1 //===- Support/FileUtilities.cpp - File System Utilities ------------------===//
3 // The LLVM Compiler Infrastructure
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
8 //===----------------------------------------------------------------------===//
10 // This file implements a family of utility functions which are useful for doing
11 // various things with files.
13 //===----------------------------------------------------------------------===//
15 #include "llvm/Support/FileUtilities.h"
16 #include "llvm/System/Path.h"
17 #include "llvm/System/MappedFile.h"
18 #include "llvm/ADT/StringExtras.h"
24 static bool isSignedChar(char C) {
25 return (C == '+' || C == '-');
28 static bool isExponentChar(char C) {
30 case 'D': // Strange exponential notation.
31 case 'd': // Strange exponential notation.
33 case 'E': return true;
34 default: return false;
38 static bool isNumberChar(char C) {
40 case '0': case '1': case '2': case '3': case '4':
41 case '5': case '6': case '7': case '8': case '9':
42 case '.': return true;
43 default: return isSignedChar(C) || isExponentChar(C);
47 static char *BackupNumber(char *Pos, char *FirstChar) {
48 // If we didn't stop in the middle of a number, don't backup.
49 if (!isNumberChar(*Pos)) return Pos;
51 // Otherwise, return to the start of the number.
52 while (Pos > FirstChar && isNumberChar(Pos[-1])) {
54 if (Pos > FirstChar && isSignedChar(Pos[0]) && !isExponentChar(Pos[-1]))
60 /// CompareNumbers - compare two numbers, returning true if they are different.
61 static bool CompareNumbers(char *&F1P, char *&F2P, char *F1End, char *F2End,
62 double AbsTolerance, double RelTolerance,
63 std::string *ErrorMsg) {
64 char *F1NumEnd, *F2NumEnd;
65 double V1 = 0.0, V2 = 0.0;
67 // If one of the positions is at a space and the other isn't, chomp up 'til
68 // the end of the space.
69 while (isspace(*F1P) && F1P != F1End)
71 while (isspace(*F2P) && F2P != F2End)
74 // If we stop on numbers, compare their difference. Note that some ugliness
75 // is built into this to permit support for numbers that use "D" or "d" as
76 // their exponential marker, e.g. "1.234D45". This occurs in 200.sixtrack in
78 if (isNumberChar(*F1P) && isNumberChar(*F2P)) {
82 V1 = strtod(F1P, &F1NumEnd);
83 V2 = strtod(F2P, &F2NumEnd);
85 if (*F1NumEnd == 'D' || *F1NumEnd == 'd') {
86 *F1NumEnd = 'e'; // Strange exponential notation!
89 if (*F2NumEnd == 'D' || *F2NumEnd == 'd') {
90 *F2NumEnd = 'e'; // Strange exponential notation!
93 } while (isDNotation);
95 // Otherwise, the diff failed.
100 if (F1NumEnd == F1P || F2NumEnd == F2P) {
102 *ErrorMsg = "FP Comparison failed, not a numeric difference between '";
104 *ErrorMsg += "' and '";
111 // Check to see if these are inside the absolute tolerance
112 if (AbsTolerance < std::abs(V1-V2)) {
113 // Nope, check the relative tolerance...
116 Diff = std::abs(V1/V2 - 1.0);
118 Diff = std::abs(V2/V1 - 1.0);
120 Diff = 0; // Both zero.
121 if (Diff > RelTolerance) {
123 *ErrorMsg = "Compared: " + ftostr(V1) + " and " + ftostr(V2) + "\n";
124 *ErrorMsg += "abs. diff = " + ftostr(std::abs(V1-V2)) +
125 " rel.diff = " + ftostr(Diff) + "\n";
126 *ErrorMsg += "Out of tolerance: rel/abs: " + ftostr(RelTolerance) +
127 "/" + ftostr(AbsTolerance);
133 // Otherwise, advance our read pointers to the end of the numbers.
134 F1P = F1NumEnd; F2P = F2NumEnd;
138 // PadFileIfNeeded - If the files are not identical, we will have to be doing
139 // numeric comparisons in here. There are bad cases involved where we (i.e.,
140 // strtod) might run off the beginning or end of the file if it starts or ends
141 // with a number. Because of this, if needed, we pad the file so that it starts
142 // and ends with a null character.
143 static void PadFileIfNeeded(char *&FileStart, char *&FileEnd, char *&FP) {
144 if (FileStart-FileEnd < 2 ||
145 isNumberChar(FileStart[0]) || isNumberChar(FileEnd[-1])) {
146 unsigned FileLen = FileEnd-FileStart;
147 char *NewFile = new char[FileLen+2];
148 NewFile[0] = 0; // Add null padding
149 NewFile[FileLen+1] = 0; // Add null padding
150 memcpy(NewFile+1, FileStart, FileLen);
151 FP = NewFile+(FP-FileStart)+1;
152 FileStart = NewFile+1;
153 FileEnd = FileStart+FileLen;
157 /// DiffFilesWithTolerance - Compare the two files specified, returning 0 if the
158 /// files match, 1 if they are different, and 2 if there is a file error. This
159 /// function differs from DiffFiles in that you can specify an absolete and
160 /// relative FP error that is allowed to exist. If you specify a string to fill
161 /// in for the error option, it will set the string to an error message if an
162 /// error occurs, allowing the caller to distinguish between a failed diff and a
163 /// file system error.
165 int llvm::DiffFilesWithTolerance(const sys::PathWithStatus &FileA,
166 const sys::PathWithStatus &FileB,
167 double AbsTol, double RelTol,
168 std::string *Error) {
169 const sys::FileStatus *FileAStat = FileA.getFileStatus(false, Error);
172 const sys::FileStatus *FileBStat = FileB.getFileStatus(false, Error);
176 // Check for zero length files because some systems croak when you try to
177 // mmap an empty file.
178 size_t A_size = FileAStat->getSize();
179 size_t B_size = FileBStat->getSize();
181 // If they are both zero sized then they're the same
182 if (A_size == 0 && B_size == 0)
185 // If only one of them is zero sized then they can't be the same
186 if ((A_size == 0 || B_size == 0)) {
188 *Error = "Files differ: one is zero-sized, the other isn't";
192 // Now its safe to mmap the files into memory becasue both files
193 // have a non-zero size.
195 if (F1.open(FileA, sys::MappedFile::READ_ACCESS, Error))
198 if (F2.open(FileB, sys::MappedFile::READ_ACCESS, Error))
205 // Okay, now that we opened the files, scan them for the first difference.
206 char *File1Start = F1.charBase();
207 char *File2Start = F2.charBase();
208 char *File1End = File1Start+A_size;
209 char *File2End = File2Start+B_size;
210 char *F1P = File1Start;
211 char *F2P = File2Start;
213 if (A_size == B_size) {
214 // Are the buffers identical?
215 if (std::memcmp(File1Start, File2Start, A_size) == 0)
218 if (AbsTol == 0 && RelTol == 0) {
220 *Error = "Files differ without tolerance allowance";
221 return 1; // Files different!
225 char *OrigFile1Start = File1Start;
226 char *OrigFile2Start = File2Start;
228 // If the files need padding, do so now.
229 PadFileIfNeeded(File1Start, File1End, F1P);
230 PadFileIfNeeded(File2Start, File2End, F2P);
232 bool CompareFailed = false;
234 // Scan for the end of file or next difference.
235 while (F1P < File1End && F2P < File2End && *F1P == *F2P)
238 if (F1P >= File1End || F2P >= File2End) break;
240 // Okay, we must have found a difference. Backup to the start of the
241 // current number each stream is at so that we can compare from the
243 F1P = BackupNumber(F1P, File1Start);
244 F2P = BackupNumber(F2P, File2Start);
246 // Now that we are at the start of the numbers, compare them, exiting if
248 if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error)) {
249 CompareFailed = true;
254 // Okay, we reached the end of file. If both files are at the end, we
256 bool F1AtEnd = F1P >= File1End;
257 bool F2AtEnd = F2P >= File2End;
258 if (!CompareFailed && (!F1AtEnd || !F2AtEnd)) {
259 // Else, we might have run off the end due to a number: backup and retry.
260 if (F1AtEnd && isNumberChar(F1P[-1])) --F1P;
261 if (F2AtEnd && isNumberChar(F2P[-1])) --F2P;
262 F1P = BackupNumber(F1P, File1Start);
263 F2P = BackupNumber(F2P, File2Start);
265 // Now that we are at the start of the numbers, compare them, exiting if
267 if (CompareNumbers(F1P, F2P, File1End, File2End, AbsTol, RelTol, Error))
268 CompareFailed = true;
270 // If we found the end, we succeeded.
271 if (F1P < File1End || F2P < File2End)
272 CompareFailed = true;
275 if (OrigFile1Start != File1Start)
276 delete[] (File1Start-1); // Back up past null byte
277 if (OrigFile2Start != File2Start)
278 delete[] (File2Start-1); // Back up past null byte
279 return CompareFailed;