pg(1): start work on a quoted strtok(3) and internal environment

Makefile

@@ -1,6 +1,7 @@
 # Copyright (c) 2023–2024 Emma Tebibyte <emma@tebibyte.media>
 # Copyright (c) 2023–2024 DTB <trinity@trinity.moe>
 # Copyright (c) 2023 Sasha Koshka <sashakoshka@tebibyte.media>
+# Copyright (c) 2024 Aaditya Aryal <aryalaadi123@gmail.com>
 # SPDX-License-Identifier: FSFAP
 #
 # Copying and distribution of this file, with or without modification, are
@@ -8,16 +9,26 @@
 # notice are preserved.  This file is offered as-is, without any warranty.
 
 .POSIX:
+
+# if using BSD make(1), remove these pragmas because they break it
 .PRAGMA: posix_202x # future POSIX standard support à la pdpmake(1)
 .PRAGMA: command_comment # breaks without this?
 
-PREFIX=/usr/local
+DESTDIR ?= dist
+PREFIX ?= /usr/local
 
-CC=cc
+SYSEXITS != printf '\043include <sysexits.h>\n' | cpp -M - | sed 's/ /\n/g' \
-RUSTC=rustc
+	| sed -n 's/sysexits\.h//p' || printf 'include\n'
+
+CC ?= cc
+RUSTC ?= rustc
+RUSTLIBS = --extern getopt=build/o/libgetopt.rlib \
+	--extern sysexits=build/o/libsysexits.rlib \
+	--extern strerror=build/o/libstrerror.rlib
+CFLAGS += -I$(SYSEXITS)
 
 .PHONY: all
-all: dj false fop hru intcmp rpn scrut str strcmp swab true
+all: dj false fop hru intcmp mm npc rpn scrut str strcmp swab true
 
 build:
 	# keep build/include until bindgen(1) has stdin support
@@ -26,36 +37,40 @@ build:
 
 .PHONY: clean
 clean:
-	rm -rf build/ dist/
+	rm -rf build dist
 
 dist: all
-	mkdir -p dist/bin dist/share/man/man1
+	mkdir -p $(DESTDIR)/$(PREFIX)/bin $(DESTDIR)/$(PREFIX)/share/man/man1
-	cp build/bin/* dist/bin/
+	cp build/bin/* $(DESTDIR)/$(PREFIX)/bin
-	cp docs/*.1 dist/share/man/man1/
+	cp docs/*.1 $(DESTDIR)/$(PREFIX)/share/man/man1
 
 .PHONY: install
 install: dist
-	mkdir -p $(PREFIX)
+	cp -r $(DESTDIR)/* /
-	cp -r dist/* $(PREFIX)/
 
 .PHONY: test
 test: build
 	tests/posix-compat.sh
 	$(RUSTC) --test src/getopt-rs/lib.rs -o build/test/getopt
 
-build/o/libsysexits.rlib: build
+.PHONY: rustlibs
+rustlibs: build/o/libsysexits.rlib build/o/libgetopt.rlib \
+	build/o/libstrerror.rlib
+
+build/o/libgetopt.rlib: build src/getopt-rs/lib.rs
+	$(RUSTC) $(RUSTFLAGS) --crate-type=lib --crate-name=getopt \
+		-o $@ src/getopt-rs/lib.rs
+
+build/o/libstrerror.rlib: build src/strerror.rs
+	$(RUSTC) $(RUSTFLAGS) --crate-type=lib -o $@ \
+		src/strerror.rs
+
+build/o/libsysexits.rlib: build $(SYSEXITS)sysexits.h
 	# bandage solution until bindgen(1) gets stdin support
-	printf '#define EXIT_FAILURE 1\n' | cat - include/sysexits.h \
+	printf '#define EXIT_FAILURE 1\n' | cat - $(SYSEXITS)sysexits.h \
 		> build/include/sysexits.h 
 	bindgen --default-macro-constant-type signed --use-core --formatter=none \
-		"$$(printf '#include <sysexits.h>\n' \
+		build/include/sysexits.h | $(RUSTC) $(RUSTFLAGS) --crate-type lib -o $@ -
-			| cpp -M -idirafter "build/include" - \
-			| sed 's/ /\n/g' | grep sysexits.h)" \
-		| $(RUSTC) $(RUSTFLAGS) --crate-type lib -o build/o/libsysexits.rlib -
-
-build/o/libgetopt.rlib: src/getopt-rs/lib.rs
-	$(RUSTC) $(RUSTFLAGS) --crate-type=lib --crate-name=getopt \
-		-o build/o/libgetopt.rlib src/getopt-rs/lib.rs
 
 .PHONY: dj
 dj: build/bin/dj
@@ -69,29 +84,38 @@ build/bin/false: src/false.c build
 
 .PHONY: fop
 fop: build/bin/fop
-build/bin/fop: src/fop.rs build build/o/libgetopt.rlib build/o/libsysexits.rlib
+build/bin/fop: src/fop.rs build rustlibs
-	$(RUSTC) $(RUSTFLAGS) --extern getopt=build/o/libgetopt.rlib \
+	$(RUSTC) $(RUSTFLAGS) $(RUSTLIBS) -o $@ src/fop.rs
-		--extern sysexits=build/o/libsysexits.rlib \
-		-o $@ src/fop.rs
 
 .PHONY: hru
 hru: build/bin/hru
-build/bin/hru: src/hru.rs build build/o/libgetopt.rlib build/o/libsysexits.rlib
+build/bin/hru: src/hru.rs build rustlibs 
-	$(RUSTC) $(RUSTFLAGS) --extern getopt=build/o/libgetopt.rlib \
+	$(RUSTC) $(RUSTFLAGS) $(RUSTLIBS) -o $@ src/hru.rs
-		--extern sysexits=build/o/libsysexits.rlib \
-		-o $@ src/hru.rs
 
 .PHONY: intcmp
 intcmp: build/bin/intcmp
 build/bin/intcmp: src/intcmp.c build
 	$(CC) $(CFLAGS) -o $@ src/intcmp.c
 
+.PHONY: mm
+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: pg
+pg: build/bin/pg
+build/bin/pg: src/pg.c build
+	$(CC) $(CFLAGS) -o $@ src/pg.c
+
 .PHONY: rpn
 rpn: build/bin/rpn
-build/bin/rpn: src/rpn.rs build build/o/libsysexits.rlib
+build/bin/rpn: src/rpn.rs build rustlibs
-	$(RUSTC) $(RUSTFLAGS) \
+	$(RUSTC) $(RUSTFLAGS) $(RUSTLIBS) -o $@ src/rpn.rs
-		--extern sysexits=build/o/libsysexits.rlib \
-		-o $@ src/rpn.rs
 
 .PHONY: scrut
 scrut: build/bin/scrut

docs/mm.1

@@ -0,0 +1,76 @@
+.\" Copyright (c) 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 mm 1
+
+.SH NAME
+
+mm \(en middleman
+
+.SH SYNOPSIS
+
+mm
+.RB ( -aenu )
+.RB ( -i
+.RB [ input ])
+.RB ( -o
+.RB [ output ])
+
+.SH DESCRIPTION
+
+Mm catenates input files and writes them to the start of each output file.
+
+.SH OPTIONS
+
+Mm, upon receiving the
+.B -a
+option, will open subsequent outputs for appending rather than updating.
+.PP
+The
+.B -i
+option opens a path as an input. Without any inputs specified mm will use
+standard input. Standard input itself can be specified by giving the path '-'.
+.PP
+The
+.B -o
+option opens a path as an output. Without any outputs specified mm will use
+standard output. Standard output itself can be specified by giving the
+path '-'. Standard error itself can be specified with the
+.B -e
+option.
+.PP
+The
+.B -u
+option ensures neither input or output will be buffered.
+.PP
+The
+.B -n
+option tells mm to ignore SIGINT signals.
+
+.SH DIAGNOSTICS
+
+If an output can no longer be written mm prints a diagnostic message, ceases
+writing to that particular output, and if there are more outputs specified,
+continues, eventually exiting unsuccessfully.
+.PP
+On error mm prints a diagnostic message and exits with the appropriate
+sysexits.h(3) status.
+
+.SH BUGS
+
+Mm does not truncate existing files, which may lead to unexpected results.
+
+.SH RATIONALE
+
+Mm was modeled after the cat and tee utilities specified in POSIX.
+
+.SH COPYRIGHT
+
+Copyright (c) 2024 DTB. License AGPLv3+: GNU AGPL version 3 or later
+<https://gnu.org/licenses/agpl.html>.
+
+.SH SEE ALSO
+
+cat(1p), dd(1), dj(1), tee(1p)

docs/pg.1

@@ -0,0 +1,41 @@
+.\" Copyright (c) 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 PG 1
+
+.SH NAME
+
+pg \(en paginate
+
+.SH SYNOPSIS
+
+pg
+.RB ( -p
+.RB [ prompt ])
+
+.SH DESCRIPTION
+
+Pg prints standard input to standard output, accepting commands between pages.
+
+.SH OPTIONS
+
+.B -p
+.RS
+Replaces the default prompt (": ") with the next argument.
+.RE
+
+.SH DIAGNOSTICS
+
+Pg returns an unsuccessful exit code if the tty couldn't be opened or if pg was
+invoked incorrectly (with any arguments).
+
+.SH RATIONALE
+
+Plan 9 from Bell Labs had p(1), a similar "cooked"-mode paginator (as opposed
+to "raw" mode, which a vast majority of paginators use).
+
+.SH SEE ALSO
+
+more(1)

src/dj.c

@@ -203,28 +203,28 @@ Io_fdseek(struct Io *io){
 	
 	if(!fdisstd(io->fd) && lseek(io->fd, io->seek, SEEK_SET) != -1)
 		return -1;
-	else if(io->fl == write_flags){
-		memset(io->buf, '\0', io->bs);
-		/* This is a dirty trick; rather than testing conditions and operating
-		 * likewise, because the parameters to read or write are going to be
-		 * the same either way, just use a function pointer to keep track of
-		 * the intended operation. */
-		op = (int (*)(int, void *, size_t))&write;
-		/* Function pointer casts are risky; this works because the difference
-		 * is in the second parameter and only that write(2) makes the buffer
-		 * const whereas read(2) does not. To avoid even the slightest
-		 * undefined behavior comment out the cast, just be ready for a
-		 * -Wincompatible-function-pointer-types if your compiler notices it.
-		 */
-	}else
-		op = &read;
 
-	/* We're going to cheat and use bufuse as the retval for write(2), which is
+	/* repeated code to get the condition out of the loop */
-	 * fine because it'll be zeroed as this function returns anyway. */
+	if(io->fl == write_flags){
-	do{	if(	(io->bufuse = (*op)(io->fd, io->buf, MIN(io->bs, io->seek))) == 0)
+		memset(io->buf, '\0', io->bs);
+		/* We're going to cheat and use bufuse as the retval for write(2),
+		 * which is fine because it'll be zeroed as this function returns
+		 * anyway. */
+		do{
+			if((io->bufuse = write(io->fd, io->buf, MIN(io->bs, io->seek)))
+					== 0)
 				/* second chance */
-			io->bufuse = (*op)(io->fd, io->buf, MIN(io->bs, io->seek));
+				io->bufuse = write(io->fd, io->buf, MIN(io->bs, io->seek));
 		}while((io->seek -= io->bufuse) > 0 && io->bufuse != 0);
+	}else if(io->fl == read_flags){
+		do{
+			if((io->bufuse = read(io->fd, io->buf, MIN(io->bs, io->seek)))
+					== 0)
+				/* second chance */
+				io->bufuse = read(io->fd, io->buf, MIN(io->bs, io->seek));
+		}while((io->seek -= io->bufuse) > 0 && io->bufuse != 0);
+	}else
+		return EX_SOFTWARE;
 
 	io->bufuse = 0;
 

src/fop.rs

@@ -22,10 +22,12 @@ use std::{
 	process::{ Command, exit, Stdio },
 };
 
-extern crate sysexits;
 extern crate getopt;
+extern crate strerror;
+extern crate sysexits;
 
 use getopt::{ Opt, Parser };
+use strerror::StrError;
 use sysexits::{ EX_DATAERR, EX_IOERR, EX_UNAVAILABLE, EX_USAGE };
 
 fn main() {
@@ -55,7 +57,7 @@ fn main() {
 	});
 
 	let index = argv[index_arg].parse::<usize>().unwrap_or_else(|e| {
-		eprintln!("{}: {}: {}.", argv[0], argv[1], e);
+		eprintln!("{}: {}: {}", argv[0], argv[1], e);
 		exit(EX_DATAERR);
 	});
 
@@ -75,13 +77,13 @@ fn main() {
 		.stdout(Stdio::piped())
 		.spawn()
 		.unwrap_or_else( |e| {
-			eprintln!("{}: {}: {}.", argv[0], argv[command_arg], e);
+			eprintln!("{}: {}: {}", argv[0], argv[command_arg], e.strerror());
 			exit(EX_UNAVAILABLE);
 		});
 
 	let field = fields.get(index).unwrap_or_else(|| {
 		eprintln!(
-			"{}: {}: No such index in input.",
+			"{}: {}: No such index in input",
 			argv[0],
 			index.to_string(),
 		);
@@ -94,7 +96,7 @@ fn main() {
 	}
 
 	let output = spawned.wait_with_output().unwrap_or_else(|e| {
-		eprintln!("{}: {}: {}.", argv[0], argv[command_arg], e);
+		eprintln!("{}: {}: {}", argv[0], argv[command_arg], e.strerror());
 		exit(EX_IOERR);
 	});
 
@@ -103,7 +105,7 @@ fn main() {
 	if replace.pop() != Some(b'\n') { replace = output.stdout; }
 
 	let new_field = String::from_utf8(replace).unwrap_or_else(|e| {
-		eprintln!("{}: {}: {}.", argv[0], argv[command_arg], e);
+		eprintln!("{}: {}: {}", argv[0], argv[command_arg], e);
 		exit(EX_IOERR);
 	});
 
@@ -112,7 +114,7 @@ fn main() {
 	stdout().write_all(
 		fields.join(&d.to_string()).as_bytes()
 	).unwrap_or_else(|e| {
-		eprintln!("{}: {}.", argv[0], e);
+		eprintln!("{}: {}", argv[0], e.strerror());
 		exit(EX_IOERR);
 	});
 }

src/hru.rs

@@ -23,8 +23,10 @@ use std::{
 	process::{ ExitCode, exit },
 };
 
+extern crate strerror;
 extern crate sysexits;
 
+use strerror::StrError;
 use sysexits::{ EX_DATAERR, EX_IOERR, EX_SOFTWARE };
 
 const LIST: [(u32, &str); 10] = [
@@ -49,7 +51,7 @@ fn convert(input: u128) -> Result<(f64, (u32, &'static str)), String> {
 		let c = match 10_u128.checked_pow(n) {
 			Some(c) => c,
 			None => {
-				return Err(format!("10^{}: Integer overflow.", n.to_string()));
+				return Err(format!("10^{}: Integer overflow", n.to_string()));
 			},
 		};
 
@@ -79,7 +81,7 @@ fn main() -> ExitCode {
 				f
 			},
 			Err(err) => {
-				eprintln!("{}: {}.", argv[0], err);
+				eprintln!("{}: {}", argv[0], err);
 				return ExitCode::from(EX_DATAERR as u8);
 			},
 		};
@@ -87,7 +89,7 @@ fn main() -> ExitCode {
 		let (number, prefix) = match convert(n) {
 			Ok(x) => x,
 			Err(err) => {
-				eprintln!("{}: {}.", argv[0], err);
+				eprintln!("{}: {}", argv[0], err);
 				return ExitCode::from(EX_SOFTWARE as u8);
 			},
 		};
@@ -98,7 +100,7 @@ fn main() -> ExitCode {
 
 		stdout().write_all(format!("{} {}\n", out, si_prefix).as_bytes())
 			.unwrap_or_else(|e| {
-				eprintln!("{}: {}.", argv[0], e);
+				eprintln!("{}: {}", argv[0], e.strerror());
 				exit(EX_IOERR);
 			});
 	}

src/mm.c

@@ -0,0 +1,224 @@
+/*
+ * Copyright (c) 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 <errno.h> /* errno */
+#include <signal.h> /* signal(2), SIG_ERR, SIG_IGN, SIGINT */
+#include <stdio.h> /* fclose(3), fopen(3), fprintf(3), getc(3), putc(3),
+                    * setvbuf(3), size_t, _IONBF, NULL */
+#include <stdlib.h> /* free(3), realloc(3) */
+#include <string.h> /* strcmp(3), strerror(3) */
+#include <unistd.h> /* getopt(3) */
+#if !defined EX_IOERR || !defined EX_OK || !defined EX_OSERR \
+		|| !defined EX_USAGE
+#	include <sysexits.h>
+#endif
+extern int errno;
+
+/* This structure is how open files are tracked. */
+struct Files{
+	size_t a;	/* allocation */
+	size_t s;	/* used size */
+	char *mode;	/* file opening mode */
+	char **names;	/* file names */
+	FILE **files;	/* file pointers */
+};
+
+/* How much to grow the allocation when it's saturated. */
+#ifndef ALLOC_INCREMENT
+#	define ALLOC_INCREMENT 1
+#endif
+
+/* How much to grow the allocation at program start. */
+#ifndef ALLOC_INITIAL
+#	define ALLOC_INITIAL 10
+#endif
+
+/* pre-allocated strings */
+static char *program_name = "<no argv[0]>";
+static char *stdin_name = "<stdin>";
+static char *stdout_name = "<stdout>";
+static char *stderr_name = "<stderr>";
+static char *(fmode[]) = { (char []){"rb"}, (char []){"rb+"} };
+static char *wharsh = "wb";
+
+/* Adds the open FILE pointer for the file at the path s to the files struct,
+ * returning the FILE if successful and NULL if not, allocating more memory in
+ * the files buffers as needed. */
+static FILE *
+Files_append(struct Files *files, FILE *file, char *name){
+
+	if(file == NULL || (files->s == files->a
+			&& ((files->files = realloc(files->files,
+					(files->a += (files->a == 0)
+						? ALLOC_INITIAL
+						: ALLOC_INCREMENT)
+					* sizeof *(files->files))) == NULL
+				|| (files->names = realloc(files->names,
+					files->a * sizeof *(files->names))) == NULL)))
+		return NULL;
+
+	files->names[files->s] = name;
+	return files->files[files->s++] = file;
+}
+
+/* Opens the file at the path p and puts it in the files struct, returning NULL
+ * if either the opening or the placement of the open FILE pointer fail. */
+#define Files_open(files, p) \
+	Files_append((files), fopen((p), (files)->mode), (p))
+
+/* Prints a diagnostic message based on errno and returns an exit status
+ * appropriate for an OS error. */
+static int
+oserr(char *s, char *r){
+
+	fprintf(stderr, "%s: %s: %s\n", s, r, strerror(errno));
+
+	return EX_OSERR;
+}
+
+/* Hijacks i and j from main and destructs the files[2] struct used by main by
+ * closing its files and freeing its files and names arrays, returning retval
+ * from main. */
+#define terminate \
+	for(i = 0; i < 2; ++i){ \
+		for(j = 0; j < files[i].s; ++j) \
+			if(files[i].files[j] != stdin \
+					&& files[i].files[j] != stdout \
+					&& files[i].files[j] != stderr) \
+				fclose(files[i].files[j]); \
+		free(files[i].files); \
+		free(files[i].names); \
+	} \
+	return retval
+
+int main(int argc, char *argv[]){
+	int c;
+	struct Files files[2]; /* {read, write} */
+	size_t i;
+	size_t j;
+	size_t k; /* loop index but also unbuffer status */
+	int retval;
+
+	/* Initializes the files structs with their default values, standard
+	 * input and standard output. If an input or an output is specified
+	 * these initial values will be overwritten, so to, say, use mm(1)
+	 * equivalently to tee(1p), -o - will need to be specified before
+	 * additional files to ensure standard output is still written. */
+	for(i = 0; i < 2; ++i){
+		files[i].a = 0;
+		files[i].s = 0;
+		files[i].mode = fmode[i];
+		files[i].files = NULL;
+		files[i].names = NULL;
+		Files_append(&files[i], i == 0 ? stdin : stdout,
+			i == 0 ? stdin_name : stdout_name);
+		files[i].s = 0;
+	}
+
+	k = 0;
+
+	if(argc > 0)
+		program_name = argv[0];
+
+	if(argc > 1)
+		while((c = getopt(argc, argv, "aehi:no:u")) != -1)
+			switch(c){
+			case 'a': /* "rb+" -> "ab" */
+				files[1].mode[0] = 'a';
+				files[1].mode[2] = '\0';
+				break;
+			case 'e':
+				if(Files_append(&files[1], stderr, stderr_name) != NULL)
+					break;
+				retval = oserr(argv[0], "-e");
+				terminate;
+			case 'i':
+				if((strcmp(optarg, "-") == 0 && Files_append(&files[0],
+							stdin, stdin_name) != NULL)
+						|| Files_open(&files[0], optarg) != NULL)
+					break;
+				retval = oserr(argv[0], optarg);
+				terminate;
+			case 'o':
+				if((strcmp(optarg, "-") == 0 && Files_append(&files[1],
+							stdout, stdout_name) != NULL)
+						|| Files_open(&files[1], optarg) != NULL)
+					break;
+				/* does not exist, so try to create it */
+				if(errno == ENOENT){
+					files[1].mode = wharsh;
+					if(Files_open(&files[1], optarg) != NULL){
+						files[1].mode = fmode[1];
+						break;
+					}
+				}
+				retval = oserr(argv[0], optarg);
+				terminate;
+			case 'n':
+				if(signal(SIGINT, SIG_IGN) != SIG_ERR)
+					break;
+				retval = oserr(argv[0], "-n");
+				terminate;
+			case 'u':
+				k = 1;
+				break;
+			default:
+				fprintf(stderr, "Usage: %s (-aenu) (-i [input])..."
+					" (-o [output])...\n", argv[0]);
+				retval = EX_USAGE;
+				terminate;
+			}
+
+	files[0].s += files[0].s == 0;
+	files[1].s += files[1].s == 0;
+	
+	/* Unbuffer files. */
+	if(k){
+		for(i = 0;
+			i < files[0].s;
+			setvbuf(files[0].files[i++], NULL, _IONBF, 0));
+		for(i = 0;
+			i < files[1].s;
+			setvbuf(files[1].files[i++], NULL, _IONBF, 0));
+	}
+
+	retval = EX_OK;
+
+	/* Actual program loop. */
+	for(i = 0; i < files[0].s; ++i) /* iterate ins */
+		while((c = getc(files[0].files[i])) != EOF) /* iterate chars */
+			for(j = 0; j < files[1].s; ++j) /* iterate outs */
+				if(putc(c, files[1].files[j]) == EOF){
+					/* notebook's full */
+					retval = EX_IOERR;
+					fprintf(stderr, "%s: %s: %s\n",
+						program_name, files[1].names[j], strerror(errno));
+					if(fclose(files[1].files[j]) == EOF)
+						fprintf(stderr, "%s: %s: %s\n",
+							program_name, files[1].names[j], strerror(errno));
+					/* massage out the tense muscle */
+					for(k = j--; k < files[1].s - 1; ++k){
+						files[1].files[k] = files[1].files[k+1];
+						files[1].names[k] = files[1].names[k+1];
+					}
+					if(--files[1].s == 0)
+						terminate;
+				}
+
+	terminate;
+}

src/pg.c

@@ -0,0 +1,260 @@
+/*
+ * Copyright (c) 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 <errno.h> /* errno */
+#include <stdio.h> /* fclose(3), feof(3), fgetc(3), fgets(3), fopen(3),
+                    * fprintf(3), fputc(3), stderr, stdin, stdout, EOF, FILE,
+                    * NULL */
+#include <stdlib.h> /* size_t */
+#include <string.h> /* strchr(3), strcmp(3), strerror(3), strtok(3) */
+#include <unistd.h> /* getopt(3) */
+
+/* Commands start with cmd_. They take an argc and NULL-terminated argv, like
+ * main, and return a status from <sysexits.h>. Return values other than EX_OK
+ * and EX_USAGE cause pg(1) to exit with that value, except EX_UNAVAILABLE,
+ * which causes pg(1) to exit with the status EX_OK. */
+#include <sysexits.h>
+
+#define CMDLINE_MAX 99+1 /* Maximum length of command line. */
+
+static char *whitespace = " \n\r\t\v";
+
+static struct {
+	size_t quantity;
+	enum { LINES = 0, BYTES = 1 } type;
+} default_page_unit = { 22, LINES } /* Plan 9 default */;
+static FILE *input;
+static char *prompt = ": ";
+static char *program_name = "pg";
+
+static char *permute_out(char *str, size_t index){
+	size_t j;
+
+	for(j = index; str[j - 1] != '\0'; ++j)
+		str[j - 1] = str[j];
+
+	return str;
+}
+
+/* strtok(3p), but supports double-quotes and escapes (but only for escaping
+ * quotes). UTF-8 is safe only in str. Unmatched quotes in str are considered
+ * literal. The behavior of strtok_quoted when '"' or '\\' are in sep is
+ * undefined. */
+/* TODO: Seems to only ever return NULL. */
+static char *strtok_quoted(char *str, char *sep){
+	static char *s;
+	size_t i;
+	size_t j;
+
+	if(str != NULL)
+		s = str;
+
+	while(strchr(sep, *s) == NULL)
+		if(*++s == '\0')
+			return NULL; /* no remaining characters except seperators */
+
+	{
+		char in_escape;
+		int in_quotes;
+
+		in_escape = 0;
+		in_quotes = -1;
+		for(i = 0; s[i] != '\0'; ++i)
+			switch(s[i]){
+			case '\\':
+				if(in_escape)
+					permute_out(s, i--);
+				in_escape = !in_escape;
+				break;
+			case '"':
+				if(in_escape){
+					s[i] = s[i - 1];
+					permute_out(s, i--);
+				}else if(in_quotes != -1){
+					permute_out(s, in_quotes);
+					--i;
+					permute_out(s, i--);
+					in_quotes = -1;
+				}else
+					in_quotes = i;
+				break;
+			case '\0':
+				return s;
+			default:
+				if(!in_escape && strchr(sep, s[i]) != NULL){
+					s[i] = '\0';
+					return s;
+				}
+			}
+	}
+
+	return s;
+}
+
+/* Page at most l bytes from f without consideration of a buffer (print them to
+ * stdout). */
+static int pg_b_u(FILE *f, size_t l){
+	int c;
+
+	while((c = fgetc(f)) != EOF)
+		if((c = fputc(c, stdout)) == EOF || --l == 0)
+			break;
+
+	return c;
+}
+
+/* Page at most l lines, which are lengths of characters terminated by nl, from
+ * f without consideration of a buffer (print them to stdout). */
+static int pg_l_u(FILE *f, size_t l, char nl){
+	int c;
+
+	while((c = fgetc(f)) != EOF)
+		if((c = fputc(c, stdout)) == EOF || (l -= (c == nl)) == 0)
+			break;
+
+	return c;
+}
+
+static int cmd_quit(int argc, char **argv, char **envp){return EX_UNAVAILABLE;}
+static int cmd_default_page(int argc, char **argv, char **envp){
+
+	if(argc > 1) /* This shouldn't be possible. */
+		return EX_USAGE;
+	if(default_page_unit.type == BYTES)
+		return pg_b_u(input, default_page_unit.quantity) == EOF
+			? EX_UNAVAILABLE
+			: EX_OK;
+	else if(default_page_unit.type == LINES)
+		return pg_l_u(input, default_page_unit.quantity, '\n') == EOF
+			? EX_UNAVAILABLE
+			: EX_OK;
+	else
+		return EX_SOFTWARE;
+}
+static int cmd_page_down_lines(int argc, char **argv, char **envp){
+
+	switch(argc){
+	case 1:
+		return cmd_default_page(argc, argv, envp);
+	case 2: /* not implemented */
+	default:
+		fprintf(stderr, "Usage: %s" /*" (lines)"*/ "\n", argv[0]);
+		return EX_USAGE;
+	}
+}
+
+/* A CmdMap must be NULL-terminated. */
+static struct CmdMap{ char *name; int (*fn)(int, char **, char **); }
+builtins[] = {
+	{ "", cmd_default_page },
+	{ "+", cmd_page_down_lines },
+
+	/* don't make the user feel trapped */
+	{ ":q", cmd_quit }, { ":q!", cmd_quit },
+	{ "exit", cmd_quit }, { "q", cmd_quit }, { "Q", cmd_quit },
+	{ "quit", cmd_quit }, { "ZZ", cmd_quit },
+	
+	{ NULL, NULL }
+};
+
+#define ARGV_MAX 10
+/* Find and execute the command in the command map, given a corresponding
+ * command line. */
+static int cmdline_exec(struct CmdMap *map, char *cmdline, char **envp){
+	/* Command line word splitting is naive and based on whitespace ONLY; no
+	 * fancy quoting or escaping here. Adding that would (ideally) entail
+	 * replacing strtok(3) with something specific to this task. */
+	static int argc;
+	static char *argv[ARGV_MAX];
+
+	if((argv[(argc = 0)] = strtok(cmdline, whitespace)) == NULL){
+		while(cmdline[0] != '\0')
+			cmdline = &cmdline[1];
+		argv[argc] = cmdline;
+		argv[++argc] = NULL;
+	}else{
+		while((argv[++argc] = strtok(NULL, whitespace)) != NULL
+		      && argc < ARGV_MAX);
+	}
+
+	for(; map->name != NULL; map = &map[1])
+		if(strcmp(map->name, argv[0]) == 0)
+			return map->fn(argc, argv, envp);
+
+	fprintf(stderr, "%s: %s: not found\n", program_name, argv[0]);
+	return EX_USAGE;
+}
+
+int usage(char *s){
+	fprintf(stderr, "Usage: %s (-p [prompt])\n", s);
+	return EX_USAGE;
+}
+
+int main(int argc, char *argv[]){
+	unsigned char cmd[CMDLINE_MAX];
+	FILE *t;
+
+	input = stdin;
+
+	if(argc < 1)
+		program_name = argv[0];
+	else{
+		int c;
+
+		while((c = getopt(argc, argv, "p:")) != -1)
+			switch(c){
+			case 'p': prompt = optarg; break;
+			default:
+				return usage(program_name);
+			}
+	}
+
+	if(argc > optind)
+		return usage(program_name);
+
+
+	if((t = fopen("/dev/tty", "rb")) == NULL){
+		fprintf(stderr, "%s: /dev/tty: %s\n", program_name, strerror(errno));
+		return EX_OSERR;
+	}
+
+	for(;;){
+		fputs(prompt, stderr);
+		fgets((char *)cmd, (sizeof cmd) / (sizeof *cmd), t);
+		if(strchr((char *)cmd, '\n') == NULL){ /* fast-forward stream */
+			int c;
+
+			while((c = fgetc(t)) != '\n')
+				if(c == EOF)
+					break;
+		}
+
+		if(feof(t))
+			return EX_OK;
+
+		{	int r;
+			switch((r = cmdline_exec(builtins, (char *)cmd, NULL))){
+			case EX_OK: case EX_USAGE: break;
+			case EX_UNAVAILABLE: return EX_OK;
+			default: return r;
+		}	}
+	}
+
+	/* UNREACHABLE */
+	return EX_SOFTWARE;
+}

src/rpn.rs

@@ -172,7 +172,7 @@ fn eval(
 					};
 				} else {
 					return Err(EvaluationError {
-						message: format!("{}: Unexpected operation.", op),
+						message: format!("{}: Unexpected operation", op),
 						code: EX_DATAERR,
 					})
 				}

src/strerror.rs

@@ -0,0 +1,31 @@
+/*
+ * Copyright (c) 2024 Emma Tebibyte <emma@tebibyte.media>
+ * SPDX-License-Identifier: FSFAP
+ *
+ * Copying and distribution of this file, with or without modification, are
+ * permitted in any medium without royalty provided the copyright notice and
+ * this notice are preserved.  This file is offered as-is, without any warranty.
+ */
+
+use std::ffi::{ c_int, c_char, CStr };
+
+pub trait StrError { fn strerror(&self) -> String; }
+
+impl StrError for std::io::Error {
+	/* wrapper function for use in Rust */
+	fn strerror(&self) -> String {
+		/* Get the raw OS error. If it’s None, what the hell is going on‽ */
+		let errno = self.raw_os_error().unwrap_or(0) as c_int;
+
+		/* Get a CStr from the error message so that it’s referenced and then
+		 * convert it to an owned value. If the string is not valid UTF-8,
+		 * return that error instead. */
+		match unsafe { CStr::from_ptr(strerror(errno)) }.to_str() {
+			Ok(s) => s.to_owned(), // yay!! :D
+			Err(e) => e.to_string(), // awww :(
+		}
+	}
+}
+
+/* binding to strerror(3p) */
+extern "C" { fn strerror(errnum: c_int) -> *mut c_char; }

src/test.rs

@@ -0,0 +1,10 @@
+extern crate strerror;
+
+use strerror::raw_message;
+
+fn main() {
+	stdout.write_all(b"meow\n").unwrap_or_else(|e| {
+		eprintln!("{}", raw_message(e));
+		std::process::exit(1);
+	});
+}