//===- Support/FileUtilities.cpp - File System Utilities ------------------===//
-//
+//
// The LLVM Compiler Infrastructure
//
-// This file was developed by the LLVM research group and is distributed under
-// the University of Illinois Open Source License. See LICENSE.TXT for details.
-//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
//===----------------------------------------------------------------------===//
//
// This file implements a family of utility functions which are useful for doing
//===----------------------------------------------------------------------===//
#include "llvm/Support/FileUtilities.h"
-#include "llvm/Support/DataTypes.h"
-#include "llvm/Config/unistd.h"
-#include "llvm/Config/fcntl.h"
-#include "llvm/Config/sys/types.h"
-#include "llvm/Config/sys/stat.h"
-#include "llvm/Config/sys/mman.h"
-#include "llvm/Config/alloca.h"
-#include <cerrno>
-#include <cstdio>
-#include <fstream>
-#include <iostream>
+#include "llvm/ADT/OwningPtr.h"
+#include "llvm/ADT/SmallString.h"
+#include "llvm/Support/MemoryBuffer.h"
+#include "llvm/Support/Path.h"
+#include "llvm/Support/raw_ostream.h"
+#include "llvm/Support/system_error.h"
+#include <cctype>
+#include <cstdlib>
+#include <cstring>
using namespace llvm;
-/// CheckMagic - Returns true IFF the file named FN begins with Magic. FN must
-/// name a readable file.
-///
-bool llvm::CheckMagic(const std::string &FN, const std::string &Magic) {
- char *buf = (char*)alloca(1 + Magic.size());
- std::ifstream f(FN.c_str());
- f.read(buf, Magic.size());
- buf[Magic.size()] = '\0';
- return Magic == buf;
-}
-
-/// IsArchive - Returns true IFF the file named FN appears to be a "ar" library
-/// archive. The file named FN must exist.
-///
-bool llvm::IsArchive(const std::string &FN) {
- // Inspect the beginning of the file to see if it contains the "ar"
- // library archive format magic string.
- return CheckMagic(FN, "!<arch>\012");
+static bool isSignedChar(char C) {
+ return (C == '+' || C == '-');
}
-/// IsBytecode - Returns true IFF the file named FN appears to be an LLVM
-/// bytecode file. The file named FN must exist.
-///
-bool llvm::IsBytecode(const std::string &FN) {
- // Inspect the beginning of the file to see if it contains the LLVM
- // bytecode format magic string.
- return CheckMagic(FN, "llvm") || CheckMagic(FN, "llvc");
-}
-
-/// IsSharedObject - Returns trus IFF the file named FN appears to be a shared
-/// object with an ELF header. The file named FN must exist.
-///
-bool llvm::IsSharedObject(const std::string &FN) {
- // Inspect the beginning of the file to see if it contains the ELF shared
- // object magic string.
- static const char elfMagic[] = { 0x7f, 'E', 'L', 'F', '\0' };
- return CheckMagic(FN, elfMagic);
+static bool isExponentChar(char C) {
+ switch (C) {
+ case 'D': // Strange exponential notation.
+ case 'd': // Strange exponential notation.
+ case 'e':
+ case 'E': return true;
+ default: return false;
+ }
}
-/// FileOpenable - Returns true IFF Filename names an existing regular
-/// file which we can successfully open.
-///
-bool llvm::FileOpenable(const std::string &Filename) {
- struct stat s;
- if (stat (Filename.c_str (), &s) == -1)
- return false; // Cannot stat file
- if (!S_ISREG (s.st_mode))
- return false; // File is not a regular file
- std::ifstream FileStream (Filename.c_str ());
- if (!FileStream)
- return false; // File is not openable
- return true;
+static bool isNumberChar(char C) {
+ switch (C) {
+ case '0': case '1': case '2': case '3': case '4':
+ case '5': case '6': case '7': case '8': case '9':
+ case '.': return true;
+ default: return isSignedChar(C) || isExponentChar(C);
+ }
}
-/// DiffFiles - Compare the two files specified, returning true if they are
-/// different or if there is a file error. If you specify a string to fill in
-/// for the error option, it will set the string to an error message if an error
-/// occurs, allowing the caller to distinguish between a failed diff and a file
-/// system error.
-///
-bool llvm::DiffFiles(const std::string &FileA, const std::string &FileB,
- std::string *Error) {
- std::ifstream FileAStream(FileA.c_str());
- if (!FileAStream) {
- if (Error) *Error = "Couldn't open file '" + FileA + "'";
- return true;
- }
+static const char *BackupNumber(const char *Pos, const char *FirstChar) {
+ // If we didn't stop in the middle of a number, don't backup.
+ if (!isNumberChar(*Pos)) return Pos;
+
+ // Otherwise, return to the start of the number.
+ bool HasPeriod = false;
+ while (Pos > FirstChar && isNumberChar(Pos[-1])) {
+ // Backup over at most one period.
+ if (Pos[-1] == '.') {
+ if (HasPeriod)
+ break;
+ HasPeriod = true;
+ }
- std::ifstream FileBStream(FileB.c_str());
- if (!FileBStream) {
- if (Error) *Error = "Couldn't open file '" + FileB + "'";
- return true;
+ --Pos;
+ if (Pos > FirstChar && isSignedChar(Pos[0]) && !isExponentChar(Pos[-1]))
+ break;
}
+ return Pos;
+}
- // Compare the two files...
- int C1, C2;
- do {
- C1 = FileAStream.get();
- C2 = FileBStream.get();
- if (C1 != C2) return true;
- } while (C1 != EOF);
-
- return false;
+/// EndOfNumber - Return the first character that is not part of the specified
+/// number. This assumes that the buffer is null terminated, so it won't fall
+/// off the end.
+static const char *EndOfNumber(const char *Pos) {
+ while (isNumberChar(*Pos))
+ ++Pos;
+ return Pos;
}
+/// CompareNumbers - compare two numbers, returning true if they are different.
+static bool CompareNumbers(const char *&F1P, const char *&F2P,
+ const char *F1End, const char *F2End,
+ double AbsTolerance, double RelTolerance,
+ std::string *ErrorMsg) {
+ const char *F1NumEnd, *F2NumEnd;
+ double V1 = 0.0, V2 = 0.0;
+
+ // If one of the positions is at a space and the other isn't, chomp up 'til
+ // the end of the space.
+ while (isspace(static_cast<unsigned char>(*F1P)) && F1P != F1End)
+ ++F1P;
+ while (isspace(static_cast<unsigned char>(*F2P)) && F2P != F2End)
+ ++F2P;
+
+ // If we stop on numbers, compare their difference.
+ if (!isNumberChar(*F1P) || !isNumberChar(*F2P)) {
+ // The diff failed.
+ F1NumEnd = F1P;
+ F2NumEnd = F2P;
+ } else {
+ // Note that some ugliness is built into this to permit support for numbers
+ // that use "D" or "d" as their exponential marker, e.g. "1.234D45". This
+ // occurs in 200.sixtrack in spec2k.
+ V1 = strtod(F1P, const_cast<char**>(&F1NumEnd));
+ V2 = strtod(F2P, const_cast<char**>(&F2NumEnd));
+
+ if (*F1NumEnd == 'D' || *F1NumEnd == 'd') {
+ // Copy string into tmp buffer to replace the 'D' with an 'e'.
+ SmallString<200> StrTmp(F1P, EndOfNumber(F1NumEnd)+1);
+ // Strange exponential notation!
+ StrTmp[static_cast<unsigned>(F1NumEnd-F1P)] = 'e';
+
+ V1 = strtod(&StrTmp[0], const_cast<char**>(&F1NumEnd));
+ F1NumEnd = F1P + (F1NumEnd-&StrTmp[0]);
+ }
-/// CopyFile - Copy the specified source file to the specified destination,
-/// overwriting destination if it exists. This returns true on failure.
-///
-bool llvm::CopyFile(const std::string &Dest, const std::string &Src) {
- FDHandle InFD(open(Src.c_str(), O_RDONLY));
- if (InFD == -1) return true;
+ if (*F2NumEnd == 'D' || *F2NumEnd == 'd') {
+ // Copy string into tmp buffer to replace the 'D' with an 'e'.
+ SmallString<200> StrTmp(F2P, EndOfNumber(F2NumEnd)+1);
+ // Strange exponential notation!
+ StrTmp[static_cast<unsigned>(F2NumEnd-F2P)] = 'e';
- FileRemover FR(Dest);
+ V2 = strtod(&StrTmp[0], const_cast<char**>(&F2NumEnd));
+ F2NumEnd = F2P + (F2NumEnd-&StrTmp[0]);
+ }
+ }
- FDHandle OutFD(open(Dest.c_str(), O_WRONLY|O_CREAT, 0666));
- if (OutFD == -1) return true;
+ if (F1NumEnd == F1P || F2NumEnd == F2P) {
+ if (ErrorMsg) {
+ *ErrorMsg = "FP Comparison failed, not a numeric difference between '";
+ *ErrorMsg += F1P[0];
+ *ErrorMsg += "' and '";
+ *ErrorMsg += F2P[0];
+ *ErrorMsg += "'";
+ }
+ return true;
+ }
- char Buffer[16*1024];
- while (ssize_t Amt = read(InFD, Buffer, 16*1024)) {
- if (Amt == -1) {
- if (errno != EINTR) return true; // Error reading the file.
- } else {
- char *BufPtr = Buffer;
- while (Amt) {
- ssize_t AmtWritten = write(OutFD, BufPtr, Amt);
- if (AmtWritten == -1) {
- if (errno != EINTR) return true; // Error writing the file.
- } else {
- Amt -= AmtWritten;
- BufPtr += AmtWritten;
- }
+ // Check to see if these are inside the absolute tolerance
+ if (AbsTolerance < std::abs(V1-V2)) {
+ // Nope, check the relative tolerance...
+ double Diff;
+ if (V2)
+ Diff = std::abs(V1/V2 - 1.0);
+ else if (V1)
+ Diff = std::abs(V2/V1 - 1.0);
+ else
+ Diff = 0; // Both zero.
+ if (Diff > RelTolerance) {
+ if (ErrorMsg) {
+ raw_string_ostream(*ErrorMsg)
+ << "Compared: " << V1 << " and " << V2 << '\n'
+ << "abs. diff = " << std::abs(V1-V2) << " rel.diff = " << Diff << '\n'
+ << "Out of tolerance: rel/abs: " << RelTolerance << '/'
+ << AbsTolerance;
}
+ return true;
}
}
- FR.releaseFile(); // Success!
+ // Otherwise, advance our read pointers to the end of the numbers.
+ F1P = F1NumEnd; F2P = F2NumEnd;
return false;
}
-
-/// MoveFileOverIfUpdated - If the file specified by New is different than Old,
-/// or if Old does not exist, move the New file over the Old file. Otherwise,
-/// remove the New file.
-///
-void llvm::MoveFileOverIfUpdated(const std::string &New,
- const std::string &Old) {
- if (DiffFiles(New, Old)) {
- if (std::rename(New.c_str(), Old.c_str()))
- std::cerr << "Error renaming '" << New << "' to '" << Old << "'!\n";
- } else {
- std::remove(New.c_str());
- }
-}
-
-/// removeFile - Delete the specified file
-///
-void llvm::removeFile(const std::string &Filename) {
- std::remove(Filename.c_str());
-}
-
-/// getUniqueFilename - Return a filename with the specified prefix. If the
-/// file does not exist yet, return it, otherwise add a suffix to make it
-/// unique.
+/// DiffFilesWithTolerance - Compare the two files specified, returning 0 if the
+/// files match, 1 if they are different, and 2 if there is a file error. This
+/// function differs from DiffFiles in that you can specify an absolete and
+/// relative FP error that is allowed to exist. If you specify a string to fill
+/// in for the error option, it will set the string to an error message if an
+/// error occurs, allowing the caller to distinguish between a failed diff and a
+/// file system error.
///
-std::string llvm::getUniqueFilename(const std::string &FilenameBase) {
- if (!std::ifstream(FilenameBase.c_str()))
- return FilenameBase; // Couldn't open the file? Use it!
-
- // Create a pattern for mkstemp...
- char *FNBuffer = new char[FilenameBase.size()+8];
- strcpy(FNBuffer, FilenameBase.c_str());
- strcpy(FNBuffer+FilenameBase.size(), "-XXXXXX");
-
- // Agree on a temporary file name to use....
-#if defined(HAVE_MKSTEMP) && !defined(_MSC_VER)
- int TempFD;
- if ((TempFD = mkstemp(FNBuffer)) == -1) {
- // FIXME: this should return an emtpy string or something and allow the
- // caller to deal with the error!
- std::cerr << "bugpoint: ERROR: Cannot create temporary file in the current "
- << " directory!\n";
- exit(1);
+int llvm::DiffFilesWithTolerance(StringRef NameA,
+ StringRef NameB,
+ double AbsTol, double RelTol,
+ std::string *Error) {
+ // Now its safe to mmap the files into memory because both files
+ // have a non-zero size.
+ OwningPtr<MemoryBuffer> F1;
+ if (error_code ec = MemoryBuffer::getFile(NameA, F1)) {
+ if (Error)
+ *Error = ec.message();
+ return 2;
}
-
- // We don't need to hold the temp file descriptor... we will trust that no one
- // will overwrite/delete the file while we are working on it...
- close(TempFD);
-#else
- // If we don't have mkstemp, use the old and obsolete mktemp function.
- if (mktemp(FNBuffer) == 0) {
- // FIXME: this should return an emtpy string or something and allow the
- // caller to deal with the error!
- std::cerr << "bugpoint: ERROR: Cannot create temporary file in the current "
- << " directory!\n";
- exit(1);
+ OwningPtr<MemoryBuffer> F2;
+ if (error_code ec = MemoryBuffer::getFile(NameB, F2)) {
+ if (Error)
+ *Error = ec.message();
+ return 2;
}
-#endif
-
- std::string Result(FNBuffer);
- delete[] FNBuffer;
- return Result;
-}
-
-static bool AddPermissionsBits (const std::string &Filename, int bits) {
- // Get the umask value from the operating system. We want to use it
- // when changing the file's permissions. Since calling umask() sets
- // the umask and returns its old value, we must call it a second
- // time to reset it to the user's preference.
- int mask = umask(0777); // The arg. to umask is arbitrary.
- umask(mask); // Restore the umask.
-
- // Get the file's current mode.
- struct stat st;
- if ((stat(Filename.c_str(), &st)) == -1)
- return false;
-
- // Change the file to have whichever permissions bits from 'bits'
- // that the umask would not disable.
- if ((chmod(Filename.c_str(), (st.st_mode | (bits & ~mask)))) == -1)
- return false;
-
- return true;
-}
-
-/// MakeFileExecutable - Make the file named Filename executable by
-/// setting whichever execute permissions bits the process's current
-/// umask would allow. Filename must name an existing file or
-/// directory. Returns true on success, false on error.
-///
-bool llvm::MakeFileExecutable(const std::string &Filename) {
- return AddPermissionsBits(Filename, 0111);
-}
-
-/// MakeFileReadable - Make the file named Filename readable by
-/// setting whichever read permissions bits the process's current
-/// umask would allow. Filename must name an existing file or
-/// directory. Returns true on success, false on error.
-///
-bool llvm::MakeFileReadable(const std::string &Filename) {
- return AddPermissionsBits(Filename, 0444);
-}
-
-/// getFileSize - Return the size of the specified file in bytes, or -1 if the
-/// file cannot be read or does not exist.
-long long llvm::getFileSize(const std::string &Filename) {
- struct stat StatBuf;
- if (stat(Filename.c_str(), &StatBuf) == -1)
- return -1;
- return StatBuf.st_size;
-}
-/// getFileTimestamp - Get the last modified time for the specified file in an
-/// unspecified format. This is useful to allow checking to see if a file was
-/// updated since that last time the timestampt was aquired. If the file does
-/// not exist or there is an error getting the time-stamp, zero is returned.
-unsigned long long llvm::getFileTimestamp(const std::string &Filename) {
- struct stat StatBuf;
- if (stat(Filename.c_str(), &StatBuf) == -1)
+ // Okay, now that we opened the files, scan them for the first difference.
+ const char *File1Start = F1->getBufferStart();
+ const char *File2Start = F2->getBufferStart();
+ const char *File1End = F1->getBufferEnd();
+ const char *File2End = F2->getBufferEnd();
+ const char *F1P = File1Start;
+ const char *F2P = File2Start;
+ uint64_t A_size = F1->getBufferSize();
+ uint64_t B_size = F2->getBufferSize();
+
+ // Are the buffers identical? Common case: Handle this efficiently.
+ if (A_size == B_size &&
+ std::memcmp(File1Start, File2Start, A_size) == 0)
return 0;
- return StatBuf.st_mtime;
-}
-
-/// ReadFileIntoAddressSpace - Attempt to map the specific file into the
-/// address space of the current process for reading. If this succeeds,
-/// return the address of the buffer and the length of the file mapped. On
-/// failure, return null.
-void *llvm::ReadFileIntoAddressSpace(const std::string &Filename,
- unsigned &Length) {
-#if defined(HAVE_MMAP_FILE) && !defined(_MSC_VER)
- Length = (unsigned)getFileSize(Filename);
- if ((int)Length == -1) return 0;
-
- FDHandle FD(open(Filename.c_str(), O_RDONLY));
- if (FD == -1) return 0;
-
- // If the file has a length of zero, mmap might return a null pointer. In
- // this case, allocate a single byte of memory and return it instead.
- if (Length == 0)
- return malloc(1);
-
- // mmap in the file all at once...
- void *Buffer = (void*)mmap(0, Length, PROT_READ, MAP_PRIVATE, FD, 0);
-
- if (Buffer == (void*)MAP_FAILED)
- return 0;
-
- return Buffer;
-#else
- // FIXME: implement with read/write
-#error Unimplemented ReadFileIntoAddressSpace - need to use read/write.
- return 0;
-#endif
-}
-/// UnmapFileFromAddressSpace - Remove the specified file from the current
-/// address space.
-void llvm::UnmapFileFromAddressSpace(void *Buffer, unsigned Length) {
-#if defined(HAVE_MMAP_FILE) && !defined(_MSC_VER)
- if (Length)
- munmap((char*)Buffer, Length);
- else
- free(Buffer); // Zero byte files are malloc(1)'s.
-#else
- free(Buffer);
-#endif
-}
+ // Otherwise, we are done a tolerances are set.
+ if (AbsTol == 0 && RelTol == 0) {
+ if (Error)
+ *Error = "Files differ without tolerance allowance";
+ return 1; // Files different!
+ }
-//===----------------------------------------------------------------------===//
-// FDHandle class implementation
-//
+ bool CompareFailed = false;
+ while (1) {
+ // Scan for the end of file or next difference.
+ while (F1P < File1End && F2P < File2End && *F1P == *F2P)
+ ++F1P, ++F2P;
+
+ if (F1P >= File1End || F2P >= File2End) break;
+
+ // Okay, we must have found a difference. Backup to the start of the
+ // current number each stream is at so that we can compare from the
+ // beginning.
+ F1P = BackupNumber(F1P, File1Start);
+ F2P = BackupNumber(F2P, File2Start);
+
+ // Now that we are at the start of the numbers, compare them, exiting if
+ // they don't match.
+ if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error)) {
+ CompareFailed = true;
+ break;
+ }
+ }
-FDHandle::~FDHandle() throw() {
- if (FD != -1) close(FD);
-}
+ // Okay, we reached the end of file. If both files are at the end, we
+ // succeeded.
+ bool F1AtEnd = F1P >= File1End;
+ bool F2AtEnd = F2P >= File2End;
+ if (!CompareFailed && (!F1AtEnd || !F2AtEnd)) {
+ // Else, we might have run off the end due to a number: backup and retry.
+ if (F1AtEnd && isNumberChar(F1P[-1])) --F1P;
+ if (F2AtEnd && isNumberChar(F2P[-1])) --F2P;
+ F1P = BackupNumber(F1P, File1Start);
+ F2P = BackupNumber(F2P, File2Start);
+
+ // Now that we are at the start of the numbers, compare them, exiting if
+ // they don't match.
+ if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error))
+ CompareFailed = true;
+
+ // If we found the end, we succeeded.
+ if (F1P < File1End || F2P < File2End)
+ CompareFailed = true;
+ }
-FDHandle &FDHandle::operator=(int fd) throw() {
- if (FD != -1) close(FD);
- FD = fd;
- return *this;
+ return CompareFailed;
}
-