mmichlol/PCCCompiler
1
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 6 Wiki Activity

String Update

Browse Source
This commit is contained in:
mmichlol
2026-02-07 19:44:03 +01:00
parent c37a122850
commit 3916b4ff7a
3 changed files with 232 additions and 244 deletions
Download Patch File Download Diff File Expand all files Collapse all files

321
PCCcompiler/codegen.cpp
View File

@@ -37,7 +37,8 @@ std::string getVarLocation(const std::string& name, const std::map<std::string,
}
std::string generateAssembly(const CompilerState& state) {
std::string result;
// 1. NAG£ÓWEK I DEKLARACJE EXTERN
std::string result = "default rel\n";
result += "global main\n";
result += "extern printf\n";
result += "extern getchar\n";
@@ -45,15 +46,21 @@ std::string generateAssembly(const CompilerState& state) {
result += "extern rand\n";
result += "extern srand\n";
result += "extern time\n";
result += "section .data\n";
result += " fmt db '%d', 10, 0\n";
result += "section .data\n";
result += " fmt_int db '%d', 10, 0\n"; // Format dla liczb
result += " fmt_str db '%s', 10, 0\n"; // Format dla stringów
// --- WYPISYWANIE STRINGÓW ---
for (const auto& pair : state.stringLiterals) { result += " " + pair.second + " db '" + pair.first + "', 0\n"; }
result += "section .text\n\n";
// 2. SEKCJA DATA (Tylko raz!)
result += "section .data\n";
result += " fmt_int db \"%lld\", 10, 0\n"; // Format dla liczb
result += " fmt_str db \"%s\", 10, 0\n"; // Format dla stringów
// Zrzucamy stringi: ETYKIETA db "TRESC", 0
for (const auto& p : state.stringLiterals) {
// p.first -> Etykieta (np. str_0)
// p.second -> TreϾ (np. Hello World)
result += " " + p.first + " db \"" + p.second + "\", 0\n";
}
// 3. SEKCJA TEXT (Kod programu)
result += "section .text\n";
for (const auto& pair : state.functions) {
const Function& func = pair.second;
@@ -66,7 +73,7 @@ std::string generateAssembly(const CompilerState& state) {
std::map<std::string, int> stackMap;
int currentStack = 8;
// 1. ARGUMENTY
// ARGUMENTY FUNKCJI
if (func.args.size() > 0) {
stackMap[func.args[0]] = currentStack;
result += " mov [rbp-" + std::to_string(currentStack) + "], rcx ; arg " + func.args[0] + "\n";
@@ -78,11 +85,10 @@ std::string generateAssembly(const CompilerState& state) {
currentStack += 8;
}
// 2. INSTRUKCJE
// GENEROWANIE INSTRUKCJI
for (const auto& instr : func.instructions) {
// Rezerwacja miejsca dla nowych zmiennych (wynikowych)
// Dodajemy tu OpType::SUB i OpType::MUL
// Rezerwacja miejsca na stosie
bool isWriteOp = (instr.type == OpType::ASSIGN ||
instr.type == OpType::ADD ||
instr.type == OpType::EQ ||
@@ -90,7 +96,7 @@ std::string generateAssembly(const CompilerState& state) {
instr.type == OpType::MUL ||
instr.type == OpType::DIV ||
instr.type == OpType::MOD ||
instr.type == OpType::LOGIC_AND ||
instr.type == OpType::LOGIC_AND ||
instr.type == OpType::LOGIC_OR);
if (isWriteOp && stackMap.find(instr.arg1) == stackMap.end() && instr.arg1 != "RAX") {
@@ -102,14 +108,23 @@ std::string generateAssembly(const CompilerState& state) {
case OpType::ASSIGN: {
std::string src = instr.arg2;
if (instr.arg3 == "STRING") {
// Przypisanie stringa: ³adujemy ADRES (LEA)
result += " lea rax, [rel " + src + "]\n";
std::string dst = getVarLocation(instr.arg1, stackMap);
// Zapisujemy adres w zmiennej lokalnej (wskaŸnik 64-bit qword)
result += " mov qword " + dst + ", rax\n";
}
else {
// Zwyk³e przypisanie liczby
std::string srcLoc = getVarLocation(instr.arg2, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
result += " mov eax, " + srcLoc + "\n";
if (isNumber(src)) {
result += " mov eax, " + src + "\n";
}
else {
result += " mov eax, " + srcLoc + "\n";
}
result += " mov " + dst + ", eax\n";
}
break;
@@ -124,29 +139,55 @@ std::string generateAssembly(const CompilerState& state) {
break;
}
case OpType::SUB: {
// a = b - c
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
result += " mov eax, " + op1 + "\n";
result += " sub eax, " + op2 + "\n"; // sub = odejmowanie
result += " sub eax, " + op2 + "\n";
result += " mov " + dst + ", eax\n";
break;
}
case OpType::MUL: {
// a = b * c
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
result += " mov eax, " + op1 + "\n";
// Mno¿enie w x86 jest specyficzne: imul eax, operand
// Wynik l¹duje w eax (i edx jeœli du¿y, ale ignorujemy nadmiar dla prostoty)
result += " imul eax, " + op2 + "\n";
result += " mov " + dst + ", eax\n";
break;
}
case OpType::DIV: {
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
result += " mov eax, " + op1 + "\n";
result += " cdq\n";
if (isdigit(op2[0]) || op2[0] == '-') {
result += " mov ecx, " + op2 + "\n";
result += " idiv ecx\n";
}
else {
result += " idiv dword " + op2 + "\n";
}
result += " mov " + dst + ", eax\n";
break;
}
case OpType::MOD: {
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
result += " mov eax, " + op1 + "\n";
result += " cdq\n";
if (isdigit(op2[0]) || op2[0] == '-') {
result += " mov ecx, " + op2 + "\n";
result += " idiv ecx\n";
}
else {
result += " idiv dword " + op2 + "\n";
}
result += " mov " + dst + ", edx\n"; // Reszta
break;
}
case OpType::EQ: {
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
@@ -158,34 +199,59 @@ std::string generateAssembly(const CompilerState& state) {
result += " mov " + dst + ", eax\n";
break;
}
case OpType::LOGIC_AND: {
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
result += " mov eax, " + op1 + "\n";
result += " cmp eax, 0\n";
result += " setne al\n";
if (isdigit(op2[0])) result += " mov ecx, " + op2 + "\n";
else result += " mov ecx, " + op2 + "\n";
result += " cmp ecx, 0\n";
result += " setne cl\n";
result += " and al, cl\n";
result += " movzx eax, al\n";
result += " mov " + dst + ", eax\n";
break;
}
case OpType::LOGIC_OR: {
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
result += " mov eax, " + op1 + "\n";
result += " cmp eax, 0\n";
result += " setne al\n";
if (isdigit(op2[0])) result += " mov ecx, " + op2 + "\n";
else result += " mov ecx, " + op2 + "\n";
result += " cmp ecx, 0\n";
result += " setne cl\n";
result += " or al, cl\n";
result += " movzx eax, al\n";
result += " mov " + dst + ", eax\n";
break;
}
case OpType::JMP_FALSE: {
// PARSOWANIE WARUNKU: np. "suma == 30" lub "test"
std::string condRaw = instr.arg2;
size_t eqPos = condRaw.find("==");
if (eqPos != std::string::npos) {
// Mamy porównanie w IFie (a == b)
std::string leftStr = condRaw.substr(0, eqPos);
std::string rightStr = condRaw.substr(eqPos + 2);
std::string op1 = getVarLocation(leftStr, stackMap);
std::string op2 = getVarLocation(rightStr, stackMap);
result += " mov eax, " + op1 + "\n";
result += " cmp eax, " + op2 + "\n";
result += " jne " + instr.arg1 + " ; jump if NOT equal\n";
result += " jne " + instr.arg1 + "\n";
}
else {
// Zwyk³a zmienna boolowska (if test)
std::string cond = getVarLocation(condRaw, stackMap);
result += " mov eax, " + cond + "\n";
result += " test eax, eax\n";
result += " jz " + instr.arg1 + " ; jump if zero\n";
result += " jz " + instr.arg1 + "\n";
}
break;
}
case OpType::JMP: {
// Skok bezwarunkowy (u¿ywany na koñcu pêtli while)
result += " jmp " + instr.arg1 + "\n";
break;
}
@@ -193,50 +259,54 @@ std::string generateAssembly(const CompilerState& state) {
result += instr.arg1 + ":\n";
break;
}
case OpType::RETURN: {
case OpType::PRINT: {
// Instrukcja PRINT zwyk³a (liczba)
std::string val = getVarLocation(instr.arg1, stackMap);
if (val.empty() || val == ";");
else {
result += " mov eax, " + val + " ; return value\n";
}
result += " leave\n";
result += " ret\n";
result += " mov edx, " + val + "\n";
result += " lea rcx, [rel fmt_int]\n";
result += " call printf\n";
break;
}
case OpType::PRINT: {
if (instr.arg2 == "STRING") {
result += " lea rdx, [rel " + instr.arg1 + "]\n";
result += " lea rcx, [rel fmt_str]\n";
result += " call printf\n";
case OpType::PRINT_STRING: {
// Instrukcja PRINT_STRING (tekst)
std::string target = instr.arg1;
if (target.find("str_") == 0) {
// Litera³: print("tekst")
result += " lea rdx, [rel " + target + "]\n";
}
else {
std::string val = getVarLocation(instr.arg1, stackMap);
result += " mov edx, " + val + "\n";
result += " lea rcx, [rel fmt_int]\n";
result += " call printf\n";
// Zmienna: print(s) -> s trzyma adres
std::string val = getVarLocation(target, stackMap);
result += " mov rdx, " + val + "\n";
}
result += " lea rcx, [rel fmt_str]\n";
result += " call printf\n";
break;
}
case OpType::CALL: {
// Parsowanie argumentów
std::string argsRaw = instr.arg2;
std::vector<std::string> callArgs;
// --- SPECJALNE FUNKCJE SYSTEMOWE ---
// 1. input() - czeka na ENTER (stare)
if (instr.arg1 == "input") {
result += " call getchar\n";
break;
}
// 2. read_key() - zwraca kod wciœniêtego klawisza (NOWOŒÆ)
if (instr.arg1 == "read_key") {
result += " call _getch\n"; // Zwraca kod znaku w EAX
// Jeœli to klawisz specjalny (strza³ki), _getch zwraca 0 lub 224,
// a potem trzeba wywo³aæ go drugi raz.
// Na razie zróbmy prosto: zwracamy to co zwróci³ pierwszy _getch.
result += " call _getch\n";
break;
}
if (instr.arg1 == "sys_seed") {
result += " mov rcx, 0\n";
result += " call time\n";
result += " mov rcx, rax\n";
result += " call srand\n";
break;
}
if (instr.arg1 == "sys_rand") {
result += " call rand\n";
break;
}
// Standardowe wywo³anie funkcji
std::string argsRaw = instr.arg2;
std::vector<std::string> callArgs;
if (!argsRaw.empty()) {
size_t comma = argsRaw.find(',');
if (comma != std::string::npos) {
@@ -247,137 +317,26 @@ std::string generateAssembly(const CompilerState& state) {
callArgs.push_back(argsRaw);
}
}
// Obs³uga RDX (arg 2)
if (callArgs.size() > 1) {
std::string val = getVarLocation(callArgs[1], stackMap);
if (isNumber(val)) {
// Jeœli liczba: mov rdx, 100
result += " mov rdx, " + val + "\n";
}
else {
// Jeœli zmienna/pamiêæ: movsxd rdx, dword [rbp-8]
result += " movsxd rdx, dword " + val + "\n";
}
if (isNumber(val)) result += " mov rdx, " + val + "\n";
else result += " movsxd rdx, dword " + val + "\n";
}
// Obs³uga RCX (arg 1)
if (callArgs.size() > 0) {
std::string val = getVarLocation(callArgs[0], stackMap);
if (isNumber(val)) {
result += " mov rcx, " + val + "\n";
}
else {
result += " movsxd rcx, dword " + val + "\n";
}
if (isNumber(val)) result += " mov rcx, " + val + "\n";
else result += " movsxd rcx, dword " + val + "\n";
}
// ... wewn¹trz case OpType::CALL ...
if (instr.arg1 == "sys_seed") {
result += " mov rcx, 0\n";
result += " call time\n"; // Pobierz czas
result += " mov rcx, rax\n"; // Czas jako argument dla srand
result += " call srand\n"; // Inicjuj generator
break;
}
if (instr.arg1 == "sys_rand") {
result += " call rand\n"; // Wynik w EAX
break;
}
result += " call " + instr.arg1 + "\n";
break;
}
case OpType::DIV: {
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
result += " mov eax, " + op1 + "\n";
result += " cdq\n"; // Rozszerza EAX na EDX:EAX (konieczne przed dzieleniem)
// IDIV nie przyjmuje sta³ej (np. 10), musi byæ rejestr
if (isdigit(op2[0]) || op2[0] == '-') {
result += " mov ecx, " + op2 + "\n";
result += " idiv ecx\n";
case OpType::RETURN: {
std::string val = getVarLocation(instr.arg1, stackMap);
if (!val.empty() && val != ";") {
result += " mov eax, " + val + "\n";
}
else {
result += " idiv dword " + op2 + "\n";
}
result += " mov " + dst + ", eax\n"; // Wynik dzielenia
break;
}
case OpType::MOD: {
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
result += " mov eax, " + op1 + "\n";
result += " cdq\n"; // Rozszerza EAX
if (isdigit(op2[0]) || op2[0] == '-') {
result += " mov ecx, " + op2 + "\n";
result += " idiv ecx\n";
}
else {
result += " idiv dword " + op2 + "\n";
}
result += " mov " + dst + ", edx\n"; // EDX to reszta z dzielenia!
break;
}
case OpType::LOGIC_AND: {
// a && b
// Algorytm: result = (op1 != 0) * (op2 != 0)
// Ale w ASM ³atwiej:
// cmp op1, 0 -> setne al
// cmp op2, 0 -> setne bl
// and al, bl
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
// Sprawdzamy op1
result += " mov eax, " + op1 + "\n";
result += " cmp eax, 0\n";
result += " setne al\n"; // al = 1 jeœli eax != 0
// Sprawdzamy op2 (u¿ywamy ECX jako pomocniczy)
if (isdigit(op2[0])) result += " mov ecx, " + op2 + "\n";
else result += " mov ecx, " + op2 + "\n";
result += " cmp ecx, 0\n";
result += " setne cl\n"; // cl = 1 jeœli ecx != 0
// Logiczne AND na bajtach
result += " and al, cl\n";
// Zapisujemy wynik (rozszerzamy bajt do dword)
result += " movzx eax, al\n";
result += " mov " + dst + ", eax\n";
break;
}
case OpType::LOGIC_OR: {
// a || b
std::string op1 = getVarLocation(instr.arg2, stackMap);
std::string op2 = getVarLocation(instr.arg3, stackMap);
std::string dst = getVarLocation(instr.arg1, stackMap);
result += " mov eax, " + op1 + "\n";
result += " cmp eax, 0\n";
result += " setne al\n";
if (isdigit(op2[0])) result += " mov ecx, " + op2 + "\n";
else result += " mov ecx, " + op2 + "\n";
result += " cmp ecx, 0\n";
result += " setne cl\n";
// Logiczne OR
result += " or al, cl\n";
result += " movzx eax, al\n";
result += " mov " + dst + ", eax\n";
result += " leave\n";
result += " ret\n";
break;
}
}
@@ -390,4 +349,4 @@ std::string generateAssembly(const CompilerState& state) {
}
return result;
}
}