Index: include/clang/Basic/ConvertUTF.h =================================================================== --- include/clang/Basic/ConvertUTF.h (revision 145028) +++ include/clang/Basic/ConvertUTF.h (working copy) @@ -135,11 +135,12 @@ const UTF8** sourceStart, const UTF8* sourceEnd, UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags); -#ifdef CLANG_NEEDS_THESE_ONE_DAY ConversionResult ConvertUTF16toUTF8 ( const UTF16** sourceStart, const UTF16* sourceEnd, UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags); +#ifdef CLANG_NEEDS_THESE_ONE_DAY + ConversionResult ConvertUTF32toUTF8 ( const UTF32** sourceStart, const UTF32* sourceEnd, UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags); Index: lib/Basic/ConvertUTF.c =================================================================== --- lib/Basic/ConvertUTF.c (revision 145028) +++ lib/Basic/ConvertUTF.c (working copy) @@ -218,74 +218,7 @@ #endif return result; } -ConversionResult ConvertUTF16toUTF8 ( - const UTF16** sourceStart, const UTF16* sourceEnd, - UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags) { - ConversionResult result = conversionOK; - const UTF16* source = *sourceStart; - UTF8* target = *targetStart; - while (source < sourceEnd) { - UTF32 ch; - unsigned short bytesToWrite = 0; - const UTF32 byteMask = 0xBF; - const UTF32 byteMark = 0x80; - const UTF16* oldSource = source; /* In case we have to back up because of target overflow. */ - ch = *source++; - /* If we have a surrogate pair, convert to UTF32 first. */ - if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_HIGH_END) { - /* If the 16 bits following the high surrogate are in the source buffer... */ - if (source < sourceEnd) { - UTF32 ch2 = *source; - /* If it's a low surrogate, convert to UTF32. */ - if (ch2 >= UNI_SUR_LOW_START && ch2 <= UNI_SUR_LOW_END) { - ch = ((ch - UNI_SUR_HIGH_START) << halfShift) - + (ch2 - UNI_SUR_LOW_START) + halfBase; - ++source; - } else if (flags == strictConversion) { /* it's an unpaired high surrogate */ - --source; /* return to the illegal value itself */ - result = sourceIllegal; - break; - } - } else { /* We don't have the 16 bits following the high surrogate. */ - --source; /* return to the high surrogate */ - result = sourceExhausted; - break; - } - } else if (flags == strictConversion) { - /* UTF-16 surrogate values are illegal in UTF-32 */ - if (ch >= UNI_SUR_LOW_START && ch <= UNI_SUR_LOW_END) { - --source; /* return to the illegal value itself */ - result = sourceIllegal; - break; - } - } - /* Figure out how many bytes the result will require */ - if (ch < (UTF32)0x80) { bytesToWrite = 1; - } else if (ch < (UTF32)0x800) { bytesToWrite = 2; - } else if (ch < (UTF32)0x10000) { bytesToWrite = 3; - } else if (ch < (UTF32)0x110000) { bytesToWrite = 4; - } else { bytesToWrite = 3; - ch = UNI_REPLACEMENT_CHAR; - } - target += bytesToWrite; - if (target > targetEnd) { - source = oldSource; /* Back up source pointer! */ - target -= bytesToWrite; result = targetExhausted; break; - } - switch (bytesToWrite) { /* note: everything falls through. */ - case 4: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6; - case 3: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6; - case 2: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6; - case 1: *--target = (UTF8)(ch | firstByteMark[bytesToWrite]); - } - target += bytesToWrite; - } - *sourceStart = source; - *targetStart = target; - return result; -} - /* --------------------------------------------------------------------- */ ConversionResult ConvertUTF32toUTF8 ( @@ -546,3 +479,76 @@ similarly unrolled loops. --------------------------------------------------------------------- */ + +ConversionResult ConvertUTF16toUTF8 ( + const UTF16** sourceStart, const UTF16* sourceEnd, + UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags) { + ConversionResult result = conversionOK; + const UTF16* source = *sourceStart; + UTF8* target = *targetStart; + while (source < sourceEnd) { + UTF32 ch; + unsigned short bytesToWrite = 0; + const UTF32 byteMask = 0xBF; + const UTF32 byteMark = 0x80; + const UTF16* oldSource = source; /* In case we have to back up because of target overflow. */ + ch = *source++; + /* If we have a surrogate pair, convert to UTF32 first. */ + if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_HIGH_END) { + /* If the 16 bits following the high surrogate are in the source buffer... */ + if (source < sourceEnd) { + UTF32 ch2 = *source; + /* If it's a low surrogate, convert to UTF32. */ + if (ch2 >= UNI_SUR_LOW_START && ch2 <= UNI_SUR_LOW_END) { + ch = ((ch - UNI_SUR_HIGH_START) << halfShift) + + (ch2 - UNI_SUR_LOW_START) + halfBase; + ++source; + } else if (flags == strictConversion) { /* it's an unpaired high surrogate */ + --source; /* return to the illegal value itself */ + result = sourceIllegal; + break; + } + } else { /* We don't have the 16 bits following the high surrogate. */ + --source; /* return to the high surrogate */ + result = sourceExhausted; + break; + } + } else if (flags == strictConversion) { + /* UTF-16 surrogate values are illegal in UTF-32 */ + if (ch >= UNI_SUR_LOW_START && ch <= UNI_SUR_LOW_END) { + --source; /* return to the illegal value itself */ + result = sourceIllegal; + break; + } + } + /* Figure out how many bytes the result will require */ + if (ch < (UTF32)0x80) + bytesToWrite = 1; + else if (ch < (UTF32)0x800) + bytesToWrite = 2; + else if (ch < (UTF32)0x10000) + bytesToWrite = 3; + else if (ch < (UTF32)0x110000) + bytesToWrite = 4; + else { + bytesToWrite = 3; + ch = UNI_REPLACEMENT_CHAR; + } + + target += bytesToWrite; + if (target > targetEnd) { + source = oldSource; /* Back up source pointer! */ + target -= bytesToWrite; result = targetExhausted; break; + } + switch (bytesToWrite) { /* note: everything falls through. */ + case 4: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6; + case 3: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6; + case 2: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6; + case 1: *--target = (UTF8)(ch | firstByteMark[bytesToWrite]); + } + target += bytesToWrite; + } + *sourceStart = source; + *targetStart = target; + return result; +} Index: lib/Basic/FileManager.cpp =================================================================== --- lib/Basic/FileManager.cpp (revision 145028) +++ lib/Basic/FileManager.cpp (working copy) @@ -17,6 +17,7 @@ // //===----------------------------------------------------------------------===// +#include "clang/Basic/ConvertUTF.h" #include "clang/Basic/FileManager.h" #include "clang/Basic/FileSystemStatCache.h" #include "llvm/ADT/SmallString.h" @@ -26,6 +27,7 @@ #include "llvm/Support/Path.h" #include "llvm/Support/system_error.h" #include "llvm/Config/config.h" +#include #include #include #include @@ -64,14 +66,43 @@ #ifdef LLVM_ON_WIN32 namespace { - static std::string GetFullPath(const char *relPath) { - char *absPathStrPtr = _fullpath(NULL, relPath, 0); - assert(absPathStrPtr && "_fullpath() returned NULL!"); + static std::string GetFullPath(const char *relPath, bool convertToLowercase) { + StringRef String(relPath); + unsigned NumBytes = String.size() + 1; + SmallVector utf16(NumBytes); + const UTF8 *FromUTF8Ptr = (UTF8 *)String.data(); + UTF16 *ToUTF16Ptr = &utf16[0]; - std::string absPath(absPathStrPtr); + ConversionResult Result = ConvertUTF8toUTF16(&FromUTF8Ptr, + FromUTF8Ptr + NumBytes, + &ToUTF16Ptr, + ToUTF16Ptr + NumBytes, + strictConversion); + assert(Result == conversionOK); + + wchar_t *absPath = _wfullpath(NULL, (wchar_t *)utf16.data(), utf16.size()); + assert(absPath && "_wfullpath() returned NULL!"); - free(absPathStrPtr); - return absPath; + unsigned NumWChars = ::wcslen(absPath); + if (convertToLowercase) + { + for (int i = 0; i != NumWChars; ++i) + absPath[i] = std::tolower(absPath[i], std::locale()); + } + + SmallVector utf8(NumBytes); + const UTF16 *FromUTF16Ptr = (UTF16 *)absPath; + UTF8 *ToUTF8Ptr = &utf8[0]; + + Result = ConvertUTF16toUTF8(&FromUTF16Ptr, FromUTF16Ptr + NumWChars, + &ToUTF8Ptr, ToUTF8Ptr + NumBytes, + strictConversion); + + assert(Result == conversionOK); + + free(absPath); + + return std::string((char *)utf8.data()); } } @@ -86,7 +117,7 @@ /// default-constructed DirectoryEntry. DirectoryEntry &getDirectory(const char *Name, const struct stat & /*StatBuf*/) { - std::string FullPath(GetFullPath(Name)); + std::string FullPath(GetFullPath(Name, false)); return UniqueDirs.GetOrCreateValue(FullPath).getValue(); } @@ -103,10 +134,8 @@ /// there is already one; otherwise create and return a /// default-constructed FileEntry. FileEntry &getFile(const char *Name, const struct stat & /*StatBuf*/) { - std::string FullPath(GetFullPath(Name)); - // Lowercase string because Windows filesystem is case insensitive. - FullPath = StringRef(FullPath).lower(); + std::string FullPath(GetFullPath(Name, true)); return UniqueFiles.GetOrCreateValue(FullPath).getValue(); } @@ -561,7 +590,7 @@ llvm::SmallString<128> FilePath(Path); FixupRelativePath(FilePath); - return ::stat(FilePath.c_str(), &StatBuf) != 0; + return llvm::sys::fs::Stat(FilePath.c_str(), &StatBuf) != 0; } void FileManager::GetUniqueIDMapping( Index: lib/Basic/FileSystemStatCache.cpp =================================================================== --- lib/Basic/FileSystemStatCache.cpp (revision 145028) +++ lib/Basic/FileSystemStatCache.cpp (working copy) @@ -13,6 +13,7 @@ #include "clang/Basic/FileSystemStatCache.h" #include "llvm/Support/Path.h" +#include "llvm/Support/FileSystem.h" #include // FIXME: This is terrible, we need this for ::close. @@ -47,7 +48,7 @@ R = Cache->getStat(Path, StatBuf, FileDescriptor); else if (isForDir) { // If this is a directory and we have no cache, just go to the file system. - R = ::stat(Path, &StatBuf) != 0 ? CacheMissing : CacheExists; + R = llvm::sys::fs::Stat(Path, &StatBuf) != 0 ? CacheMissing : CacheExists; } else { // Otherwise, we have to go to the filesystem. We can always just use // 'stat' here, but (for files) the client is asking whether the file exists @@ -60,7 +61,7 @@ #ifdef O_BINARY OpenFlags |= O_BINARY; // Open input file in binary mode on win32. #endif - *FileDescriptor = ::open(Path, OpenFlags); + *FileDescriptor = llvm::sys::fs::Open(Path, OpenFlags); if (*FileDescriptor == -1) { // If the open fails, our "stat" fails. Index: lib/Basic/SourceManager.cpp =================================================================== --- lib/Basic/SourceManager.cpp (revision 145028) +++ lib/Basic/SourceManager.cpp (working copy) @@ -19,6 +19,7 @@ #include "llvm/ADT/Optional.h" #include "llvm/ADT/STLExtras.h" #include "llvm/Support/Compiler.h" +#include "llvm/Support/FileSystem.h" #include "llvm/Support/MemoryBuffer.h" #include "llvm/Support/raw_ostream.h" #include "llvm/Support/Path.h" @@ -1332,7 +1333,7 @@ return llvm::Optional(); struct stat StatBuf; - if (::stat(File->getName(), &StatBuf)) + if (llvm::sys::fs::Stat(File->getName(), &StatBuf)) return llvm::Optional(); return StatBuf.st_ino; Index: lib/Driver/Compilation.cpp =================================================================== --- lib/Driver/Compilation.cpp (revision 145028) +++ lib/Driver/Compilation.cpp (working copy) @@ -17,6 +17,7 @@ #include "clang/Driver/ToolChain.h" #include "llvm/ADT/STLExtras.h" +#include "llvm/Support/FileSystem.h" #include "llvm/Support/raw_ostream.h" #include "llvm/Support/Program.h" #include @@ -124,7 +125,8 @@ // FIXME: Grumble, P.exists() is broken. PR3837. struct stat buf; - if (::stat(P.c_str(), &buf) == 0 ? (buf.st_mode & S_IFMT) == S_IFREG : + if (llvm::sys::fs::Stat(P.c_str(), &buf) == 0 ? + (buf.st_mode & S_IFMT) == S_IFREG : (errno != ENOENT)) { if (IssueErrors) getDriver().Diag(clang::diag::err_drv_unable_to_remove_file) Index: lib/Serialization/ASTReader.cpp =================================================================== --- lib/Serialization/ASTReader.cpp (revision 145028) +++ lib/Serialization/ASTReader.cpp (working copy) @@ -2389,7 +2389,7 @@ // The stat info from the FileEntry came from the cached stat // info of the PCH, so we cannot trust it. struct stat StatBuf; - if (::stat(File->getName(), &StatBuf) != 0) { + if (llvm::sys::fs::Stat(File->getName(), &StatBuf) != 0) { StatBuf.st_size = File->getSize(); StatBuf.st_mtime = File->getModificationTime(); } Index: tools/driver/driver.cpp =================================================================== --- tools/driver/driver.cpp (revision 145028) +++ tools/driver/driver.cpp (working copy) @@ -241,17 +241,17 @@ } } -static void ExpandArgv(int argc, const char **argv, +static void ExpandArgv(const std::vector &argv, SmallVectorImpl &ArgVector, std::set &SavedStrings) { - for (int i = 0; i < argc; ++i) { - const char *Arg = argv[i]; - if (Arg[0] != '@') { - ArgVector.push_back(SaveStringInSet(SavedStrings, std::string(Arg))); + for (std::vector::const_iterator I = argv.begin(), + E = argv.end(); I != E; ++I) { + if (I->front() != '@') { + ArgVector.push_back(SaveStringInSet(SavedStrings, *I)); continue; } - ExpandArgsFromBuf(Arg, ArgVector, SavedStrings); + ExpandArgsFromBuf(I->c_str(), ArgVector, SavedStrings); } } @@ -343,7 +343,7 @@ std::set SavedStrings; SmallVector argv; - ExpandArgv(argc_, argv_, argv, SavedStrings); + ExpandArgv(llvm::sys::fs::GetArgvAsUTF8(argc_, argv_), argv, SavedStrings); // Handle -cc1 integrated tools. if (argv.size() > 1 && StringRef(argv[1]).startswith("-cc1")) {