascii: false, blank: false, cntrl: false, digit: false, lower: false,
+		upper: false, inuse: false, extra: String::new()
+	};
+
+	loop {
+		match opts.next() {
+			None => break,
+			Some(opt) => {
+				match opt {
+					Ok(Opt('7', None)) => reqs.ascii = true,
+					Ok(Opt('b', None)) => reqs.blank = true,
+					Ok(Opt('c', None)) => reqs.cntrl = true,
+					Ok(Opt('d', None)) => reqs.digit = true,
+					Ok(Opt('i', Some(arg))) => reqs.extra = arg,
+					Ok(Opt('l', None)) => reqs.lower = true,
+					Ok(Opt('u', None)) => reqs.upper = true,
+					_ => { return usage(&argv[0]); }
+				}
+				reqs.inuse = true;
+			}
+		}
+	}
+
+	if argv.len() == opts.index() {
+		return usage(&argv[0]);
+	}
+
+	drop(argv);
+
+	if reqs.inuse {
+		for arg in args().skip(opts.index()) {
+			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
+}