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   try {
150     // Check for zero length files because some systems croak when you try to
151     // mmap an empty file.
152     size_t A_size = FileA.getSize();
153     size_t B_size = FileB.getSize();
154 
155     // If they are both zero sized then they're the same
156     if (A_size == 0 && B_size == 0)
157       return 0;
158     // If only one of them is zero sized then they can't be the same
159     if ((A_size == 0 || B_size == 0))
160       return 1;
161 
162     // Now its safe to mmap the files into memory becasue both files
163     // have a non-zero size.
164     sys::MappedFile F1(FileA);
165     sys::MappedFile F2(FileB);
166     F1.map();
167     F2.map();
168 
169     // Okay, now that we opened the files, scan them for the first difference.
170     char *File1Start = F1.charBase();
171     char *File2Start = F2.charBase();
172     char *File1End = File1Start+A_size;
173     char *File2End = File2Start+B_size;
174     char *F1P = File1Start;
175     char *F2P = File2Start;
176 
177     if (A_size == B_size) {
178       // Are the buffers identical?
179       if (std::memcmp(File1Start, File2Start, A_size) == 0)
180         return 0;
181 
182       if (AbsTol == 0 && RelTol == 0)
183         return 1;   // Files different!
184     }
185 
186     char *OrigFile1Start = File1Start;
187     char *OrigFile2Start = File2Start;
188 
189     // If the files need padding, do so now.
190     PadFileIfNeeded(File1Start, File1End, F1P);
191     PadFileIfNeeded(File2Start, File2End, F2P);
192 
193     bool CompareFailed = false;
194     while (1) {
195       // Scan for the end of file or next difference.
196       while (F1P < File1End && F2P < File2End && *F1P == *F2P)
197         ++F1P, ++F2P;
198 
199       if (F1P >= File1End || F2P >= File2End) break;
200 
201       // Okay, we must have found a difference.  Backup to the start of the
202       // current number each stream is at so that we can compare from the
203       // beginning.
204       F1P = BackupNumber(F1P, File1Start);
205       F2P = BackupNumber(F2P, File2Start);
206 
207       // Now that we are at the start of the numbers, compare them, exiting if
208       // they don't match.
209       if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error)) {
210         CompareFailed = true;
211         break;
212       }
213     }
214 
215     // Okay, we reached the end of file.  If both files are at the end, we
216     // succeeded.
217     bool F1AtEnd = F1P >= File1End;
218     bool F2AtEnd = F2P >= File2End;
219     if (!CompareFailed && (!F1AtEnd || !F2AtEnd)) {
220       // Else, we might have run off the end due to a number: backup and retry.
221       if (F1AtEnd && isNumberChar(F1P[-1])) --F1P;
222       if (F2AtEnd && isNumberChar(F2P[-1])) --F2P;
223       F1P = BackupNumber(F1P, File1Start);
224       F2P = BackupNumber(F2P, File2Start);
225 
226       // Now that we are at the start of the numbers, compare them, exiting if
227       // they don't match.
228       if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error))
229         CompareFailed = true;
230 
231       // If we found the end, we succeeded.
232       if (F1P < File1End || F2P < File2End)
233         CompareFailed = true;
234     }
235 
236     if (OrigFile1Start != File1Start)
237       delete[] (File1Start-1);   // Back up past null byte
238     if (OrigFile2Start != File2Start)
239       delete[] (File2Start-1);   // Back up past null byte
240     return CompareFailed;
241   } catch (const std::string &Msg) {
242     if (Error) *Error = Msg;
243     return 2;
244   }
245 }
246