mm(1): improves comments

Makefile

@@ -16,9 +16,12 @@
 DESTDIR ?= dist
 PREFIX ?= /usr/local
 
-MANDIR != [ $(PREFIX) = / ] && printf '/usr/share/man\n' \
+# normalized prefix
+PREFIX_N != (test -d $(PREFIX) && [ '-' != $(PREFIX) ] \
+	&& CDPATH= cd -P -- $(PREFIX) && pwd -P)
+MANDIR != [ $(PREFIX_N) = / ] && printf '/usr/share/man\n' \
 	|| printf '/share/man\n'
-SYSEXITS != printf '\043include <sysexits.h>\n' | cpp -M - | sed 's/ /\n/g' \
+SYSEXITS != printf '\043include <sysexits.h>\n' | cpp -M - | tr ' ' '\n' \
 	| sed -n 's/sysexits\.h//p' || printf 'include\n'
 
 CC ?= cc
@@ -29,7 +32,7 @@ RUSTLIBS = --extern getopt=build/o/libgetopt.rlib \
 CFLAGS += -I$(SYSEXITS)
 
 .PHONY: all
-all: docs dj false fop hru intcmp mm npc rpn scrut str strcmp swab true
+all: dj false fop hru intcmp mm npc rpn scrut str strcmp swab true
 
 # keep build/include until bindgen(1) has stdin support
 # https://github.com/rust-lang/rust-bindgen/issues/2703
@@ -40,8 +43,8 @@ build:
 clean:
 	rm -rf build dist
 
-dist: all
-	mkdir -p $(DESTDIR)/$(PREFIX)/bin $(DESTDIR)/$(PREFIX)/share/man/man1
+dist: all docs
+	mkdir -p $(DESTDIR)/$(PREFIX)/bin $(DESTDIR)/$(PREFIX)/$(MANDIR)/man1
 	cp build/bin/* $(DESTDIR)/$(PREFIX)/bin
 	cp build/docs/*.1 $(DESTDIR)/$(PREFIX)/$(MANDIR)/man1
 
@@ -111,8 +114,8 @@ build/bin/intcmp: src/intcmp.c build
 
 .PHONY: mm
 mm: build/bin/mm
-build/bin/mm: src/mm.c build
-	$(CC) $(CFLAGS) -o $@ src/mm.c
+build/bin/mm: src/mm.rs build rustlibs 
+	$(RUSTC) $(RUSTFLAGS) $(RUSTLIBS) -o $@ src/mm.rs
 
 .PHONY: npc
 npc: build/bin/npc

docs/dj.1

@@ -4,7 +4,7 @@
 .\" 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 DJ 1 2024-06-29 "Harakit X.X.X"
+.TH DJ 1 2024-07-03 "Harakit X.X.X"
 .SH NAME
 dj \(en disk jockey
 .\"
@@ -34,8 +34,13 @@ respectively. This language is inherited from the
 .BR dd (1p)
 utility and used here to decrease ambiguity.
 
-When seeking or skipping to a byte, writing or reading starts at the byte
-immediately subsequent to the specified byte.
+The offset used when skipping or seeking refers to how many bytes are skipped
+or sought. Running
+.BR dj (1)
+with a skip offset of 1 skips one byte into the input and reads from the second
+byte onwards. A programmer may think of a file as a zero-indexed array of
+bytes; in this analogy, the offset given is the index of the byte at which to
+start reading or writing.
 .\"
 .SH OPTIONS
 
@@ -80,6 +85,54 @@ input file is \(lq-\(rq.
 The standard output shall be used as an output if no inputs are specified or if
 the output file is \(lq-\(rq.
 .\"
+.SH EXAMPLES
+
+The following
+.BR sh (1p)
+line:
+
+.RS
+printf 'Hello, world!\(rsn' | dj -c 1 -b 7 -s 7 2>/dev/null
+.RE
+
+Produces the following output:
+
+.RS
+world!
+.RE
+
+The following
+.BR sh (1p)
+lines run sequentially:
+
+.RS
+tr '\(rs0' 0    </dev/zero | dj -c 1 -b 6 -o hello.txt
+
+tr '\(rs0' H    </dev/zero | dj -c 1 -b 1 -o hello.txt
+
+tr '\(rs0' e    </dev/zero | dj -c 1 -b 1 -o hello.txt -S 1
+
+tr '\(rs0' l    </dev/zero | dj -c 1 -b 2 -o hello.txt -S 2
+
+tr '\(rs0' o    </dev/zero | dj -c 1 -b 1 -o hello.txt -S 4
+
+tr '\(rs0' '\(rsn' </dev/zero | dj -c 1 -b 1 -o hello.txt -S 5
+
+dj -i hello.txt
+.RE
+
+Produce the following output:
+
+.RS
+Hello
+.RE
+
+It may be particularly illuminating to print the contents of the example
+.B hello.txt
+after each
+.BR dj (1)
+invocation.
+.\"
 .SH DIAGNOSTICS
 
 On a partial or empty read, a diagnostic message is printed. Then, the program
@@ -128,9 +181,20 @@ option is specified, this could make written data nonsensical.
 
 Existing files are not truncated on ouput and are instead overwritten.
 
-Many lowercase options have capitalized variants and vice-versa which can be
-confusing. Capitalized options tend to affect output or are more intense
-versions of lowercase options.
+The options
+.B -b
+and
+.B -B
+could be confused for each other, and so could
+.B -s
+and
+.BR -S .
+The lowercase option affects input and the capitalized option affects output.
+
+The skipped or sought bytes while processing irregular files, such as streams,
+are reported in the diagnostic output, because they were actually read or
+written. This is as opposed to bytes skipped while processing regular files,
+which are not reported.
 .\"
 .SH RATIONALE
 
@@ -151,3 +215,4 @@ Copyright \(co 2023 DTB. License AGPLv3+: GNU AGPL version 3 or later
 .\"
 .SH SEE ALSO
 .BR dd (1p)
+.BR lseek (3p)

docs/mm.1

@@ -3,14 +3,14 @@
 .\" 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 2024-06-17 "Harakit X.X.X"
+.TH MM 1 2024-07-14 "Harakit X.X.X"
 .SH NAME
 mm \(en middleman
 .\"
 .SH SYNOPSIS
 
 mm
-.RB [ -aenu ]
+.RB [ -aetu ]
 .RB [ -i\ input ]
 .RB [ -o\ output ]
 .\"
@@ -21,19 +21,25 @@ Catenate input files and write them to the start of each output file or stream.
 .SH OPTIONS
 
 .IP \fB-a\fP
-Opens subsequent outputs for appending rather than updating.
+Opens outputs for appending rather than updating.
 .IP \fB-e\fP
 Use the standard error as an output.
-.IP \fB-i\fP\ \fIinput\fP
-Opens a path as an input. If one or more of the input files is \(lq-\(rq or if
-no inputs are specified, the standard input shall be used.
-.IP \fB-o\fP\ \fIoutput\fP
-Opens a path as an output. If one or more of the output files is \(lq-\(rq or if
-no outputs are specified, the standard output shall be used.
+.IP \fB-t\fP
+Causes outputs to be overwritten instead of being truncated.
 .IP \fB-u\fP
 Ensures neither input or output will be buffered.
-.IP \fB-n\fP
-Causes SIGINT signals to be ignored.
+.IP \fB-i\fP\ \fIinput\fP
+Opens a path as an input. If one or more of the input files is \(lq-\(rq or if
+no inputs are specified, the standard input shall be used. If specified as the
+last option and if there are trailing arguments to the program, they shall be
+appended to the list of files to use as inputs.
+.IP \fB-o\fP\ \fIoutput\fP
+Opens a path as an output. If one or more of the output files is \(lq-\(rq or if
+no outputs are specified and the
+.B -e
+option is not specified, the standard output shall be used. If specified as the
+last option and if there are trailing arguments to the program, they shall be
+appended to the list of files to use as outputs.
 .\"
 .SH DIAGNOSTICS
 
@@ -45,10 +51,6 @@ exits with the appropriate
 .BR sysexits.h (3)
 status.
 .\"
-.SH CAVEATS
-
-Existing files are not truncated on ouput and are instead overwritten.
-.\"
 .SH RATIONALE
 
 The

docs/swab.1

@@ -11,7 +11,6 @@ swab \(en swap bytes
 .SH SYNOPSIS
 
 swab
-.RB [ -f ]
 .RB [ -w\ word_size ]
 .\"
 .SH DESCRIPTION
@@ -20,8 +19,6 @@ Swap the latter and former halves of a block of bytes.
 .\"
 .SH OPTIONS
 
-.IP \fB-f\fP
-Ignore SIGINT signal.
 .IP \fB-w\fP\ \fIword_size\fP
 Configures the word size; that is, the size in bytes of the block size on which
 to operate. The default word size is 2. The word size must be cleanly divisible

src/dj.c

@@ -16,12 +16,15 @@
  * along with this program. If not, see https://www.gnu.org/licenses/.
  */
 
+#include <assert.h> /* assert(3) */
 #include <errno.h> /* errno */
 #include <fcntl.h> /* open(2) */
 #include <stdio.h> /* fprintf(3), stderr */
-#include <stdlib.h> /* free(3), malloc(3), strtol(3), size_t */
+#include <stdlib.h> /* malloc(3), strtol(3), size_t */
 #include <string.h> /* memcpy(3), memmove(3), memset(3) */
-#include <sysexits.h> /* EX_OK, EX_USAGE */
+#if !defined EX_OK || !defined EX_OSERR || !defined EX_USAGE
+#	include <sysexits.h>
+#endif
 #include <unistd.h> /* close(2), getopt(3), lseek(2), read(2), write(2),
                      * optarg, optind, STDIN_FILENO, STDOUT_FILENO */
 #include <sys/stat.h> /* S_IRGRP, S_IROTH, S_IRUSR, S_IWGRP, S_IWOTH,
@@ -34,198 +37,90 @@ char *program_name = "dj";
  * writing ends of its jockeyed "pipe". User-configurable members are noted
  * with their relevant options. */
 struct Io{
-	int bs;			/* buffer size (-bB) */
-	size_t bufuse;	/* buffer usage */
 	char *buf;	  	/* buffer */
-	int bytes;		/* bytes processed */
+	char *fn;	  	/* file name (-io) */
+	size_t bs;    	/* buffer size (-bB) */
+	size_t bufuse;	/* buffer usage */
+	size_t bytes; 	/* bytes processed */
+	size_t prec;  	/* partial records processed */
+	size_t rec;   	/* records processed */
+	long seek;  	/* remaining bytes to seek/skip (-sS) */
+	int error;  	/* errno */
 	int fd;		  	/* file descriptor */
 	int fl;		  	/* file opening flags */
-	char *fn;		/* file name (may be stdin_name or stdout_name) (-io) */
-	int prec;		/* partial records processed */
-	int rec;		/* records processed */
-	long seek;		/* bytes to seek/skip (will be 0 after skippage) (-sS) */
 };
 
-/* To be assigned to main:fmt_output and used with output(). */
+/* To be assigned to main:fmt and used with printio(). */
 static char *fmt_asv   = "%d\037%d\036%d\037%d\035%d\036%d\034";
 static char *fmt_human = "%d+%d > %d+%d; %d > %d\n";
 
 static char *stdin_name = "<stdin>";
 static char *stdout_name = "<stdout>";
 
+static int creat_mode = S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP | S_IROTH
+	| S_IWOTH; /* Consistent with touch(1p). */
 static int read_flags = O_RDONLY; /* Consistent with Busybox dd(1). */
 static int write_flags = O_WRONLY | O_CREAT;
 
 #define MIN(a, b) (((a) < (b)) ? (a) : (b))
 
-/* Macro to check if fd is a std* file, e.g. stdin. */
-#define fdisstd(fd) \
-		((fd) == STDIN_FILENO \
-		|| (fd) == STDOUT_FILENO \
-		|| (fd) == STDERR_FILENO)
+/* Macro to check if fd is stdin or stdout */
+#define fdisstd(fd) ((fd) == STDIN_FILENO || (fd) == STDOUT_FILENO)
 
-/* Macro to call the cleanup functions that operate on struct io on the
- * particular io[2] used in main. Error conditions are not checked because this
- * is only used when the program is about to terminate (hence its name). */
-#define terminate(io) do{ \
-	free((io[0]).buf); \
-	free((io[1]).buf); \
-	Io_fdclose(&(io)[0]); \
-	Io_fdclose(&(io)[1]); }while(0)
-
-/* Allocates *io's buffer. Returns NULL if unsuccessful. */
-static void *
-Io_bufalloc(struct Io *io){
-
-	return (io->buf = malloc(io->bs * (sizeof *io->buf)));
-}
-
-/* Fills the unused portion of io's buffer with padding, updating io->bufuse.
- * Returns io. */
-static struct Io *
-Io_bufrpad(struct Io *io, int padding){
-	
-	memset(io->buf + io->bufuse, padding, io->bs - io->bufuse);
-	io->bufuse = io->bs;
-
-	return io;
-}
-
-/* Copies from the buffer in src as much as possible to the free space in the
- * dest buffer, removing the copied units from src and permuting the remaining
- * units in the src buffer to the start of the buffer, modifying both the src
- * and dest bufuse and returning dest. */
-static struct Io*
-Io_bufxapp(struct Io *dest, struct Io *src){
-	int n;
-
-	n = MIN(src->bufuse, dest->bs - dest->bufuse);
-	memcpy(dest->buf + dest->bufuse, src->buf, n);
-	dest->bufuse += n;
-	memmove(src->buf, src->buf + n, src->bs - n);
-	src->bufuse -= n;
-
-	return dest;
-}
-
-/* Copies from the buffer in src to the buffer in dest no more than n units,
- * removing the copied units from src and permuting the remaining units in the
- * src buffer to the start of the buffer, modifying both the src and dest
- * bufuse and returning dest. */
-static struct Io*
-Io_bufxfer(struct Io *dest, struct Io *src, int n){
-
-	memcpy(dest->buf, src->buf, (dest->bufuse = n));
-	memmove(src->buf, src->buf + n, (src->bufuse -= n));
-
-	return dest;
-}
-
-/* Closes io->fn and returns -1 on error, otherwise io->fd. */
-static int
-Io_fdclose(struct Io *io){
-
-	return fdisstd(io->fd)
-		? 0
-		: close(io->fd);
-}
-
-/* Opens io->fn and saves the file descriptor into io->fd. Returns io->fd,
- * which will be -1 if an error occured. */
-static int
-Io_fdopen(struct Io *io, char *fn){
-	int fd;
-	
-	if((fd = open(fn, io->fl,
-					/* these are the flags used by touch(1p) */
-					S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP | S_IROTH | S_IWOTH))
-				!= -1
-			&& Io_fdclose(io) == 0){
-		io->fd = fd;
-		io->fn = fn;
-	}
-
-	return fd;
-}
-
-/* Seeks io->seek bytes through *io's file descriptor, subtracting the number
- * of sought bytes from io->seek. This procedure leaves garbage in io->buf. */
-static void
-Io_fdseek(struct Io *io){
-	
-	if(io->seek != 0
-			|| (!fdisstd(io->fd) && lseek(io->fd, io->seek, SEEK_SET) != -1))
-		return;
-
-	if(io->fl == write_flags)
-		memset(io->buf, '\0', io->bs);
-
-	if(io->fl == write_flags){
-		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 = 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);
-	}
-
-	io->bufuse = 0;
-
-	return;
-}
-
-/* Reads io->bs bytes from *io's file descriptor into io->buf, storing the
- * number of read bytes in io->bufuse and updating io->bytes. If io->bufuse is
- * 0, errno will probably be set. Returns io. */
 static struct Io *
 Io_read(struct Io *io){
+	int t;
 
-	io->bytes += (io->bufuse = read(io->fd, io->buf, io->bs));
+	assert(io->bs > 0);
+	assert(io->bufuse < io->bs);
+
+	if((t = read(io->fd, &(io->buf)[io->bufuse], io->bs - io->bufuse)) < 0){
+		io->error = errno;
+		t = 0;
+	}
+
+	io->bufuse += t;
+	io->bytes  += t;
+	io->prec   += (0 < io->bufuse && io->bufuse < io->bs);
+	io->rec    += (io->bufuse == io->bs);
+
+	assert(io->bufuse <= io->bs);
 
 	return io;
 }
 
-/* Writes io->bufuse units from io->buf to io->fd, permuting any unwritten
- * bytes to the start of io->buf and updating io->bufuse. If io->bufuse doesn't
- * change, errno will probably be set. Returns io. */
 static struct Io *
 Io_write(struct Io *io){
 	int t;
 
-	if((t = write(io->fd, io->buf, io->bufuse)) > 0)
-		memmove(io->buf, io->buf + t, (io->bufuse -= t));
+	assert(io->bufuse > 0);
+	assert(io->bufuse <= io->bs);
+
+	if((t = write(io->fd, io->buf, io->bufuse)) < 0){
+		io->error = errno;
+		t = 0;
+	}else if(t > 0)
+		memmove(io->buf, &(io->buf)[t], (io->bufuse -= t));
+
 	io->bytes += t;
+	io->prec  += (t > 0 && io->bufuse > 0);
+	io->rec   += (t > 0 && io->bufuse == 0);
 
 	return io;
 }
 
-/* Prints an error message suitable for the event of an operating system error,
- * with the error itself to be described in the string s. */
 static int
-oserr(char *s){
-
-	fprintf(stderr, "%s: %s: %s\n", program_name, s, strerror(errno));
-
+oserr(char *e, int n){
+	fprintf(stderr, "%s: %s: %s\n", program_name, e, strerror(n));
 	return EX_OSERR;
 }
 
 /* Prints statistics regarding the use of dj, particularly partially and
  * completely read and written records. */
 static void
-output(struct Io io[2], char *fmt){
+fprintio(FILE *stream, char *fmt, struct Io io[2]){
 
-	fprintf(stderr, fmt,
+	fprintf(stream, fmt,
 		io[0].rec, io[0].prec, io[1].rec, io[1].prec,
 		io[0].bytes, io[1].bytes);
 
@@ -260,22 +155,24 @@ usage(char *s){
 int main(int argc, char *argv[]){
  	int align;    /* low 8b used, negative if no alignment is being done */
 	int count;    /* 0 if dj(1) runs until no more reads are possible */
-	char *fmt_output; /* == fmt_asv (default) or fmt_human (-H) */
+	char *fmt;    /* == fmt_asv (default) or fmt_human (-H) */
 	size_t i;     /* side of io being modified */
-	struct Io io[2];
 	char noerror; /* 0=exits (default) 1=retries on partial reads or writes */
+	struct Io io[2 /* { in, out } */];
 
 	/* Set defaults. */
 	align = -1;
 	count = 0;
-	fmt_output = fmt_asv;
+	fmt = fmt_asv;
 	noerror = 0;
-	for(i = 0; i < 2; ++i){
+	for(i = 0; i < (sizeof io) / (sizeof *io); ++i){
 		io[i].bs = 1024 /* 1 KiB */; /* GNU dd(1) default; POSIX says 512B */
+		io[i].bufuse = 0;
 		io[i].bytes = 0;
-		io[i].fd = i ? STDIN_FILENO : STDOUT_FILENO;
-		io[i].fn = i ? stdin_name   : stdout_name;
-		io[i].fl = i ? read_flags   : write_flags;
+		io[i].fd = i == 0 ? STDIN_FILENO : STDOUT_FILENO;
+		io[i].fn = i == 0 ? stdin_name   : stdout_name;
+		io[i].fl = i == 0 ? read_flags   : write_flags;
+		io[i].error = 0;
 		io[i].prec = 0;
 		io[i].rec = 0;
 		io[i].seek = 0;
@@ -285,19 +182,26 @@ int main(int argc, char *argv[]){
 		int c;
 
 		program_name = argv[0];
-		while((c = getopt(argc, argv, "a:b:B:c:i:hHns:S:o:")) != -1)
+		while((c = getopt(argc, argv, ":a:b:B:c:i:hHns:S:o:")) != -1)
 			switch(c){
 			case 'i': case 'o': i = (c == 'o');
 				if(optarg[0] == '-' && optarg[1] == '\0'){ /* optarg == "-" */
-					io[i].fd = i ? STDIN_FILENO : STDOUT_FILENO;
-					io[i].fn = i ? stdin_name   : stdout_name;
+					io[i].fd = i == 0 ? STDIN_FILENO : STDOUT_FILENO;
+					io[i].fn = i == 0 ? stdin_name   : stdout_name;
 					break;
-				}else if(Io_fdopen(&io[i], optarg) != -1)
+				}else{
+					int fd;
+
+					if((fd = open(optarg, io[i].fl, creat_mode)) != -1
+							&& (fdisstd(io[i].fd) || close(io[i].fd) == 0)){
+						io[i].fd = fd;
+						io[i].fn = optarg;
 						break;
-				terminate(io);
-				return oserr(optarg);
+					}
+				}
+				return oserr(optarg, errno);
 			case 'n': noerror = 1;     break;
-			case 'H': fmt_output = fmt_human; break;
+			case 'H': fmt = fmt_human; break;
 			case 'a':
 				if(optarg[0] == '\0' || optarg[1] == '\0'){
 					align = optarg[0];
@@ -307,87 +211,143 @@ int main(int argc, char *argv[]){
 			case 'c': case 'b': case 's': case 'B': case 'S': /* numbers */
 				if(c == 'c' && (count = parse(optarg)) >= 0)
 					break;
-				i = (c >= 'A' && c <= 'Z'); /* uppercase changes output */
-				c &= 0x20 /* 0b 0010 0000 */; /* (ASCII) make lowercase */
+				i = (c >= 'A' && c <= 'Z');
+				c |= 0x20 /* 0b 0010 0000 */; /* (ASCII) make lowercase */
 				if((c == 'b' && (io[i].bs = parse(optarg)) > 0)
 						|| (c == 's' && (io[i].seek = parse(optarg)) >= 0))
 					break;
 				/* FALLTHROUGH */
 			default:
-				terminate(io);
 				return usage(program_name);
 			}
 	}
 
-	if(argc > optind){
-		terminate(io);
-		return usage(program_name);
-	}
+	assert(io->fd != STDIN_FILENO  || io->fl == read_flags);
+	assert(io->fd != STDOUT_FILENO || io->fl == write_flags);
 
-	for(i = 0; i < 2; ++i){
-		if(Io_bufalloc(&io[i]) == NULL){
-			fprintf(stderr, "%s: Failed to allocate %d bytes\n",
+	if(argc > optind)
+		return usage(program_name);
+
+	for(i = 0; i < (sizeof io) / (sizeof *io); ++i){
+		/* buffer allocation */
+		if((io[i].buf = malloc(io[i].bs * (sizeof *(io[i].buf)))) == NULL){
+			fprintf(stderr, "%s: Failed to allocate %zd bytes\n",
 				program_name, io[i].bs);
-			terminate(io);
 			return EX_OSERR;
-		}else if(io[i].seek > 0)
-			Io_fdseek(&io[i]);
-			if(io[i].seek > 0){
-				terminate(io);
-				return oserr(io[i].fn);
 		}
+		/* easy seeking */
+		if(!fdisstd(io[i].fd) && lseek(io[i].fd, io[i].seek, SEEK_SET) != -1)
+			io[i].seek = 0;
 	}
 
-	do{	/* read */
-		Io_read(&io[0]);
-		if(!noerror && io[0].bufuse == 0)
+	/* hard seeking */
+	if(io[1].seek > 0){
+		size_t t;
+		do{
+			memset(io[1].buf, '\0',
+				(t = io[1].bufuse = MIN(io[1].bs, io[1].seek)));
+			if(Io_write(&io[1])->bufuse == t && !noerror && io[1].error == 0)
+				Io_write(&io[1]); /* second chance */
+			if(io[1].error != 0)
+				return oserr(io[1].fn, io[1].error);
+		}while((io[1].seek -= (t - io[1].bufuse)) > 0 && io[1].bufuse != t);
+		io[1].bufuse = 0;
+	}
+
+	if(io[1].seek > 0){
+		fprintio(stderr, fmt, io);
+		return oserr(io[1].fn, errno);
+	}
+
+	do{
+		assert(io[0].bufuse == 0);
+
+		{ /* read */
+			long skipping;
+			size_t t;
+
+			/* hack to intentionally get a partial read from Io_read */
+			if((skipping = MIN(io[0].seek, io[0].bs)) > 0)
+				io[0].bufuse = io[0].bs - (size_t)skipping;
+
+			t = io[0].bufuse;
+			if(Io_read(&io[0])->bufuse == t && !noerror && io[0].error == 0)
 				Io_read(&io[0]); /* second chance */
-		if(io[0].bufuse == 0) /* that's all she wrote */
+			assert(io[0].bufuse >= t);
+			if(io[0].bufuse == t) /* that's all she wrote */
 				break;
-		else if(io[0].bufuse < io[0].bs){
-			++io[0].prec;
+
+			if(/* t < io[0].bufuse && */ io[0].bufuse < io[0].bs){
 				fprintf(stderr, "%s: Partial read:\n\t", program_name);
-			output(io, fmt_output);
+				fprintio(stderr, fmt, io);
 				if(!noerror)
 					count = 1;
-			if(align >= 0)
-				Io_bufrpad(&io[0], align);
-		}else
-			++io[0].rec;
+				if(align >= 0){
+					/* fill the rest of the ibuf with padding */
+					memset(&(io[0].buf)[io[0].bufuse], align,
+						io[0].bs - io[0].bufuse);
+					io->bufuse = io->bs;
+				}
+			}
+
+			if(skipping > 0){
+				io[0].seek -= skipping;
+				io[0].bufuse = 0;
+				count += (count != 0);
+				continue;
+			}
+		}
 
 		/* write */
 		do{
 			int t;
 
-			if(io[1].bs > io[0].bs){
-				Io_bufxapp(&io[1], &io[0]);
+			if(io[0].bs <= io[1].bs){
+				int n;
+
+				/* saturate obuf */
+				memcpy(io[1].buf, io[0].buf,
+				       (io[1].bufuse = (n = MIN(io[0].bufuse, io[1].bs))));
+				/* permute the copied units out of ibuf */
+				memmove(io[0].buf, &(io[0].buf)[n], (io[0].bufuse -= n));
+			}else /* if(io[0].bs < io[1].bs) */ {
+				int n;
+
+				/* drain what we can from ibuf */
+				memcpy(&(io[1].buf)[io[1].bufuse], io[0].buf,
+				       (n = MIN(io[0].bufuse, io[1].bs - io[1].bufuse)));
+				io[1].bufuse += n;
+				/* permute out the copied units */
+				memmove(io[0].buf, &(io[0].buf)[n], io[0].bs - n);
+				io[0].bufuse -= n;
+
 				if(io[0].bs + io[1].bufuse <= io[1].bs && count != 1)
-					continue; /* we could write more */
-			}else
-				Io_bufxfer(&io[1], &io[0], MIN(io[0].bufuse, io[1].bs));
+					continue; /* obuf not saturated - we could write more */
+			}
 
 			t = io[1].bufuse;
-			Io_write(&io[1]);
-			if(!noerror && io[1].bufuse == t)
+			if(Io_write(&io[1])->bufuse == t && !noerror && io[1].error == 0)
 				Io_write(&io[1]); /* second chance */
-			if(t == io[1].bufuse){ /* no more love */
+			assert(io[1].bufuse <= t);
+			if(io[1].bufuse == t){ /* no more love */
 				count = 1;
 				break;
-			}else if(t > io[1].bufuse && io[1].bufuse > 0){
-				io[1].prec += 1;
+			}
+
+			if(0 < io[1].bufuse /* && io[1].bufuse < t */){
 				fprintf(stderr, "%s: Partial write:\n\t", program_name);
-				output(io, fmt_output);
+				fprintio(stderr, fmt, io);
 				if(!noerror)
 					count = 1;
-			}else if(io[1].bufuse == 0 && t < io[1].bs)
-				++io[1].prec;
-			else
-				++io[1].rec;
+			}
 		}while(io[0].bufuse > 0);
 	}while(count == 0 || --count > 0);
 
-	output(io, fmt_output);
-	terminate(io);
+	fprintio(stderr, fmt, io);
+
+	for(i = 0; i < (sizeof io) / (sizeof *io); ++i)
+		if(io[i].error)
+			return oserr(io[i].fn, io[i].error);
 
 	return EX_OK;
 }

src/mm.c

@@ -1,236 +0,0 @@
-/*
- * 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
-
-/* 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} */
-	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:
-				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;
-	
-	/* 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/mm.rs

@@ -0,0 +1,185 @@
+/*
+ * Copyright (c) 2024 Emma Tebibyte <emma@tebibyte.media>
+ * 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/.
+ */
+
+use std::{
+	env::args,
+	fs::File,
+	io::{ stdin, stdout, stderr, BufWriter, Read, Write },
+	os::fd::{ AsRawFd, FromRawFd },
+	process::{ exit, ExitCode },
+};
+
+extern crate getopt;
+extern crate strerror;
+extern crate sysexits;
+
+use getopt::GetOpt;
+use strerror::StrError;
+use sysexits::{ EX_IOERR, EX_USAGE };
+
+use ArgMode::*;
+
+enum ArgMode { In, Out }
+
+fn main() -> ExitCode {
+	let argv = args().collect::<Vec<_>>();
+	let usage = format!("Usage: {} [-aetu] [-i input] [-o output]", argv[0]);
+
+	let mut a = false; /* append to the file */
+	let mut e = false; /* use stderr as an output */
+	let mut t = true; /* do not truncate the file before writing */
+	let mut u = false; /* unbuffer i/o */
+	let mut ins = Vec::new(); /* initial input file path vector */
+	let mut outs = Vec::new(); /* initial output file path vector */
+	let mut mode: Option<ArgMode> = None; /* mode set by last-used option */
+	let mut optind = 0;
+
+	while let Some(opt) = argv.getopt("aei:o:tu") {
+		match opt.opt() {
+			Ok("a") => a = true,
+			Ok("e") => e = true, 
+			Ok("u") => u = true,
+			Ok("t") => t = false,
+			Ok("i") => { /* add inputs */
+				let input = opt.arg().unwrap();
+				ins.push(input);
+				mode = Some(In); /* latest argument == -i */
+			},
+			Ok("o") => { /* add output */
+				let output = opt.arg().unwrap(); 
+				outs.push(output);
+				mode = Some(Out); /* latest argument == -o */
+			},
+			Err(_) | Ok(_) => {
+				eprintln!("{}", usage);
+				return ExitCode::from(EX_USAGE as u8);
+			},
+		};
+
+		optind = opt.ind();
+	}
+
+	let remaining = argv.iter().skip(optind);
+
+	/* check the last flag specified */
+	if let Some(m) = mode {
+		for arg in remaining {
+			/* move the subsequent arguments to the list of inputs or outputs */
+			match m {
+				In => ins.push(arg.to_string()),
+				Out => outs.push(arg.to_string()),
+			};
+		}
+	} else {
+		eprintln!("{}", usage);
+		return ExitCode::from(EX_USAGE as u8);
+	}
+
+	/* use stdin if no inputs are specified */
+	if ins.is_empty() { ins.push("-".to_string()); }
+
+	/* use stdout if no outputs are specified */
+	if outs.is_empty() && !e { outs.push("-".to_string()); }
+	
+	/* map all path strings to files */
+	let inputs = ins.iter().map(|file| {
+		/* if a file is “-”, it is stdin */
+		if *file == "-" {
+			/* portable way to access stdin as a file */
+			return unsafe { File::from_raw_fd(stdin().as_raw_fd()) };
+		}
+
+		match File::open(file) {
+			Ok(f) => f,
+			Err(e) => {
+				eprintln!("{}: {}: {}", argv[0], file, e.strerror());
+				exit(EX_IOERR);
+			},
+		}
+	}).collect::<Vec<_>>();
+
+	/* map all path strings to files */
+	let mut outputs = outs.iter().map(|file| {
+		/* of a file is “-”, it is stdout */
+		if *file == "-" {
+			/* portable way to access stdout as a file */
+			return unsafe { File::from_raw_fd(stdout().as_raw_fd()) }; 
+		}
+
+		let options = File::options()
+			/* don’t truncate if -t is specified, append if -a is specified */
+			.truncate(t)
+			.append(a)
+			/* enable the ability to create and write to files */
+			.create(true)
+			.write(true)
+			/* finally, open the file! */
+			.open(file);
+
+		match options {
+			Ok(f) => return f,
+			Err(e) => {
+				eprintln!("{}: {}: {}", argv[0], file, e.strerror());
+				exit(EX_IOERR);
+			},
+		};
+	}).collect::<Vec<_>>();
+
+	/* if -e is specified, use stderr */
+	if e {
+		/* portable way to access stderr as a file */
+		outputs.push(unsafe { File::from_raw_fd(stderr().as_raw_fd()) });
+	}
+
+	let mut outputs = outputs.iter().map(|o| {
+		if u {
+			/* unbuffered writing through a buffer of capacity 0 */
+			BufWriter::with_capacity(0, o)
+		} else {
+			/* theoretically buffered writing */
+			BufWriter::new(o)
+		}
+	}).collect::<Vec<_>>();
+
+	for file in inputs {
+		for byte in file.bytes().map(|b| {
+			b.unwrap_or_else(|e| {
+				eprintln!("{}: {}", argv[0], e.strerror());
+				exit(EX_IOERR);
+			})
+		}) {
+			for out in &mut outputs {
+				if let Err(e) = out.write(&[byte]) {
+					eprintln!("{}: {}", argv[0], e.strerror());
+					return ExitCode::from(EX_IOERR as u8);
+				}
+
+				if u {
+					/* immediately flush the output for -u */
+					if let Err(e) = out.flush() { 
+						eprintln!("{}: {}", argv[0], e.strerror());
+						return ExitCode::from(EX_IOERR as u8);
+					}
+				}
+			}
+		}
+	}
+
+	ExitCode::SUCCESS
+}

src/swab.rs

@@ -1,5 +1,6 @@
 /*
  * Copyright (c) 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
@@ -18,7 +19,7 @@
 
 use std::{
 	env::args,
-	io::{ stdin, stdout, Error, ErrorKind, Read, Write },
+	io::{ stdin, stdout, Error, Read, Write },
 	process::ExitCode,
 	vec::Vec
 };
@@ -29,13 +30,16 @@ use getopt::GetOpt;
 extern crate sysexits;
 use sysexits::{ EX_OK, EX_OSERR, EX_USAGE };
 
+extern crate strerror;
+use strerror::StrError;
+
 fn oserr(s: &str, e: Error) -> ExitCode {
-	eprintln!("{}: {}", s, e);
+	eprintln!("{}: {}", s, e.strerror());
 	ExitCode::from(EX_OSERR as u8)
 }
 
 fn usage(s: &str) -> ExitCode {
-	eprintln!("Usage: {} [-f] [-w word_size]", s);
+	eprintln!("Usage: {} [-w word_size]", s);
 	ExitCode::from(EX_USAGE as u8)
 }
 
@@ -45,24 +49,26 @@ fn main() -> ExitCode {
 	let mut input = stdin();
 	let mut output = stdout().lock();
 
-	let mut force = false;
-	let mut wordsize: usize = 2;
+	let mut optind: usize = 1;   // argv[0]
+	let mut wordsize: usize = 2; // Equivalent to dd(1p).
 
-	while let Some(opt) = argv.getopt("fw:") {
+	while let Some(opt) = argv.getopt("w:") {
 		match opt.opt() {
-			Ok("f") => force = true,
 			Ok("w") => {
-				if let Some(arg) = opt.arg() {
-					match arg.parse::<usize>() {
-						Ok(w) if w % 2 == 0 => { wordsize = w; () },
+					match opt.arg().unwrap().parse::<usize>() {
+						Ok(w) if w % 2 == 0 => { wordsize = w; },
 						_ => { return usage(&argv[0]); },
 					}
-				}
+				optind = opt.ind();
 			},
 			_ => { return usage(&argv[0]); }
 		}
 	}
 
+	if optind < argv.len() {
+		return usage(&argv[0]);
+	}
+
 	buf.resize(wordsize, 0);
 
 	loop {
@@ -80,7 +86,6 @@ fn main() -> ExitCode {
 					break oserr(&argv[0], e)
 				}
 			},
-			Err(e) if e.kind() == ErrorKind::Interrupted && force => continue,
 			Err(e) => break oserr(&argv[0], e)
 		}
 	}