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/SmallString.h"
  16 #include "llvm/Support/ErrorOr.h"
  17 #include "llvm/Support/MemoryBuffer.h"
  18 #include "llvm/Support/raw_ostream.h"
  19 #include <cctype>
  20 #include <cmath>
  21 #include <cstdint>
  22 #include <cstdlib>
  23 #include <cstring>
  24 #include <memory>
  25 #include <system_error>
  26
  27 using namespace llvm;
  28
  29 static bool isSignedChar(char C) {
  30   return (C == '+' || C == '-');
  31 }
  32
  33 static bool isExponentChar(char C) {
  34   switch (C) {
  35   case 'D':  // Strange exponential notation.
  36   case 'd':  // Strange exponential notation.
  37   case 'e':
  38   case 'E': return true;
  39   default: return false;
  40   }
  41 }
  42
  43 static bool isNumberChar(char C) {
  44   switch (C) {
  45   case '0': case '1': case '2': case '3': case '4':
  46   case '5': case '6': case '7': case '8': case '9':
  47   case '.': return true;
  48   default: return isSignedChar(C) || isExponentChar(C);
  49   }
  50 }
  51
  52 static const char *BackupNumber(const char *Pos, const char *FirstChar) {
  53   // If we didn't stop in the middle of a number, don't backup.
  54   if (!isNumberChar(*Pos)) return Pos;
  55
  56   // Otherwise, return to the start of the number.
  57   bool HasPeriod = false;
  58   while (Pos > FirstChar && isNumberChar(Pos[-1])) {
  59     // Backup over at most one period.
  60     if (Pos[-1] == '.') {
  61       if (HasPeriod)
  62         break;
  63       HasPeriod = true;
  64     }
  65
  66     --Pos;
  67     if (Pos > FirstChar && isSignedChar(Pos[0]) && !isExponentChar(Pos[-1]))
  68       break;
  69   }
  70   return Pos;
  71 }
  72
  73 /// EndOfNumber - Return the first character that is not part of the specified
  74 /// number.  This assumes that the buffer is null terminated, so it won't fall
  75 /// off the end.
  76 static const char *EndOfNumber(const char *Pos) {
  77   while (isNumberChar(*Pos))
  78     ++Pos;
  79   return Pos;
  80 }
  81
  82 /// CompareNumbers - compare two numbers, returning true if they are different.
  83 static bool CompareNumbers(const char *&F1P, const char *&F2P,
  84                            const char *F1End, const char *F2End,
  85                            double AbsTolerance, double RelTolerance,
  86                            std::string *ErrorMsg) {
  87   const char *F1NumEnd, *F2NumEnd;
  88   double V1 = 0.0, V2 = 0.0;
  89
  90   // If one of the positions is at a space and the other isn't, chomp up 'til
  91   // the end of the space.
  92   while (isspace(static_cast<unsigned char>(*F1P)) && F1P != F1End)
  93     ++F1P;
  94   while (isspace(static_cast<unsigned char>(*F2P)) && F2P != F2End)
  95     ++F2P;
  96
  97   // If we stop on numbers, compare their difference.
  98   if (!isNumberChar(*F1P) || !isNumberChar(*F2P)) {
  99     // The diff failed.
 100     F1NumEnd = F1P;
 101     F2NumEnd = F2P;
 102   } else {
 103     // Note that some ugliness is built into this to permit support for numbers
 104     // that use "D" or "d" as their exponential marker, e.g. "1.234D45".  This
 105     // occurs in 200.sixtrack in spec2k.
 106     V1 = strtod(F1P, const_cast<char**>(&F1NumEnd));
 107     V2 = strtod(F2P, const_cast<char**>(&F2NumEnd));
 108
 109     if (*F1NumEnd == 'D' || *F1NumEnd == 'd') {
 110       // Copy string into tmp buffer to replace the 'D' with an 'e'.
 111       SmallString<200> StrTmp(F1P, EndOfNumber(F1NumEnd)+1);
 112       // Strange exponential notation!
 113       StrTmp[static_cast<unsigned>(F1NumEnd-F1P)] = 'e';
 114
 115       V1 = strtod(&StrTmp[0], const_cast<char**>(&F1NumEnd));
 116       F1NumEnd = F1P + (F1NumEnd-&StrTmp[0]);
 117     }
 118
 119     if (*F2NumEnd == 'D' || *F2NumEnd == 'd') {
 120       // Copy string into tmp buffer to replace the 'D' with an 'e'.
 121       SmallString<200> StrTmp(F2P, EndOfNumber(F2NumEnd)+1);
 122       // Strange exponential notation!
 123       StrTmp[static_cast<unsigned>(F2NumEnd-F2P)] = 'e';
 124
 125       V2 = strtod(&StrTmp[0], const_cast<char**>(&F2NumEnd));
 126       F2NumEnd = F2P + (F2NumEnd-&StrTmp[0]);
 127     }
 128   }
 129
 130   if (F1NumEnd == F1P || F2NumEnd == F2P) {
 131     if (ErrorMsg) {
 132       *ErrorMsg = "FP Comparison failed, not a numeric difference between '";
 133       *ErrorMsg += F1P[0];
 134       *ErrorMsg += "' and '";
 135       *ErrorMsg += F2P[0];
 136       *ErrorMsg += "'";
 137     }
 138     return true;
 139   }
 140
 141   // Check to see if these are inside the absolute tolerance
 142   if (AbsTolerance < std::abs(V1-V2)) {
 143     // Nope, check the relative tolerance...
 144     double Diff;
 145     if (V2)
 146       Diff = std::abs(V1/V2 - 1.0);
 147     else if (V1)
 148       Diff = std::abs(V2/V1 - 1.0);
 149     else
 150       Diff = 0;  // Both zero.
 151     if (Diff > RelTolerance) {
 152       if (ErrorMsg) {
 153         raw_string_ostream(*ErrorMsg)
 154           << "Compared: " << V1 << " and " << V2 << '\n'
 155           << "abs. diff = " << std::abs(V1-V2) << " rel.diff = " << Diff << '\n'
 156           << "Out of tolerance: rel/abs: " << RelTolerance << '/'
 157           << AbsTolerance;
 158       }
 159       return true;
 160     }
 161   }
 162
 163   // Otherwise, advance our read pointers to the end of the numbers.
 164   F1P = F1NumEnd;  F2P = F2NumEnd;
 165   return false;
 166 }
 167
 168 /// DiffFilesWithTolerance - Compare the two files specified, returning 0 if the
 169 /// files match, 1 if they are different, and 2 if there is a file error.  This
 170 /// function differs from DiffFiles in that you can specify an absolete and
 171 /// relative FP error that is allowed to exist.  If you specify a string to fill
 172 /// in for the error option, it will set the string to an error message if an
 173 /// error occurs, allowing the caller to distinguish between a failed diff and a
 174 /// file system error.
 175 ///
 176 int llvm::DiffFilesWithTolerance(StringRef NameA,
 177                                  StringRef NameB,
 178                                  double AbsTol, double RelTol,
 179                                  std::string *Error) {
 180   // Now its safe to mmap the files into memory because both files
 181   // have a non-zero size.
 182   ErrorOr<std::unique_ptr<MemoryBuffer>> F1OrErr = MemoryBuffer::getFile(NameA);
 183   if (std::error_code EC = F1OrErr.getError()) {
 184     if (Error)
 185       *Error = EC.message();
 186     return 2;
 187   }
 188   MemoryBuffer &F1 = *F1OrErr.get();
 189
 190   ErrorOr<std::unique_ptr<MemoryBuffer>> F2OrErr = MemoryBuffer::getFile(NameB);
 191   if (std::error_code EC = F2OrErr.getError()) {
 192     if (Error)
 193       *Error = EC.message();
 194     return 2;
 195   }
 196   MemoryBuffer &F2 = *F2OrErr.get();
 197
 198   // Okay, now that we opened the files, scan them for the first difference.
 199   const char *File1Start = F1.getBufferStart();
 200   const char *File2Start = F2.getBufferStart();
 201   const char *File1End = F1.getBufferEnd();
 202   const char *File2End = F2.getBufferEnd();
 203   const char *F1P = File1Start;
 204   const char *F2P = File2Start;
 205   uint64_t A_size = F1.getBufferSize();
 206   uint64_t B_size = F2.getBufferSize();
 207
 208   // Are the buffers identical?  Common case: Handle this efficiently.
 209   if (A_size == B_size &&
 210       std::memcmp(File1Start, File2Start, A_size) == 0)
 211     return 0;
 212
 213   // Otherwise, we are done a tolerances are set.
 214   if (AbsTol == 0 && RelTol == 0) {
 215     if (Error)
 216       *Error = "Files differ without tolerance allowance";
 217     return 1;   // Files different!
 218   }
 219
 220   bool CompareFailed = false;
 221   while (true) {
 222     // Scan for the end of file or next difference.
 223     while (F1P < File1End && F2P < File2End && *F1P == *F2P) {
 224       ++F1P;
 225       ++F2P;
 226     }
 227
 228     if (F1P >= File1End || F2P >= File2End) break;
 229
 230     // Okay, we must have found a difference.  Backup to the start of the
 231     // current number each stream is at so that we can compare from the
 232     // beginning.
 233     F1P = BackupNumber(F1P, File1Start);
 234     F2P = BackupNumber(F2P, File2Start);
 235
 236     // Now that we are at the start of the numbers, compare them, exiting if
 237     // they don't match.
 238     if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error)) {
 239       CompareFailed = true;
 240       break;
 241     }
 242   }
 243
 244   // Okay, we reached the end of file.  If both files are at the end, we
 245   // succeeded.
 246   bool F1AtEnd = F1P >= File1End;
 247   bool F2AtEnd = F2P >= File2End;
 248   if (!CompareFailed && (!F1AtEnd || !F2AtEnd)) {
 249     // Else, we might have run off the end due to a number: backup and retry.
 250     if (F1AtEnd && isNumberChar(F1P[-1])) --F1P;
 251     if (F2AtEnd && isNumberChar(F2P[-1])) --F2P;
 252     F1P = BackupNumber(F1P, File1Start);
 253     F2P = BackupNumber(F2P, File2Start);
 254
 255     // Now that we are at the start of the numbers, compare them, exiting if
 256     // they don't match.
 257     if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error))
 258       CompareFailed = true;
 259
 260     // If we found the end, we succeeded.
 261     if (F1P < File1End || F2P < File2End)
 262       CompareFailed = true;
 263   }
 264
 265   return CompareFailed;
 266 }