listamsb: use atascii.c instead of piping to a8cat.

4 files changed, 94 insertions, 79 deletions

diff --git a/Makefile b/Makefile
index c46103e..90f22c8 100644
--- a/Makefile
+++ b/Makefile
@@ -79,6 +79,8 @@ bas2aplus: bas.o
 listbas: listbas.c bas.o bcdfp.o tokens.o atables.o turbo_tokens.o aplus_tokens.o bxl_tokens.o bxe_tokens.o int_disk_tokens.o int_tokens.o
 	$(CC) $(CFLAGS) -o listbas listbas.c bas.o bcdfp.o tokens.o atables.o turbo_tokens.o aplus_tokens.o bxl_tokens.o bxe_tokens.o int_disk_tokens.o int_tokens.o -lm
 
+listamsb: listamsb.o atascii.o atables.o
+
 bas.o: bas.c bas.h
 
 wtable.o: wtable.c wtable.h
diff --git a/listamsb.1 b/listamsb.1
index 009b5a8..6559659 100644
--- a/listamsb.1
+++ b/listamsb.1
@@ -27,7 +27,7 @@ level margin: \\n[rst2man-indent\\n[rst2man-indent-level]]
 .\" new: \\n[rst2man-indent\\n[rst2man-indent-level]]
 .in \\n[rst2man-indent\\n[rst2man-indent-level]]u
 ..
-.TH "LISTAMSB" 1 "2025-03-17" "0.2.2" "Urchlay's Atari 8-bit Tools"
+.TH "LISTAMSB" 1 "2025-04-01" "0.2.2" "Urchlay's Atari 8-bit Tools"
 .SH NAME
 listamsb \- List the source of a tokenized Atari Microsoft BASIC program
 .SH SYNOPSIS
@@ -57,14 +57,14 @@ If no \fBoutput\-file\fP is given, output is to \fBstdout\fP\&. When output
 is to a file, the file will be overwritten without prompting if it
 already exists.
 .sp
-When listing a program to \fBstdout\fP, output is piped through
-\fBa8cat\fP(1), which converts the ATASCII output to UTF\-8 that\(aqs
-human\-readable on modern terminals. This means \fBa8cat\fP must be
-available in your \fBPATH\fP\&. If this is not convenient, you can set
-the location of \fBa8cat\fP via the \fBA8CAT\fP environment variable.
-When listing in color (when the \fB\-M\fP option is not used), \fBa8cat\fP\(aqs
-output is piped through \fBcolorize\-amsb\fP, which also must be available
-in your \fBPATH\fP\&.
+By default, output is modern UTF\-8 equivalents of the ATASCII
+character set, with ANSI control sequences for inverse video and
+color. See the \fB\-M\fP, \fB\-a\fP, \fB\-i\fP, \fB\-u\fP, \fB\-t\fP, \fB\-m\fP, \fB\-s\fP
+options to change the output style.
+.sp
+When listing in color (when the \fB\-M\fP option is not used),
+\fBlistamsb\fP\(aqs output is piped through \fBcolorize\-amsb\fP, which must
+be available in your \fBPATH\fP\&.
 .sp
 \fBlistamsb\fP will not write binary data (ATASCII or tokenized BASIC)
 directly to a terminal, under any circumstances.
@@ -121,15 +121,28 @@ or \fB\-U\fP\&.
 .B \fB\-a\fP
 Output raw ATASCII. This option must be used with an
 \fBoutput\-file\fP, a pipe, or redirection: \fBlistamsb\fP will not write
-ATASCII to a terminal. \fBa8cat\fP and \fBcolorize\-amsb\fP are not used, with this option.
+ATASCII to a terminal. \fBcolorize\-amsb\fP is not used, with this option.
 .TP
 .B \fB\-c\fP
 Check only. No output on \fBstdout\fP\&. Diagnostics are still printed on stderr,
 and the exit status is unchanged. It\(aqs an error to give an \fBoutput\-file\fP with
 this option.
 .TP
-.B \fB\-i\fP, \fB\-u\fP, \fB\-t\fP, \fB\-m\fP, \fB\-s\fP
-These options are passed to \fBa8cat\fP\&. See its man page for details.
+.B \fB\-i\fP
+Input uses Atari XL/XE International Character Set encoding, rather than
+ATASCII graphics.
+.TP
+.B \fB\-u\fP
+Use underlining for inverse video (see \fBa8cat\fP(1) for more detail).
+.TP
+.B \fB\-t\fP
+Text mode (see \fBa8cat\fP(1) for more detail).
+.TP
+.B \fB\-m\fP
+Magazine listing mode (see \fBa8cat\fP(1) for more detail).
+.TP
+.B \fB\-s\fP
+Strip bit 7 (inverse video bit; see \fBa8cat\fP(1) for more detail).
 .TP
 .B \fB\-M\fP
 Monochrome listing: disable color syntax highlighting, meaning \fBcolorize\-amsb\fP
@@ -282,16 +295,6 @@ you see this warning, you\(aqll also see \fIN\fP <unknown token \fIXX\fP>
 messages in the actual listing.
 .UNINDENT
 .SH ENVIRONMENT
-.sp
-\fBA8CAT\fP
-.INDENT 0.0
-.INDENT 3.5
-The path to the \fBa8cat\fP(1) binary. Can be a full path or just
-the executable name, in which case it will be searched for with the
-normal \fBPATH\fP mechanism. If not set, \fIa8cat\fP is used. If there are
-spaces or other metacharacters, they must be escaped.
-.UNINDENT
-.UNINDENT
 .SH LIMITATIONS
 .sp
 \fBlistamsb\fP may behave strangely when fed something other than a
diff --git a/listamsb.c b/listamsb.c
index b7aad8a..a14c336 100644
--- a/listamsb.c
+++ b/listamsb.c
@@ -7,6 +7,7 @@
 #include <ctype.h>
 #include <sys/stat.h>
 
+#include "atascii.h"
 #include "amsbtok.h"
 
 /* this should always be defined in <stdio.h>, but just in case... */
@@ -80,7 +81,7 @@
 
 const char *self;
 
-char pipe_command[BUFSIZ + 1] = { "a8cat" };
+atascii_ctx actx;
 
 int verbosity =   0;           /* -v */
 int raw_output =  0;           /* -a */
@@ -198,7 +199,7 @@ void finish(int rv) {
 		int got = pclose(outfile);
 		verbose(1, "return value from pipe is %d", got);
 		if(got != 0) {
-			os_err("a8cat child process failed, do you have a8cat on your PATH?");
+			os_err("colorize-amsb child process failed, do you have it on your PATH?");
 		}
 	} else {
 		fclose(outfile);
@@ -301,7 +302,13 @@ void unknown_token(int tok) {
 }
 
 void list_char(unsigned char c) {
-	if(printing) fputc(c, outfile);
+	if(!printing) return;
+	if(raw_output) {
+		fputc(c, outfile);
+	} else {
+		char buf[20];
+		fputs(atascii_a2utf(&actx, c, buf), outfile);
+	}
 }
 
 int expand_token(int t, unsigned char *buf);
@@ -324,7 +331,13 @@ void list_lineno(int l) {
 void start_listing(void) {
 	/* AMSB always prints a blank line when it LISTs, so we do, too.
 	   unless it's disabled with -n, of course. */
-	if(initial_eol) fputc(EOL, outfile);
+	if(initial_eol) {
+		if(raw_output) {
+			fputc(0x9b, outfile);
+		} else {
+			fputc('\n', outfile);
+		}
+	}
 }
 
 int read_token(int literal);
@@ -466,6 +479,11 @@ int next_line(void) {
 
 	list_char(EOL);
 
+	if(!raw_output && (actx.inv)) {
+		char buf[20];
+		fputs(atascii_a2utf(&actx, ATA_CHR_FINISH, buf), outfile);
+	}
+
 	return 1;
 }
 
@@ -808,9 +826,9 @@ void print_help(void) {
 	puts("  -r: only list lines numbered from *start* to *end*");
 	puts("  --help, -h: print this help and exit");
 	puts("  --version:  print version number and exit");
-	puts("  -i -u -t -m -s: passed to a8cat (try 'a8cat -h')");
+	puts("  -i -u -t -m -s: same as a8cat (try 'a8cat -h')");
 	puts("file must be a tokenized (SAVEd) AMSB file. if not given, reads from stdin.");
-	puts("if outfile not given, writes to stdout (via pipe to a8cat)");
+	puts("if outfile not given, writes to stdout");
 }
 
 void version(void) {
@@ -849,14 +867,8 @@ void get_line_range(const char *arg) {
 }
 
 void parse_args(int argc, char **argv) {
-	const char *a8cat;
-	char tmp[10];
-	int opt, a8catopt = 0, ropt = 0;
-
-	a8cat = getenv("A8CAT");
-	if(!a8cat) a8cat = "a8cat";
-
-	strncpy(pipe_command, a8cat, BUFSIZ);
+	int opt, ropt = 0;
+	int ata_mode = ATA_MODE_UTF8, ata_flags = ATA_FLAG_NONE, a8catopt = 0;
 
 	if(argc >= 2) {
 		if(strcmp(argv[1], "--help") == 0) {
@@ -883,17 +895,11 @@ void parse_args(int argc, char **argv) {
 			case 'M': color =       0; break;
 			case 'h': print_help(); exit(0);
 			case 'r': get_line_range(optarg); ropt++; break;
-			case 'i':
-			case 'u':
-			case 't':
-			case 'm':
-			case 's':
-				a8catopt++;
-				if(strlen(pipe_command) > (BUFSIZ - 10))
-					os_err("too many a8cat options");
-				sprintf(tmp, " -%c", opt);
-				strcat(pipe_command, tmp);
-				break;
+			case 'i': ata_flags |= ATA_FLAG_ICS; a8catopt++; break;
+			case 'u': ata_flags |= ATA_FLAG_UNDERLINE; a8catopt++; break;
+			case 't': ata_flags |= ATA_FLAG_ICS; a8catopt++; break;
+			case 'm': ata_mode = ATA_MODE_MAGAZINE; a8catopt++; break;
+			case 's': ata_flags |= ATA_FLAG_STRIP_INVERSE; a8catopt++; break;
 			default: print_help(); exit(1);
 		}
 	}
@@ -910,6 +916,8 @@ void parse_args(int argc, char **argv) {
 		if(!initial_eol || check_only || raw_output || ropt || a8catopt || !color) {
 			warn("-[acimMnrstu] options ignored with -U, -C, or -D");
 		}
+	} else if(!raw_output) {
+		atascii_context_init(&actx, ata_mode, ata_flags);
 	}
 
 	if(optind >= argc) {
@@ -956,22 +964,18 @@ void open_output() {
 		outfile = stdout;
 		verbose(1, "using stdout for output");
 	} else {
-		const char *a8cat;
-		if((a8cat = getenv("A8CAT"))) {
-			verbose(1, "read A8CAT=%s from environment", a8cat);
+		if(color) {
+			verbose(1, "using pipe for output: %s", "colorize-amsb");
+			outfile = popen("colorize-amsb", "w"); /* "wb" not allowed! */
+			/* we probably never see this error. popen() fails only if
+				we feed it an invalid mode, or if fork() or pipe() fails,
+				or I suppose if some idjit has done a "rm -f /bin/sh"...
+				all other errors are caught at pclose(). */
+			if(!outfile) os_err("|%s: %s", "colorize-amsb", strerror(errno));
+			need_pclose = 1;
 		} else {
-			verbose(1, "A8CAT not set in environment, using 'a8cat'");
+			outfile = stdout;
 		}
-		if(color)
-			strcat(pipe_command, " | colorize-amsb");
-		verbose(1, "using pipe for output: %s", pipe_command);
-		outfile = popen(pipe_command, "w"); /* "wb" not allowed! */
-		/* we probably never see this error. popen() fails only if
-		   we feed it an invalid mode, or if fork() or pipe() fails,
-		   or I suppose if some idjit has done a "rm -f /bin/sh"...
-		   all other errors are caught at pclose(). */
-		if(!outfile) os_err("|%s: %s", pipe_command, strerror(errno));
-		need_pclose = 1;
 	}
 }
 
diff --git a/listamsb.rst b/listamsb.rst
index 3753bdf..2a86a23 100644
--- a/listamsb.rst
+++ b/listamsb.rst
@@ -38,14 +38,14 @@ If no **output-file** is given, output is to **stdout**. When output
 is to a file, the file will be overwritten without prompting if it
 already exists.
 
-When listing a program to **stdout**, output is piped through
-**a8cat**\(1), which converts the ATASCII output to UTF-8 that's
-human-readable on modern terminals. This means **a8cat** must be
-available in your **PATH**\. If this is not convenient, you can set
-the location of **a8cat** via the **A8CAT** environment variable.
-When listing in color (when the **-M** option is not used), **a8cat**\'s
-output is piped through **colorize-amsb**, which also must be available
-in your **PATH**\.
+By default, output is modern UTF-8 equivalents of the ATASCII
+character set, with ANSI control sequences for inverse video and
+color. See the **-M**, **-a**, **-i**, **-u**, **-t**, **-m**, **-s**
+options to change the output style.
+
+When listing in color (when the **-M** option is not used),
+**listamsb**\'s output is piped through **colorize-amsb**, which must
+be available in your **PATH**\.
 
 **listamsb** will not write binary data (ATASCII or tokenized BASIC)
 directly to a terminal, under any circumstances.
@@ -105,15 +105,28 @@ or **-U**.
 **-a**
   Output raw ATASCII. This option must be used with an
   **output-file**, a pipe, or redirection: **listamsb** will not write
-  ATASCII to a terminal. **a8cat** and **colorize-amsb** are not used, with this option.
+  ATASCII to a terminal. **colorize-amsb** is not used, with this option.
 
 **-c**
   Check only. No output on **stdout**. Diagnostics are still printed on stderr,
   and the exit status is unchanged. It's an error to give an **output-file** with
   this option.
 
-**-i**\, **-u**\, **-t**\, **-m**\, **-s**
-  These options are passed to **a8cat**. See its man page for details.
+**-i**
+  Input uses Atari XL/XE International Character Set encoding, rather than
+  ATASCII graphics.
+
+**-u**
+  Use underlining for inverse video (see **a8cat**\(1) for more detail).
+
+**-t**
+  Text mode (see **a8cat**\(1) for more detail).
+
+**-m**
+  Magazine listing mode (see **a8cat**\(1) for more detail).
+
+**-s**
+  Strip bit 7 (inverse video bit; see **a8cat**\(1) for more detail).
 
 **-M**
   Monochrome listing: disable color syntax highlighting, meaning **colorize-amsb**
@@ -270,13 +283,6 @@ continues processing.
 ENVIRONMENT
 ===========
 
-**A8CAT**
-
-  The path to the **a8cat**\(1) binary. Can be a full path or just
-  the executable name, in which case it will be searched for with the
-  normal **PATH** mechanism. If not set, *a8cat* is used. If there are
-  spaces or other metacharacters, they must be escaped.
-
 LIMITATIONS
 ===========