From 62b605c6c0ea13e9763c0a0115464007b13a8831 Mon Sep 17 00:00:00 2001 From: _ov4 Date: Tue, 20 Oct 2026 11:57:16 +0000 Subject: refactor(strings): replace compiler builtins with the STRING library Construct literals as STRING values and route string operations through ordinary library functions and operators. Preserve value ownership in records, arrays, constants and file I/O. Co-Authored-By: Claude Opus 5.5 --- include/cps/CodeGen.h | 16 ++++++------- include/cps/FileHandler.h | 2 +- include/cps/StringConversionHandler.h | 33 --------------------------- include/cps/StringHandler.h | 42 ----------------------------------- include/cps/TypeSystem.h | 5 +++-- 5 files changed, 12 insertions(+), 86 deletions(-) delete mode 100644 include/cps/StringConversionHandler.h delete mode 100644 include/cps/StringHandler.h (limited to 'include') diff --git a/include/cps/CodeGen.h b/include/cps/CodeGen.h index 30f4135..e6281c5 100644 --- a/include/cps/CodeGen.h +++ b/include/cps/CodeGen.h @@ -11,8 +11,6 @@ #include "cps/TypeSystem.h" #include "cps/ArithmeticHandler.h" -#include "cps/StringHandler.h" -#include "cps/StringConversionHandler.h" #include #include @@ -93,6 +91,12 @@ class CodeGen { bool prepareSymbolDeclaration(const std::string &Name, bool IsConstant = false); bool requireWritable(const SymbolInfo &Info, const std::string &Name); void emitConstantStmt(ConstantStmtAST *Stmt); + std::map StringConstants; + const TypeInfo *requireStringType(); + llvm::Value *emitStringFromCString(llvm::Value *Text); + llvm::Value *emitStringCString(llvm::Value *Value); + llvm::Value *emitCStringExpr(ExprAST *Expr); + bool containsString(const TypeInfo *Type) const; std::unique_ptr Types; std::unique_ptr Arrays; @@ -101,14 +105,11 @@ class CodeGen { std::unique_ptr FuncGen; std::unique_ptr ArithHandler; - std::unique_ptr StrHandler; - std::unique_ptr StrConvHandler; std::unique_ptr Files; std::unique_ptr Memory; llvm::FunctionCallee PrintfFunc; llvm::FunctionCallee ScanfFunc; - llvm::FunctionCallee MallocFunc; llvm::Value *PrintfFormatStr; // %lld\n llvm::Value *PrintfFloatFormatStr; // %f\n @@ -117,7 +118,6 @@ class CodeGen { llvm::Value *ScanfFormatStr; // %lld llvm::Value *ScanfFloatFormatStr; // %lf - llvm::Value *ScanfStringFormatStr; // %s llvm::Value *TrueStr; // "TRUE" llvm::Value *FalseStr; // "FALSE" @@ -157,7 +157,7 @@ class CodeGen { // Resolves a variable / array element / field chain / deref to an address // (emitting index and null-pointer checks along the way). bool emitLValue(ExprAST *E, LValueInfo &Out); - // The rvalue read: load + the STRING null->"" guard. + // Read a value from resolved storage. llvm::Value *loadFromLValue(const LValueInfo &LV); llvm::Value *emitAllocationSize(ExprAST *Expr); llvm::Value *emitCall(const std::string &Callee, @@ -209,7 +209,7 @@ public: llvm::Value *emitExpr(ExprAST *Expr); void emitStmt(StmtAST *Stmt); - // STRING slots zero-init to null; normalize null to "" at load/print sites. + // Normalize a raw C string pointer to "" when explicitly casting it to STRING. llvm::Value *emitStringNullGuard(llvm::Value *StrVal); // Name-level admission gate for Enum/Record/Pointer plus the existing diff --git a/include/cps/FileHandler.h b/include/cps/FileHandler.h index 50593de..20ece16 100644 --- a/include/cps/FileHandler.h +++ b/include/cps/FileHandler.h @@ -29,7 +29,7 @@ class FileHandler { // libc — declared lazily on first file use llvm::FunctionCallee FopenFunc, FcloseFunc, FgetcFunc, UngetcFunc, FputsFunc, FputcFunc, MallocFunc, ReallocFunc, - StrcmpFunc; + StrcmpFunc, StrlenFunc, MemcpyFunc, FreeFunc; // open-file table llvm::GlobalVariable *FileNames = nullptr; // [TableSize x ptr] STRING passed to OPENFILE diff --git a/include/cps/StringConversionHandler.h b/include/cps/StringConversionHandler.h deleted file mode 100644 index c4879aa..0000000 --- a/include/cps/StringConversionHandler.h +++ /dev/null @@ -1,33 +0,0 @@ -#pragma once -#include "llvm/IR/IRBuilder.h" -#include "llvm/IR/LLVMContext.h" -#include "llvm/IR/Module.h" - -namespace cps { - -class StringConversionHandler { - llvm::LLVMContext &Context; - llvm::IRBuilder<> &Builder; - llvm::Module &Module; - - llvm::FunctionCallee SnprintfFunc; - llvm::FunctionCallee StrtollFunc; - llvm::FunctionCallee StrtodFunc; - llvm::FunctionCallee MallocFunc; - -public: - StringConversionHandler(llvm::LLVMContext &Ctx, llvm::IRBuilder<> &B, llvm::Module &M); - - void setupExternalFunctions(); - - // NUM TO STR(x) RETURNS STRING - llvm::Value *emitNumToStr(llvm::Value *Num, bool IsReal); - - // STR TO NUM(x) RETURNS NUMBER - llvm::Value *emitStrToNum(llvm::Value *Str, bool AsReal); - - // IS NUM(ThisString) RETURNS BOOLEAN - llvm::Value *emitIsNum(llvm::Value *Str); -}; - -} // namespace cps diff --git a/include/cps/StringHandler.h b/include/cps/StringHandler.h deleted file mode 100644 index 0b2447c..0000000 --- a/include/cps/StringHandler.h +++ /dev/null @@ -1,42 +0,0 @@ -#pragma once -#include "llvm/IR/IRBuilder.h" -#include "llvm/IR/LLVMContext.h" -#include "llvm/IR/Module.h" -#include - -namespace cps { - -class StringHandler { - llvm::LLVMContext &Context; - llvm::IRBuilder<> &Builder; - llvm::Module &Module; - - llvm::FunctionCallee MallocFunc; - llvm::FunctionCallee StrLenFunc; - llvm::FunctionCallee MemCpyFunc; - llvm::FunctionCallee StrCpyFunc; - llvm::FunctionCallee StrCatFunc; - llvm::FunctionCallee StrCmpFunc; - llvm::FunctionCallee ToUpperFunc; - llvm::FunctionCallee ToLowerFunc; - - llvm::Value *emitCaseConvert(llvm::Value *Str, llvm::FunctionCallee CaseFn); - llvm::Value *allocCopySubstring(llvm::Value *Src, llvm::Value *StartIdx, llvm::Value *Len); - -public: - StringHandler(llvm::LLVMContext &Ctx, llvm::IRBuilder<> &B, llvm::Module &M); - - void setupExternalFunctions(); - llvm::Value *createLiteral(const std::string &Val); - - llvm::Value *emitLength(llvm::Value *Str); - llvm::Value *emitConcat(llvm::Value *LHS, llvm::Value *RHS); - llvm::Value *emitCompare(int Op, llvm::Value *LHS, llvm::Value *RHS); - llvm::Value *emitMid(llvm::Value *Str, llvm::Value *Start, llvm::Value *Len); - llvm::Value *emitRight(llvm::Value *Str, llvm::Value *Len); - llvm::Value *emitLeft(llvm::Value *Str, llvm::Value *Len); - llvm::Value *emitLCase(llvm::Value *Str); - llvm::Value *emitUCase(llvm::Value *Str); -}; - -} diff --git a/include/cps/TypeSystem.h b/include/cps/TypeSystem.h index 1391345..e4663eb 100644 --- a/include/cps/TypeSystem.h +++ b/include/cps/TypeSystem.h @@ -26,7 +26,6 @@ enum class TypeKind { Integer, Real, Boolean, - String, Char, Void, Enum, @@ -109,10 +108,12 @@ struct SymbolInfo { llvm::Value *Storage = nullptr; std::string TypeName; bool IsArray = false; - llvm::Constant *ConstantValue = nullptr; // Non-null for CONSTANT; no storage. + llvm::Constant *ConstantValue = nullptr; // Compile-time scalar CONSTANT value, if available. unsigned ScopeID = 0; bool IsReference = false; // Storage is a slot containing the current object address. bool IsGlobal = false; + bool ReadOnly = false; // Owned STRING constants use storage instead of ConstantValue. + bool isConstant() const { return ConstantValue || ReadOnly; } }; class TypeSystem { -- cgit v1.3.1-4-g156e