Index: include/clang/Basic/ConvertUTFWrapper.h =================================================================== --- include/clang/Basic/ConvertUTFWrapper.h (revision 0) +++ include/clang/Basic/ConvertUTFWrapper.h (revision 0) @@ -0,0 +1,27 @@ +//===-- ConvertUTFWrapper.h - Wrap ConvertUTF.h with clang data types -------=== +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_CLANG_BASIC_CONVERT_UTF_WRAPPER_H +#define LLVM_CLANG_BASIC_CONVERT_UTF_WRAPPER_H + +#include "llvm/ADT/StringRef.h" + +namespace clang { + +/// Convert an UTF8 StringRef to UTF8, UTF16, or UTF32 depending on +/// WideCharWidth. The converted data is written to ResultPtr, which needs to +/// point to at least WideCharWidth * (Source.Size() + 1) bytes. On success, +/// ResultPtr will point one after the end of the copied string. +/// \returns true on success. +bool ConvertUTF8toWide(unsigned WideCharWidth, llvm::StringRef Source, + char *&ResultPtr); + +} + +#endif // LLVM_CLANG_BASIC_CONVERT_UTF_WRAPPER_H Index: lib/Basic/ConvertUTFWrapper.cpp =================================================================== --- lib/Basic/ConvertUTFWrapper.cpp (revision 0) +++ lib/Basic/ConvertUTFWrapper.cpp (revision 0) @@ -0,0 +1,55 @@ +//===-- ConvertUTFWrapper.cpp - Wrap ConvertUTF.h with clang data types -----=== +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "clang/Basic/ConvertUTFWrapper.h" +#include "clang/Basic/ConvertUTF.h" +#include "clang/Basic/LLVM.h" + +namespace clang { + +bool ConvertUTF8toWide(unsigned WideCharWidth, llvm::StringRef Source, + char *&ResultPtr) { + assert(WideCharWidth == 1 || WideCharWidth == 2 || WideCharWidth == 4); + ConversionResult result = conversionOK; + // Copy the character span over. + if (WideCharWidth == 1) { + if (!isLegalUTF8String(reinterpret_cast(Source.begin()), + reinterpret_cast(Source.end()))) + result = sourceIllegal; + memcpy(ResultPtr, Source.data(), Source.size()); + ResultPtr += Source.size(); + } else if (WideCharWidth == 2) { + const UTF8 *sourceStart = (const UTF8*)Source.data(); + // FIXME: Make the type of the result buffer correct instead of + // using reinterpret_cast. + UTF16 *targetStart = reinterpret_cast(ResultPtr); + ConversionFlags flags = strictConversion; + result = ConvertUTF8toUTF16( + &sourceStart, sourceStart + Source.size(), + &targetStart, targetStart + 2*Source.size(), flags); + if (result == conversionOK) + ResultPtr = reinterpret_cast(targetStart); + } else if (WideCharWidth == 4) { + const UTF8 *sourceStart = (const UTF8*)Source.data(); + // FIXME: Make the type of the result buffer correct instead of + // using reinterpret_cast. + UTF32 *targetStart = reinterpret_cast(ResultPtr); + ConversionFlags flags = strictConversion; + result = ConvertUTF8toUTF32( + &sourceStart, sourceStart + Source.size(), + &targetStart, targetStart + 4*Source.size(), flags); + if (result == conversionOK) + ResultPtr = reinterpret_cast(targetStart); + } + assert((result != targetExhausted) + && "ConvertUTF8toUTFXX exhausted target buffer"); + return result == conversionOK; +} + +} Index: lib/Basic/CMakeLists.txt =================================================================== --- lib/Basic/CMakeLists.txt (revision 159393) +++ lib/Basic/CMakeLists.txt (working copy) @@ -3,6 +3,7 @@ add_clang_library(clangBasic Builtins.cpp ConvertUTF.c + ConvertUTFWrapper.cpp Diagnostic.cpp DiagnosticIDs.cpp FileManager.cpp Index: lib/Lex/LiteralSupport.cpp =================================================================== --- lib/Lex/LiteralSupport.cpp (revision 159393) +++ lib/Lex/LiteralSupport.cpp (working copy) @@ -17,6 +17,7 @@ #include "clang/Lex/LexDiagnostic.h" #include "clang/Basic/TargetInfo.h" #include "clang/Basic/ConvertUTF.h" +#include "clang/Basic/ConvertUTFWrapper.h" #include "llvm/ADT/StringExtras.h" #include "llvm/Support/ErrorHandling.h" using namespace clang; @@ -1330,45 +1331,10 @@ } } - /// copyStringFragment - This function copies from Start to End into ResultPtr. /// Performs widening for multi-byte characters. bool StringLiteralParser::CopyStringFragment(StringRef Fragment) { - assert(CharByteWidth==1 || CharByteWidth==2 || CharByteWidth==4); - ConversionResult result = conversionOK; - // Copy the character span over. - if (CharByteWidth == 1) { - if (!isLegalUTF8String(reinterpret_cast(Fragment.begin()), - reinterpret_cast(Fragment.end()))) - result = sourceIllegal; - memcpy(ResultPtr, Fragment.data(), Fragment.size()); - ResultPtr += Fragment.size(); - } else if (CharByteWidth == 2) { - UTF8 const *sourceStart = (UTF8 const *)Fragment.data(); - // FIXME: Make the type of the result buffer correct instead of - // using reinterpret_cast. - UTF16 *targetStart = reinterpret_cast(ResultPtr); - ConversionFlags flags = strictConversion; - result = ConvertUTF8toUTF16( - &sourceStart,sourceStart + Fragment.size(), - &targetStart,targetStart + 2*Fragment.size(),flags); - if (result==conversionOK) - ResultPtr = reinterpret_cast(targetStart); - } else if (CharByteWidth == 4) { - UTF8 const *sourceStart = (UTF8 const *)Fragment.data(); - // FIXME: Make the type of the result buffer correct instead of - // using reinterpret_cast. - UTF32 *targetStart = reinterpret_cast(ResultPtr); - ConversionFlags flags = strictConversion; - result = ConvertUTF8toUTF32( - &sourceStart,sourceStart + Fragment.size(), - &targetStart,targetStart + 4*Fragment.size(),flags); - if (result==conversionOK) - ResultPtr = reinterpret_cast(targetStart); - } - assert((result != targetExhausted) - && "ConvertUTF8toUTFXX exhausted target buffer"); - return result != conversionOK; + return !ConvertUTF8toWide(CharByteWidth, Fragment, ResultPtr); } bool StringLiteralParser::DiagnoseBadString(const Token &Tok) { Index: lib/CodeGen/CGExpr.cpp =================================================================== --- lib/CodeGen/CGExpr.cpp (revision 159393) +++ lib/CodeGen/CGExpr.cpp (working copy) @@ -21,7 +21,7 @@ #include "TargetInfo.h" #include "clang/AST/ASTContext.h" #include "clang/AST/DeclObjC.h" -#include "clang/Basic/ConvertUTF.h" +#include "clang/Basic/ConvertUTFWrapper.h" #include "clang/Frontend/CodeGenOptions.h" #include "llvm/Intrinsics.h" #include "llvm/LLVMContext.h" @@ -1729,47 +1729,12 @@ return GV; } -// FIXME: Mostly copied from StringLiteralParser::CopyStringFragment static void ConvertUTF8ToWideString(unsigned CharByteWidth, StringRef Source, SmallString<32>& Target) { Target.resize(CharByteWidth * (Source.size() + 1)); char* ResultPtr = &Target[0]; - - assert(CharByteWidth==1 || CharByteWidth==2 || CharByteWidth==4); - ConversionResult result = conversionOK; - // Copy the character span over. - if (CharByteWidth == 1) { - if (!isLegalUTF8String(reinterpret_cast(&*Source.begin()), - reinterpret_cast(&*Source.end()))) - result = sourceIllegal; - memcpy(ResultPtr, Source.data(), Source.size()); - ResultPtr += Source.size(); - } else if (CharByteWidth == 2) { - UTF8 const *sourceStart = (UTF8 const *)Source.data(); - // FIXME: Make the type of the result buffer correct instead of - // using reinterpret_cast. - UTF16 *targetStart = reinterpret_cast(ResultPtr); - ConversionFlags flags = strictConversion; - result = ConvertUTF8toUTF16( - &sourceStart,sourceStart + Source.size(), - &targetStart,targetStart + 2*Source.size(),flags); - if (result==conversionOK) - ResultPtr = reinterpret_cast(targetStart); - } else if (CharByteWidth == 4) { - UTF8 const *sourceStart = (UTF8 const *)Source.data(); - // FIXME: Make the type of the result buffer correct instead of - // using reinterpret_cast. - UTF32 *targetStart = reinterpret_cast(ResultPtr); - ConversionFlags flags = strictConversion; - result = ConvertUTF8toUTF32( - &sourceStart,sourceStart + Source.size(), - &targetStart,targetStart + 4*Source.size(),flags); - if (result==conversionOK) - ResultPtr = reinterpret_cast(targetStart); - } - assert((result != targetExhausted) - && "ConvertUTF8toUTFXX exhausted target buffer"); - assert(result == conversionOK); + bool success = ConvertUTF8toWide(CharByteWidth, Source, ResultPtr); + assert(success); Target.resize(ResultPtr - &Target[0]); }