llvm/lib/Support/FileUtilities.cpp

   1 //===- Support/FileUtilities.cpp - File System Utilities ------------------===//
   2 //
   3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
   4 // See https://llvm.org/LICENSE.txt for license information.
   5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
   6 //
   7 //===----------------------------------------------------------------------===//
   8 //
   9 // This file implements a family of utility functions which are useful for doing
  10 // various things with files.
  11 //
  12 //===----------------------------------------------------------------------===//
  13
  14 #include "llvm/Support/FileUtilities.h"
  15 #include "llvm/ADT/ScopeExit.h"
  16 #include "llvm/ADT/SmallString.h"
  17 #include "llvm/ADT/StringExtras.h"
  18 #include "llvm/Support/Error.h"
  19 #include "llvm/Support/ErrorOr.h"
  20 #include "llvm/Support/MemoryBuffer.h"
  21 #include "llvm/Support/Path.h"
  22 #include "llvm/Support/raw_ostream.h"
  23 #include <cctype>
  24 #include <cmath>
  25 #include <cstdint>
  26 #include <cstdlib>
  27 #include <cstring>
  28 #include <memory>
  29 #include <system_error>
  30
  31 using namespace llvm;
  32
  33 static bool isSignedChar(char C) {
  34   return (C == '+' || C == '-');
  35 }
  36
  37 static bool isExponentChar(char C) {
  38   switch (C) {
  39   case 'D':  // Strange exponential notation.
  40   case 'd':  // Strange exponential notation.
  41   case 'e':
  42   case 'E': return true;
  43   default: return false;
  44   }
  45 }
  46
  47 static bool isNumberChar(char C) {
  48   switch (C) {
  49   case '0': case '1': case '2': case '3': case '4':
  50   case '5': case '6': case '7': case '8': case '9':
  51   case '.': return true;
  52   default: return isSignedChar(C) || isExponentChar(C);
  53   }
  54 }
  55
  56 static const char *BackupNumber(const char *Pos, const char *FirstChar) {
  57   // If we didn't stop in the middle of a number, don't backup.
  58   if (!isNumberChar(*Pos)) return Pos;
  59
  60   // Otherwise, return to the start of the number.
  61   bool HasPeriod = false;
  62   while (Pos > FirstChar && isNumberChar(Pos[-1])) {
  63     // Backup over at most one period.
  64     if (Pos[-1] == '.') {
  65       if (HasPeriod)
  66         break;
  67       HasPeriod = true;
  68     }
  69
  70     --Pos;
  71     if (Pos > FirstChar && isSignedChar(Pos[0]) && !isExponentChar(Pos[-1]))
  72       break;
  73   }
  74   return Pos;
  75 }
  76
  77 /// EndOfNumber - Return the first character that is not part of the specified
  78 /// number.  This assumes that the buffer is null terminated, so it won't fall
  79 /// off the end.
  80 static const char *EndOfNumber(const char *Pos) {
  81   while (isNumberChar(*Pos))
  82     ++Pos;
  83   return Pos;
  84 }
  85
  86 /// CompareNumbers - compare two numbers, returning true if they are different.
  87 static bool CompareNumbers(const char *&F1P, const char *&F2P,
  88                            const char *F1End, const char *F2End,
  89                            double AbsTolerance, double RelTolerance,
  90                            std::string *ErrorMsg) {
  91   const char *F1NumEnd, *F2NumEnd;
  92   double V1 = 0.0, V2 = 0.0;
  93
  94   // If one of the positions is at a space and the other isn't, chomp up 'til
  95   // the end of the space.
  96   while (isSpace(static_cast<unsigned char>(*F1P)) && F1P != F1End)
  97     ++F1P;
  98   while (isSpace(static_cast<unsigned char>(*F2P)) && F2P != F2End)
  99     ++F2P;
 100
 101   // If we stop on numbers, compare their difference.
 102   if (!isNumberChar(*F1P) || !isNumberChar(*F2P)) {
 103     // The diff failed.
 104     F1NumEnd = F1P;
 105     F2NumEnd = F2P;
 106   } else {
 107     // Note that some ugliness is built into this to permit support for numbers
 108     // that use "D" or "d" as their exponential marker, e.g. "1.234D45".  This
 109     // occurs in 200.sixtrack in spec2k.
 110     V1 = strtod(F1P, const_cast<char**>(&F1NumEnd));
 111     V2 = strtod(F2P, const_cast<char**>(&F2NumEnd));
 112
 113     if (*F1NumEnd == 'D' || *F1NumEnd == 'd') {
 114       // Copy string into tmp buffer to replace the 'D' with an 'e'.
 115       SmallString<200> StrTmp(F1P, EndOfNumber(F1NumEnd)+1);
 116       // Strange exponential notation!
 117       StrTmp[static_cast<unsigned>(F1NumEnd-F1P)] = 'e';
 118
 119       V1 = strtod(&StrTmp[0], const_cast<char**>(&F1NumEnd));
 120       F1NumEnd = F1P + (F1NumEnd-&StrTmp[0]);
 121     }
 122
 123     if (*F2NumEnd == 'D' || *F2NumEnd == 'd') {
 124       // Copy string into tmp buffer to replace the 'D' with an 'e'.
 125       SmallString<200> StrTmp(F2P, EndOfNumber(F2NumEnd)+1);
 126       // Strange exponential notation!
 127       StrTmp[static_cast<unsigned>(F2NumEnd-F2P)] = 'e';
 128
 129       V2 = strtod(&StrTmp[0], const_cast<char**>(&F2NumEnd));
 130       F2NumEnd = F2P + (F2NumEnd-&StrTmp[0]);
 131     }
 132   }
 133
 134   if (F1NumEnd == F1P || F2NumEnd == F2P) {
 135     if (ErrorMsg) {
 136       *ErrorMsg = "FP Comparison failed, not a numeric difference between '";
 137       *ErrorMsg += F1P[0];
 138       *ErrorMsg += "' and '";
 139       *ErrorMsg += F2P[0];
 140       *ErrorMsg += "'";
 141     }
 142     return true;
 143   }
 144
 145   // Check to see if these are inside the absolute tolerance
 146   if (AbsTolerance < std::abs(V1-V2)) {
 147     // Nope, check the relative tolerance...
 148     double Diff;
 149     if (V2)
 150       Diff = std::abs(V1/V2 - 1.0);
 151     else if (V1)
 152       Diff = std::abs(V2/V1 - 1.0);
 153     else
 154       Diff = 0;  // Both zero.
 155     if (Diff > RelTolerance) {
 156       if (ErrorMsg) {
 157         raw_string_ostream(*ErrorMsg)
 158           << "Compared: " << V1 << " and " << V2 << '\n'
 159           << "abs. diff = " << std::abs(V1-V2) << " rel.diff = " << Diff << '\n'
 160           << "Out of tolerance: rel/abs: " << RelTolerance << '/'
 161           << AbsTolerance;
 162       }
 163       return true;
 164     }
 165   }
 166
 167   // Otherwise, advance our read pointers to the end of the numbers.
 168   F1P = F1NumEnd;  F2P = F2NumEnd;
 169   return false;
 170 }
 171
 172 /// DiffFilesWithTolerance - Compare the two files specified, returning 0 if the
 173 /// files match, 1 if they are different, and 2 if there is a file error.  This
 174 /// function differs from DiffFiles in that you can specify an absolete and
 175 /// relative FP error that is allowed to exist.  If you specify a string to fill
 176 /// in for the error option, it will set the string to an error message if an
 177 /// error occurs, allowing the caller to distinguish between a failed diff and a
 178 /// file system error.
 179 ///
 180 int llvm::DiffFilesWithTolerance(StringRef NameA,
 181                                  StringRef NameB,
 182                                  double AbsTol, double RelTol,
 183                                  std::string *Error) {
 184   // Now its safe to mmap the files into memory because both files
 185   // have a non-zero size.
 186   ErrorOr<std::unique_ptr<MemoryBuffer>> F1OrErr = MemoryBuffer::getFile(NameA);
 187   if (std::error_code EC = F1OrErr.getError()) {
 188     if (Error)
 189       *Error = EC.message();
 190     return 2;
 191   }
 192   MemoryBuffer &F1 = *F1OrErr.get();
 193
 194   ErrorOr<std::unique_ptr<MemoryBuffer>> F2OrErr = MemoryBuffer::getFile(NameB);
 195   if (std::error_code EC = F2OrErr.getError()) {
 196     if (Error)
 197       *Error = EC.message();
 198     return 2;
 199   }
 200   MemoryBuffer &F2 = *F2OrErr.get();
 201
 202   // Okay, now that we opened the files, scan them for the first difference.
 203   const char *File1Start = F1.getBufferStart();
 204   const char *File2Start = F2.getBufferStart();
 205   const char *File1End = F1.getBufferEnd();
 206   const char *File2End = F2.getBufferEnd();
 207   const char *F1P = File1Start;
 208   const char *F2P = File2Start;
 209   uint64_t A_size = F1.getBufferSize();
 210   uint64_t B_size = F2.getBufferSize();
 211
 212   // Are the buffers identical?  Common case: Handle this efficiently.
 213   if (A_size == B_size &&
 214       std::memcmp(File1Start, File2Start, A_size) == 0)
 215     return 0;
 216
 217   // Otherwise, we are done a tolerances are set.
 218   if (AbsTol == 0 && RelTol == 0) {
 219     if (Error)
 220       *Error = "Files differ without tolerance allowance";
 221     return 1;   // Files different!
 222   }
 223
 224   bool CompareFailed = false;
 225   while (true) {
 226     // Scan for the end of file or next difference.
 227     while (F1P < File1End && F2P < File2End && *F1P == *F2P) {
 228       ++F1P;
 229       ++F2P;
 230     }
 231
 232     if (F1P >= File1End || F2P >= File2End) break;
 233
 234     // Okay, we must have found a difference.  Backup to the start of the
 235     // current number each stream is at so that we can compare from the
 236     // beginning.
 237     F1P = BackupNumber(F1P, File1Start);
 238     F2P = BackupNumber(F2P, File2Start);
 239
 240     // Now that we are at the start of the numbers, compare them, exiting if
 241     // they don't match.
 242     if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error)) {
 243       CompareFailed = true;
 244       break;
 245     }
 246   }
 247
 248   // Okay, we reached the end of file.  If both files are at the end, we
 249   // succeeded.
 250   bool F1AtEnd = F1P >= File1End;
 251   bool F2AtEnd = F2P >= File2End;
 252   if (!CompareFailed && (!F1AtEnd || !F2AtEnd)) {
 253     // Else, we might have run off the end due to a number: backup and retry.
 254     if (F1AtEnd && isNumberChar(F1P[-1])) --F1P;
 255     if (F2AtEnd && isNumberChar(F2P[-1])) --F2P;
 256     F1P = BackupNumber(F1P, File1Start);
 257     F2P = BackupNumber(F2P, File2Start);
 258
 259     // Now that we are at the start of the numbers, compare them, exiting if
 260     // they don't match.
 261     if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error))
 262       CompareFailed = true;
 263
 264     // If we found the end, we succeeded.
 265     if (F1P < File1End || F2P < File2End)
 266       CompareFailed = true;
 267   }
 268
 269   return CompareFailed;
 270 }
 271
 272 void llvm::AtomicFileWriteError::log(raw_ostream &OS) const {
 273   OS << "atomic_write_error: ";
 274   switch (Error) {
 275   case atomic_write_error::failed_to_create_uniq_file:
 276     OS << "failed_to_create_uniq_file";
 277     return;
 278   case atomic_write_error::output_stream_error:
 279     OS << "output_stream_error";
 280     return;
 281   case atomic_write_error::failed_to_rename_temp_file:
 282     OS << "failed_to_rename_temp_file";
 283     return;
 284   }
 285   llvm_unreachable("unknown atomic_write_error value in "
 286                    "failed_to_rename_temp_file::log()");
 287 }
 288
 289 llvm::Error llvm::writeFileAtomically(StringRef TempPathModel,
 290                                       StringRef FinalPath, StringRef Buffer) {
 291   return writeFileAtomically(TempPathModel, FinalPath,
 292                              [&Buffer](llvm::raw_ostream &OS) {
 293                                OS.write(Buffer.data(), Buffer.size());
 294                                return llvm::Error::success();
 295                              });
 296 }
 297
 298 llvm::Error llvm::writeFileAtomically(
 299     StringRef TempPathModel, StringRef FinalPath,
 300     std::function<llvm::Error(llvm::raw_ostream &)> Writer) {
 301   SmallString<128> GeneratedUniqPath;
 302   int TempFD;
 303   if (sys::fs::createUniqueFile(TempPathModel.str(), TempFD,
 304                                 GeneratedUniqPath)) {
 305     return llvm::make_error<AtomicFileWriteError>(
 306         atomic_write_error::failed_to_create_uniq_file);
 307   }
 308   llvm::FileRemover RemoveTmpFileOnFail(GeneratedUniqPath);
 309
 310   raw_fd_ostream OS(TempFD, /*shouldClose=*/true);
 311   if (llvm::Error Err = Writer(OS)) {
 312     return Err;
 313   }
 314
 315   OS.close();
 316   if (OS.has_error()) {
 317     OS.clear_error();
 318     return llvm::make_error<AtomicFileWriteError>(
 319         atomic_write_error::output_stream_error);
 320   }
 321
 322   if (sys::fs::rename(/*from=*/GeneratedUniqPath.c_str(),
 323                       /*to=*/FinalPath.str().c_str())) {
 324     return llvm::make_error<AtomicFileWriteError>(
 325         atomic_write_error::failed_to_rename_temp_file);
 326   }
 327
 328   RemoveTmpFileOnFail.releaseFile();
 329   return Error::success();
 330 }
 331
 332 char llvm::AtomicFileWriteError::ID;