temporarely disabled variable size reading

src/VM/VM.c

@@ -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/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;\
+    T* data;\
-        u32 len;\
+    u32 len;\
-    } Array_##T;\
+} Array_##T;\
-    \
+\
-    static inline Array_##T Array_##T##_construct(T* data_ptr, u32 len) {\
+static inline Array_##T Array_##T##_alloc(u32 len){\
-        return (Array_##T){ .data = data_ptr, .len = len };\
+    return Array_construct(T, (T*)malloc(len * sizeof(T)), len);\
-    }\
+}\
-    \
+static inline void Array_##T##_realloc(Array_##T* ptr, u32 new_len){\
-    static inline Array_##T Array_##T##_alloc(u32 len){\
+    ptr->data = (T*)realloc(ptr->data, new_len * sizeof(T));\
-        return Array_##T##_construct((T*)malloc(len * sizeof(T)), len);\
+    ptr->len = new_len;\
-    }\
+}
-    static inline void Array_##T##_realloc(Array_##T* ptr, u32 new_len){\
+
-        ptr->data = (T*)realloc(ptr->data, new_len * sizeof(T));\
+Array_declare(u8)
-        ptr->len = new_len;\
+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.h

@@ -20,6 +20,7 @@
     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)\
@@ -29,17 +30,22 @@
         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);\
-        return List_##T##_construct((T*)malloc(initial_len * sizeof(T)), 0, 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 > ptr->max_len){\
+        while(ptr->len > expanded_max_len){\
             expanded_max_len *= 2;\
         }\
-        ptr->data = (T*)realloc(ptr->data, expanded_max_len * sizeof(T));\
+        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;\
     }\
     \
@@ -52,6 +58,17 @@
         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;\
+    }\
 
 
 List_declare(u32);

src/compiler/AST.c

@@ -9,14 +9,14 @@ static str _ArgumentType_str[] = {
     STR("Unset"),
     STR("Register"),
     STR("ConstValue"),
-    STR("DataName"),
+    STR("VarDataName"),
-    STR("NamedDataPointer"),
+    STR("ConstDataPointer"),
-    STR("NamedDataSize"),
+    STR("ConstDataSize"),
 };
 
 str ArgumentType_toString(ArgumentType t){
     if(t >= ARRAY_SIZE(_ArgumentType_str))
-        return STR("!!INDEX_ERROR!!");
+        return STR("!!ArgumentType INDEX_ERROR!!");
     return _ArgumentType_str[t];
 }
 
@@ -28,8 +28,14 @@ void Section_init(Section* sec, str name){
 }
 
 void Section_free(Section* sec){
-    free(sec->name.data);
+    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);
 }
 

src/compiler/AST.h

@@ -2,23 +2,28 @@
 #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_DataName,
+    ArgumentType_VarDataName,
-    ArgumentType_NamedDataPointer,
+    ArgumentType_ConstDataPointer,
-    ArgumentType_NamedDataSize,
+    ArgumentType_ConstDataSize,
 } ArgumentType;
 
 str ArgumentType_toString(ArgumentType t);
 
-
 typedef struct Argument {
     ArgumentType type;
-    u32 value;
+    union {
+        i64 i;
+        f64 f;
+        str data_name;
+        RegisterCode register_code;
+    } value;
 } Argument;
 
 List_declare(Argument);
@@ -26,7 +31,7 @@ List_declare(Argument);
 
 typedef struct Operation {
     List_Argument args;
-    Opcode op;
+    Opcode opcode;
 } Operation;
 
 List_declare(Operation);

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

@@ -1,11 +1,14 @@
 #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){
@@ -13,6 +16,7 @@ void Compiler_free(Compiler* cmp){
     free(cmp->tokens.data);
     free(cmp->line_lengths.data);
     AST_free(&cmp->ast);
+    BinaryObject_free(&cmp->binary);
 }
 
 CodePos Compiler_getLineAndColumn(Compiler* cmp, u32 pos){
@@ -60,10 +64,156 @@ str Compiler_constructTokenStr(Compiler* cmp, Token t){
     return s;
 }
 
-static bool compileFile(Compiler* cmp, FILE* f){
+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;
 }
 
@@ -98,17 +248,21 @@ bool Compiler_compile(Compiler* cmp, cstr source_file_name, cstr out_file_name,
     }
 
     if(debug_log){
-        printf("----------------------------------[%s]---------------------------------\n", source_file_name);
+        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");
+        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];
@@ -116,24 +270,92 @@ bool Compiler_compile(Compiler* cmp, cstr source_file_name, cstr out_file_name,
             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);
+                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;
     }
 
-    success = Compiler_parse(cmp);
+    if(debug_log)
-    if(!success){
+        printf("==================================[compiling]==================================\n");
-        fclose(f);
+    success = writeBinaryFile(cmp, f);
-        return false;
-    }
-
-    success = compileFile(cmp, f);
     fclose(f);
     if(success){
         cmp->state = CompilerState_Success;

src/compiler/Compiler.h

@@ -2,8 +2,9 @@
 #include "../std.h"
 #include "../string/str.h"
 #include "../collections/List.h"
+#include "../collections/HashMap.h"
 #include "Token.h"
-#include "AST.h"
+#include "Binary.h"
 
 typedef enum CompilerState {
     CompilerState_Initial,
@@ -14,16 +15,24 @@ typedef enum CompilerState {
     CompilerState_Success
 } CompilerState;
 
+typedef Section* SectionPtr;
+HashMap_declare(SectionPtr);
+
 typedef struct Compiler {
+    /* 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);

src/compiler/Parser.c

@@ -17,6 +17,12 @@
     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"
@@ -28,7 +34,7 @@ static void List_u8_pushBytes(List_u8* l, void* value, u32 startIndex, u32 count
     }
 }
 
-static inline bool isVarSizeBits(u32 B) { return (B == 8 && B == 16 && B == 32 && B == 64); }
+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);
@@ -83,30 +89,44 @@ static void parseDataDefinition(Compiler* cmp, str instr_name, DataDefinition* d
     }
     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_Unset:
-                setError(Error_TokenUnset);
-                return;
             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(tok_str.data);
+                    f64 f = atof(processed_str.data);
                     List_u8_pushBytes(&ddf->data, &f, 8 - ddf->element_size, ddf->element_size);
                 }
                 else {
-                    i64 i = atoll(tok_str.data);
+                    i64 i = atoll(processed_str.data);
                     List_u8_pushBytes(&ddf->data, &i, 8 - ddf->element_size, ddf->element_size);
                 }
                 free(processed_str.data);
@@ -132,18 +152,84 @@ static void parseDataDefinition(Compiler* cmp, str instr_name, DataDefinition* d
                 List_u8_pushBytes(&ddf->data, processed_str.data, 0, processed_str.len);
                 free(processed_str.data);
                 break;
-            case TokenType_OperationEnd:
-                return;
-            default:
-                setError_unexpectedToken(tok);
-                return;
         }
     }
 }
 
 
 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){

src/compiler/Token.c

@@ -13,11 +13,12 @@ static str _TokenType_str[] = {
     STR("String"),
     STR("Name"),
     STR("NamedDataPointer"),
-    STR("NamedDataSize")
+    STR("NamedDataSize"),
+    STR("OperationEnd"),
 };
 
 str TokenType_toString(TokenType t){
     if(t >= ARRAY_SIZE(_TokenType_str))
-        return STR("!!INDEX_ERROR!!");
+        return STR("!!TokenType INDEX_ERROR!!");
     return _TokenType_str[t];
 }

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/MOV.c

@@ -4,7 +4,7 @@
 i32 MOV_impl(VM* vm){
     u8 dst_register_i = 0;
     readRegisterVar(dst_register_i);
-    u8  src_register_i = 0;
+    u8 src_register_i = 0;
     readRegisterVar(src_register_i);
     if(dst_register_i == src_register_i){
         VM_setError(vm, "dst_register_i == src_register_i (%x) ", src_register_i);

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;
+    /*u8 value_size = 0;
-    readValueSizeVar(value_size);
+    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;\
+    /*u8 value_size = 0;\
-    readValueSizeVar(value_size);\
+    readValueSizeVar(value_size);*/\
+    u8 value_size = 4;\
     \
     switch(value_size){\
         case 1: \

src/instructions/instructions.c

@@ -1,4 +1,5 @@
 #include "instructions.h"
+#include "../collections/HashMap.h"
 
 i32 NOP_impl(VM* vm);
 i32 PUSH_impl(VM* vm);
@@ -13,8 +14,8 @@ i32 EXIT_impl(VM* vm);
 i32 JMP_impl(VM* vm);
 i32 CALL_impl(VM* vm);
 
-
+Array_declare(Instruction);
-const Instruction instructions[] = {
+static const Array_Instruction instructions_array = ARRAY(Instruction, {
     Instruction_construct(NOP),
     Instruction_construct(PUSH),
     Instruction_construct(MOV),
@@ -27,11 +28,38 @@ const Instruction instructions[] = {
     Instruction_construct(EXIT),
     // Instruction_construct(JMP),
     // Instruction_construct(CALL),
-};
+});
 
 const Instruction* Instruction_getByOpcode(Opcode opcode){
-    if(opcode >= ARRAY_SIZE(instructions))
+    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

@@ -34,3 +34,5 @@ typedef struct Instruction {
 /// @param opcode any byte
 /// @return ptr to struct or NULL
 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

@@ -92,6 +92,8 @@ i32 main(const i32 argc, cstr* argv){
         exit_code = bootFromImage(image_file);
     }
 
+    // frees global variables to supress valgrind memory leak errors
+    Instruction_freeSearchStructs();
     return exit_code;
 }
 

src/std.h

@@ -26,7 +26,13 @@ typedef u8 bool;
 
 typedef const char* cstr;
 
-#define ARRAY_SIZE(A) sizeof(A)/sizeof(A[0])
+#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( \

src/string/StringBuilder.c

@@ -36,18 +36,18 @@ void StringBuilder_append_cstr(StringBuilder* b, char* s){
 
 void StringBuilder_append_i64(StringBuilder* b, i64 n){
     char buf[32];
-    sprintf_s(buf, sizeof(buf), "%llu", n);
+    sprintf(buf, IFWIN("%lli", "%li"), n);
     StringBuilder_append_cstr(b, buf);
 }
 
 void StringBuilder_append_u64(StringBuilder* b, u64 n){
     char buf[32];
-    sprintf_s(buf, sizeof(buf), "%llu", n);
+    sprintf(buf, IFWIN("%llu", "%lu"), n);
     StringBuilder_append_cstr(b, buf);
 }
 
 void StringBuilder_append_f64(StringBuilder* b, f64 n){
     char buf[32];
-    sprintf_s(buf, sizeof(buf), "%lf", n);
+    sprintf(buf, "%lf", n);
     StringBuilder_append_cstr(b, buf);
 }

src/string/str.c

@@ -97,3 +97,29 @@ bool str_endsWith(str src, str fragment){
     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

@@ -8,7 +8,7 @@ typedef struct str {
     bool isZeroTerminated;
 } str;
 
-// creates str from a string literal
+/// 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 })
@@ -32,3 +32,10 @@ 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);