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

Makefile

@@ -102,11 +102,15 @@ 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) $(CFLAGAS) -o $@ src/npc.c
+	$(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

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/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;
+}