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/ADT/StringExtras.h"
17#include "llvm/Support/Error.h"
18#include "llvm/Support/ErrorOr.h"
19#include "llvm/Support/MemoryBuffer.h"
20#include "llvm/Support/Process.h"
21#include "llvm/Support/raw_ostream.h"
22#include <cmath>
23#include <cstdint>
24#include <cstdlib>
25#include <cstring>
26#include <memory>
27#include <system_error>
28
29using namespace llvm;
30
31static bool isSignedChar(char C) {
32 return (C == '+' || C == '-');
33}
34
35static bool isExponentChar(char C) {
36 switch (C) {
37 case 'D': // Strange exponential notation.
38 case 'd': // Strange exponential notation.
39 case 'e':
40 case 'E': return true;
41 default: return false;
42 }
43}
44
45static bool isNumberChar(char C) {
46 switch (C) {
47 case '0': case '1': case '2': case '3': case '4':
48 case '5': case '6': case '7': case '8': case '9':
49 case '.': return true;
50 default: return isSignedChar(C) || isExponentChar(C);
51 }
52}
53
54static const char *BackupNumber(const char *Pos, const char *FirstChar) {
55 // If we didn't stop in the middle of a number, don't backup.
56 if (!isNumberChar(C: *Pos)) return Pos;
57
58 // Otherwise, return to the start of the number.
59 bool HasPeriod = false;
60 while (Pos > FirstChar && isNumberChar(C: Pos[-1])) {
61 // Backup over at most one period.
62 if (Pos[-1] == '.') {
63 if (HasPeriod)
64 break;
65 HasPeriod = true;
66 }
67
68 --Pos;
69 if (Pos > FirstChar && isSignedChar(C: Pos[0]) && !isExponentChar(C: Pos[-1]))
70 break;
71 }
72 return Pos;
73}
74
75/// EndOfNumber - Return the first character that is not part of the specified
76/// number. This assumes that the buffer is null terminated, so it won't fall
77/// off the end.
78static const char *EndOfNumber(const char *Pos) {
79 while (isNumberChar(C: *Pos))
80 ++Pos;
81 return Pos;
82}
83
84/// CompareNumbers - compare two numbers, returning true if they are different.
85static bool CompareNumbers(const char *&F1P, const char *&F2P,
86 const char *F1End, const char *F2End,
87 double AbsTolerance, double RelTolerance,
88 std::string *ErrorMsg) {
89 const char *F1NumEnd, *F2NumEnd;
90 double V1 = 0.0, V2 = 0.0;
91
92 // If one of the positions is at a space and the other isn't, chomp up 'til
93 // the end of the space.
94 while (isSpace(C: static_cast<unsigned char>(*F1P)) && F1P != F1End)
95 ++F1P;
96 while (isSpace(C: static_cast<unsigned char>(*F2P)) && F2P != F2End)
97 ++F2P;
98
99 // If we stop on numbers, compare their difference.
100 if (!isNumberChar(C: *F1P) || !isNumberChar(C: *F2P)) {
101 // The diff failed.
102 F1NumEnd = F1P;
103 F2NumEnd = F2P;
104 } else {
105 // Note that some ugliness is built into this to permit support for numbers
106 // that use "D" or "d" as their exponential marker, e.g. "1.234D45". This
107 // occurs in 200.sixtrack in spec2k.
108 V1 = strtod(nptr: F1P, endptr: const_cast<char**>(&F1NumEnd));
109 V2 = strtod(nptr: F2P, endptr: const_cast<char**>(&F2NumEnd));
110
111 if (*F1NumEnd == 'D' || *F1NumEnd == 'd') {
112 // Copy string into tmp buffer to replace the 'D' with an 'e'.
113 SmallString<200> StrTmp(F1P, EndOfNumber(Pos: F1NumEnd)+1);
114 // Strange exponential notation!
115 StrTmp[static_cast<unsigned>(F1NumEnd-F1P)] = 'e';
116
117 V1 = strtod(nptr: &StrTmp[0], endptr: const_cast<char**>(&F1NumEnd));
118 F1NumEnd = F1P + (F1NumEnd-&StrTmp[0]);
119 }
120
121 if (*F2NumEnd == 'D' || *F2NumEnd == 'd') {
122 // Copy string into tmp buffer to replace the 'D' with an 'e'.
123 SmallString<200> StrTmp(F2P, EndOfNumber(Pos: F2NumEnd)+1);
124 // Strange exponential notation!
125 StrTmp[static_cast<unsigned>(F2NumEnd-F2P)] = 'e';
126
127 V2 = strtod(nptr: &StrTmp[0], endptr: const_cast<char**>(&F2NumEnd));
128 F2NumEnd = F2P + (F2NumEnd-&StrTmp[0]);
129 }
130 }
131
132 if (F1NumEnd == F1P || F2NumEnd == F2P) {
133 if (ErrorMsg) {
134 *ErrorMsg = "FP Comparison failed, not a numeric difference between '";
135 *ErrorMsg += F1P[0];
136 *ErrorMsg += "' and '";
137 *ErrorMsg += F2P[0];
138 *ErrorMsg += "'";
139 }
140 return true;
141 }
142
143 // Check to see if these are inside the absolute tolerance
144 if (AbsTolerance < std::abs(x: V1-V2)) {
145 // Nope, check the relative tolerance...
146 double Diff;
147 if (V2)
148 Diff = std::abs(x: V1/V2 - 1.0);
149 else if (V1)
150 Diff = std::abs(x: V2/V1 - 1.0);
151 else
152 Diff = 0; // Both zero.
153 if (Diff > RelTolerance) {
154 if (ErrorMsg) {
155 raw_string_ostream(*ErrorMsg)
156 << "Compared: " << V1 << " and " << V2 << '\n'
157 << "abs. diff = " << std::abs(x: V1-V2) << " rel.diff = " << Diff << '\n'
158 << "Out of tolerance: rel/abs: " << RelTolerance << '/'
159 << AbsTolerance;
160 }
161 return true;
162 }
163 }
164
165 // Otherwise, advance our read pointers to the end of the numbers.
166 F1P = F1NumEnd; F2P = F2NumEnd;
167 return false;
168}
169
170/// DiffFilesWithTolerance - Compare the two files specified, returning 0 if the
171/// files match, 1 if they are different, and 2 if there is a file error. This
172/// function differs from DiffFiles in that you can specify an absolute and
173/// relative FP error that is allowed to exist. If you specify a string to fill
174/// in for the error option, it will set the string to an error message if an
175/// error occurs, allowing the caller to distinguish between a failed diff and a
176/// file system error.
177///
178int llvm::DiffFilesWithTolerance(StringRef NameA,
179 StringRef NameB,
180 double AbsTol, double RelTol,
181 std::string *Error) {
182 // Now its safe to mmap the files into memory because both files
183 // have a non-zero size.
184 ErrorOr<std::unique_ptr<MemoryBuffer>> F1OrErr = MemoryBuffer::getFile(Filename: NameA);
185 if (std::error_code EC = F1OrErr.getError()) {
186 if (Error)
187 *Error = EC.message();
188 return 2;
189 }
190 MemoryBuffer &F1 = *F1OrErr.get();
191
192 ErrorOr<std::unique_ptr<MemoryBuffer>> F2OrErr = MemoryBuffer::getFile(Filename: NameB);
193 if (std::error_code EC = F2OrErr.getError()) {
194 if (Error)
195 *Error = EC.message();
196 return 2;
197 }
198 MemoryBuffer &F2 = *F2OrErr.get();
199
200 // Okay, now that we opened the files, scan them for the first difference.
201 const char *File1Start = F1.getBufferStart();
202 const char *File2Start = F2.getBufferStart();
203 const char *File1End = F1.getBufferEnd();
204 const char *File2End = F2.getBufferEnd();
205 const char *F1P = File1Start;
206 const char *F2P = File2Start;
207 uint64_t A_size = F1.getBufferSize();
208 uint64_t B_size = F2.getBufferSize();
209
210 // Are the buffers identical? Common case: Handle this efficiently.
211 if (A_size == B_size &&
212 std::memcmp(s1: File1Start, s2: File2Start, n: A_size) == 0)
213 return 0;
214
215 // Otherwise, we are done a tolerances are set.
216 if (AbsTol == 0 && RelTol == 0) {
217 if (Error)
218 *Error = "Files differ without tolerance allowance";
219 return 1; // Files different!
220 }
221
222 bool CompareFailed = false;
223 while (true) {
224 // Scan for the end of file or next difference.
225 while (F1P < File1End && F2P < File2End && *F1P == *F2P) {
226 ++F1P;
227 ++F2P;
228 }
229
230 if (F1P >= File1End || F2P >= File2End) break;
231
232 // Okay, we must have found a difference. Backup to the start of the
233 // current number each stream is at so that we can compare from the
234 // beginning.
235 F1P = BackupNumber(Pos: F1P, FirstChar: File1Start);
236 F2P = BackupNumber(Pos: F2P, FirstChar: File2Start);
237
238 // Now that we are at the start of the numbers, compare them, exiting if
239 // they don't match.
240 if (CompareNumbers(F1P, F2P, F1End: File1End, F2End: File2End, AbsTolerance: AbsTol, RelTolerance: RelTol, ErrorMsg: Error)) {
241 CompareFailed = true;
242 break;
243 }
244 }
245
246 // Okay, we reached the end of file. If both files are at the end, we
247 // succeeded.
248 bool F1AtEnd = F1P >= File1End;
249 bool F2AtEnd = F2P >= File2End;
250 if (!CompareFailed && (!F1AtEnd || !F2AtEnd)) {
251 // Else, we might have run off the end due to a number: backup and retry.
252 if (F1AtEnd && isNumberChar(C: F1P[-1])) --F1P;
253 if (F2AtEnd && isNumberChar(C: F2P[-1])) --F2P;
254 F1P = BackupNumber(Pos: F1P, FirstChar: File1Start);
255 F2P = BackupNumber(Pos: F2P, FirstChar: File2Start);
256
257 // Now that we are at the start of the numbers, compare them, exiting if
258 // they don't match.
259 if (CompareNumbers(F1P, F2P, F1End: File1End, F2End: File2End, AbsTolerance: AbsTol, RelTolerance: RelTol, ErrorMsg: Error))
260 CompareFailed = true;
261
262 // If we found the end, we succeeded.
263 if (F1P < File1End || F2P < File2End)
264 CompareFailed = true;
265 }
266
267 return CompareFailed;
268}
269
270Expected<FilePermissionsApplier>
271FilePermissionsApplier::create(StringRef InputFilename) {
272 sys::fs::file_status Status;
273
274 if (InputFilename != "-") {
275 if (auto EC = sys::fs::status(path: InputFilename, result&: Status))
276 return createFileError(F: InputFilename, EC);
277 } else {
278 Status.permissions(p: static_cast<sys::fs::perms>(0777));
279 }
280
281 return FilePermissionsApplier(InputFilename, Status);
282}
283
284Error FilePermissionsApplier::apply(
285 StringRef OutputFilename, bool CopyDates,
286 std::optional<sys::fs::perms> OverwritePermissions) {
287 sys::fs::file_status Status = InputStatus;
288
289 if (OverwritePermissions)
290 Status.permissions(p: *OverwritePermissions);
291
292 int FD = 0;
293
294 // Writing to stdout should not be treated as an error here, just
295 // do not set access/modification times or permissions.
296 if (OutputFilename == "-")
297 return Error::success();
298
299 if (std::error_code EC = sys::fs::openFileForWrite(Name: OutputFilename, ResultFD&: FD,
300 Disp: sys::fs::CD_OpenExisting))
301 return createFileError(F: OutputFilename, EC);
302
303 if (CopyDates)
304 if (std::error_code EC = sys::fs::setLastAccessAndModificationTime(
305 FD, AccessTime: Status.getLastAccessedTime(), ModificationTime: Status.getLastModificationTime()))
306 return createFileError(F: OutputFilename, EC);
307
308 sys::fs::file_status OStat;
309 if (std::error_code EC = sys::fs::status(FD, Result&: OStat))
310 return createFileError(F: OutputFilename, EC);
311 if (OStat.type() == sys::fs::file_type::regular_file) {
312#ifndef _WIN32
313 // Keep ownership if llvm-objcopy is called under root.
314 if (OutputFilename == InputFilename && OStat.getUser() == 0)
315 sys::fs::changeFileOwnership(FD, Owner: Status.getUser(), Group: Status.getGroup());
316#endif
317
318 sys::fs::perms Perm = Status.permissions();
319 if (OutputFilename != InputFilename)
320 Perm = static_cast<sys::fs::perms>(Perm & ~sys::fs::getUmask() & ~06000);
321#ifdef _WIN32
322 if (std::error_code EC = sys::fs::setPermissions(OutputFilename, Perm))
323#else
324 if (std::error_code EC = sys::fs::setPermissions(FD, Permissions: Perm))
325#endif
326 return createFileError(F: OutputFilename, EC);
327 }
328
329 if (std::error_code EC = sys::Process::SafelyCloseFileDescriptor(FD))
330 return createFileError(F: OutputFilename, EC);
331
332 return Error::success();
333}
334