lib/Support/FileUtilities.cpp

   1 //===- Support/FileUtilities.cpp - File System Utilities ------------------===//
   2 //
   3 //                     The LLVM Compiler Infrastructure
   4 //
   5 // This file was developed by the LLVM research group and is distributed under
   6 // the University of Illinois Open Source License. See LICENSE.TXT for details.
   7 //
   8 //===----------------------------------------------------------------------===//
   9 //
  10 // This file implements a family of utility functions which are useful for doing
  11 // various things with files.
  12 //
  13 //===----------------------------------------------------------------------===//
  14
  15 #include "llvm/Support/FileUtilities.h"
  16 #include "llvm/System/Path.h"
  17 #include "llvm/System/MappedFile.h"
  18 #include "llvm/ADT/StringExtras.h"
  19 #include <cmath>
  20 #include <cstring>
  21 #include <cctype>
  22 using namespace llvm;
  23
  24 static bool isNumberChar(char C) {
  25   switch (C) {
  26   case '0': case '1': case '2': case '3': case '4':
  27   case '5': case '6': case '7': case '8': case '9':
  28   case '.': case '+': case '-':
  29   case 'D':  // Strange exponential notation.
  30   case 'd':  // Strange exponential notation.
  31   case 'e':
  32   case 'E': return true;
  33   default: return false;
  34   }
  35 }
  36
  37 static char *BackupNumber(char *Pos, char *FirstChar) {
  38   // If we didn't stop in the middle of a number, don't backup.
  39   if (!isNumberChar(*Pos)) return Pos;
  40
  41   // Otherwise, return to the start of the number.
  42   while (Pos > FirstChar && isNumberChar(Pos[-1]))
  43     --Pos;
  44   return Pos;
  45 }
  46
  47 /// CompareNumbers - compare two numbers, returning true if they are different.
  48 static bool CompareNumbers(char *&F1P, char *&F2P, char *F1End, char *F2End,
  49                            double AbsTolerance, double RelTolerance,
  50                            std::string *ErrorMsg) {
  51   char *F1NumEnd, *F2NumEnd;
  52   double V1 = 0.0, V2 = 0.0;
  53
  54   // If one of the positions is at a space and the other isn't, chomp up 'til
  55   // the end of the space.
  56   while (isspace(*F1P) && F1P != F1End)
  57     ++F1P;
  58   while (isspace(*F2P) && F2P != F2End)
  59     ++F2P;
  60
  61   // If we stop on numbers, compare their difference.  Note that some ugliness
  62   // is built into this to permit support for numbers that use "D" or "d" as
  63   // their exponential marker, e.g. "1.234D45".  This occurs in 200.sixtrack in
  64   // spec2k.
  65   if (isNumberChar(*F1P) && isNumberChar(*F2P)) {
  66     bool isDNotation;
  67     do {
  68       isDNotation = false;
  69       V1 = strtod(F1P, &F1NumEnd);
  70       V2 = strtod(F2P, &F2NumEnd);
  71
  72       if (*F1NumEnd == 'D' || *F1NumEnd == 'd') {
  73         *F1NumEnd = 'e';  // Strange exponential notation!
  74         isDNotation = true;
  75       }
  76       if (*F2NumEnd == 'D' || *F2NumEnd == 'd') {
  77         *F2NumEnd = 'e';  // Strange exponential notation!
  78         isDNotation = true;
  79       }
  80     } while (isDNotation);
  81   } else {
  82     // Otherwise, the diff failed.
  83     F1NumEnd = F1P;
  84     F2NumEnd = F2P;
  85   }
  86
  87   if (F1NumEnd == F1P || F2NumEnd == F2P) {
  88     if (ErrorMsg) *ErrorMsg = "Comparison failed, not a numeric difference.";
  89     return true;
  90   }
  91
  92   // Check to see if these are inside the absolute tolerance
  93   if (AbsTolerance < std::abs(V1-V2)) {
  94     // Nope, check the relative tolerance...
  95     double Diff;
  96     if (V2)
  97       Diff = std::abs(V1/V2 - 1.0);
  98     else if (V1)
  99       Diff = std::abs(V2/V1 - 1.0);
 100     else
 101       Diff = 0;  // Both zero.
 102     if (Diff > RelTolerance) {
 103       if (ErrorMsg) {
 104         *ErrorMsg = "Compared: " + ftostr(V1) + " and " + ftostr(V2) +
 105                     ": diff = " + ftostr(Diff) + "\n";
 106         *ErrorMsg += "Out of tolerance: rel/abs: " + ftostr(RelTolerance) +
 107                      "/" + ftostr(AbsTolerance);
 108       }
 109       return true;
 110     }
 111   }
 112
 113   // Otherwise, advance our read pointers to the end of the numbers.
 114   F1P = F1NumEnd;  F2P = F2NumEnd;
 115   return false;
 116 }
 117
 118 // PadFileIfNeeded - If the files are not identical, we will have to be doing
 119 // numeric comparisons in here.  There are bad cases involved where we (i.e.,
 120 // strtod) might run off the beginning or end of the file if it starts or ends
 121 // with a number.  Because of this, if needed, we pad the file so that it starts
 122 // and ends with a null character.
 123 static void PadFileIfNeeded(char *&FileStart, char *&FileEnd, char *&FP) {
 124   if (FileStart-FileEnd < 2 ||
 125       isNumberChar(FileStart[0]) || isNumberChar(FileEnd[-1])) {
 126     unsigned FileLen = FileEnd-FileStart;
 127     char *NewFile = new char[FileLen+2];
 128     NewFile[0] = 0;              // Add null padding
 129     NewFile[FileLen+1] = 0;      // Add null padding
 130     memcpy(NewFile+1, FileStart, FileLen);
 131     FP = NewFile+(FP-FileStart)+1;
 132     FileStart = NewFile+1;
 133     FileEnd = FileStart+FileLen;
 134   }
 135 }
 136
 137 /// DiffFilesWithTolerance - Compare the two files specified, returning 0 if the
 138 /// files match, 1 if they are different, and 2 if there is a file error.  This
 139 /// function differs from DiffFiles in that you can specify an absolete and
 140 /// relative FP error that is allowed to exist.  If you specify a string to fill
 141 /// in for the error option, it will set the string to an error message if an
 142 /// error occurs, allowing the caller to distinguish between a failed diff and a
 143 /// file system error.
 144 ///
 145 int llvm::DiffFilesWithTolerance(const sys::Path &FileA,
 146                                  const sys::Path &FileB,
 147                                  double AbsTol, double RelTol,
 148                                  std::string *Error) {
 149   sys::FileStatus FileAStat, FileBStat;
 150   if (FileA.getFileStatus(FileAStat, Error) ||
 151       FileB.getFileStatus(FileBStat, Error))
 152     return 2;
 153   // Check for zero length files because some systems croak when you try to
 154   // mmap an empty file.
 155   size_t A_size = FileAStat.getSize();
 156   size_t B_size = FileBStat.getSize();
 157
 158   // If they are both zero sized then they're the same
 159   if (A_size == 0 && B_size == 0)
 160     return 0;
 161   // If only one of them is zero sized then they can't be the same
 162   if ((A_size == 0 || B_size == 0))
 163     return 1;
 164
 165   try {
 166     // Now its safe to mmap the files into memory becasue both files
 167     // have a non-zero size.
 168     sys::MappedFile F1(FileA);
 169     sys::MappedFile F2(FileB);
 170     F1.map();
 171     F2.map();
 172
 173     // Okay, now that we opened the files, scan them for the first difference.
 174     char *File1Start = F1.charBase();
 175     char *File2Start = F2.charBase();
 176     char *File1End = File1Start+A_size;
 177     char *File2End = File2Start+B_size;
 178     char *F1P = File1Start;
 179     char *F2P = File2Start;
 180
 181     if (A_size == B_size) {
 182       // Are the buffers identical?
 183       if (std::memcmp(File1Start, File2Start, A_size) == 0)
 184         return 0;
 185
 186       if (AbsTol == 0 && RelTol == 0)
 187         return 1;   // Files different!
 188     }
 189
 190     char *OrigFile1Start = File1Start;
 191     char *OrigFile2Start = File2Start;
 192
 193     // If the files need padding, do so now.
 194     PadFileIfNeeded(File1Start, File1End, F1P);
 195     PadFileIfNeeded(File2Start, File2End, F2P);
 196
 197     bool CompareFailed = false;
 198     while (1) {
 199       // Scan for the end of file or next difference.
 200       while (F1P < File1End && F2P < File2End && *F1P == *F2P)
 201         ++F1P, ++F2P;
 202
 203       if (F1P >= File1End || F2P >= File2End) break;
 204
 205       // Okay, we must have found a difference.  Backup to the start of the
 206       // current number each stream is at so that we can compare from the
 207       // beginning.
 208       F1P = BackupNumber(F1P, File1Start);
 209       F2P = BackupNumber(F2P, File2Start);
 210
 211       // Now that we are at the start of the numbers, compare them, exiting if
 212       // they don't match.
 213       if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error)) {
 214         CompareFailed = true;
 215         break;
 216       }
 217     }
 218
 219     // Okay, we reached the end of file.  If both files are at the end, we
 220     // succeeded.
 221     bool F1AtEnd = F1P >= File1End;
 222     bool F2AtEnd = F2P >= File2End;
 223     if (!CompareFailed && (!F1AtEnd || !F2AtEnd)) {
 224       // Else, we might have run off the end due to a number: backup and retry.
 225       if (F1AtEnd && isNumberChar(F1P[-1])) --F1P;
 226       if (F2AtEnd && isNumberChar(F2P[-1])) --F2P;
 227       F1P = BackupNumber(F1P, File1Start);
 228       F2P = BackupNumber(F2P, File2Start);
 229
 230       // Now that we are at the start of the numbers, compare them, exiting if
 231       // they don't match.
 232       if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error))
 233         CompareFailed = true;
 234
 235       // If we found the end, we succeeded.
 236       if (F1P < File1End || F2P < File2End)
 237         CompareFailed = true;
 238     }
 239
 240     if (OrigFile1Start != File1Start)
 241       delete[] (File1Start-1);   // Back up past null byte
 242     if (OrigFile2Start != File2Start)
 243       delete[] (File2Start-1);   // Back up past null byte
 244     return CompareFailed;
 245   } catch (const std::string &Msg) {
 246     if (Error) *Error = Msg;
 247     return 2;
 248   } catch (...) {
 249     *Error = "Unknown Exception Occurred";
 250     return 2;
 251   }
 252 }