path: root/compiler.c

#include <stddef.h>
#include <stdio.h>
#include <stdlib.h>
#include <stdbool.h>
#include <string.h>



// Vector utils

#define VECT_MAX_GROW 100
#define VECT_MIN_SIZE 4

typedef struct {
	int _el_sz, count, size;
	void *data;
} Vector;

Vector vect_init(size_t item_size) {
	Vector out = {0};
	
	out._el_sz = item_size;
	out.size = VECT_MIN_SIZE;
	out.count = 0;
	out.data = malloc(out.size * out._el_sz);

	return out;
}


void _vect_grow(Vector *v) {
	if (v->size / 2 > VECT_MAX_GROW) {
		v->size += VECT_MAX_GROW;
	} else {
		v->size += v->size / 2;
	}
	v->data = realloc(v->data, v->size * v->_el_sz);
}

void _vect_shrink(Vector *v) {
	if (v->size / 2 > VECT_MIN_SIZE) {
		v->size = v->size / 2;
		v->data = realloc(v->data, v->size * v->_el_sz);
	}
}


bool vect_remove(Vector *v, size_t index) {
	if (index >= v->count) {
		return false;
	}

	char *remove = v->data + (index * v->_el_sz);
	char *override = v->data + (index + 1) * v->_el_sz;

	for(int i = 0; i < (v->count - 1) * v->_el_sz; i++) {
		remove[i] = override[i];
	}

	v->count -= 1;

	if (v->count < v->size / 4) {
		_vect_shrink(v);
	}

	return true;
}

void vect_pop(Vector *v) {
	vect_remove(v, v->count - 1);
}


bool vect_insert(Vector *v, size_t index, void *el) {
	if (index > v->count) {
		return false;
	}

	char *new_spot = v->data + (v->count + 1) * v->_el_sz;
	char *old_spot = v->data + v->count * v->_el_sz;

	for (int i = 0; i > (index * v->_el_sz); i--) {
		new_spot[i] = old_spot[i];
	}

	for (int i = 0; i < v->_el_sz; i++) {
		old_spot[i] = ((char *)el)[i];
	}

	v->count += 1;

	if (v->count == v->size - 1) {
		_vect_grow(v);
	}

	return true;
}

void vect_push(Vector *v, void *el) {
	vect_insert(v, v->count, el);
}


void *vect_get(Vector *v, size_t index) {
	if (index >= v->count) {
		return NULL;
	}

	return v->data + (v->_el_sz * index);
}


Vector vect_clone(Vector *v) {
	Vector out = {0};
	
	out._el_sz = v->_el_sz;
	out.count = 0;
	out.size = v->count + 1;

	out.data = malloc((out.count + 1) * out._el_sz);

	char *former = v->data;
	char *latter = out.data;

	for(int i = 0; i < out.count * out._el_sz; i++) {
		latter[i] = former[i];
	}

	return out;
}

Vector vect_from_string(char *s) {
	Vector out = vect_init(1);

	size_t i = 0;
	while(s[i] != 0) {
		vect_push(&out, s + i);
	}

	return out;
}

// Returns the vector data as a null-terminated string
// do NOT free this pointer. Not safe to use this string
// at the same time as you are adding or removing from the
// vector. Instead consider cloning the vector if you must
// have both, or want an independant copy of the string.
char *vect_as_string(Vector *v) {
	((char*)v->data)[v->count * v->_el_sz] = 0;
	return v->data;
}


void vect_end(Vector *v) {
	v->_el_sz = 0;
	v->count = 0;
	v->size = 0;
	free(v->data);
	v->data = NULL;
}



// Artifacts (vect of strings)

typedef Vector Artifact;

/* Splits the string via the given character, and
 * stores the split strings in an artifact
 */
Artifact art_from_str(const char *str, char split) {
	Artifact out = vect_init(sizeof(char *));

	char *cur = malloc(1);
	cur[0] = 0;
	int cur_len = 0;

	for (int i = 0; str[i] != 0; i++) {
		if (str[i] == split) {
			cur[cur_len] = 0;
			vect_push(&out, &cur);
			cur = malloc(1);
			cur_len = 0;
		} else {
			cur_len += 1;
			cur = realloc(cur, cur_len + 1);
			cur[cur_len - 1] = str[i];
		}
	}

	if (cur_len > 0) {
		cur[cur_len] = 0;
		vect_push(&out, &cur);
	} else {
		free(cur);
	}

	return out;
}

// Joins the string together with the provided character in between
// must free the returned data after use.
char *art_to_str(Artifact *art, char join) {
	char *out = malloc(1);
	int out_len = 0;

	for (int i = 0; i < art->count; i++) {
		char ** cpy = vect_get(art, i);
		for(int j = 0; cpy[j] != 0; j++) {
			out[out_len] = (*cpy)[j];
			out_len += 1;
			out = realloc(out, out_len + 1);
		}
	}

	out[out_len] = 0;

	return out;
}

// Pops a string off the end of the artifact,
// freeing the data associated
void art_pop_str(Artifact *art) {
	if (art->count == 0)
		return;
	
	char ** to_free = vect_get(art, art->count - 1);
	free(*to_free);
	vect_pop(art);
}

// Copies a string onto the artifact,
// you must free the original string
// manually if it was malloc-ed
void art_add_str(Artifact *art, char *str) {
	Vector copy = vect_from_string(str);
	char * copy_ptr = vect_as_string(&copy);
	vect_push(art, &copy_ptr);
}



// Types

typedef struct {
	char *name;       // Name of the type
	int size;         // Size (bytes) of the type
	Vector members;   // Member variables (Stored as variables)
	void *module;     // Module (for methods and member-type resolution) to tie the type to.
} Type;

typedef struct {
	char *name;
	Type *type;
	Vector ptr_chain;
	int location; // negative for on stack, positive or zero for in register
} Variable;

#define PTYPE_PTR 0
#define PTYPE_REF 1
#define PTYPE_ARR 2



// Copies the name, does not copy the module.
// Types should be freed at the end of the second pass,
// as they are shared among all variable structs
Type typ_init(char *name, void *module) {
	Type out = {0};

	Vector name_cpy = vect_from_string(name);
	out.name = vect_as_string(&name_cpy);
	out.members = vect_init(sizeof(Variable));
	out.module = module;
	out.size = 0;

	return out;
}

// Only for use by type_end on variable clean up
// at the end of the second pass
void var_deep_end(Variable *var);

// Deep end, will free all memory associated with the
// struct, including name and sub-member variables
void typ_end(Type *t) {
	free(t->name);
	t->module = NULL;

	for (int i = 0; i < t->members.count; i++) {
		Variable *to_end = vect_get(&(t->members), i);
		var_deep_end(to_end);
	}

	vect_end(&(t->members));
}



// Variables

// Initializes the variable, copying name, not copying type.
Variable var_init(char *name, Type *type) {
	Variable out = {0};
	
	Vector name_cpy = vect_from_string(name);
	out.name = vect_as_string(&name_cpy);
	out.type = type;
	out.ptr_chain = vect_init(sizeof(int));
	out.location = 0;

	return out;
}

void var_deep_end(Variable *var) {
	Variable *v = var;
	free(v->name);
	vect_end(&(v->ptr_chain));
	typ_end(v->type);
}

// Simple cleanup for variables while the second pass is ongoing.
void var_end(Variable *v) {
	free(v->name);
	vect_end(&(v->ptr_chain));
}

// Variable operations

// Type coercion engine
// TODO: all
Variable _op_coerce(Variable *base, Variable *to_coerce) {
	Variable out = {0};
	return out;
}



// Functions

typedef struct {
	char *name;
	Vector inputs, outputs;
	void *module;
} Function;



// Modules

typedef struct {
	char *name;
	bool exported;
	Vector vars, funcs, submods;
} Module;



// Whatev

// Tokenizer
typedef struct {
	char *data;
	int line, col;
	int type;
} Token;

#define TT_DEFWORD 0
#define TT_KEYWORD 1
#define TT_KEYTYPE 2
#define TT_LITERAL 3
#define TT_AUGMENT 4
#define TT_DELIMIT 5
#define TT_SPLITTR 6

char *KEYWORDS = "module,export,asm,if,else,loop,label,goto,continue,break,return,import,as,using,struct,method,interface,implements,operator,len,is";
char *KEYTYPES = "uint8,uint16,uint32,uint64,uint,int8,int16,int32,int64,int,float32,float64,float,comp64,comp,bool,vect,void,type";
char *RESERVED = "~`!@#$%^&*()[]{}+-=\"\'\\|:;/?>.<,";
char *MULTI_DELIMS = ";:#";

int token_type(char*data) {
	return TT_DEFWORD;
}


void help() {
	printf("\n");
	printf("Usage:\n");
	printf("\tctc - The TNSL compiler (written in c)\n\n");
	printf("\tctc [file in]              - compile the given file, writing output assembly in out.asm\n");
	printf("\tctc [file in] [file out]   - same as before, but write the output assembly to the given filename\n");
	printf("\t    -h                     - print this output message\n");
	printf("\n");
}

int main(int argc, char ** argv) {
	if (argc < 2 || strcmp(argv[1], "-h") == 0) {
		help();
		return 1;
	}

	for (int i = 0; i < argc; i++) {
		printf("%s\n", argv[i]);
	}

	return 0;
}