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 'e':
  30   case 'E': return true;
  31   default: return false;
  32   }
  33 }
  34
  35 static char *BackupNumber(char *Pos, char *FirstChar) {
  36   // If we didn't stop in the middle of a number, don't backup.
  37   if (!isNumberChar(*Pos)) return Pos;
  38
  39   // Otherwise, return to the start of the number.
  40   while (Pos > FirstChar && isNumberChar(Pos[-1]))
  41     --Pos;
  42   return Pos;
  43 }
  44
  45 /// CompareNumbers - compare two numbers, returning true if they are different.
  46 static bool CompareNumbers(char *&F1P, char *&F2P, char *F1End, char *F2End,
  47                            double AbsTolerance, double RelTolerance,
  48                            std::string *ErrorMsg) {
  49   char *F1NumEnd, *F2NumEnd;
  50   double V1 = 0.0, V2 = 0.0;
  51
  52   // If one of the positions is at a space and the other isn't, chomp up 'til
  53   // the end of the space.
  54   while (isspace(*F1P) && F1P != F1End)
  55     ++F1P;
  56   while (isspace(*F2P) && F2P != F2End)
  57     ++F2P;
  58
  59   // If we stop on numbers, compare their difference.
  60   if (isNumberChar(*F1P) && isNumberChar(*F2P)) {
  61     V1 = strtod(F1P, &F1NumEnd);
  62     V2 = strtod(F2P, &F2NumEnd);
  63   } else {
  64     // Otherwise, the diff failed.
  65     F1NumEnd = F1P;
  66     F2NumEnd = F2P;
  67   }
  68
  69   if (F1NumEnd == F1P || F2NumEnd == F2P) {
  70     if (ErrorMsg) *ErrorMsg = "Comparison failed, not a numeric difference.";
  71     return true;
  72   }
  73
  74   // Check to see if these are inside the absolute tolerance
  75   if (AbsTolerance < std::abs(V1-V2)) {
  76     // Nope, check the relative tolerance...
  77     double Diff;
  78     if (V2)
  79       Diff = std::abs(V1/V2 - 1.0);
  80     else if (V1)
  81       Diff = std::abs(V2/V1 - 1.0);
  82     else
  83       Diff = 0;  // Both zero.
  84     if (Diff > RelTolerance) {
  85       if (ErrorMsg) {
  86         *ErrorMsg = "Compared: " + ftostr(V1) + " and " + ftostr(V2) +
  87                     ": diff = " + ftostr(Diff) + "\n";
  88         *ErrorMsg += "Out of tolerance: rel/abs: " + ftostr(RelTolerance) +
  89                      "/" + ftostr(AbsTolerance);
  90       }
  91       return true;
  92     }
  93   }
  94
  95   // Otherwise, advance our read pointers to the end of the numbers.
  96   F1P = F1NumEnd;  F2P = F2NumEnd;
  97   return false;
  98 }
  99
 100 // PadFileIfNeeded - If the files are not identical, we will have to be doing
 101 // numeric comparisons in here.  There are bad cases involved where we (i.e.,
 102 // strtod) might run off the beginning or end of the file if it starts or ends
 103 // with a number.  Because of this, if needed, we pad the file so that it starts
 104 // and ends with a null character.
 105 static void PadFileIfNeeded(char *&FileStart, char *&FileEnd, char *&FP) {
 106   if (FileStart-FileEnd < 2 ||
 107       isNumberChar(FileStart[0]) || isNumberChar(FileEnd[-1])) {
 108     unsigned FileLen = FileEnd-FileStart;
 109     char *NewFile = new char[FileLen+2];
 110     NewFile[0] = 0;              // Add null padding
 111     NewFile[FileLen+1] = 0;      // Add null padding
 112     memcpy(NewFile+1, FileStart, FileLen);
 113     FP = NewFile+(FP-FileStart)+1;
 114     FileStart = NewFile+1;
 115     FileEnd = FileStart+FileLen;
 116   }
 117 }
 118
 119 /// DiffFilesWithTolerance - Compare the two files specified, returning 0 if the
 120 /// files match, 1 if they are different, and 2 if there is a file error.  This
 121 /// function differs from DiffFiles in that you can specify an absolete and
 122 /// relative FP error that is allowed to exist.  If you specify a string to fill
 123 /// in for the error option, it will set the string to an error message if an
 124 /// error occurs, allowing the caller to distinguish between a failed diff and a
 125 /// file system error.
 126 ///
 127 int llvm::DiffFilesWithTolerance(const sys::Path &FileA,
 128                                  const sys::Path &FileB,
 129                                  double AbsTol, double RelTol,
 130                                  std::string *Error) {
 131   try {
 132     // Check for zero length files because some systems croak when you try to
 133     // mmap an empty file.
 134     size_t A_size = FileA.getSize();
 135     size_t B_size = FileB.getSize();
 136
 137     // If they are both zero sized then they're the same
 138     if (A_size == 0 && B_size == 0)
 139       return 0;
 140     // If only one of them is zero sized then they can't be the same
 141     if ((A_size == 0 || B_size == 0))
 142       return 1;
 143
 144     // Now its safe to mmap the files into memory becasue both files
 145     // have a non-zero size.
 146     sys::MappedFile F1(FileA);
 147     sys::MappedFile F2(FileB);
 148     F1.map();
 149     F2.map();
 150
 151     // Okay, now that we opened the files, scan them for the first difference.
 152     char *File1Start = F1.charBase();
 153     char *File2Start = F2.charBase();
 154     char *File1End = File1Start+A_size;
 155     char *File2End = File2Start+B_size;
 156     char *F1P = File1Start;
 157     char *F2P = File2Start;
 158
 159     if (A_size == B_size) {
 160       // Are the buffers identical?
 161       if (std::memcmp(File1Start, File2Start, A_size) == 0)
 162         return 0;
 163
 164       if (AbsTol == 0 && RelTol == 0)
 165         return 1;   // Files different!
 166     }
 167
 168     char *OrigFile1Start = File1Start;
 169     char *OrigFile2Start = File2Start;
 170
 171     // If the files need padding, do so now.
 172     PadFileIfNeeded(File1Start, File1End, F1P);
 173     PadFileIfNeeded(File2Start, File2End, F2P);
 174
 175     bool CompareFailed = false;
 176     while (1) {
 177       // Scan for the end of file or next difference.
 178       while (F1P < File1End && F2P < File2End && *F1P == *F2P)
 179         ++F1P, ++F2P;
 180
 181       if (F1P >= File1End || F2P >= File2End) break;
 182
 183       // Okay, we must have found a difference.  Backup to the start of the
 184       // current number each stream is at so that we can compare from the
 185       // beginning.
 186       F1P = BackupNumber(F1P, File1Start);
 187       F2P = BackupNumber(F2P, File2Start);
 188
 189       // Now that we are at the start of the numbers, compare them, exiting if
 190       // they don't match.
 191       if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error)) {
 192         CompareFailed = true;
 193         break;
 194       }
 195     }
 196
 197     // Okay, we reached the end of file.  If both files are at the end, we
 198     // succeeded.
 199     bool F1AtEnd = F1P >= File1End;
 200     bool F2AtEnd = F2P >= File2End;
 201     if (!CompareFailed && (!F1AtEnd || !F2AtEnd)) {
 202       // Else, we might have run off the end due to a number: backup and retry.
 203       if (F1AtEnd && isNumberChar(F1P[-1])) --F1P;
 204       if (F2AtEnd && isNumberChar(F2P[-1])) --F2P;
 205       F1P = BackupNumber(F1P, File1Start);
 206       F2P = BackupNumber(F2P, File2Start);
 207
 208       // Now that we are at the start of the numbers, compare them, exiting if
 209       // they don't match.
 210       if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error))
 211         CompareFailed = true;
 212
 213       // If we found the end, we succeeded.
 214       if (F1P < File1End || F2P < File2End)
 215         CompareFailed = true;
 216     }
 217
 218     if (OrigFile1Start != File1Start)
 219       delete[] (File1Start-1);   // Back up past null byte
 220     if (OrigFile2Start != File2Start)
 221       delete[] (File2Start-1);   // Back up past null byte
 222     return CompareFailed;
 223   } catch (const std::string &Msg) {
 224     if (Error) *Error = Msg;
 225     return 2;
 226   }
 227 }