temporarely disabled variable size reading

.vscode/launch.json

@@ -7,7 +7,7 @@
             "request": "launch",
             "program": "${workspaceFolder}/bin/tcpu",
             "windows": { "program": "${workspaceFolder}/bin/tcpu.exe" },
-            "args": [ "-c", "s.tasm", "o.bin" ],
+            "args": [ "-c", "../examples/s.tasm", "o.bin", "--debug" ],
             "cwd": "${workspaceFolder}/bin",
             "preLaunchTask": "build_exec_dbg",
             "stopAtEntry": false,

examples/s.tasm

@@ -0,0 +1,16 @@
+/*
+"hello world" program in my assembly language
+*/
+
+.data:
+// named array of 8-bit values
+const8 msg "Hello, World :3\0"
+
+.main:
+push ax 1; // sys_write
+push bx 1; // stdout
+push cx @msg; // address of msg data
+push dx #msg; // size of msg data
+sys
+push ax 0
+exit

src/VM/VM.c

@@ -49,7 +49,7 @@ i32 VM_boot(VM* vm){
     while (vm->current_pos < vm->data_size){
         u8 opcode = vm->data[vm->current_pos];
 
-        const Instruction* instr = Instruction_getFromOpcode(opcode);
+        const Instruction* instr = Instruction_getByOpcode(opcode);
         // printfe("[at 0x%x] %02X %s\n", (u32)vm->current_pos, opcode, instr->name);
         if(instr == NULL){
             VM_setError(vm, "unknown opcode %02X", opcode);
@@ -77,7 +77,9 @@ i32 VM_boot(VM* vm){
 
 bool VM_dataRead(VM* vm, void* dst, size_t pos, size_t size){
     if(pos + size >= vm->data_size){
-        VM_setError(vm, "can't read %lli bytes from 0x%x, because only %lli are avaliable",
+        VM_setError(vm, 
+            "can't read " IFWIN("%lli", "%li") " bytes from 0x%x, because only "
+                IFWIN("%lli", "%li") " are avaliable",
             size, (u32)pos, vm->data_size - size);
         return false;
     }

src/VM/VM.h

@@ -1,5 +1,6 @@
 #pragma once
 #include "../std.h"
+#include "../string/str.h"
 
 typedef union Register {
     u32 u32v;

src/collections/Array.h

@@ -1,20 +1,24 @@
 #pragma once
 #include "../std.h"
 
+#define Array_construct(T, DATA, LEN) ((Array_##T){ .data = DATA, .len = LEN })
+
+/// creates Array_##T from a const array
+#define ARRAY(T, A...) Array_construct(T, ((T[])A), ARRAY_SIZE(((T[])A)))
+
 #define Array_declare(T)\
-    typedef struct Array_##T {\
+typedef struct Array_##T {\
     T* data;\
     u32 len;\
-    } Array_##T;\
-    \
-    static inline Array_##T Array_##T##_construct(T* data_ptr, u32 len) {\
-        return (Array_##T){ .data = data_ptr, .len = len };\
-    }\
-    \
-    static inline Array_##T Array_##T##_alloc(u32 len){\
-        return Array_##T##_construct(malloc(len * sizeof(T)), len);\
-    }\
-    static inline void Array_##T##_realloc(Array_##T* ptr, u32 new_len){\
-        ptr->data = realloc(ptr->data, new_len * sizeof(T));\
+} Array_##T;\
+\
+static inline Array_##T Array_##T##_alloc(u32 len){\
+    return Array_construct(T, (T*)malloc(len * sizeof(T)), len);\
+}\
+static inline void Array_##T##_realloc(Array_##T* ptr, u32 new_len){\
+    ptr->data = (T*)realloc(ptr->data, new_len * sizeof(T));\
     ptr->len = new_len;\
-    }
+}
+
+Array_declare(u8)
+Array_declare(u32)

src/collections/HashMap.h

@@ -0,0 +1,140 @@
+#pragma once
+#include "../std.h"
+#include "../string/str.h"
+#include "Array.h"
+#include "List.h"
+
+//TODO: sorting of bucket and binary search
+//TODO: delayed deletion
+
+#define __HashMap_HASH_FUNC str_hash32
+#define __HashMapBucket_MAX_LEN 16
+
+#define HashMap_DESTROY_VALUE_FUNC_NULL ((void (*)(void*))NULL) 
+
+/// call this in a header file
+///@param T Value type
+#define HashMap_declare(T)\
+typedef struct KeyValue_##T {\
+    str key;\
+    T value;\
+    u32 hash;\
+} KeyValue_##T;\
+\
+List_declare(KeyValue_##T);\
+\
+typedef struct HashMapBucket_##T {\
+    List_KeyValue_##T kvs;\
+} HashMapBucket_##T;\
+\
+typedef struct HashMap_##T {\
+    HashMapBucket_##T* table;\
+    u32 height;\
+    u16 height_n;\
+} HashMap_##T;\
+\
+void HashMap_##T##_alloc(HashMap_##T* ptr);\
+void HashMap_##T##_free(HashMap_##T* ptr);\
+T* NULLABLE(HashMap_##T##_tryGetPtr)(HashMap_##T* ptr, str key);\
+bool HashMap_##T##_tryPush(HashMap_##T* ptr, str key, T value);\
+bool HashMap_##T##_tryDelete(HashMap_##T* ptr, str key);\
+
+
+
+/// call this in a source code file
+///@param T Value type
+///@param DESTROY_VALUE_FUNC `void foo (T*)` or HashMap_DESTROY_VALUE_FUNC_NULL
+#define HashMap_define(T, DESTROY_VALUE_FUNC)\
+List_define(KeyValue_##T);\
+\
+static const Array_u32 __HashMap_##T##_heights = ARRAY(u32, {\
+    17, 31, 61, 127, 257, 521, 1021, 2053, 4099, 8191, 16381, 32771,\
+    65521, 131071, 262147, 524287, 1048583, 2097169, 4194319,\
+    8388617, 16777213, 33554467, 67108859, 134217757, 268435493\
+});\
+\
+void HashMap_##T##_alloc(HashMap_##T* ptr){\
+    ptr->height_n = 0;\
+    ptr->height = __HashMap_##T##_heights.data[0];\
+    ptr->table = (HashMapBucket_##T*)malloc(ptr->height * sizeof(HashMapBucket_##T));\
+    memset(ptr->table, 0, ptr->height * sizeof(HashMapBucket_##T));\
+}\
+\
+void HashMap_##T##_free(HashMap_##T* ptr){\
+    for(u32 i = 0; i < ptr->height; i++){\
+            for(u32 j = 0; j < ptr->table[i].kvs.len; j++){\
+                KeyValue_##T* kv_ptr = &ptr->table[i].kvs.data[j];\
+                if(DESTROY_VALUE_FUNC){\
+                    DESTROY_VALUE_FUNC(&kv_ptr->value);\
+                }\
+                free(kv_ptr->key.data);\
+            }\
+        \
+        free(ptr->table[i].kvs.data);\
+    }\
+    \
+    free(ptr->table);\
+}\
+\
+T* NULLABLE(HashMap_##T##_tryGetPtr)(HashMap_##T* ptr, str key){\
+    u32 hash = __HashMap_HASH_FUNC(key);\
+    HashMapBucket_##T* bu = &ptr->table[hash % ptr->height];\
+    for(u32 i = 0; i < bu->kvs.len; i++){\
+        if(bu->kvs.data[i].hash == hash && str_equals(bu->kvs.data[i].key, key)){\
+            return &bu->kvs.data[i].value;\
+        }\
+    }\
+    \
+    return NULL;\
+}\
+\
+bool HashMap_##T##_tryPush(HashMap_##T* ptr, str key, T value){\
+    u32 hash = __HashMap_HASH_FUNC(key);\
+    HashMapBucket_##T* bu = &ptr->table[hash % ptr->height];\
+    for(u32 i = 0; i < bu->kvs.len; i++){\
+        if(bu->kvs.data[i].hash == hash && str_equals(bu->kvs.data[i].key, key)){\
+            return false;\
+        }\
+    }\
+    \
+    if(bu->kvs.len >= __HashMapBucket_MAX_LEN){\
+        u32 height_expanded_n = ptr->height_n + 1;\
+        if(height_expanded_n >= __HashMap_##T##_heights.len){\
+            printf("ERROR: HashMap_" #T " IS FULL\n");\
+            return false;\
+        }\
+        \
+        u32 height_expanded = __HashMap_##T##_heights.data[height_expanded_n];\
+        HashMapBucket_##T* table_expanded = (HashMapBucket_##T*)malloc(height_expanded * sizeof(HashMapBucket_##T));\
+        memset(table_expanded, 0, height_expanded * sizeof(HashMapBucket_##T));\
+        for(u32 i = 0; i < height_expanded; i++){\
+            for(u32 j = 0; j < ptr->table[i].kvs.len; j++){\
+                KeyValue_##T kv = ptr->table[i].kvs.data[j];\
+                List_KeyValue_##T##_push(&table_expanded[kv.hash % height_expanded].kvs, kv);\
+            }\
+            \
+            free(ptr->table[i].kvs.data);\
+        }\
+        free(ptr->table);\
+        ptr->table = table_expanded;\
+        ptr->height = height_expanded;\
+        ptr->height_n = height_expanded_n;\
+        bu = &ptr->table[hash % ptr->height];\
+    }\
+    \
+    KeyValue_##T kv = { .key = str_copy(key), .value = value, .hash = hash };\
+    List_KeyValue_##T##_push(&bu->kvs, kv);\
+    return true;\
+}\
+\
+bool HashMap_##T##_tryDelete(HashMap_##T* ptr, str key){\
+    u32 hash = __HashMap_HASH_FUNC(key);\
+    HashMapBucket_##T* bu = &ptr->table[hash % ptr->height];\
+    for(u32 i = 0; i < bu->kvs.len; i++){\
+        if(bu->kvs.data[i].hash == hash && str_equals(bu->kvs.data[i].key, key)){\
+            return List_KeyValue_##T##_tryRemoveAt(&bu->kvs, i);\
+        }\
+    }\
+    \
+    return false;\
+}

src/collections/List.c

@@ -1,5 +1,4 @@
 #include "List.h"
 
-List_define(cstr);
 List_define(u32);
 List_define(u8);

src/collections/List.h

@@ -1,6 +1,9 @@
 #pragma once
 #include "../std.h"
 
+// minimal max_len after initial (0)
+#define __List_min_size 16
+
 #define List_declare(T)\
     typedef struct List_##T {\
         T* data;\
@@ -12,34 +15,61 @@
         return (List_##T){ .data = data_ptr, .len = len, .max_len = max_len };\
     }\
     \
-    static inline List_##T List_##T##_alloc(u32 len){\
-        return List_##T##_construct(len > 0 ? malloc(len * sizeof(T)) : NULL, 0, 0);\
-    }\
+    List_##T List_##T##_alloc(u32 initial_len);\
     \
-    void List_##T##_push(List_##T* ptr, T value);
+    T* List_##T##_expand(List_##T* ptr, u32 count);\
+    void List_##T##_push(List_##T* ptr, T value);\
+    void List_##T##_pushMany(List_##T* ptr, T* values, u32 count);\
+    bool List_##T##_tryRemoveAt(List_##T* ptr, u32 i);\
 
 
 #define List_define(T)\
-    void List_##T##_push(List_##T* ptr, T value){\
-        u32 max_len = ptr->max_len;\
-        if(ptr->len == max_len){\
-            max_len = max_len * 1.5;\
-            max_len += __List_padding_in_sizeof_T(T);\
+    List_##T List_##T##_alloc(u32 initial_len){\
+        if(initial_len == 0)\
+            return List_##T##_construct((T*)NULL, 0, 0);\
+        u32 max_len = ALIGN_TO(initial_len, sizeof(void*)/sizeof(T));\
         /* branchless version of max(max_len, __List_min_size) */\
         max_len += (max_len < __List_min_size) * (__List_min_size - max_len);\
-            ptr->data = realloc(ptr->data, max_len * sizeof(T));\
-            ptr->max_len = max_len;\
+        return List_##T##_construct((T*)malloc(max_len * sizeof(T)), 0, max_len);\
+    }\
+    \
+    T* List_##T##_expand(List_##T* ptr, u32 count){\
+        u32 occupied_len = ptr->len;\
+        u32 expanded_max_len = ptr->max_len;\
+        expanded_max_len += (expanded_max_len < __List_min_size) * (__List_min_size - expanded_max_len);\
+        ptr->len += count;\
+        while(ptr->len > expanded_max_len){\
+            expanded_max_len *= 2;\
+        }\
+        u32 alloc_size = expanded_max_len * sizeof(T);\
+        if(ptr->data == NULL)\
+            ptr->data = (T*)malloc(alloc_size);\
+        else ptr->data = (T*)realloc(ptr->data, alloc_size);\
+        ptr->max_len = expanded_max_len;\
+        return ptr->data + occupied_len;\
+    }\
+    \
+    void List_##T##_push(List_##T* ptr, T value){\
+        T* empty_cell_ptr = List_##T##_expand(ptr, 1);\
+        *empty_cell_ptr = value;\
+    }\
+    \
+    void List_##T##_pushMany(List_##T* ptr, T* values, u32 count){\
+        T* empty_cell_ptr = List_##T##_expand(ptr, count);\
+        memcpy(empty_cell_ptr, values, count * sizeof(T));\
+    }\
+    \
+    bool List_##T##_tryRemoveAt(List_##T* ptr, u32 i){\
+        if(ptr->len == 0 || i >= ptr->len)\
+            return false;\
+        \
+        ptr->len--;\
+        for(; i < ptr->len; i++){\
+            ptr->data[i] = ptr->data[i + 1];\
+        }\
+        return true;\
     }\
-        ptr->data[ptr->len++] = value;\
-    }
 
-#define __List_min_size 16
 
-// sizeof(T) == 1  - padding is 7 of sizeof(T)
-// sizeof(T) == 2  - padding is 3 of sizeof(T)
-// sizeof(T) == 4  - padding is 1 of sizeof(T)
-#define __List_padding_in_sizeof_T(T) ((8 - sizeof(T) % 8) / sizeof(T) )
-
-List_declare(cstr);
 List_declare(u32);
 List_declare(u8);

src/compiler/AST.c

@@ -0,0 +1,52 @@
+#include "AST.h"
+
+List_define(Argument);
+List_define(Operation);
+List_define(DataDefinition);
+List_define(Section);
+
+static str _ArgumentType_str[] = {
+    STR("Unset"),
+    STR("Register"),
+    STR("ConstValue"),
+    STR("VarDataName"),
+    STR("ConstDataPointer"),
+    STR("ConstDataSize"),
+};
+
+str ArgumentType_toString(ArgumentType t){
+    if(t >= ARRAY_SIZE(_ArgumentType_str))
+        return STR("!!ArgumentType INDEX_ERROR!!");
+    return _ArgumentType_str[t];
+}
+
+
+void Section_init(Section* sec, str name){
+    sec->name = name;
+    sec->data = List_DataDefinition_alloc(256);
+    sec->code = List_Operation_alloc(1024);
+}
+
+void Section_free(Section* sec){
+    for(u32 i = 0; i < sec->data.len; i++){
+        free(sec->data.data[i].data.data);
+    }
+    free(sec->data.data);
+
+    for(u32 i = 0; i < sec->code.len; i++){
+        free(sec->code.data[i].args.data);
+    }
+    free(sec->code.data);
+}
+
+
+void AST_init(AST* ast){
+    ast->sections = List_Section_alloc(32);
+}
+
+void AST_free(AST* ast){
+    for(u32 i = 0; i != ast->sections.len; i++){
+        Section_free(&ast->sections.data[i]);
+    }
+    free(ast->sections.data);
+}

src/compiler/AST.h

@@ -0,0 +1,65 @@
+#pragma once
+#include "../std.h"
+#include "../string/str.h"
+#include "../instructions/instructions.h"
+#include "../instructions/registers.h"
+#include "../collections/List.h"
+
+typedef enum ArgumentType {
+    ArgumentType_Unset,
+    ArgumentType_Register,
+    ArgumentType_ConstValue,
+    ArgumentType_VarDataName,
+    ArgumentType_ConstDataPointer,
+    ArgumentType_ConstDataSize,
+} ArgumentType;
+
+str ArgumentType_toString(ArgumentType t);
+
+typedef struct Argument {
+    ArgumentType type;
+    union {
+        i64 i;
+        f64 f;
+        str data_name;
+        RegisterCode register_code;
+    } value;
+} Argument;
+
+List_declare(Argument);
+
+
+typedef struct Operation {
+    List_Argument args;
+    Opcode opcode;
+} Operation;
+
+List_declare(Operation);
+
+
+typedef struct DataDefinition {
+    str name;
+    List_u8 data;
+    u32 element_size;
+} DataDefinition;
+
+List_declare(DataDefinition);
+
+
+typedef struct Section {
+    str name;
+    List_DataDefinition data;
+    List_Operation code;
+} Section;
+
+List_declare(Section);
+
+void Section_init(Section* Section, str name);
+void Section_free(Section* Section);
+
+typedef struct AST {
+    List_Section sections;
+} AST;
+
+void AST_init(AST* ast);
+void AST_free(AST* ast);

src/compiler/Binary.c

@@ -0,0 +1,40 @@
+#include "Binary.h"
+
+List_define(ConstDataProps);
+HashMap_define(ConstDataProps, HashMap_DESTROY_VALUE_FUNC_NULL);
+List_define(NamedRef);
+List_define(CompiledSection);
+HashMap_define(CompiledSectionPtr, HashMap_DESTROY_VALUE_FUNC_NULL);
+
+
+void CompiledSection_construct(CompiledSection* ptr, str name){
+    ptr->name = name;
+    ptr->next = NULL;
+    ptr->offset = 0;
+    ptr->const_data_props_list = List_ConstDataProps_construct(NULL, 0, 0);
+    ptr->named_refs = List_NamedRef_construct(NULL, 0, 0);
+    ptr->bytes = List_u8_alloc(64);
+}
+
+void CompiledSection_free(CompiledSection* ptr){
+    free(ptr->const_data_props_list.data);
+    free(ptr->named_refs.data);
+    free(ptr->bytes.data);
+}
+
+
+void BinaryObject_construct(BinaryObject* ptr){
+    ptr->section_list = List_CompiledSection_alloc(64);
+    HashMap_CompiledSectionPtr_alloc(&ptr->section_map);
+    HashMap_ConstDataProps_alloc(&ptr->const_data_map);
+}
+
+void BinaryObject_free(BinaryObject* ptr){
+    for(u32 i = 0; i < ptr->section_list.len; i++){
+        CompiledSection_free(&ptr->section_list.data[i]);
+    }
+    free(ptr->section_list.data);
+    
+    HashMap_CompiledSectionPtr_free(&ptr->section_map);
+    HashMap_ConstDataProps_free(&ptr->const_data_map);
+}

src/compiler/Binary.h

@@ -0,0 +1,65 @@
+#pragma once
+#include "../std.h"
+#include "../string/str.h"
+#include "../instructions/instructions.h"
+#include "../instructions/registers.h"
+#include "../collections/List.h"
+#include "../collections/HashMap.h"
+#include "AST.h"
+
+typedef struct CompiledSection CompiledSection;
+typedef struct ConstDataProps {
+    str name;
+    u32 size;   // size in bytes
+    u32 offset; // offset in bytes from section start
+} ConstDataProps;
+
+#define ConstDataProps_construct(NAME, SIZE, OFFSET) ((ConstDataProps){ .name = NAME, .size = SIZE, .offset = OFFSET})
+
+List_declare(ConstDataProps);
+HashMap_declare(ConstDataProps);
+
+
+typedef enum NamedRefType {
+    NamedRefType_Unset,
+    NamedRefType_Ptr,
+    NamedRefType_Size,
+} NamedRefType;
+
+typedef struct NamedRef {
+    str name;
+    NamedRefType type;
+    u32 offset; // offset in bytes from section start
+} NamedRef;
+
+#define NamedRef_construct(NAME, TYPE, OFFSET) ((NamedRef){ .name = NAME, .type = TYPE, .offset = OFFSET})
+
+List_declare(NamedRef);
+
+
+typedef struct CompiledSection {
+    str name;
+    CompiledSection* next;
+    u32 offset;
+    List_ConstDataProps const_data_props_list;
+    List_NamedRef named_refs;
+    List_u8 bytes;
+} CompiledSection;
+
+void CompiledSection_construct(CompiledSection* ptr, str name);
+void CompiledSection_free(CompiledSection* ptr);
+
+List_declare(CompiledSection);
+typedef CompiledSection* CompiledSectionPtr;
+HashMap_declare(CompiledSectionPtr);
+
+
+typedef struct BinaryObject {
+    List_CompiledSection section_list;
+    HashMap_CompiledSectionPtr section_map;
+    HashMap_ConstDataProps const_data_map;
+    u32 total_size;
+} BinaryObject;
+
+void BinaryObject_construct(BinaryObject* ptr);
+void BinaryObject_free(BinaryObject* ptr);

src/compiler/Compiler.c

@@ -0,0 +1,365 @@
+#include "Compiler_internal.h"
+
+HashMap_define(SectionPtr, HashMap_DESTROY_VALUE_FUNC_NULL);
+
+void Compiler_init(Compiler* cmp){
+    memset(cmp, 0, sizeof(Compiler));
+    cmp->state = CompilerState_Initial;
+    cmp->tokens = List_Token_alloc(4096);
+    cmp->line_lengths = List_u32_alloc(1024);
+    AST_init(&cmp->ast);
+    BinaryObject_construct(&cmp->binary);
+}
+
+void Compiler_free(Compiler* cmp){
+    free(cmp->code.data);
+    free(cmp->tokens.data);
+    free(cmp->line_lengths.data);
+    AST_free(&cmp->ast);
+    BinaryObject_free(&cmp->binary);
+}
+
+CodePos Compiler_getLineAndColumn(Compiler* cmp, u32 pos){
+    u32 prev_lines_len = 0;
+    if(pos >= cmp->code.len)
+        return CodePos_create(0, 0);
+    
+    for(u32 i = 0; i < cmp->line_lengths.len; i++){
+        u32 line_len = cmp->line_lengths.data[i];
+        if(prev_lines_len + line_len > pos)
+            return CodePos_create(i + 1, pos + 1 - prev_lines_len);
+        prev_lines_len += line_len;
+    }
+
+    return CodePos_create(0, 0);
+}
+
+void _Compiler_setError(Compiler* cmp, cstr context, cstr format, ...){
+    // happens at the end of file
+    if(cmp->pos >= cmp->code.len)
+        cmp->pos = cmp->code.len - 1;
+    char position_str[32];
+    CodePos code_pos = Compiler_getLineAndColumn(cmp, cmp->pos);
+    sprintf(position_str, "[at %u:%u][", code_pos.line, code_pos.column);
+    char* real_format = strcat_malloc(position_str, context, "] ", format);
+    va_list argv;
+    va_start(argv, format);
+    char* NULLABLE(buf) = vsprintf_malloc(512, real_format, argv);
+    va_end(argv);
+    free(real_format);
+    if(buf == NULL){
+        buf = malloc(16);
+        strcpy(buf, "SPRINTF FAILED");
+    }
+    cmp->state = CompilerState_Error;
+    cmp->error_message = buf;
+}
+
+#define setError(FORMAT, ...) {\
+    Compiler_setError(cmp, FORMAT, ##__VA_ARGS__);\
+}
+
+str Compiler_constructTokenStr(Compiler* cmp, Token t){
+    str s = str_construct((char*)(cmp->code.data + t.begin), t.length, false);
+    return s;
+}
+
+static bool compileSection(Compiler* cmp, Section* sec){
+    CompiledSection* cs = List_CompiledSection_expand(&cmp->binary.section_list, 1);
+    CompiledSection_construct(cs, sec->name);
+    if(!HashMap_CompiledSectionPtr_tryPush(&cmp->binary.section_map, cs->name, cs)){
+        returnError("duplicate section '%s'", str_copy(sec->name));
+    }
+
+    // compile code
+    u8 zeroes[8] = {0, 0, 0, 0, 0, 0, 0, 0}; 
+    for(u32 i = 0; i < sec->code.len; i++){
+        Operation* op = &sec->code.data[i];
+        List_u8_pushMany(&cs->bytes, (void*)&op->opcode, sizeof(op->opcode));
+        for(u32 j = 0; j < op->args.len; j++){
+            Argument* arg = &op->args.data[j];
+            switch(arg->type){
+                case ArgumentType_VarDataName:
+                    returnError("argument type 'VarDataName' is not supported yet");
+                case ArgumentType_Unset:
+                    returnError("ArgumentType is not set");
+                default:
+                    returnError("invalid ArgumentType %i", arg->type);
+                
+                case ArgumentType_Register:
+                    List_u8_push(&cs->bytes, arg->value.register_code);
+                    break;
+                case ArgumentType_ConstValue:
+                    //TODO: add const value size parsing
+                    List_u8_pushMany(&cs->bytes, (void*)&arg->value.i, 4);
+                    break;
+                case ArgumentType_ConstDataPointer:
+                    List_NamedRef_push(&cs->named_refs, NamedRef_construct(
+                            arg->value.data_name, 
+                            NamedRefType_Ptr, 
+                            cs->bytes.len));
+                    List_u8_pushMany(&cs->bytes, zeroes, 4);
+                    break;
+                case ArgumentType_ConstDataSize:
+                    List_NamedRef_push(&cs->named_refs, NamedRef_construct(
+                            arg->value.data_name, 
+                            NamedRefType_Size, 
+                            cs->bytes.len));
+                    List_u8_pushMany(&cs->bytes, zeroes, 4);
+                    break;
+            }
+        }
+    }
+
+    // compile data
+    for(u32 i = 0; i < sec->data.len; i++){
+        DataDefinition* dd = &sec->data.data[i];
+        List_ConstDataProps_push(&cs->const_data_props_list, ConstDataProps_construct(dd->name, dd->data.len, cs->bytes.len));
+        List_u8_pushMany(&cs->bytes, dd->data.data, dd->data.len);
+    }
+
+    // TODO: push padding
+
+    return true;
+}
+
+static bool compileBinary(Compiler* cmp){
+    for(u32 i = 0; i < cmp->ast.sections.len; i++){
+        SectionPtr sec = &cmp->ast.sections.data[i];
+        if(!compileSection(cmp, sec)){
+            return false;
+        }
+    }
+
+    // find main section
+    str main_sec_name = STR("main");
+    CompiledSection** main_sec_ptrptr = HashMap_CompiledSectionPtr_tryGetPtr(&cmp->binary.section_map, main_sec_name);
+    if(main_sec_ptrptr == NULL){
+        returnError("no 'main' section was defined");
+    }
+
+    // create linked list of CompiledSection where main is the first    
+    CompiledSection* prev_sec = *main_sec_ptrptr;
+    u32 total_size = 0;
+    for(u32 i = 0; i < cmp->binary.section_list.len; i++){
+        CompiledSection* sec = &cmp->binary.section_list.data[i];
+        total_size += sec->bytes.len;
+        if(str_equals(sec->name, main_sec_name))
+           continue;
+        prev_sec->next = sec;
+        sec->offset = prev_sec->offset + prev_sec->bytes.len;
+
+        ConstDataProps cd = ConstDataProps_construct(sec->name, sec->bytes.len, sec->offset);
+        if(!HashMap_ConstDataProps_tryPush(&cmp->binary.const_data_map, cd.name, cd)){
+            returnError("duplicate named data '%s'", str_copy(cd.name).data);
+        }
+        for(u32 j = 0; j < sec->const_data_props_list.len; j++){
+            cd = sec->const_data_props_list.data[j];
+            cd.offset += sec->offset;
+            if(!HashMap_ConstDataProps_tryPush(&cmp->binary.const_data_map, cd.name, cd)){
+                returnError("duplicate named data '%s'", str_copy(cd.name).data);
+            }
+        }
+    }
+
+    // insert calculated offsets into sections
+    for(u32 i = 0; i < cmp->binary.section_list.len; i++){
+        CompiledSection* sec = &cmp->binary.section_list.data[i];
+        for(u32 j = 0; j < sec->named_refs.len; j++){
+            NamedRef* ref = &sec->named_refs.data[j];
+            
+            ConstDataProps* target_data = HashMap_ConstDataProps_tryGetPtr(
+                        &cmp->binary.const_data_map, ref->name);
+            if(target_data == NULL){
+                returnError("can't find named data '%s'", str_copy(ref->name).data);
+            }
+
+            void* ref_value_ptr = sec->bytes.data + ref->offset;
+            
+            switch(ref->type){
+                default:
+                    returnError("invalid NamedRefType %i", ref->type);
+                case NamedRefType_Size:
+                    *((u32*)ref_value_ptr) = target_data->size;
+                    break;
+                case NamedRefType_Ptr:
+                    *((u32*)ref_value_ptr) = target_data->offset;
+                    break;
+            }
+        } 
+    }
+
+    cmp->binary.total_size = total_size;
+    return true;
+}
+
+static bool writeBinaryFile(Compiler* cmp, FILE* f){
+    returnErrorIf_auto(cmp->state != CompilerState_Parsing);
+    cmp->state = CompilerState_Compiling;
+
+    if(!compileBinary(cmp)){
+        return false;
+    }    
+
+    
+    CompiledSection** main_sec_ptrptr = HashMap_CompiledSectionPtr_tryGetPtr(&cmp->binary.section_map, STR("main"));
+    if(main_sec_ptrptr == NULL){
+        returnError("no 'main' section was defined");
+    }
+    CompiledSection* sec = *main_sec_ptrptr;
+    while(sec){
+        fwrite(sec->bytes.data, 1, sec->bytes.len, f);
+        sec = sec->next;
+    }
+
+    //TODO: print warnings for unused sections
+
+    return true;
+}
+
+bool Compiler_compile(Compiler* cmp, cstr source_file_name, cstr out_file_name, bool debug_log){
+    FILE* f = fopen(source_file_name, "rb");
+    if(f == NULL)
+        returnError("ERROR: can't open file '%s'", source_file_name);
+    
+    StringBuilder sb = StringBuilder_alloc(64 * 1024);
+    int ret;
+    while((ret = fgetc(f)) != EOF) {
+        StringBuilder_append_char(&sb, ret);
+    }
+    if(ferror(f)){
+        StringBuilder_free(&sb);
+        fclose(f);
+        returnError("can't read file '%s'", source_file_name);
+    }
+    fclose(f);
+
+    if(sb.buffer.len == 0){
+        StringBuilder_free(&sb);
+        returnError("soucre file is empty");
+    }
+
+    cmp->code = str_copy(StringBuilder_getStr(&sb));
+    StringBuilder_free(&sb);
+
+    f = fopen(out_file_name, "wb");
+    if(f == NULL){
+        returnError("ERROR: can't open file '%s'", out_file_name);
+    }
+
+    if(debug_log){
+        printf("===========================[%s]===========================\n", source_file_name);
+        fputs(cmp->code.data, stdout);
+        fputc('\n', stdout);
+    }
+    
+    if(debug_log)
+        printf("===================================[lexing]===================================\n");
+    bool success = Compiler_lex(cmp);
+
+    if(debug_log){
+        printf("------------------------------------[lines]------------------------------------\n");
+        for(u32 i = 0; i < cmp->line_lengths.len; i++){
+            printf("[%u] length: %u\n", i+1, cmp->line_lengths.data[i]);
+        }
+
+        printf("------------------------------------[tokens]-----------------------------------\n");
+        for(u32 i = 0; i < cmp->tokens.len; i++){
+            Token t = cmp->tokens.data[i];
+            CodePos pos = Compiler_getLineAndColumn(cmp, t.begin);
+            char* tokstr = malloc(4096);
+            strncpy(tokstr, cmp->code.data + t.begin, t.length);
+            tokstr[t.length] = 0;
+            char* tokstr_stripped = tokstr;
+            while(*tokstr_stripped == '\r' || *tokstr_stripped == '\n'){
+                tokstr_stripped++;
+            }
+            printf("[l:%3u, c:%3u] %s '%s'\n", 
+                pos.line, pos.column, 
+                TokenType_toString(t.type).data, tokstr_stripped);
+            free(tokstr);
+        }
+    }
+
+    if(!success){
+        fclose(f);
+        return false;
+    }
+
+    if(debug_log)
+        printf("===================================[parsing]===================================\n");
+    success = Compiler_parse(cmp);
+    if (debug_log){
+        printf("-------------------------------------[AST]-------------------------------------\n");
+        for(u32 i = 0; i < cmp->ast.sections.len; i++){
+            Section* sec = &cmp->ast.sections.data[i];
+            str tmpstr = str_copy(sec->name);
+            printf("section '%s'\n", tmpstr.data);
+            free(tmpstr.data);
+
+            for(u32 j = 0; j < sec->data.len; j++){
+                DataDefinition* dd = &sec->data.data[j];
+                tmpstr = str_copy(dd->name);
+                printf("    const%u %s (len %u)\n", dd->element_size * 8, tmpstr.data, dd->data.len/dd->element_size);
+                free(tmpstr.data);
+            }
+
+            
+            for(u32 j = 0; j < sec->code.len; j++){
+                Operation* op = &sec->code.data[j];
+                const Instruction* instr = Instruction_getByOpcode(op->opcode);
+                printf("    %s", instr->name.data);
+                for(u32 k = 0; k < op->args.len; k++){
+                    Argument* arg = &op->args.data[k];
+                    printf(" %s(", ArgumentType_toString(arg->type).data);
+                    
+                    switch(arg->type){
+                        default:
+                            fclose(f);
+                            returnError("invalid argument type %i", arg->type);
+                        case ArgumentType_Register:
+                            const char* register_names[] = {"null", "ax", "bx", "cx", "dx"};
+                            printf("%s", register_names[arg->value.register_code]);
+                            break;
+                        case ArgumentType_ConstValue:
+                            printf(IFWIN("%lli", "%li"), arg->value.i);
+                            break;
+                        case ArgumentType_ConstDataPointer:
+                            tmpstr = str_copy(arg->value.data_name);
+                            printf("@%s", tmpstr.data);
+                            free(tmpstr.data);
+                            break;
+                        case ArgumentType_ConstDataSize:
+                            tmpstr = str_copy(arg->value.data_name);
+                            printf("#%s", tmpstr.data);
+                            free(tmpstr.data);
+                            break;
+                        case ArgumentType_VarDataName:
+                            tmpstr = str_copy(arg->value.data_name);
+                            printf("%s", tmpstr.data);
+                            free(tmpstr.data);
+                            break;
+                    }
+
+                    printf(")");
+                }
+
+                printf("\n");
+            }
+        }
+    }
+    if(!success){
+        fclose(f);
+        return false;
+    }
+
+    if(debug_log)
+        printf("==================================[compiling]==================================\n");
+    success = writeBinaryFile(cmp, f);
+    fclose(f);
+    if(success){
+        cmp->state = CompilerState_Success;
+    }
+    
+    return success;
+}

src/compiler/Compiler.h

@@ -1,9 +1,10 @@
 #pragma once
 #include "../std.h"
+#include "../string/str.h"
 #include "../collections/List.h"
-#include "token.h"
-
-List_declare(Token);
+#include "../collections/HashMap.h"
+#include "Token.h"
+#include "Binary.h"
 
 typedef enum CompilerState {
     CompilerState_Initial,
@@ -14,15 +15,24 @@ typedef enum CompilerState {
     CompilerState_Success
 } CompilerState;
 
+typedef Section* SectionPtr;
+HashMap_declare(SectionPtr);
+
 typedef struct Compiler {
-    char* code;
-    u32 code_len;
+    /* general fields */
+    str code;
     u32 column;     // > 0 if code parsing started
     u32 pos;
     CompilerState state;
     NULLABLE(char* error_message);
+    /* lexer fields */
     List_Token tokens;
     List_u32 line_lengths;
+    /* parser fields */
+    AST ast;
+    u32 tok_i;
+    /* compiler fields */
+    BinaryObject binary;
 } Compiler;
 
 void Compiler_init(Compiler* cmp);
@@ -30,18 +40,4 @@ void Compiler_free(Compiler* cmp);
 
 /// @brief compile assembly language code to machine code
 /// @return true if no errors, false if any error occured (check cmp->error_message)
-bool Compiler_compileTasm(Compiler* cmp, cstr source_file_name, cstr out_file_name, bool debug);
-
-#define Compiler_setError(cmp, format, ...) _Compiler_setError(cmp, __func__, format ,##__VA_ARGS__)
-void _Compiler_setError(Compiler* cmp, cstr context, cstr format, ...) __attribute__((__format__(__printf__, 3, 4)));
-
-typedef struct CodePos {
-    u32 line; // 0 on error
-    u32 column; // 0 on error
-} CodePos;
-
-#define CodePos_create(L, C) ((CodePos){ .line = L, .column = C })
-
-/// @param pos index in code buffer
-CodePos Compiler_getLineAndColumn(Compiler* cmp, u32 pos);
-
+bool Compiler_compile(Compiler* cmp, cstr source_file_name, cstr out_file_name, bool debug);

src/compiler/Compiler_internal.h

@@ -0,0 +1,30 @@
+#include "Compiler.h"
+#include "../string/StringBuilder.h"
+
+void _Compiler_setError(Compiler* cmp, cstr context, cstr format, ...) __attribute__((__format__(__printf__, 3, 4)));
+
+#define Compiler_setError(cmp, format, ...) _Compiler_setError(cmp, __func__, format ,##__VA_ARGS__)
+
+#define returnError(FORMAT, ...) {\
+    setError(FORMAT, ##__VA_ARGS__);\
+    return false;\
+}
+
+#define returnErrorIf(STATEMENT, FORMAT, ...) if(STATEMENT) returnError(FORMAT, ##__VA_ARGS__)
+
+#define returnErrorIf_auto(STATEMENT) returnErrorIf(STATEMENT, #STATEMENT)
+
+typedef struct CodePos {
+    u32 line; // 0 on error
+    u32 column; // 0 on error
+} CodePos;
+
+#define CodePos_create(L, C) ((CodePos){ .line = L, .column = C })
+
+/// @param pos index in code buffer
+CodePos Compiler_getLineAndColumn(Compiler* cmp, u32 pos);
+
+str Compiler_constructTokenStr(Compiler* cmp, Token t);
+
+bool Compiler_lex(Compiler* cmp);
+bool Compiler_parse(Compiler* cmp);

src/compiler/Lexer.c

@@ -0,0 +1,269 @@
+#include "Compiler_internal.h"
+
+#define setError(FORMAT, ...) {\
+    completeLine(cmp);\
+    Compiler_setError(cmp, FORMAT, ##__VA_ARGS__);\
+}
+
+#define Error_unexpectedCharacter(C) "unexpected character '%c'", C
+#define Error_endOfFile "unexpected end of file"
+
+static void completeLine(Compiler* cmp){
+    List_u32_push(&cmp->line_lengths, cmp->column);
+    cmp->column = 0;
+}
+
+static void readCommentSingleLine(Compiler* cmp){
+    char c; // '/'
+    Token tok = Token_construct(TokenType_SingleLineComment, cmp->pos - 1, 0);
+    cmp->column++;
+    cmp->pos++;
+    
+    while(cmp->pos < cmp->code.len){
+        c = cmp->code.data[cmp->pos];
+        // end of line
+        if(c == '\r' || c == '\n'){
+            tok.length = cmp->pos - tok.begin;
+            List_Token_push(&cmp->tokens, tok);
+            // cmp->line will be increased in lex()
+            return;
+        }
+        
+        cmp->column++;
+        cmp->pos++;
+    }
+    
+    // end of file
+    tok.length = cmp->pos - tok.begin;
+    List_Token_push(&cmp->tokens, tok);
+}
+
+static void readCommentMultiLine(Compiler* cmp){
+    char c; // '*'
+    Token tok = Token_construct(TokenType_MultiLineComment, cmp->pos - 1, 0);
+    cmp->column++;
+    cmp->pos++;
+    
+    while(cmp->pos < cmp->code.len){
+        c = cmp->code.data[cmp->pos];
+        // closing comment
+        if(cmp->pos > tok.begin + 3 && c == '/' && cmp->code.data[cmp->pos - 1] == '*') {
+            tok.length = cmp->pos - tok.begin + 1;
+            List_Token_push(&cmp->tokens, tok);
+            return;
+        }
+
+        if(c == '\n')
+            completeLine(cmp);
+        cmp->column++;
+        cmp->pos++;
+    }
+    
+    // end of file
+    setError(Error_endOfFile);
+}
+
+static void readComment(Compiler* cmp){
+    char c; // '/'
+    if(cmp->pos + 1 == cmp->code.len){
+        setError(Error_endOfFile);
+        return;
+    }
+
+    c = cmp->code.data[cmp->pos + 1];
+    if(c == '\r' || c == '\n'){
+        setError(Error_unexpectedCharacter(cmp->code.data[--cmp->pos]));
+        return;
+    }
+
+    cmp->pos++;
+    cmp->column++;
+    if(c == '/')
+        readCommentSingleLine(cmp);
+    else if(c == '*')
+        readCommentMultiLine(cmp);
+    else setError(Error_unexpectedCharacter(c));
+}
+
+static void readLabel(Compiler* cmp){
+    char c; // '.'
+    cmp->pos++;
+    cmp->column++;
+    Token tok = Token_construct(TokenType_Label, cmp->pos, 0);
+    
+    while(cmp->pos < cmp->code.len){
+        c = cmp->code.data[cmp->pos];
+        // end of line
+        if(c == ':' || c == '\r' || c == '\n'){
+            tok.length = cmp->pos - tok.begin;
+            if(tok.length > 0)
+                List_Token_push(&cmp->tokens, tok);
+            else setError(Error_unexpectedCharacter(cmp->code.data[--cmp->pos]));
+            // cmp->line will be increased in lex()
+            return;
+        }
+
+        if(!isAlphabeticalLower(c) && !isAlphabeticalUpper(c) && !isDigit(c) &&
+            c != '_' && c != '.'){
+            setError(Error_unexpectedCharacter(c));
+            return;
+        }
+        
+        cmp->column++;
+        cmp->pos++;
+    }
+    
+    // end of file
+    tok.length = cmp->pos - tok.begin;
+    if(tok.length > 0)
+        List_Token_push(&cmp->tokens, tok);
+    else setError(Error_endOfFile);
+}
+
+static void readArguments(Compiler* cmp){
+    char c; // space
+    Token tok = Token_construct(TokenType_Unset, cmp->pos, 0);
+    char quot = '\0'; // quotation character of a string value
+
+    while(cmp->pos < cmp->code.len){
+        c = cmp->code.data[cmp->pos];
+
+        // string argument reading
+        if(quot != '\0'){
+            if(c == quot && cmp->code.data[cmp->pos - 1] != '\\'){
+                quot = '\0';
+            }
+            else if(c == '\r' || c == '\n'){
+                setError("line end reached but string hasn't been closed yet");
+                return;
+            }
+        }
+
+        // end of operation
+        else if(c == '\r' || c == '\n' || c == ';'){
+            tok.length = cmp->pos - tok.begin;
+            if(tok.length > 0)
+                List_Token_push(&cmp->tokens, tok);
+            // cmp->line will be increased in lex()
+            return;
+        }
+
+        // new argument begins
+        else if(c == ' ' || c == '\t'){
+            tok.length = cmp->pos - tok.begin;
+            if(tok.length > 0)
+                List_Token_push(&cmp->tokens, tok);
+            tok = Token_construct(TokenType_Unset, cmp->pos + 1, 0);
+        }
+
+        else if(tok.type == TokenType_Unset){
+            if(c == '\''){
+                tok.type = TokenType_Char;
+                quot = c;
+            }
+            else if(c == '"'){
+                tok.type = TokenType_String;
+                quot = c;
+            }
+            else if(c == '@')
+                tok.type = TokenType_NamedDataPointer;
+            else if(c == '#')
+                tok.type = TokenType_NamedDataSize;
+            else if(isDigit(c))
+                tok.type = TokenType_Number;
+            else tok.type = TokenType_Name;
+        }
+
+        cmp->column++;
+        cmp->pos++;
+    }
+    
+    // end of file
+    tok.length = cmp->pos - tok.begin;
+    if(tok.length > 0)
+        List_Token_push(&cmp->tokens, tok);
+}
+
+static void readInstruction(Compiler* cmp){
+    Token tok = Token_construct(TokenType_Instruction, cmp->pos, 0);
+    cmp->pos++;
+    cmp->column++;
+    
+    while(cmp->pos < cmp->code.len){
+        char c = cmp->code.data[cmp->pos];
+        // end of line
+        if(c == '\r' || c == '\n' || c == ';'){
+            tok.length = cmp->pos - tok.begin;
+            List_Token_push(&cmp->tokens, tok);
+            tok = Token_construct(TokenType_OperationEnd, cmp->pos, 1);
+            List_Token_push(&cmp->tokens, tok);
+            // cmp->line will be increased in lex()
+            return;
+        }
+
+        // arguments begin
+        if(c == ' ' || c == '\t'){
+            tok.length = cmp->pos - tok.begin;
+            List_Token_push(&cmp->tokens, tok);
+            readArguments(cmp);
+            tok = Token_construct(TokenType_OperationEnd, cmp->pos, 1);
+            List_Token_push(&cmp->tokens, tok);
+            return;
+        }
+
+        if(!isAlphabeticalLower(c) && !isAlphabeticalUpper(c) && !isDigit(c)){
+            setError(Error_unexpectedCharacter(c));
+            return;
+        }
+
+        cmp->column++;
+        cmp->pos++;
+    }
+    
+    // end of file
+    tok.length = cmp->pos - tok.begin;
+    List_Token_push(&cmp->tokens, tok);
+    tok = Token_construct(TokenType_OperationEnd, cmp->pos, 1);
+    List_Token_push(&cmp->tokens, tok);
+}
+
+bool Compiler_lex(Compiler* cmp){
+    returnErrorIf_auto(cmp->state != CompilerState_Initial);
+    cmp->state = CompilerState_Lexing;
+    cmp->column = 1;
+
+    while(cmp->pos < cmp->code.len){
+        char c = cmp->code.data[cmp->pos];
+        switch(c){
+            // skip blank characters
+            case ' ': case '\t': case '\r': case '\n':
+                break;
+            // try read comment
+            case '/':
+                readComment(cmp);
+                break;
+            // try read label
+            case '.':
+                readLabel(cmp);
+                break;
+            default:
+                // try read instruction
+                if(isAlphabeticalLower(c) || isAlphabeticalUpper(c))
+                    readInstruction(cmp);
+                else returnError(Error_unexpectedCharacter(c));
+                break;
+        }
+
+        if(cmp->state == CompilerState_Error)
+            return false;
+        
+        c = cmp->code.data[cmp->pos];
+        if(c == '\n')
+            completeLine(cmp);
+        cmp->column++;
+        cmp->pos++;
+    }
+
+    completeLine(cmp);
+    return true;
+}

src/compiler/Parser.c

@@ -0,0 +1,281 @@
+#include "Compiler_internal.h"
+
+#define setError(FORMAT, ...) {\
+    cmp->pos = cmp->tokens.data[cmp->tok_i].begin;\
+    Compiler_setError(cmp, FORMAT, ##__VA_ARGS__);\
+}
+
+#define setError_unexpectedToken(T) {\
+    str tok_str = str_copy(Compiler_constructTokenStr(cmp, T));\
+    cmp->pos = T.begin;\
+    Compiler_setError(cmp, "unexpected token '%s'", tok_str.data);\
+    free(tok_str.data);\
+}
+
+#define setError_unexpectedTokenChar(T, I) {\
+    cmp->pos = T.begin + I;\
+    Compiler_setError(cmp, "unexpected token '%c'", cmp->code.data[cmp->pos]);\
+}
+
+#define setError_unexpectedInstruction(T) {\
+    str tok_str = str_copy(Compiler_constructTokenStr(cmp, T));\
+    cmp->pos = T.begin;\
+    Compiler_setError(cmp, "unexpected instruction '%s'", tok_str.data);\
+    free(tok_str.data);\
+}
+
+#define Error_TokenUnset "token of undefined type"
+#define Error_BitSize "invalid size in bits"
+
+static void List_u8_pushBytes(List_u8* l, void* value, u32 startIndex, u32 count){
+    u8* v = value;
+    for(u32 byte_i = startIndex; byte_i < startIndex + count; byte_i++){
+        List_u8_push(l, v[byte_i]);
+    }
+}
+
+static inline bool isVarSizeBits(u32 B) { return (B == 8 || B == 16 || B == 32 || B == 64); }
+
+static NULLABLE(str) resolveEscapeSequences(Compiler* cmp, str src){
+    StringBuilder sb = StringBuilder_alloc(src.len);
+    char c;
+    bool escaped = false;
+    for(u32 i = 0; i < src.len; i++){
+        c = src.data[i];
+        if(c == '\\'){
+            escaped = !escaped;
+            continue;
+        }
+        
+        if(!escaped){
+            StringBuilder_append_char(&sb, c);
+            continue;
+        }
+
+        // escape codes
+        switch(c){
+            case '0':
+                StringBuilder_append_char(&sb, '\0');
+                break;
+            case 'n':
+                StringBuilder_append_char(&sb, '\n');
+                break;
+            case 'r':
+                StringBuilder_append_char(&sb, '\r');
+                break;
+            case 't':
+                StringBuilder_append_char(&sb, '\t');
+                break;
+            case 'e':
+                StringBuilder_append_char(&sb, '\e');
+                break;
+            default:
+                setError_unexpectedTokenChar(cmp->tokens.data[cmp->tok_i], i);
+                StringBuilder_free(&sb);
+                return str_null;
+        }
+    }
+
+    return StringBuilder_getStr(&sb);
+}
+
+static void parseDataDefinition(Compiler* cmp, str instr_name, DataDefinition* ddf){
+    i32 _element_size_bits;
+    str _instr_name_zero_terminated = str_copy(instr_name);
+    if(sscanf(_instr_name_zero_terminated.data, "const%i", &_element_size_bits) != 1 || !isVarSizeBits(_element_size_bits)){
+        free(_instr_name_zero_terminated.data);
+        setError(Error_BitSize);
+        return;
+    }
+    free(_instr_name_zero_terminated.data);
+    ddf->element_size = _element_size_bits / 8;
+    ddf->data = List_u8_alloc(32);
+
+    Token tok = cmp->tokens.data[++cmp->tok_i];
+    if(tok.type != TokenType_Name){
+        setError_unexpectedToken(tok);
+        return;
+    }
+
+    str tok_str = Compiler_constructTokenStr(cmp, tok);
+    str processed_str = str_null;
+    ddf->name = tok_str;
+    
+    while(++cmp->tok_i < cmp->tokens.len){
+        tok = cmp->tokens.data[cmp->tok_i];
+        switch(tok.type){
+            case TokenType_SingleLineComment:
+            case TokenType_MultiLineComment:
+                // skip comments
+                break;
+            
+            case TokenType_OperationEnd:
+                return;
+            case TokenType_Unset:
+                setError(Error_TokenUnset);
+                return;
+            default:
+                setError_unexpectedToken(tok);
+                return;
+            
+            case TokenType_Number:
+                tok_str = Compiler_constructTokenStr(cmp, tok);
+                processed_str = str_copy(tok_str);
+                if(str_seekChar(tok_str, '.', 0) != -1){
+                    f64 f = atof(processed_str.data);
+                    List_u8_pushBytes(&ddf->data, &f, 8 - ddf->element_size, ddf->element_size);
+                }
+                else {
+                    i64 i = atoll(processed_str.data);
+                    List_u8_pushBytes(&ddf->data, &i, 8 - ddf->element_size, ddf->element_size);
+                }
+                free(processed_str.data);
+                break;
+            case TokenType_Char:
+                tok.begin += 1;
+                tok.length -= 2;
+                tok_str = Compiler_constructTokenStr(cmp, tok);
+                processed_str = resolveEscapeSequences(cmp, tok_str);
+                
+                if(processed_str.len != ddf->element_size){
+                    setError("can't fit char of size %i in %u bit variable", processed_str.len, _element_size_bits);
+                    return;
+                }
+                List_u8_pushBytes(&ddf->data, processed_str.data, 0, processed_str.len);
+                free(processed_str.data);
+                break;
+            case TokenType_String:
+                tok.begin += 1;
+                tok.length -= 2;
+                tok_str = Compiler_constructTokenStr(cmp, tok);
+                processed_str = resolveEscapeSequences(cmp, tok_str);
+                List_u8_pushBytes(&ddf->data, processed_str.data, 0, processed_str.len);
+                free(processed_str.data);
+                break;
+        }
+    }
+}
+
+
+static void parseOperation(Compiler* cmp, str instr_name, Operation* operPtr){
+    Token tok = cmp->tokens.data[cmp->tok_i];
+    const Instruction* instr = Instruction_getByName(instr_name);
+    if(instr == NULL){
+        setError_unexpectedInstruction(tok);
+        return;
+    }
+
+    operPtr->opcode = instr->opcode;
+    operPtr->args = List_Argument_alloc(8);
+    Argument arg = (Argument){ .type = ArgumentType_Unset, .value.i = 0 };
+    str tok_str = str_null;
+    str processed_str = str_null;
+    while(++cmp->tok_i < cmp->tokens.len){
+        tok = cmp->tokens.data[cmp->tok_i];
+        switch(tok.type){
+            case TokenType_SingleLineComment:
+            case TokenType_MultiLineComment:
+                // skip comments
+                break;
+            
+            case TokenType_OperationEnd:
+                return;
+            case TokenType_Unset:
+                setError(Error_TokenUnset);
+                return;
+            default:
+                setError_unexpectedToken(tok);
+                return;
+            
+            case TokenType_Number:
+                arg.type = ArgumentType_ConstValue;
+                tok_str = Compiler_constructTokenStr(cmp, tok);
+                processed_str = str_copy(tok_str);
+                if(str_seekChar(tok_str, '.', 0) != -1){
+                    arg.value.f = atof(processed_str.data);
+                }
+                else {
+                    arg.value.i = atoll(processed_str.data);
+                }
+                free(processed_str.data);
+                List_Argument_push(&operPtr->args, arg);
+                break;
+            case TokenType_Name:
+                tok_str = Compiler_constructTokenStr(cmp, tok);
+                arg.value.register_code = RegisterCode_parse(tok_str);
+                if(arg.value.register_code != RegisterCode_Unset){
+                    arg.type = ArgumentType_Register;
+                }
+                else {
+                    arg.type = ArgumentType_VarDataName;
+                    arg.value.data_name = tok_str;
+                }
+                List_Argument_push(&operPtr->args, arg);
+                break;
+            case TokenType_NamedDataPointer:
+                tok_str = Compiler_constructTokenStr(cmp, tok);
+                tok_str.data++;
+                tok_str.len--;
+                arg.type = ArgumentType_ConstDataPointer;
+                arg.value.data_name = tok_str;
+                List_Argument_push(&operPtr->args, arg);
+                break;
+            case TokenType_NamedDataSize:
+                tok_str = Compiler_constructTokenStr(cmp, tok);
+                tok_str.data++;
+                tok_str.len--;
+                arg.type = ArgumentType_ConstDataSize;
+                arg.value.data_name = tok_str;
+                List_Argument_push(&operPtr->args, arg);
+                break;
+        }
+    }
+}
+
+bool Compiler_parse(Compiler* cmp){
+    returnErrorIf_auto(cmp->state != CompilerState_Lexing);
+    cmp->state = CompilerState_Parsing;
+    Token tok;
+    Section* sec = NULL;
+    
+    while(cmp->tok_i < cmp->tokens.len){
+        tok = cmp->tokens.data[cmp->tok_i];
+        switch(tok.type){
+            case TokenType_Unset:
+                returnError(Error_TokenUnset);
+            case TokenType_SingleLineComment:
+            case TokenType_MultiLineComment:
+                // skip comments
+                break;
+            case TokenType_Label:
+                // create new section
+                sec = List_Section_expand(&cmp->ast.sections, 1);
+                Section_init(sec, Compiler_constructTokenStr(cmp, tok));
+                break;
+            case TokenType_Instruction:
+                if(sec == NULL)
+                    returnError("no section");
+                str instr_name = Compiler_constructTokenStr(cmp, tok);
+                // data definition starts with const
+                if(str_startsWith(instr_name, STR("const"))){
+                    DataDefinition* dataDefPtr = List_DataDefinition_expand(&sec->data, 1);
+                    parseDataDefinition(cmp, instr_name, dataDefPtr);
+                }
+                else {
+                    Operation* operPtr = List_Operation_expand(&sec->code, 1);
+                    parseOperation(cmp, instr_name, operPtr);
+                }
+                break;
+            default:
+                setError_unexpectedToken(tok);
+                return false;
+        }
+
+        if(cmp->state == CompilerState_Error)
+            return false;
+
+        cmp->tok_i++;
+    }
+
+    return true;
+}

src/compiler/Token.c

@@ -0,0 +1,24 @@
+#include "Token.h"
+
+List_define(Token);
+
+static str _TokenType_str[] = {
+    STR("Unset"),
+    STR("SingleLineComment"),
+    STR("MultiLineComment"),
+    STR("Instruction"),
+    STR("Label"),
+    STR("Number"),
+    STR("Char"),
+    STR("String"),
+    STR("Name"),
+    STR("NamedDataPointer"),
+    STR("NamedDataSize"),
+    STR("OperationEnd"),
+};
+
+str TokenType_toString(TokenType t){
+    if(t >= ARRAY_SIZE(_TokenType_str))
+        return STR("!!TokenType INDEX_ERROR!!");
+    return _TokenType_str[t];
+}

src/compiler/Token.h

@@ -0,0 +1,31 @@
+#pragma once
+#include "../std.h"
+#include "../string/str.h"
+#include "../collections/List.h"
+
+typedef enum TokenType {
+    TokenType_Unset,                // initial value
+    TokenType_SingleLineComment,    // //comment
+    TokenType_MultiLineComment,     // /* comment */
+    TokenType_Instruction,          // abc
+    TokenType_Label,                // .abc:
+    TokenType_Number,               // 0123
+    TokenType_Char,                 // 'A'
+    TokenType_String,               // "aaaa"
+    TokenType_Name,                 // xyz
+    TokenType_NamedDataPointer,     // @xyz
+    TokenType_NamedDataSize,        // #xyz
+    TokenType_OperationEnd,         // EOL or EOF or ;
+} TokenType;
+
+str TokenType_toString(TokenType t);
+
+typedef struct Token {
+    u32 begin;          // some index in Compiler->code
+    u32 length : 24;    // length in characters (24 bits)
+    TokenType type : 8; // type of token (8 bits)
+} Token;
+
+List_declare(Token);
+
+#define Token_construct(TYPE, BEGIN, LEN) ((Token){ .type = TYPE, .begin = BEGIN, .length = LEN })

src/compiler/compiler.c

@@ -1,470 +0,0 @@
-#include "compiler.h"
-
-List_define(Token);
-
-CodePos Compiler_getLineAndColumn(Compiler* cmp, u32 pos){
-    u32 prev_lines_len = 0;
-    if(pos >= cmp->code_len)
-        return CodePos_create(0, 0);
-    
-    for(u32 i = 0; i < cmp->line_lengths.len; i++){
-        u32 line_len = cmp->line_lengths.data[i];
-        if(prev_lines_len + line_len > pos)
-            return CodePos_create(i + 1, pos + 1 - prev_lines_len);
-        prev_lines_len += line_len;
-    }
-
-    return CodePos_create(0, 0);
-}
-
-static void completeLine(Compiler* cmp){
-    List_u32_push(&cmp->line_lengths, cmp->column);
-    cmp->column = 0;
-}
-
-void _Compiler_setError(Compiler* cmp, cstr context, cstr format, ...){
-    completeLine(cmp);
-    // happens at the end of file
-    if(cmp->pos >= cmp->code_len)
-        cmp->pos = cmp->code_len - 1;
-    char position_str[32];
-    CodePos code_pos = Compiler_getLineAndColumn(cmp, cmp->pos);
-    sprintf(position_str, "[at %u:%u][", code_pos.line, code_pos.column);
-    char* real_format = strcat_malloc(position_str, context, "] ", format);
-    va_list argv;
-    va_start(argv, format);
-    char* NULLABLE(buf) = vsprintf_malloc(512, real_format, argv);
-    va_end(argv);
-    free(real_format);
-    if(buf == NULL){
-        buf = malloc(16);
-        strcpy(buf, "SPRINTF FAILED");
-    }
-    cmp->state = CompilerState_Error;
-    cmp->error_message = buf;
-}
-
-
-#define setError(FORMAT, ...) Compiler_setError(cmp, FORMAT, ##__VA_ARGS__)
-
-#define returnError(FORMAT, ...) {\
-    setError(FORMAT, ##__VA_ARGS__);\
-    return false;\
-}
-
-#define returnErrorIf(STATEMENT, FORMAT, ...) if(STATEMENT) returnError(FORMAT, ##__VA_ARGS__)
-
-#define returnErrorIf_auto(STATEMENT) returnErrorIf(STATEMENT, #STATEMENT)
-
-#define Error_unexpectedCharacter(C) "unexpected character '%c'", C
-#define Error_endOfFile "unexpected end of file"
-
-void Compiler_init(Compiler* cmp){
-    memset(cmp, 0, sizeof(Compiler));
-    cmp->state = CompilerState_Initial;
-    cmp->tokens = List_Token_alloc(4096);
-    cmp->line_lengths = List_u32_alloc(1024);
-}
-
-void Compiler_free(Compiler* cmp){
-    free(cmp->code);
-    free(cmp->tokens.data);
-    free(cmp->line_lengths.data);
-}
-
-static void readCommentSingleLine(Compiler* cmp){
-    char c; // '/'
-    Token tok = Token_construct(TokenType_SingleLineComment, cmp->pos - 1, 0);
-    cmp->column++;
-    cmp->pos++;
-    
-    while(cmp->pos < cmp->code_len){
-        c = cmp->code[cmp->pos];
-        // end of line
-        if(c == '\r' || c == '\n'){
-            tok.length = cmp->pos - tok.begin;
-            List_Token_push(&cmp->tokens, tok);
-            // cmp->line will be increased in lex()
-            return;
-        }
-        
-        cmp->column++;
-        cmp->pos++;
-    }
-    
-    // end of file
-    tok.length = cmp->pos - tok.begin;
-    List_Token_push(&cmp->tokens, tok);
-}
-
-static void readCommentMultiLine(Compiler* cmp){
-    char c; // '*'
-    Token tok = Token_construct(TokenType_MultiLineComment, cmp->pos - 1, 0);
-    cmp->column++;
-    cmp->pos++;
-    
-    while(cmp->pos < cmp->code_len){
-        c = cmp->code[cmp->pos];
-        // closing comment
-        if(cmp->pos > tok.begin + 3 && c == '/' && cmp->code[cmp->pos - 1] == '*') {
-            tok.length = cmp->pos - tok.begin + 1;
-            List_Token_push(&cmp->tokens, tok);
-            return;
-        }
-
-        if(c == '\n')
-            completeLine(cmp);
-        cmp->column++;
-        cmp->pos++;
-    }
-    
-    // end of file
-    setError(Error_endOfFile);
-}
-
-static void readComment(Compiler* cmp){
-    char c; // '/'
-    if(cmp->pos + 1 == cmp->code_len){
-        setError(Error_endOfFile);
-        return;
-    }
-
-    c = cmp->code[cmp->pos + 1];
-    if(c == '\r' || c == '\n'){
-        setError(Error_unexpectedCharacter(cmp->code[--cmp->pos]));
-        return;
-    }
-
-    cmp->pos++;
-    cmp->column++;
-    if(c == '/')
-        readCommentSingleLine(cmp);
-    else if(c == '*')
-        readCommentMultiLine(cmp);
-    else setError(Error_unexpectedCharacter(c));
-}
-
-static void readLabel(Compiler* cmp){
-    char c; // '.'
-    cmp->pos++;
-    cmp->column++;
-    Token tok = Token_construct(TokenType_Label, cmp->pos, 0);
-    
-    while(cmp->pos < cmp->code_len){
-        c = cmp->code[cmp->pos];
-        // end of line
-        if(c == ':' || c == '\r' || c == '\n'){
-            tok.length = cmp->pos - tok.begin;
-            if(tok.length > 0)
-                List_Token_push(&cmp->tokens, tok);
-            else setError(Error_unexpectedCharacter(cmp->code[--cmp->pos]));
-            // cmp->line will be increased in lex()
-            return;
-        }
-
-        if(!isAlphabeticalLower(c) && !isAlphabeticalUpper(c) && !isDigit(c) &&
-            c != '_' && c != '.'){
-            setError(Error_unexpectedCharacter(c));
-            return;
-        }
-        
-        cmp->column++;
-        cmp->pos++;
-    }
-    
-    // end of file
-    tok.length = cmp->pos - tok.begin;
-    if(tok.length > 0)
-        List_Token_push(&cmp->tokens, tok);
-    else setError(Error_endOfFile);
-}
-
-static void readArguments(Compiler* cmp){
-    char c; // space
-    cmp->pos++;
-    cmp->column++;
-    Token tok = Token_construct(TokenType_Argument, cmp->pos, 0);
-    
-    while(cmp->pos < cmp->code_len){
-        c = cmp->code[cmp->pos];
-        // end of line
-        if(c == '\r' || c == '\n' || c == ';'){
-            tok.length = cmp->pos - tok.begin;
-            if(tok.length > 0)
-                List_Token_push(&cmp->tokens, tok);
-            // cmp->line will be increased in lex()
-            return;
-        }
-
-        // new argument begins
-        if(c == ' ' || c == '\t'){
-            tok.length = cmp->pos - tok.begin;
-            if(tok.length > 0)
-                List_Token_push(&cmp->tokens, tok);
-            tok.begin = cmp->pos + 1;
-        }
-        
-
-        cmp->column++;
-        cmp->pos++;
-    }
-    
-    // end of file
-    tok.length = cmp->pos - tok.begin;
-    if(tok.length > 0)
-        List_Token_push(&cmp->tokens, tok);
-}
-
-static void readInstruction(Compiler* cmp){
-    Token tok = Token_construct(TokenType_Instruction, cmp->pos, 0);
-    cmp->pos++;
-    cmp->column++;
-    
-    while(cmp->pos < cmp->code_len){
-        char c = cmp->code[cmp->pos];
-        // end of line
-        if(c == '\r' || c == '\n' || c == ';'){
-            tok.length = cmp->pos - tok.begin;
-            List_Token_push(&cmp->tokens, tok);
-            // cmp->line will be increased in lex()
-            return;
-        }
-
-        // arguments begin
-        if(c == ' ' || c == '\t'){
-            tok.length = cmp->pos - tok.begin;
-            List_Token_push(&cmp->tokens, tok);
-            readArguments(cmp);
-            return;
-        }
-
-        if(!isAlphabeticalLower(c) && !isAlphabeticalUpper(c) && !isDigit(c)){
-            setError(Error_unexpectedCharacter(c));
-            return;
-        }
-
-        cmp->column++;
-        cmp->pos++;
-    }
-    
-    // end of file
-    tok.length = cmp->pos - tok.begin;
-    List_Token_push(&cmp->tokens, tok);
-}
-
-static void readChar(Compiler* cmp){
-    Token tok = Token_construct(TokenType_Char, cmp->pos, 0);
-    cmp->pos++;
-    cmp->column++;
-    
-    while(cmp->pos < cmp->code_len){
-        char c = cmp->code[cmp->pos];
-        // end of line
-        if(c == '\r' || c == '\n'){
-            setError(Error_unexpectedCharacter(cmp->code[--cmp->pos]));
-            return;
-        }
-
-        if(c == '\''){
-            tok.length = cmp->pos - tok.begin + 1;
-            List_Token_push(&cmp->tokens, tok);
-            return;
-        }
-
-        cmp->column++;
-        cmp->pos++;
-    }
-    
-    // end of file
-    setError(Error_endOfFile);
-}
-
-static void readString(Compiler* cmp){
-    Token tok = Token_construct(TokenType_String, cmp->pos, 0);
-    cmp->pos++;
-    cmp->column++;
-    
-    while(cmp->pos < cmp->code_len){
-        char c = cmp->code[cmp->pos];
-        // end of line
-        if(c == '\r' || c == '\n'){
-            setError(Error_unexpectedCharacter(cmp->code[--cmp->pos]));
-            return;
-        }
-
-        if(c == '"'){
-            tok.length = cmp->pos - tok.begin + 1;
-            List_Token_push(&cmp->tokens, tok);
-            return;
-        }
-
-        cmp->column++;
-        cmp->pos++;
-    }
-    
-    // end of file
-    setError(Error_endOfFile);
-}
-
-static void readNumber(Compiler* cmp){
-    Token tok = Token_construct(TokenType_Number, cmp->pos, 0);
-    cmp->pos++;
-    cmp->column++;
-    
-    while(cmp->pos < cmp->code_len){
-        char c = cmp->code[cmp->pos];
-
-        if(c == '\r' || c == '\n' || c == ' ' || c == '\t' || c == ',' || c == ';'){
-            tok.length = cmp->pos - tok.begin;
-            List_Token_push(&cmp->tokens, tok);
-            return;
-        }
-
-        cmp->column++;
-        cmp->pos++;
-    }
-    
-    // end of file
-    tok.length = cmp->pos - tok.begin;
-    List_Token_push(&cmp->tokens, tok);
-}
-
-static bool lex(Compiler* cmp){
-    returnErrorIf_auto(cmp->state != CompilerState_Initial);
-    cmp->state = CompilerState_Lexing;
-    cmp->column = 1;
-
-    while(cmp->pos < cmp->code_len){
-        char c = cmp->code[cmp->pos];
-        switch(c){
-            // skip blank characters
-            case ' ': case '\t': case '\r': case '\n':
-                break;
-            // try read comment
-            case '/':
-                readComment(cmp);
-                break;
-            // try read label
-            case '.':
-                readLabel(cmp);
-                break;
-            case '"':
-                readString(cmp);
-                break;
-            case '\'':
-                readChar(cmp);
-                break;
-            default:
-                // try read instruction
-                if(isAlphabeticalLower(c) || isAlphabeticalUpper(c))
-                    readInstruction(cmp);
-                else if(isDigit(c))
-                    readNumber(cmp);
-                else returnError(Error_unexpectedCharacter(c));
-                break;
-        }
-
-        if(cmp->state == CompilerState_Error)
-            return false;
-        
-        c = cmp->code[cmp->pos];
-        if(c == '\n')
-            completeLine(cmp);
-        cmp->column++;
-        cmp->pos++;
-    }
-
-    completeLine(cmp);
-    return true;
-}
-
-static bool parse(Compiler* cmp){
-    returnErrorIf_auto(cmp->state != CompilerState_Lexing);
-    cmp->state = CompilerState_Parsing;
-
-    return true;
-}
-static bool compile(Compiler* cmp, FILE* f){
-    returnErrorIf_auto(cmp->state != CompilerState_Parsing);
-    cmp->state = CompilerState_Compiling;
-
-    return true;
-}
-
-bool Compiler_compileTasm(Compiler* cmp, cstr source_file_name, cstr out_file_name, bool debug){
-    FILE* f = fopen(source_file_name, "rb");
-    if(f == NULL)
-        returnError("ERROR: can't open file '%s'", source_file_name);
-    
-    List_u8 buf = List_u8_alloc(64 * 1024);
-    int ret;
-    while((ret = fgetc(f)) != EOF) {
-        List_u8_push(&buf, ret);
-    }
-    if(ferror(f)){
-        free(buf.data);
-        fclose(f);
-        returnError("can't read file '%s'", source_file_name);
-    }
-    fclose(f);
-
-    if(buf.len == 0){
-        free(buf.data);
-        fclose(f);
-        returnError("soucre file is empty");
-    }
-
-    cmp->code = (char*)buf.data;
-    cmp->code_len = buf.len;
-    List_u8_push(&buf, 0);
-
-    f = fopen(out_file_name, "wb");
-    if(f == NULL){
-        free(buf.data);
-        returnError("ERROR: can't open file '%s'", out_file_name);
-    }
-
-    if(debug){
-        printf("----------------------------------[%s]---------------------------------\n", source_file_name);
-        fputs(cmp->code, stdout);
-        fputc('\n', stdout);
-    }
-    
-    bool success = lex(cmp);
-    if(debug){
-        printf("------------------------------------[lines]-----------------------------------\n");
-        for(u32 i = 0; i < cmp->line_lengths.len; i++){
-            printf("[%u] length: %u\n", i+1, cmp->line_lengths.data[i]);
-        }
-        printf("------------------------------------[tokens]-----------------------------------\n");
-        for(u32 i = 0; i < cmp->tokens.len; i++){
-            Token t = cmp->tokens.data[i];
-            CodePos pos = Compiler_getLineAndColumn(cmp, t.begin);
-            char* tokstr = malloc(4096);
-            strncpy(tokstr, cmp->code + t.begin, t.length);
-            tokstr[t.length] = 0;
-            printf("[l:%3u, c:%3u] %s '%s'\n", 
-                pos.line, pos.column,
-                TokenType_toString(t.type), tokstr);
-            free(tokstr);
-        }
-    }
-    if(!success){
-        fclose(f);
-        return false;
-    }
-
-    success = parse(cmp);
-    if(!success){
-        fclose(f);
-        return false;
-    }
-
-    success = compile(cmp, f);
-    fclose(f);
-    if(success){
-        cmp->state = CompilerState_Success;
-    }
-    
-    return success;
-}

src/compiler/token.c

@@ -1,19 +0,0 @@
-#include "token.h"
-
-static cstr _TokenType_str[] = {
-    "Unset",
-    "SingleLineComment",
-    "MultiLineComment",
-    "Label",
-    "Instruction",
-    "Argument",
-    "Number",
-    "Char",
-    "String",
-};
-
-cstr TokenType_toString(TokenType t){
-    if(t >= sizeof(_TokenType_str) / sizeof(cstr))
-        return "!!INDEX_ERROR!!";
-    return _TokenType_str[t];
-}

src/compiler/token.h

@@ -1,24 +0,0 @@
-#pragma once
-#include "../std.h"
-
-typedef enum TokenType {
-    TokenType_Unset,
-    TokenType_SingleLineComment,
-    TokenType_MultiLineComment,
-    TokenType_Label,
-    TokenType_Instruction,
-    TokenType_Argument,
-    TokenType_Number,
-    TokenType_Char,
-    TokenType_String,
-} TokenType;
-
-cstr TokenType_toString(TokenType t);
-
-typedef struct Token {
-    u32 begin;          // some index in Compiler->code
-    u32 length : 24;    // length in characters (24 bits)
-    TokenType type : 8; // type of token (8 bits)
-} Token;
-
-#define Token_construct(TYPE, BEGIN, END) ((Token){ .type = TYPE, .begin = BEGIN, .length = END })

src/cstr.c

@@ -43,7 +43,7 @@ char* NULLABLE(sprintf_malloc)(size_t buffer_size, cstr format, ...){
 
 char* NULLABLE(vsprintf_malloc)(size_t buffer_size, cstr format, va_list argv){
     char* buf = malloc(buffer_size);
-    int r = vsprintf_s(buf, buffer_size, format, argv);
+    int r = vsprintf(buf, format, argv);
     if(r < 0){
         free(buf);
         return NULL;

src/instructions/impl/PUSH.c

@@ -4,8 +4,9 @@
 i32 PUSH_impl(VM* vm){
     u8 dst_register_i = 0;
     readRegisterVar(dst_register_i);
-    u8 value_size = 0;
-    readValueSizeVar(value_size);
+    /*u8 value_size = 0;
+    readValueSizeVar(value_size);*/
+    u8 value_size = 4;\
     
     vm->registers[dst_register_i].u32v = 0;
     if(!VM_dataRead(vm, &vm->registers[dst_register_i].u32v, vm->current_pos, value_size))

src/instructions/impl/impl_macros.h

@@ -1,5 +1,6 @@
 #pragma once
 #include "../instructions.h"
+#include "../registers.h"
 
 #define readVar(VAR) {\
     if(!VM_dataRead(vm, &VAR, vm->current_pos, sizeof(VAR))) \
@@ -8,7 +9,7 @@
 }
 
 #define validateRegisterIndex(VAR) {\
-    if(VAR > sizeof(vm->registers)){\
+    if(VAR> sizeof(vm->registers)){\
         VM_setError(vm, "invalid register index (%x)", VAR);\
         return -1;\
     }\
@@ -16,9 +17,11 @@
 
 #define readRegisterVar(VAR) {\
     readVar(VAR);\
+    VAR -= 1;\
     validateRegisterIndex(VAR);\
 }
 
+/*
 #define validateValueSize(VAR) {\
     if(VAR < 1 || VAR > 4){\
         VM_setError(vm, "invalid value_size (%x)", VAR);\
@@ -30,3 +33,4 @@
     readVar(VAR);\
     validateValueSize(VAR);\
 }
+*/

src/instructions/impl/math_operators.c

@@ -4,8 +4,9 @@
     u8 dst_register_i = 0, src_register_i = 0;\
     readRegisterVar(dst_register_i);\
     readRegisterVar(src_register_i);\
-    u8 value_size = 0;\
-    readValueSizeVar(value_size);\
+    /*u8 value_size = 0;\
+    readValueSizeVar(value_size);*/\
+    u8 value_size = 4;\
     \
     switch(value_size){\
         case 1: \

src/instructions/instructions.c

@@ -1,6 +1,21 @@
 #include "instructions.h"
+#include "../collections/HashMap.h"
 
-const Instruction instructions[] = {
+i32 NOP_impl(VM* vm);
+i32 PUSH_impl(VM* vm);
+i32 MOV_impl(VM* vm);
+i32 ADD_impl(VM* vm);
+i32 SUB_impl(VM* vm);
+i32 MUL_impl(VM* vm);
+i32 DIV_impl(VM* vm);
+i32 MOD_impl(VM* vm);
+i32 SYS_impl(VM* vm);
+i32 EXIT_impl(VM* vm);
+i32 JMP_impl(VM* vm);
+i32 CALL_impl(VM* vm);
+
+Array_declare(Instruction);
+static const Array_Instruction instructions_array = ARRAY(Instruction, {
     Instruction_construct(NOP),
     Instruction_construct(PUSH),
     Instruction_construct(MOV),
@@ -13,12 +28,38 @@ const Instruction instructions[] = {
     Instruction_construct(EXIT),
     // Instruction_construct(JMP),
     // Instruction_construct(CALL),
-};
-const size_t instructions_count = sizeof(instructions)/sizeof(instructions[0]);
+});
 
-const Instruction* NULLABLE(Instruction_getFromOpcode)(u8 opcode){
-    if(opcode >= instructions_count)
+const Instruction* Instruction_getByOpcode(Opcode opcode){
+    if(opcode >= instructions_array.len)
         return NULL;
     
-    return instructions + opcode;
+    return instructions_array.data + opcode;
+}
+
+HashMap_declare(Instruction);
+HashMap_define(Instruction, HashMap_DESTROY_VALUE_FUNC_NULL);
+
+static HashMap_Instruction* instructions_map = NULL;
+
+const Instruction* Instruction_getByName(str name){
+    if(instructions_map == NULL){
+        instructions_map = malloc(sizeof(HashMap_Instruction));
+        HashMap_Instruction_alloc(instructions_map);
+        for(u32 i = 0; i < instructions_array.len; i++){
+            HashMap_Instruction_tryPush(instructions_map, instructions_array.data[i].name, instructions_array.data[i]);
+        }
+    }
+
+    str name_upper = str_toUpper(name);
+    Instruction* iptr = HashMap_Instruction_tryGetPtr(instructions_map, name_upper);
+    free(name_upper.data);
+    return iptr;
+}
+
+void Instruction_freeSearchStructs(){
+    if(instructions_map != NULL){
+        HashMap_Instruction_free(instructions_map);
+        free(instructions_map);
+    }
 }

src/instructions/instructions.h

@@ -5,30 +5,34 @@
 ///@returns number of bytes read
 typedef i32 (*InstructionImplFunc_t)(VM* vm);
 
+typedef enum __attribute__((__packed__)) Opcode {
+    Opcode_NOP,
+    Opcode_PUSH,
+    Opcode_MOV,
+    Opcode_ADD,
+    Opcode_SUB,
+    Opcode_MUL,
+    Opcode_DIV,
+    Opcode_MOD,
+    Opcode_SYS,
+    Opcode_EXIT,
+} Opcode;
+
 typedef struct Instruction {
-    cstr name;
+    str name;
     InstructionImplFunc_t implementation;
+    Opcode opcode;
 } Instruction;
 
 #define Instruction_construct(NAME) {\
-    .name = #NAME, \
-    .implementation = NAME##_impl \
+    .name = STR(#NAME), \
+    .implementation = NAME##_impl, \
+    .opcode = Opcode_##NAME\
 }
 
 /// @brief get instruction info from table
 /// @param opcode any byte
 /// @return ptr to struct or NULL
-const Instruction* NULLABLE(Instruction_getFromOpcode)(u8 opcode);
-
-i32 NOP_impl(VM* vm);
-i32 PUSH_impl(VM* vm);
-i32 MOV_impl(VM* vm);
-i32 ADD_impl(VM* vm);
-i32 SUB_impl(VM* vm);
-i32 MUL_impl(VM* vm);
-i32 DIV_impl(VM* vm);
-i32 MOD_impl(VM* vm);
-i32 SYS_impl(VM* vm);
-i32 EXIT_impl(VM* vm);
-i32 JMP_impl(VM* vm);
-i32 CALL_impl(VM* vm);
+const Instruction* NULLABLE(Instruction_getByOpcode)(Opcode opcode);
+const Instruction* NULLABLE(Instruction_getByName)(str name);
+void Instruction_freeSearchStructs();

src/instructions/registers.c

@@ -0,0 +1,13 @@
+#include "registers.h"
+
+RegisterCode RegisterCode_parse(str r){
+    if(str_equals(r, STR("ax")))
+        return RegisterCode_ax;
+    if(str_equals(r, STR("bx")))
+        return RegisterCode_bx;
+    if(str_equals(r, STR("cx")))
+        return RegisterCode_cx;
+    if(str_equals(r, STR("dx")))
+        return RegisterCode_dx;
+    return RegisterCode_Unset;
+}

src/instructions/registers.h

@@ -0,0 +1,13 @@
+#pragma once
+#include "../std.h"
+#include "../string/str.h"
+
+typedef enum RegisterCode {
+    RegisterCode_Unset,
+    RegisterCode_ax,
+    RegisterCode_bx,
+    RegisterCode_cx,
+    RegisterCode_dx
+} RegisterCode;
+
+RegisterCode RegisterCode_parse(str register_name);

src/main.c

@@ -1,11 +1,11 @@
 #include "VM/VM.h"
 #include "instructions/instructions.h"
 #include "collections/List.h"
-#include "compiler/compiler.h"
+#include "compiler/Compiler.h"
 
 #define arg_is(STR) (strcmp(argv[argi], STR) == 0)
 
-i32 compileSources(cstr source_file, cstr out_file);
+i32 compileSources(cstr source_file, cstr out_file, bool debug_log);
 i32 bootFromImage(cstr image_file);
 
 i32 main(const i32 argc, cstr* argv){
@@ -21,6 +21,8 @@ i32 main(const i32 argc, cstr* argv){
     cstr NULLABLE(out_file) = NULL;
     cstr NULLABLE(source_file) = NULL;
 
+    bool debug_log = false;
+
     for(i32 argi = 1; argi < argc; argi++){
         if(arg_is("-h") || arg_is("--help")){
             printf(
@@ -28,14 +30,15 @@ i32 main(const i32 argc, cstr* argv){
                 "-op, --opcodes                             Show list of all instructions.\n"
                 "-i, --image [FILE]                         Boot VM using image file.\n"
                 "-c, --compile [SOURCE_FILE] [OUT_FILE]     Compile assembly source files to machine code.\n"
+                "-d, --debug                                Enable debug log.\n"
             );
             return 0;
         }
         else if(arg_is("-op") || arg_is("--opcodes")){
             for(u8 opcode = 0; opcode < 255; opcode++){
-                const Instruction* instr = Instruction_getFromOpcode(opcode);
+                const Instruction* instr = Instruction_getByOpcode(opcode);
                 if(instr != NULL){
-                    printf("%02X %s\n", opcode, instr->name);
+                    printf("%02X %s\n", opcode, instr->name.data);
                 }
             }
             return 0;
@@ -72,6 +75,9 @@ i32 main(const i32 argc, cstr* argv){
             }
             out_file = argv[argi];
         }
+        else if(arg_is("-d") || arg_is("--debug")){
+            debug_log = true;
+        }
         else {
             printfe("ERROR: unknown argument '%s'\n", argv[argi]);
             return 1;
@@ -80,12 +86,14 @@ i32 main(const i32 argc, cstr* argv){
 
     i32 exit_code = 0;
     if(compile){
-        exit_code = compileSources(source_file, out_file);
+        exit_code = compileSources(source_file, out_file, debug_log);
     }
     if(exit_code == 0 && boot){
         exit_code = bootFromImage(image_file);
     }
 
+    // frees global variables to supress valgrind memory leak errors
+    Instruction_freeSearchStructs();
     return exit_code;
 }
 
@@ -131,19 +139,20 @@ i32 bootFromImage(cstr image_file){
     return exit_code;
 }
 
-i32 compileSources(cstr source_file, cstr out_file){
+i32 compileSources(cstr source_file, cstr out_file, bool debug_log){
     Compiler cmp;
     Compiler_init(&cmp);
-    bool success = Compiler_compileTasm(&cmp, source_file, out_file, true);
-    Compiler_free(&cmp);
+    bool success = Compiler_compile(&cmp, source_file, out_file, debug_log);
     if(!success){
         if(cmp.error_message){
             printfe("COMPILER ERROR: %s\n", cmp.error_message);
             free(cmp.error_message);
         }
         else printfe("COMPILER ERROR: unknown (error_message is null)\n");
+        Compiler_free(&cmp);
         return 111;
     }
     
+    Compiler_free(&cmp);
     return 0;
 }

src/std.h

@@ -26,6 +26,15 @@ typedef u8 bool;
 
 typedef const char* cstr;
 
+#if defined(_WIN64) || defined(_WIN32)
+    #define IFWIN(YES, NO) YES
+#else
+    #define IFWIN(YES, NO) NO
+#endif
+
+#define ARRAY_SIZE(A) (sizeof(A)/sizeof(A[0]))
+#define ALIGN_TO(_SIZE,_ALIGN)        (((_SIZE) + ((_ALIGN) - 1)) & ~((_ALIGN) - 1))
+
 #define __count_args( \
     a0, a1, a2, a3, a4, a5, a6, a7 , a8, a9, a10,a11,a12,a13,a14,a15, \
     a16,a17,a18,a19,a20,a21,a22,a23, a24,a25,a26,a27,a28,a29,a30,a31, \

src/string/StringBuilder.c

@@ -0,0 +1,53 @@
+#include "StringBuilder.h"
+
+void StringBuilder_free(StringBuilder* b){
+    free(b->buffer.data);
+    b->buffer = List_u8_construct(NULL, 0, 0);
+}
+
+str StringBuilder_getStr(StringBuilder* b){
+    List_u8_push(&b->buffer, '\0');
+    str result = str_construct((char*)b->buffer.data, b->buffer.len - 1, true);
+    return result;
+}
+
+void StringBuilder_removeFromEnd(StringBuilder* b, u32 count){
+    if(count < b->buffer.len){
+        b->buffer.len -= count;
+    }
+    else{
+        b->buffer.len = 0;
+    }
+}
+
+
+void StringBuilder_append_char(StringBuilder* b, char c){
+    List_u8_push(&b->buffer, c);
+}
+
+
+void StringBuilder_append_string(StringBuilder* b, str s){
+    List_u8_pushMany(&b->buffer, (u8*)s.data, s.len);
+}
+
+void StringBuilder_append_cstr(StringBuilder* b, char* s){
+    StringBuilder_append_string(b, str_construct(s, strlen(s), true));
+}
+
+void StringBuilder_append_i64(StringBuilder* b, i64 n){
+    char buf[32];
+    sprintf(buf, IFWIN("%lli", "%li"), n);
+    StringBuilder_append_cstr(b, buf);
+}
+
+void StringBuilder_append_u64(StringBuilder* b, u64 n){
+    char buf[32];
+    sprintf(buf, IFWIN("%llu", "%lu"), n);
+    StringBuilder_append_cstr(b, buf);
+}
+
+void StringBuilder_append_f64(StringBuilder* b, f64 n){
+    char buf[32];
+    sprintf(buf, "%lf", n);
+    StringBuilder_append_cstr(b, buf);
+}

src/string/StringBuilder.h

@@ -0,0 +1,25 @@
+#pragma once
+
+#include "../collections/List.h"
+#include "str.h"
+
+typedef struct StringBuilder {
+	List_u8 buffer;
+} StringBuilder;
+
+static inline StringBuilder StringBuilder_alloc(u32 initial_size) { 
+    return (StringBuilder){ .buffer = List_u8_alloc(initial_size) }; 
+}
+void StringBuilder_free(StringBuilder* b);
+
+/// @param count set to -1 to clear StringBuilder
+void StringBuilder_removeFromEnd(StringBuilder* b, u32 count);
+void StringBuilder_append_char(StringBuilder* b, char c);
+void StringBuilder_append_cstr(StringBuilder* b, char* s);
+void StringBuilder_append_string(StringBuilder* b, str s);
+void StringBuilder_append_i64(StringBuilder* b, i64 a);
+void StringBuilder_append_u64(StringBuilder* b, u64 a);
+void StringBuilder_append_f64(StringBuilder* b, f64 a);
+
+// adds '\0' to the buffer and returns pointer to buffer content
+str StringBuilder_getStr(StringBuilder* b);

src/string/str.c

@@ -0,0 +1,125 @@
+#include "str.h"
+
+str str_copy(str src){
+    if(src.data == NULL || src.len == 0)
+        return src;
+    
+    str nstr = str_construct((char*)malloc(src.len + 1), src.len, true);
+    memcpy(nstr.data, src.data, src.len);
+    nstr.data[nstr.len] = '\0';
+    return nstr;
+}
+
+bool str_equals(str s0, str s1){
+    if(s0.len != s1.len)
+        return false;
+    
+    for(u32 i = 0; i < s0.len; i++)
+        if(s0.data[i] != s1.data[i]) 
+            return false;
+    
+    return true;
+}
+
+str str_reverse(str s){
+    if(s.data == NULL || s.len == 0)
+        return s;
+    
+    str r = str_construct(malloc(s.len), s.len, s.isZeroTerminated);
+    for(u32 i = 0; i < s.len; i++ )
+        r.data[i] = s.data[s.len - i - 1];
+    return r;
+}
+
+i32 str_seek(str src, str fragment, u32 startIndex){
+    if(src.len == 0 || fragment.len == 0)
+        return -1;
+    
+    for(u32 i = startIndex; i < src.len - fragment.len + 1; i++){
+        for(u32 j = 0;; j++){
+            if(j == fragment.len)
+                return i;
+            if(src.data[i + j] != fragment.data[j])
+                break;
+        }
+    }
+    return -1;
+}
+
+i32 str_seekReverse(str src, str fragment, u32 startIndex){
+    if(src.len == 0 || fragment.len == 0)
+        return -1;
+    
+    if(startIndex > src.len - 1)
+        startIndex = src.len - 1;
+    for(u32 i = startIndex; i >= fragment.len - 1; i--){
+        for(u32 j = 0;; j++){
+            if(j == fragment.len)
+                return i - j + 1;
+            if(src.data[i - j] != fragment.data[fragment.len - 1 - j])
+                break;
+        }
+    }
+    return -1;
+}
+
+i32 str_seekChar(str src, char c, u32 startIndex){
+    for(u32 i = startIndex; i < src.len; i++){
+        if(src.data[i] == c)
+            return i;
+    }
+    return -1;
+}
+
+i32 str_seekCharReverse(str src, char c, u32 startIndex){
+    if(startIndex > src.len - 1)
+        startIndex = src.len - 1;
+    for(u32 i = startIndex; i != (u32)-1; i--){
+        if(src.data[i] == c)
+            return i;
+    }
+    return -1;
+}
+
+bool str_startsWith(str src, str fragment){
+    if(src.len < fragment.len)
+        return false;
+    
+    src.len = fragment.len;
+    return str_equals(src, fragment);
+}
+
+bool str_endsWith(str src, str fragment){
+    if(src.len < fragment.len)
+        return false;
+    
+    src.data = (char*)(src.data + src.len - fragment.len);
+    src.len = fragment.len;
+    return str_equals(src, fragment);
+}
+
+u32 str_hash32(str s){
+    u8* ubuf = (u8*)s.data;
+    u32 hash=0;
+    for (u32 i = 0; i < s.len; i++)
+        hash = (hash<<6) + (hash<<16) - hash + ubuf[i];
+    return hash;
+}
+
+str str_toUpper(str src){
+    str r = str_copy(src);
+    for (u32 i = 0; i < r.len; i++){
+        if(isAlphabeticalLower(r.data[i]))
+            r.data[i] = r.data[i] - 'a' + 'A';
+    }
+    return r;
+}
+
+str str_toLower(str src){
+    str r = str_copy(src);
+    for (u32 i = 0; i < r.len; i++){
+        if(isAlphabeticalUpper(r.data[i]))
+            r.data[i] = r.data[i] - 'A' + 'a';
+    }
+    return r;
+}

src/string/str.h

@@ -0,0 +1,41 @@
+#pragma once
+
+#include "../std.h"
+
+typedef struct str {
+    char* data;
+    u32 len;
+    bool isZeroTerminated;
+} str;
+
+/// creates str from a string literal
+#define STR(LITERAL) str_construct(LITERAL, ARRAY_SIZE(LITERAL) - 1, true)
+
+#define str_construct(DATA, LEN, ZERO_TERMINATED) ((str){ .data = DATA, .len = LEN, .isZeroTerminated = ZERO_TERMINATED })
+
+static const str str_null = str_construct(NULL, 0, 0);
+
+/// copies src content to new string and adds \0 at the end
+str str_copy(str src);
+
+/// compares two strings, NullPtr-friendly
+bool str_equals(str str0, str str1);
+
+/// allocates new string which is reversed variant of <s>
+str str_reverse(str s);
+
+i32 str_seek(str src, str fragment, u32 startIndex);
+i32 str_seekReverse(str src, str fragment, u32 startIndex);
+
+i32 str_seekChar(str src, char c, u32 startIndex);
+i32 str_seekCharReverse(str src, char c, u32 startIndex);
+
+bool str_startsWith(str src, str fragment);
+bool str_endsWith(str src, str fragment);
+
+/// @brief calculates string hash using sdbm32 algorythm (something like lightweight crc32)
+/// @return non-cryptografic hash of the string
+u32 str_hash32(str s);
+
+str str_toUpper(str src);
+str str_toLower(str src);