spl

commit c6cc2ba557076fd1385da67a6381e3a427a2124b
Author: thing1 <thing1@seacrossedlovers.xyz>
Date:   Wed, 19 Nov 2025 21:38:32 +0000

init commit

Diffstat:
A
.gitignore
 | 
2
++
A
Makefile
 | 
12
++++++++++++
A
lexer.c
 | 
62
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
A
lexer.h
 | 
29
+++++++++++++++++++++++++++++
A
note.md
 | 
99
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
A
spl.c
 | 
68
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

6 files changed, 272 insertions(+), 0 deletions(-)
diff --git a/.gitignore b/.gitignore
@@ -0,0 +1,2 @@
+*.o
+spl
diff --git a/Makefile b/Makefile
@@ -0,0 +1,12 @@
+CFLAGS=-ggdb
+
+SRC  = spl.c lexer.c
+OBJ  = ${SRC:.c=.o}
+EXEC = spl
+
+${EXEC}: ${OBJ}
+
+${OBJ}: ${SRC}
+
+clean:
+	rm -rf ${OBJ} ${EXEC}
diff --git a/lexer.c b/lexer.c
@@ -0,0 +1,62 @@
+#include <stdlib.h>
+#include <stdio.h>
+#include <ctype.h>
+
+#include "lexer.h"
+
+void 
+lexErr(lex *l) {
+	fprintf(stderr, "unexpected token '%c'\n", *l->ptr); 
+	exit(1);
+}
+
+int
+lexNum(lex *l) {
+	int n = atoi(l->ptr);
+	while (isdigit(*l->ptr))
+		l->ptr++;
+	return n;
+}
+
+tok
+next(lex *l) {
+	tok t = {0}; 
+	switch (*l->ptr) {
+	case 0:
+		t.op = LEOF;	
+		return t;
+
+	case '\n': case ' ': case '\t': case '\v':
+		l->ptr++;	
+		return next(l);
+
+	case ADD: case SUB: case MUL: case DIV:
+		t.op = *l->ptr++;
+		return t;
+	}
+	if (isdigit(*l->ptr)) {
+		t.op = INT;
+		t.n = lexNum(l);
+	}
+	else lexErr(l);
+
+	return t;
+}
+
+void
+printTok(tok *t) {
+	switch (t->op) {
+	case INT: printf("INT: %d\n", t->n); break;
+	case LEOF: printf("$\n"); break;
+	case ADD: printf("ADD\n"); break;
+	case SUB: printf("SUB\n"); break;
+	case MUL: printf("MUL\n"); break;
+	case DIV: printf("DIV\n"); break;
+	}
+}
+
+lex 
+mklexer(char *input) {
+	lex l = (lex){input, input};
+	return l;
+}
diff --git a/lexer.h b/lexer.h
@@ -0,0 +1,29 @@
+#ifndef __LEXER_H_
+#define __LEXER_H_
+
+enum ops {
+	NOP = 0,
+	INT = -1,
+	ADD = '+',
+	SUB = '-',
+	MUL = '*',
+	DIV = '/',
+	LEOF = '$',
+};
+
+typedef struct tok {
+	short op;
+	int n;
+} tok;
+
+typedef struct lex {
+	char *input, *ptr;
+} lex;
+
+void lexErr(lex *l); 
+int lexNum(lex *l);
+tok next(lex *l); 
+void printTok(tok *t); 
+lex mklexer(char *input); 
+
+#endif
diff --git a/note.md b/note.md
@@ -0,0 +1,99 @@
+pivot parser
+============
+
+a parser could work like so
+
+2 + 4 * 3 - 1 = 13
+
+loop through all tokens and find the first token with the lowest presidence
+
+2 + 4 * 3 - 1
+  ^    
+
+split the list
+
+2
++
+4 * 3 - 1
+
+repeat
+
+4 * 3
+-
+1
+
+4
+*
+3
+
+assemble the tree
+
+      +
+     / \
+    2   -
+       / \
+      *   1
+     / \
+    4   3
+
+2 + ((4 * 3) - 1)
+2 + (12 - 1)
+2 + 11
+13
+
+another example
+
+4 * 2 + (4 - 1) = 11
+
+4 * 2
++
+(4 - 1)
+
+4
+*
+2
+
+        +
+       / \
+      * (4 - 1)
+     / \
+    4   2
+
+4 - 1
+
+4
+-
+1
+
+        +
+       / \
+      *   \
+     / \   \
+    4   2   \
+             -
+            / \
+           4   1
+
+(4 * 2) + (4 - 1)
+8 + 3
+11
+
+
+func(a) * (10 - 4)
+
+func(a)
+*
+(10 - 4)
+
+            *
+           / \
+      func(a) \
+           (10 - 4)
+
+            *
+           / \
+      func(a) -
+             / \
+            10  4
+
+func(a) * (10 - 4)
diff --git a/spl.c b/spl.c
@@ -0,0 +1,68 @@
+#include <stdlib.h>
+#include <stdio.h>
+#include <ctype.h>
+
+#include "lexer.h"
+
+typedef struct rexpr {
+	char op;
+	
+	struct rexpr *expr[2];
+	int n[2];
+} rexpr;
+
+void
+parserErr(const char *msg) {
+	fprintf(stderr, "%s\n", msg);
+	exit(1);
+}
+
+int
+getpres(rexpr r) {
+	switch (r.op) {
+	case INT: return 0;
+
+	case ADD: return 1;
+	case SUB: return 1;
+	
+	case MUL: return 2;
+	case DIV: return 2;
+
+	case LEOF: return -1;
+	}
+}
+
+int
+isop(tok t) {
+	return (t.op == ADD || t.op == SUB || t.op == MUL || t.op == DIV);
+}
+
+rexpr *
+parseRexpr(lex *l) {
+	rexpr *e = malloc(sizeof(rexpr)), *e2;
+	tok t1 = next(l), t2 = next(l);
+
+	if (t1.op != INT) 
+		parserErr("Expected INT in rexpr");
+	e->n[0] = t1.n;
+	if (isop(t2)) {
+		e->op = t2.op;
+		e2 = parseRexpr(l);
+		if (e2->op == INT) e->n[1] = e2->n[0];
+		else e->expr[1] = e2;
+	}
+	else if (t2.op == LEOF) {
+		e->op = INT;
+		return e;
+	}
+	else
+		parserErr("Unexpected token in rexpr");
+}
+
+int
+main() {
+	lex l = mklexer("5 + 2");
+
+	rexpr *e = parseRexpr(&l);
+
+}