(uxnasm) Housekeeping

src/uxnasm.c

@@ -35,15 +35,16 @@ typedef struct {
 
 typedef struct {
 	Uint8 data[LENGTH];
-	Uint8 lambda_stack[0x100], lambda_ptr, lambda_count;
-	char scope[0x40], lambda[0x10], *location, *entry;
-	unsigned int ptr, length;
-	Uint16 label_len, macro_len, refs_len;
+	Uint8 lambda_stack[0x100], lambda_ptr, lambda_len;
+	Uint16 ptr, length, label_len, macro_len, refs_len;
+	char scope[0x40], lambda_name[0x05], *location;
 	Label labels[0x400];
 	Macro macros[0x100];
 	Reference refs[0x1000];
 } Program;
 
+char token[0x40];
+
 Program p;
 
 /* clang-format off */
@@ -56,6 +57,7 @@ static char ops[][4] = {
 };
 
 static char *runes = "|$@&,_.-;=!?#\"%~";
+static char *hexad = "0123456789abcdef";
 
 static int   scmp(char *a, char *b, int len) { int i = 0; while(a[i] == b[i]) if(!a[i] || ++i >= len) return 1; return 0; } /* string compare */
 static int   sihx(char *s) { int i = 0; char c; while((c = s[i++])) if(!(c >= '0' && c <= '9') && !(c >= 'a' && c <= 'f')) return 0; return i > 1; } /* string is hexadecimal */
@@ -77,9 +79,9 @@ error_top(const char *name, const char *msg)
 }
 
 static int
-error_asm(const char *name, const char *msg)
+error_asm(const char *name)
 {
-	fprintf(stderr, "%s: %s in @%s, %s:%d.\n", name, msg, p.scope, p.location, 123);
+	fprintf(stderr, "%s: %s in @%s, %s:%d.\n", name, token, p.scope, p.location, 123);
 	return 0;
 }
 
@@ -90,16 +92,6 @@ setlocation(char *name)
 	return name;
 }
 
-static char *
-sublabel(char *src, char *scope, char *name)
-{
-	if(slen(scope) + slen(name) >= 0x3f) {
-		error_asm("Sublabel length too long", name);
-		return NULL;
-	}
-	return scat(scat(scpy(scope, src, 0x40), "/"), name);
-}
-
 static Macro *
 findmacro(char *name)
 {
@@ -125,19 +117,20 @@ findopcode(char *s)
 {
 	int i;
 	for(i = 0; i < 0x20; i++) {
-		int m = 0;
+		int m = 3;
 		if(!scmp(ops[i], s, 3))
 			continue;
-		if(!i) i |= (1 << 7); /* force keep for LIT */
-		while(s[3 + m]) {
-			if(s[3 + m] == '2')
-				i |= (1 << 5); /* mode: short */
-			else if(s[3 + m] == 'r')
-				i |= (1 << 6); /* mode: return */
-			else if(s[3 + m] == 'k')
-				i |= (1 << 7); /* mode: keep */
+		if(!i)
+			i |= (1 << 7);
+		while(s[m]) {
+			if(s[m] == '2')
+				i |= (1 << 5);
+			else if(s[m] == 'r')
+				i |= (1 << 6);
+			else if(s[m] == 'k')
+				i |= (1 << 7);
 			else
-				return 0; /* failed to match */
+				return 0;
 			m++;
 		}
 		return i;
@@ -145,33 +138,6 @@ findopcode(char *s)
 	return 0;
 }
 
-static int
-makemacro(char *name, FILE *f)
-{
-	Macro *m;
-	char word[0x40];
-	if(findmacro(name))
-		return error_asm("Macro duplicate", name);
-	if(sihx(name) && slen(name) % 2 == 0)
-		return error_asm("Macro name is hex number", name);
-	if(findopcode(name) || scmp(name, "BRK", 4) || !slen(name))
-		return error_asm("Macro name is invalid", name);
-	if(p.macro_len == 0x100)
-		return error_asm("Macros limit exceeded", name);
-	m = &p.macros[p.macro_len++];
-	scpy(name, m->name, 0x40);
-	while(fscanf(f, "%63s", word) == 1) {
-		if(word[0] == '{') continue;
-		if(word[0] == '}') break;
-		if(word[0] == '%')
-			return error_asm("Macro error", name);
-		if(m->len >= 0x40)
-			return error_asm("Macro size exceeded", name);
-		scpy(word, m->items[m->len++], 0x40);
-	}
-	return 1;
-}
-
 static int
 isrune(char c)
 {
@@ -181,20 +147,46 @@ isrune(char c)
 	return 0;
 }
 
+static int
+isopcode(char *s)
+{
+	return findopcode(s) || scmp(s, "BRK", 4);
+}
+
+static int
+makemacro(char *name, FILE *f)
+{
+	Macro *m;
+	char word[0x40];
+	if(!slen(name)) return error_asm("Macro is empty");
+	if(findmacro(name)) return error_asm("Macro is duplicate");
+	if(sihx(name)) return error_asm("Macro is hex number");
+	if(isopcode(name)) return error_asm("Macro is opcode");
+	if(p.macro_len == 0x100) return error_asm("Macros limit exceeded");
+	m = &p.macros[p.macro_len++];
+	scpy(name, m->name, 0x40);
+	while(fscanf(f, "%63s", word) == 1) {
+		if(word[0] == '{') continue;
+		if(word[0] == '}') break;
+		if(word[0] == '%')
+			return error_asm("Macro error");
+		if(m->len >= 0x40)
+			return error_asm("Macro size exceeded");
+		scpy(word, m->items[m->len++], 0x40);
+	}
+	return 1;
+}
+
 static int
 makelabel(char *name)
 {
 	Label *l;
-	if(findlabel(name))
-		return error_asm("Label duplicate", name);
-	if(sihx(name) && (slen(name) == 2 || slen(name) == 4))
-		return error_asm("Label name is hex number", name);
-	if(findopcode(name) || scmp(name, "BRK", 4) || !slen(name))
-		return error_asm("Label name is invalid", name);
-	if(isrune(name[0]))
-		return error_asm("Label name is runic", name);
-	if(p.label_len == 0x400)
-		return error_asm("Labels limit exceeded", name);
+	if(!slen(name)) return error_asm("Label is empty");
+	if(findlabel(name)) return error_asm("Label is duplicate");
+	if(sihx(name)) return error_asm("Label is hex number");
+	if(isopcode(name)) return error_asm("Label is opcode");
+	if(isrune(name[0])) return error_asm("Label name is runic");
+	if(p.label_len == 0x400) return error_asm("Labels limit exceeded");
 	l = &p.labels[p.label_len++];
 	l->addr = p.ptr;
 	l->refs = 0;
@@ -205,26 +197,37 @@ makelabel(char *name)
 static char *
 makelambda(int id)
 {
-	scpy("lambda", p.lambda, 0x07);
-	p.lambda[6] = '0' + (id >> 0x4);
-	p.lambda[7] = '0' + (id & 0xf);
-	return p.lambda;
+	p.lambda_name[0] = (char)0xce;
+	p.lambda_name[1] = (char)0xbb;
+	p.lambda_name[2] = hexad[id >> 0x4];
+	p.lambda_name[3] = hexad[id & 0xf];
+	return p.lambda_name;
+}
+
+static char *
+makesublabel(char *src, char *scope, char *name)
+{
+	if(slen(scope) + slen(name) >= 0x3f) {
+		error_asm("Sublabel length too long");
+		return NULL;
+	}
+	return scat(scat(scpy(scope, src, 0x40), "/"), name);
 }
 
 static int
-makereference(char *scope, char *label, char rune, Uint16 addr)
+addref(char *scope, char *label, char rune, Uint16 addr)
 {
 	char subw[0x40], parent[0x40];
 	Reference *r;
 	if(p.refs_len >= 0x1000)
-		return error_asm("References limit exceeded", label);
+		return error_asm("References limit exceeded");
 	r = &p.refs[p.refs_len++];
 	if(label[0] == '{') {
-		p.lambda_stack[p.lambda_ptr++] = p.lambda_count;
-		scpy(makelambda(p.lambda_count++), r->name, 0x40);
+		p.lambda_stack[p.lambda_ptr++] = p.lambda_len;
+		scpy(makelambda(p.lambda_len++), r->name, 0x40);
 	} else if(label[0] == '&' || label[0] == '/') {
-		if(!sublabel(subw, scope, label + 1))
-			return error_asm("Invalid sublabel", label);
+		if(!makesublabel(subw, scope, label + 1))
+			return error_asm("Invalid sublabel");
 		scpy(subw, r->name, 0x40);
 	} else {
 		int pos = spos(label, '/');
@@ -244,34 +247,20 @@ static int
 writebyte(Uint8 b)
 {
 	if(p.ptr < TRIM)
-		return error_asm("Writing in zero-page", "");
-	else if(p.ptr > 0xffff)
-		return error_asm("Writing after the end of RAM", "");
+		return error_asm("Writing in zero-page");
+	else if(p.ptr == 0xffff)
+		return error_asm("Writing outside memory");
 	else if(p.ptr < p.length)
-		return error_asm("Memory overwrite", "");
+		return error_asm("Writing rewind");
 	p.data[p.ptr++] = b;
 	p.length = p.ptr;
 	return 1;
 }
 
-static int
-writeopcode(char *w)
-{
-	return writebyte(findopcode(w));
-}
-
 static int
 writeshort(Uint16 s, int lit)
 {
-	if(lit)
-		if(!writebyte(findopcode("LIT2"))) return 0;
-	return writebyte(s >> 8) && writebyte(s & 0xff);
-}
-
-static int
-writelitbyte(Uint8 b)
-{
-	return writebyte(findopcode("LIT")) && writebyte(b);
+	return (lit ? writebyte(findopcode("LIT2")) : 1) && writebyte(s >> 8) && writebyte(s & 0xff);
 }
 
 static int
@@ -280,10 +269,10 @@ doinclude(char *filename)
 	FILE *f;
 	char w[0x40];
 	if(!(f = fopen(setlocation(filename), "r")))
-		return error_asm("Include missing", filename);
+		return error_top("Include missing", filename);
 	while(fscanf(f, "%63s", w) == 1)
 		if(!parse(w, f))
-			return error_asm("Unknown token", w);
+			return error_top("Unknown token", w);
 	fclose(f);
 	return 1;
 }
@@ -296,7 +285,7 @@ parse(char *w, FILE *f)
 	Label *l;
 	Macro *m;
 	if(slen(w) >= 63)
-		return error_asm("Invalid token", w);
+		return error_asm("Invalid token");
 	switch(w[0]) {
 	case '(': /* comment */
 		if(slen(w) != 1) fprintf(stderr, "-- Malformed comment: %s\n", w);
@@ -312,22 +301,22 @@ parse(char *w, FILE *f)
 		break;
 	case '~': /* include */
 		if(!doinclude(w + 1))
-			return error_asm("Invalid include", w);
+			return error_asm("Invalid include");
 		break;
 	case '%': /* macro */
 		if(!makemacro(w + 1, f))
-			return error_asm("Invalid macro", w);
+			return error_asm("Invalid macro");
 		break;
 	case '|': /* pad-absolute */
 		if(sihx(w + 1))
 			p.ptr = shex(w + 1);
 		else if(w[1] == '&') {
-			if(!sublabel(subw, p.scope, w + 2) || !(l = findlabel(subw)))
-				return error_asm("Invalid sublabel", w);
+			if(!makesublabel(subw, p.scope, w + 2) || !(l = findlabel(subw)))
+				return error_asm("Invalid sublabel");
 			p.ptr = l->addr;
 		} else {
 			if(!(l = findlabel(w + 1)))
-				return error_asm("Invalid label", w);
+				return error_asm("Invalid label");
 			p.ptr = l->addr;
 		}
 		break;
@@ -335,52 +324,52 @@ parse(char *w, FILE *f)
 		if(sihx(w + 1))
 			p.ptr += shex(w + 1);
 		else if(w[1] == '&') {
-			if(!sublabel(subw, p.scope, w + 2) || !(l = findlabel(subw)))
-				return error_asm("Invalid sublabel", w);
+			if(!makesublabel(subw, p.scope, w + 2) || !(l = findlabel(subw)))
+				return error_asm("Invalid sublabel");
 			p.ptr += l->addr;
 		} else {
 			if(!(l = findlabel(w + 1)))
-				return error_asm("Invalid label", w);
+				return error_asm("Invalid label");
 			p.ptr += l->addr;
 		}
 		break;
 	case '@': /* label */
 		if(!makelabel(w + 1))
-			return error_asm("Invalid label", w);
+			return error_asm("Invalid label");
 		i = 0;
 		while(w[i + 1] != '/' && i < 0x3e && (p.scope[i] = w[i + 1]))
 			i++;
 		p.scope[i] = '\0';
 		break;
 	case '&': /* sublabel */
-		if(!sublabel(subw, p.scope, w + 1) || !makelabel(subw))
-			return error_asm("Invalid sublabel", w);
+		if(!makesublabel(subw, p.scope, w + 1) || !makelabel(subw))
+			return error_asm("Invalid sublabel");
 		break;
 	case '#': /* literals hex */
 		if(sihx(w + 1) && slen(w) == 3)
-			return writelitbyte(shex(w + 1));
+			return writebyte(findopcode("LIT")) && writebyte(shex(w + 1));
 		else if(sihx(w + 1) && slen(w) == 5)
 			return writeshort(shex(w + 1), 1);
 		else
-			return error_asm("Invalid hex literal", w);
+			return error_asm("Invalid hex literal");
 		break;
 	case '_': /* raw byte relative */
-		return makereference(p.scope, w + 1, w[0], p.ptr) && writebyte(0xff);
+		return addref(p.scope, w + 1, w[0], p.ptr) && writebyte(0xff);
 	case ',': /* literal byte relative */
-		return makereference(p.scope, w + 1, w[0], p.ptr + 1) && writelitbyte(0xff);
+		return addref(p.scope, w + 1, w[0], p.ptr + 1) && writebyte(findopcode("LIT")) && writebyte(0xff);
 	case '-': /* raw byte absolute */
-		return makereference(p.scope, w + 1, w[0], p.ptr) && writebyte(0xff);
+		return addref(p.scope, w + 1, w[0], p.ptr) && writebyte(0xff);
 	case '.': /* literal byte zero-page */
-		return makereference(p.scope, w + 1, w[0], p.ptr + 1) && writelitbyte(0xff);
+		return addref(p.scope, w + 1, w[0], p.ptr + 1) && writebyte(findopcode("LIT")) && writebyte(0xff);
 	case ':': fprintf(stderr, "Deprecated rune %s, use =%s\n", w, w + 1);
 	case '=': /* raw short absolute */
-		return makereference(p.scope, w + 1, w[0], p.ptr) && writeshort(0xffff, 0);
+		return addref(p.scope, w + 1, w[0], p.ptr) && writeshort(0xffff, 0);
 	case ';': /* literal short absolute */
-		return makereference(p.scope, w + 1, w[0], p.ptr + 1) && writeshort(0xffff, 1);
+		return addref(p.scope, w + 1, w[0], p.ptr + 1) && writeshort(0xffff, 1);
 	case '?': /* JCI */
-		return makereference(p.scope, w + 1, w[0], p.ptr + 1) && writebyte(0x20) && writeshort(0xffff, 0);
+		return addref(p.scope, w + 1, w[0], p.ptr + 1) && writebyte(0x20) && writeshort(0xffff, 0);
 	case '!': /* JMI */
-		return makereference(p.scope, w + 1, w[0], p.ptr + 1) && writebyte(0x40) && writeshort(0xffff, 0);
+		return addref(p.scope, w + 1, w[0], p.ptr + 1) && writebyte(0x40) && writeshort(0xffff, 0);
 	case '"': /* raw string */
 		i = 0;
 		while((c = w[++i]))
@@ -388,15 +377,15 @@ parse(char *w, FILE *f)
 		break;
 	case '}': /* lambda end */
 		if(!makelabel(makelambda(p.lambda_stack[--p.lambda_ptr])))
-			return error_asm("Invalid label", w);
+			return error_asm("Invalid label");
 		break;
 	case '[':
 	case ']':
 		if(slen(w) == 1) break; /* else fallthrough */
 	default:
 		/* opcode */
-		if(findopcode(w) || scmp(w, "BRK", 4))
-			return writeopcode(w);
+		if(isopcode(w))
+			return writebyte(findopcode(w));
 		/* raw byte */
 		else if(sihx(w) && slen(w) == 2)
 			return writebyte(shex(w));
@@ -410,7 +399,7 @@ parse(char *w, FILE *f)
 					return 0;
 			return 1;
 		} else
-			return makereference(p.scope, w, ' ', p.ptr + 1) && writebyte(0x60) && writeshort(0xffff, 0);
+			return addref(p.scope, w, ' ', p.ptr + 1) && writebyte(0x60) && writeshort(0xffff, 0);
 	}
 	return 1;
 }
@@ -430,7 +419,7 @@ resolve(void)
 				return error_top("Unknown relative reference", r->name);
 			p.data[r->addr] = (Sint8)(l->addr - r->addr - 2);
 			if((Sint8)p.data[r->addr] != (l->addr - r->addr - 2))
-				return error_asm("Relative reference is too far", r->name);
+				return error_top("Relative reference is too far", r->name);
 			l->refs++;
 			break;
 		case '-':
@@ -467,12 +456,11 @@ resolve(void)
 static int
 assemble(FILE *f)
 {
-	char w[0x40];
 	p.ptr = 0x100;
 	scpy("on-reset", p.scope, 0x40);
-	while(fscanf(f, "%62s", w) == 1)
-		if(slen(w) > 0x3d || !parse(w, f))
-			return error_asm("Invalid token", w);
+	while(fscanf(f, "%62s", token) == 1)
+		if(slen(token) > 0x3d || !parse(token, f))
+			return error_asm("Invalid token");
 	return resolve();
 }
 
@@ -481,9 +469,7 @@ review(char *filename)
 {
 	int i;
 	for(i = 0; i < p.label_len; i++)
-		if(p.labels[i].name[0] >= 'A' && p.labels[i].name[0] <= 'Z')
-			continue; /* Ignore capitalized labels(devices) */
-		else if(!p.labels[i].refs)
+		if(p.labels[i].name[0] - 'A' > 25 && !p.labels[i].refs)
 			fprintf(stdout, "-- Unused label: %s\n", p.labels[i].name);
 	fprintf(stdout,
 		"Assembled %s in %d bytes(%.2f%% used), %d labels, %d macros.\n",
@@ -518,23 +504,14 @@ int
 main(int argc, char *argv[])
 {
 	FILE *src, *dst;
-	if(argc == 1)
-		return error_top("usage", "uxnasm [-v] input.tal output.rom");
-	if(argv[1][0] == '-' && argv[1][1] == 'v')
-		return !fprintf(stdout, "Uxnasm - Uxntal Assembler, 7 Mar 2024.\n");
-	if(!(src = fopen(setlocation(argv[1]), "r")))
-		return !error_top("Invalid input", argv[1]);
-	p.entry = argv[1];
-	if(!assemble(src))
-		return !error_top("Assembly", "Failed to assemble rom.");
-	if(!(dst = fopen(argv[2], "wb")))
-		return !error_top("Invalid Output", argv[2]);
-	if(p.length <= TRIM)
-		return !error_top("Assembly", "Output rom is empty.");
+	if(argc == 1) return error_top("usage", "uxnasm [-v] input.tal output.rom");
+	if(scmp(argv[1], "-v", 2)) return !fprintf(stdout, "Uxnasm - Uxntal Assembler, 25 Mar 2024.\n");
+	if(!(src = fopen(setlocation(argv[1]), "r"))) return !error_top("Invalid input", argv[1]);
+	if(!assemble(src)) return !error_top("Assembly", "Failed to assemble rom.");
+	if(!(dst = fopen(argv[2], "wb"))) return !error_top("Invalid Output", argv[2]);
+	if(p.length <= TRIM) return !error_top("Assembly", "Output rom is empty.");
+	review(argv[2]);
 	fwrite(p.data + TRIM, p.length - TRIM, 1, dst);
-	if(!scmp(argv[2], "-", 2)) {
-		review(argv[2]);
-		writesym(argv[2]);
-	}
+	writesym(argv[2]);
 	return 0;
 }